Handbook of Personality Psychology

HANDBOOK OF PERSONALITY PSYCHOLOGY PSYCHOLOGY HANDBOOK This Page Intentionally Left Blank HANDBOOK OF HANDBOOK OF P...

Author: Robert Hogan | John Johnson | Stephen Briggs

1024 downloads 6491 Views 20MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

HANDBOOK OF PERSONALITY PSYCHOLOGY PSYCHOLOGY HANDBOOK

This Page Intentionally Left Blank

HANDBOOK OF HANDBOOK OF PERSONALITY PERSONALITY PSYCHOLOGY PSYCHOLOGY Edited by

ROBERT HO HOGAN ROBERT GAN UNIVERSITY UNIVERSITY OF OF TULSA TULSA TULSA, OKLAHOMA OKLAHOMA TULSA,

JOHN JOHNSON JOHN JOHNSON PENNSYLVANIA PENNSYLVANIA STATE STATE UNIVERSITY UNIVERSITY DuBOIS, PENNSYLVANIA DuBois, PENNSYLVANIA

STE PHEN B RIGGS STEPHEN BRIGGS UNIVERSITY UNIVERSITY OF OF TULSA TULSA TULSA, TULSA, OKLAHOMA OKLAHOMA

ACADEMIC ACADEMIC PRESS PRESS An An Imprint Imprint of of Elsevier Elsevier

SAN SAN DIEGO DIEGO

LONDON LONDON

BOSTON BOSTON

NEW NEW YORK YORK

SYDNEY SYDNEY

TOKYO TOKYO

TORONTO TORONTO

e

book is printed on on acid-free acid-free paper. @ This book Copyright e © 1997 1997 by ACADEMIC PRESS All Rights Reserved. No part of of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information without permission in writing from the publisher. storage and retrieval system, without Permissions may be sought directly from Elsevier's Elsevler'a Science and Technology Rights

Department In in Oxford, Oxford, UK. UK. Phone: Phone: (44) (44) 1865843830, 1865 843830, Fax: Fax: (44) (44) 1885853333, 1865 853333. e-mail: e-mail: Department DElnn lsslon8@e!sevler.co,yk, You may also complete your request on-lIne pemilssions^elsevier.co.ul^. on-line via the Elsevier homepage: bttp:llwww,elsev ler,corn by selecting ·Customer homepage: http://www.el8evier.com "Customer Support" and then "Obtaining Permissions'*. ·Obtalnlng Permissions·,

Academic Press, of Elsevier An Imprint of

S2S B Street, Suite 1900, 525 1900, San Diego, California 92101-449S, 92101-4495, USA USA http://www.apnet.com http://www.apnet.com

Academic Press Limited 24-28 NWl 7DX, UK 24- 28 Oval Road, London NWI http://www.hbuk.co.uk/ap/ http://www.hbuk,co,uklapl Library of of Congress Cataloging-in-Publication Data Handbook of of personality psychology I/ edited by Robert Hogan, John Johnson, Stephen Briggs, Briggs. cm. p p.. Includes index. ISBN-13: 978-0-12-134645-4 978-0-12-134645-4

ISBN-13: ISBN-13: 978-0-12-134646-1 978-0-12-134646-1 ISBN-lO: 0-12-134645-5 (hard) ISBN-10: (hard) ISBN-lO: ISBN-10: 0-12-134646-3 0-12-134646-3 (soft) (soft)

1. 1. Personality. I. L Hogan, Robert. in. Briggs, Stephen R. III. BF698.H3345 B F698,H3345 11995 995 I155.2"dc20 SS.2--dc20

II. IL Johnson, John (John A.)

94-39181 94-39181

CIP CIP

ISBN-13: 978-0-12-134645-4 978-0-12-134645-4 ISBN-lO: ISBN-IO: 0-12-134645-5 0-12-134645-5 (hard) (hard) ISBN-13: 978-0-12-134646-1 ISBN-I0: ISBN-10: 0-12-134646-3 0-12-134646-3 (soft) (soft)

PRINTED PRINTED IN INmE THEUNITED UNTIEDSTATES STATESOF OFAMERICA AMERICA 05 06 07 08 09 EB II 11 10 9 8 7

CONTENTS CONTENTS

xix CONTRIBUTORS CONTRIBUTORS xix PREFACE xxiii PREFACE xxiii

PART. PART I

INTRODUCTION INTRODUCTION NATURE OF PERSONALITY PERSONALITY PSYCHOLOGY NATURE OF PSYCHOLOGY

CHAPTER CHAPTER 11

A A CONCEPTUAL HISTORY OF PERSONALITY PSYCHOLOGY DAN DAN P. P. McADAMS MCADAMS

I. PERSONALnY PERSONALTTY PSYCHOLOGY'S PSYCHOLOGY'S UNIQUE UNIQUE FEATURES: HOLISM, MOTIVATION, AND INDMDUAL INDIVIDUAL DIFFERENCES

44

II. HISTORICAL ROOTS: BEFORE THE 1930s 1930s III. THE FORMATION OF SYSTEMS: 1930-1950 1930-1950

55 7

IV. THE ELABORATION OF CONSTRUCTS: 1950-1970 1950-1970

13 13

V. DOUBT AND A A RENEWAL OF CONFIDENCE: CONHDENCE: 1970 1970 TO

19 19

THE PRESENT

VI. CONCLUSIONS: PROGRESS AND STAGNATION REFERENCES

29 29

,

27 27

vi VI

CONTENTS CONTENTS

CHAPTER CHAPTER 2 2

STUDYING LIVES: PSYCHOBIOGRAPHY AND THE CONCEPTUAL STRUCTURE OF PERSONALITY PSYCHOLOGY WILLIAM WILLIAM MCKINLEY MCKINLEY RUNYAN RUNYAN 41 41

I. INTRODucnON INTRODUCTION I.

II. THE CONCEPTUAL CONCEPTUAL STRUCTURE STRUCTURE OF PERSONALITY PSYCHOLOGY II.

43 43

47 47

III. III. PROGRESS IN PSYCHOBIOGRAPHICAL PSYCHOBIOGRAPHICAL INQUIRY

IV. RELATIONSHIPS RELATIONSHIPS BETWEEN THE STUDY STUDY OF LIVES AND OrnER OTHER AREAS AREAS OF IV.

53 53

PERSONALITY PERSONALITY PSYCHOLOGY PSYCHOLOGY

V. A "SOFT V. THE THE STUDY STUDY OF OF LIVES LIVES AS AS A "Son SYNTHESIS" SYNTHESIS" IN IN PSYCHOLOGY PSYCHOLOGY 63 VI. CONCLUSION VI. CONCLUSION 63 REFERENCES REFERENCES

PARTll PARTII

64 64

CON CONCEPTUAL AND MEASUREMENT MEASUREMENT ISSUES ISSUES CEPTUAL AND

IN IN CHAPTER3 CHAPTER 3

PERSONALITY PERSONALITY

UNITS OF ANALYSIS FOR THE DESCRIPTION AND EXPLANATION OF PERSONALITY JOHN JOHN A. A. JOHNSON JOHNSON 73 73

I. THE NEED FOR I. FOR UNITS OF ANALYSIS 74 74

II. TRAITS AS II. AS UNITS OF ANALYSIS

III. III. ISSUES IN THE ASSESSMENT OF TRAITS IV. ALTERNATIVES TO TRAITS IV.

REFERENCES CHAPTER 4 4

59 59

79 79

86 86

89 89

IN DEFENSE OF TRAITS JE RRY S. JERRY S. WIGGINS WIGGINS

I. I. THEORIES AND AND VIEWPOINTS

97 97

II. TRAITS AS AS ATTRIBUTES AHRIBUTES OF BEHAVIOR BEHAVIOR II. III. TRAITS AS III. AS ATTRIBUTES OF PERSONS

PREDiaoRS OF BEHAVIOR IV. TRAITS AS PREDICTORS OF BEHAVIOR

99 99 102 102

108 108

vii vii

CONTENTS CONTENTS

V. TRAITS AS AS EXPLANATIONS OF OF BEHAVIOR V. REFERENCES

110 110

113 113

CHAPTER 5 INDIVIDUALS THE CHAFFERS I N D I V I D U A L S AND AND T H E DIFFERENCES D I F F E R E N C E S BETWEEN B E T W E E N THEM THEM

JAMES JAMES T. T . LAMlELL LAMIELL I. WHY INDIVIDUAL DIFFERENCES RESEARCH CANNOT ADVANCE PERSONALITY THEORY

118 118

II. SOME ISSUES IN NEED OF OF CLARIFICATION

123 123

III. "IDIOTHETIC" INQUIRY INQUIRY AS AS AN AN ALTERNATIVE TO TRADITIONAL ALTERNATIVE TO TRADITIONAL "NOMOTHETICISM" "NOMOTHETICISM" III. "IDIOTHETIC" REFERENCES

CHAPTER 66

130 130

138 138

PERSONALITY MEASUREMENT: RELIABILITY AND VALIDITY ISSUES AND STEPHEN STEPHEN G. G . WEST WEST JOHN JOHN F. F . FINCH FINCH

I. THE NATURE OF PERSONALITY CONSTRUCTS: CoNSTRuas: BASIC ISSUES II. RELIABILITY III. III. VALIDITY

145 145

150 150

IV. IV. CONCLUSION

159 159

REFERENCES

160 160

CHAPTER 7 7 CHAPTER

143

SITUATIONS PERSONALITY INFLUENCES ON THE CHOICE OF SITUATIONS WILLIAM WILLIAM ICKES ICKES MARK MARK SNYDER SNYDER STELLA STELLA GARCIA GARCIA 166 166

I. WHAT DOES "CHOOSING SITUATIONS" MEAN? II. CONCEPTUAL CONCEPTUAL AND AND METHODOLOGICAL ISSUES ISSUES II.

167 167

III. ASSESSING THE THE DIMENSIONS OF OF SOCIAL SITUATIONS III.

172 172

IV. OF SITUATIONAL SITUATIONAL CHOICES CHOICES IV. UNDERSTANDING INDIVIDUALS IN TERMS OF V. CONCLUSION CONCLUSION V.

185 185

REFERENCES

187 187

175 175

viii viii

CONTENTS CONRNTS

PART m PARTIII CHArrER CHAPTERS8

DEVELOPMENTAL DEVELOPMENTAL ISSUES ISSUES STAGES OF PERSONALITY DEVELOPMENT JANE JANE LoEVINGER LOEVINGER 199 199

I. I. MEASUREMENT OF PERSONALI'IY PERSONALITY

200 200

II. PERSONALITY STAGES AND II. PERSONALITY AND TYPES

III. KOHLBERG: MORAL JUDGMENT III. KOHLBERG; STAGES OF MORAL IV. THEORY: EGO EGO DEVELOPMENT IV. A A STAGE-TypE STAGE-TYPE THEORY: DEVELOPMENT 205 205

VI. SOURCES OF ERROR VI.

VII. STAGES VERSUS FACTORS VERSUS FACTORS

CHAPTER 9 9 CHArrER

201 201

202 202

V. V. STAGES OF EGO DEVELOPMENT

REFERENCES

201 201

206 206

207 207

THE EMOTIONAL BASIS OF EARLY PERSONALITY THE DEVELOPMENT: IMPLICATIONS FOR mE SELF-CONCEPT EMERGENT SELF·CONCEPT REBECCA REBECCA A. A . EDER EDER SARAH SARAH C. C . MANGELSDORF MANGELSDORF 209 209

I. THE EMOTIONAL BASIS OF EARLY PERSONALI1Y PERSONALITY DEVELOPMENT I.

II. INFANT PERSONALITY DEVELOPMEN1: DEVELOPMENT: FIRST SIGNS OF THE SELF-CONCEPT II.

210 210

III. THE TRANSmON TRANSITION BElWEEN BETWEEN INFANCY AND CmLDHOOD: CHILDHOOD: TODDLERS' III.

SOCIAL UNDERSTANDING SOCIAL UNDERSTANDING

221 221

N. PERSONALITY DEVELOPMENT DEVELOPMENT IN 221 IV. PERSONALITY IN EARLY EARLY CmLDHooD CHILDHOOD 221 V. V. PurnNG PUTTING It IT ALL A L L TOGETHER: TOGETHER: A A MODEL MODEL OF OF PERSONALITY PERSONALITY DEVELOPMENT DEVELOPMENT VI. VI. CONCLUSION CONCLUSION

234 234

REFERENCES REFERENCES

234 234

10 CHAPTER 10 CHArrER

228 228

FAM^Y INFLUENCES ON PERSONALITY DEVELOPMENT FAMILY CHARLES CHARLES F. F . HALVERSON, HALVERSON, JR. JR. K A R E N S. S . WAMPLER WAMPLER KAREN

I. INTRODUCTION I.

241 241

II. II. THREE CRISES FOR FOR PERSONALI1Y PERSONALITY DEVELOPMENT RESEARCH REFERENCES

260 260

242 242

ix IX

CONTENTS CONTENTS

CHAPTERll CHAPTER 11

LONGITUDINAL LONGITUDINAL STABILITY STABILITY OF ADULT ADULT PERSONALITY PERSONALITY PAUL P A U L T. T . COSTA, COSTA, JR. JR. ROBERT ROBERT R. R . MCCRAE MCCRAE

I. THE DEFINmON DEFINITIGN AND ASSESSMENT OF PERSONALITY I.

II. II. STABILITY OF OF MEAN MEAN LEVELS

269 269

271 271

III. OF INDIVIDUAL DIFFERENCES III. STABILITY OF

275 275

IV. ADULT IV. A D U L T DEVELOPMENT: CHANGE IN THE 220s 0S

279 279

V. ALTERNATIVE V. ALTERNATIVE ApPROACHES APPROACHES TO STABILITY OR CHANGE VI. IMPLICATIONS OF PERSONALITY STABILITY VI. SOME IMPLICATIONS REFERENCES REFERENCES

12 CHAPTER 12

280 280

283 283

285 285

THERE ADULT ADULT DEVELOPMENT DEVELOPMENT DISTINCTIVE DISTINCTIVE Is THERE TO WOMEN? TO WOMEN? RAVENNA R A V E N N A HELSON HELSON JENNIFER JENNIFER PALS PALS MARJORIE MARJORIE SOLOMON SOLOMON

I. ASSUMPTIONS I. SOME ORGANIZING ORGANIZING ASSUMmoNs

293 293

II. SOME IDEAS AND THEORIES ABOUT WOMEN'S WOMEN'S A D U L T DEVELOPMENT DEVELOPMENT II. ADULT

III. WOMEN'S TO THE WOMEN'S LIVES FROM THE 1920s 1920s TO THE 1990s 1990s

297 297

IV. SOME POSSIBLE GENERAL GENERAL FEATURES FEATURES OF WOMEN'S WOMEN'S ADULT A D U L T DEVELOPMENT DEVELOPMENT IV. V. RELATIONAL AND VOCATIONAL VOCATIONAL GOALS OF YOUNG YOUNG WOMEN WOMEN V. RELATIONAL

VI. GENDER-RELATED VI. GENDER-RELATED PERSONALITY PERSONALITY CHANGE CHANGE VII. OVERVIEW VII. OVERVIEW AND AND IMPLICATIONS IMPUCATIONS

VIII. ApPENDIX APPENDIX REFERENCES REFERENCES

PART IV PARTIV

CHAPTER CHAPTER 13 13

293 293 299 299

299 299

304 304

307 307

309 309 310 310

BIOLOGICAL RMINANTS OF BIOLOGICAL DETE DETERMINANTS OF PERSONALITY PERSONALITY EVOLUTIONARY EVOLUTIONARY FOUNDATIONS FOUNDATIONS OF PERSONALITY PERSONALITY D A V I D M. M . Buss BUSS DAVID

I. EVOLUTION: FACT, PATH, THEORY, AND HYPOTHESIS HYPOTHESIS I. II. OF EVOLUTIONARY PSYCHOLOGY II. BASICS OF

321 321

318 318

x

CONTENTS

III. FOR PERSONALITY PSYCHOLOGY III. FOUNDATIONAL IMPLICATIONS FOR

IV. IV. CONCLUSIONS CONCLUSIONS REFERENCES

CHAPTER CHAPTER 14 14

340 340

EVOLUTIONARY PERSPECTIVES PERSPECTIVES ON PERSONALITY PERSONALITY TRAITS TRAITS ARNOLD ARNOLD H. H . Buss BUSS 346 346

I. THE COMMON HERITAGE

354 354

11. EVOLUTIONARY TRENDS II. EVOLUTIONARY

III. HUMAN TENDENCIES III. DISTINCTIVELY HUMAN IV. THREE PERSPECTIVES IV.

REFERENCES

CHAPTER 15 15 CHAPTER

327 327

339 339

355 355

363 363

364 364

TEMPERAMENT, AND AND PERSONALITY PERSONALITY GENETICS, TEMPERAMENT, D A V I D C. C . ROWE ROWE DAVID

I. SEPARATING NATURE AND AND NURTURE

369 369

II. II. GENETIC VARIATION IN TEMPERAMENTAL AND AND PERSONALITY TRAITS

III. III. FURTHER TOPICS OF INTEREST REFERENCES

CHAPTER 16 16 CHAPTER

374 374

380 380

384 384

PSYCHOPHYSIOLOGICAL ApPROACHES APPROACHES TO TO PERSONALITY PERSONALITY RUSSELL RUSSELL G. G . GEEN GEEN 387 387

I. INTRODUCTION

II. II. THEORIES OF AROUSABILITY

388 388

III. CRITICISM CRmCISM OF THE THE AROUSABILITY HYPOTHESIS III. IV. OF PERSONALITY PSYCHOPHYSIOLOGICAL STUDY OF IV. PSYCHOPHYSIOLOGICAL V. V. SUMMARY

REFERENCES

406 406 408 408

392 392 393 393

CONTENTS

xi xi

P ARTVV SO SOCIAL PART DETERMINANTS OF PERSONALITY CIAL DETERMINANTS 17 CHAPTER 17

AND SOCIAL STRUCTURE: SOCIAL SOCLVL PERSONALITY AND PSYCHOLOGICAL CONTRIBUTIONS THOMAS F. F . PETIIGREW PETTIGREW THOMAS 417 417

INTRODUCTION I. INTRODUCflON

II. AN APPARENT PARADOX A N ApPARENT

420 420

III. III. PERSONALITY SHAPES SOCIAL STRUCTURE

425 425

IV. IV. SOCIAL STRUCTURE SHAPES PERSONALITY

428 428

V. PERSONALITY AND SOCIAL STRUCTURE SHAPE TOGETHER V.

433 433

VI. ADVANTAGES OF THE PERSONALITY PERSONALITY AND SOCIAL STRUCTURE PERSPECflVE PERSPECTIVE VI. REFERENCES CHAPTER CHAPTER 18 18

434 434

CROSS-CULTURAL PERSPECTIVES ON PERSONALITY HARRY HARRY C. C . TRIANDIS TRIANDIS 440 440

CONSIDERATIONS I. PRELIMINARY CONSIDERATIONS

442 442

II. DIMENSIONS OF CULTURAL VARIATION III. DIMENSIONS DIMENSIONS OF SOCIAL BEHAVIOR

450 450

IV. FRAMEWORK FOR STUDIES OF CULTURE AND PERSONALITY IV. A A FRAMEWORK V. SUMMARY V. REFERENCES CHAPTER 19 19

433 433

456 456

459 459 459 459

THE PSYCHOLOGY OF GETTING ALONG TRUST AND AND BETRAYAL: THE AND AND GETTING AHEAD WARREN WARREN H. H . JONES JONES LAURIE LAURIE COUCH COUCH SUSAN SUSAN SCOTI SCOTT

I. PSYCHOLOGY OF PERSONALITY II. THE PSYCHOLOGY OF TRUST III. THE PSYCHOLOGY OF BETRAYAL IV. IV. CONCLUSION

480 480

REFERENCES

481 481

466 466 468 468

475 475

xii xii

CONTENTS CONTENTS

PART VI PART VI CHAPTER 20 CHAPTER 20

DYNAMI PROCESSES DYNAMIC PERSONALITY PROCESSES C PERSONALITY MOTIVES AND GOALS ROBERT ROBERT A. A. EMMONS EMMONS 486 486

I. THE I. THE MOTIVE MOTIVE DISPOsmON DISPOSITION ApPROACH APPROACH

489 489

II. II. METATHEORETICAL METATHEORETICAL PERSPECTIVES PERSPECTIVES ON ON GOALS GOALS 492 492

OF PERSONALITY IIII. I I GOAL GOAL ApPROACHES APPROACHES OF PERSONALITY IV. CONCLUSIONS AND DIRECTIONS IV. CONCLUSIONS AND FUTURE FUTURE DIRECTIONS 506 506

REFERENCES

21 CHAPTER 21

505 505

A N INTEGRATIVE ApPROACH APPROACH THE EMOTIONS: AN JAMES JAMES R. R . AVERILL AVERILL 513 513

I. INTRODUCTION I.

513 513

II. II. THE DOMAIN OF OF EMOTION

III. A FRAMEWORK FOR THE ANALYSIS OF EMOTION III. A

524 524

IV. IV. COMPONENT COMPONENT RESPONSES RESPONSES 533 533

V. V. EMOTIONAL STATES

VI. EMOTIONAL VI. EMOTIONAL SYNDROMES SYNDROMES

534 534

VII. EMOTIONAL VII. EMOTIONAL POTENTIALS POTENTIALS AND AND CAPAcmES CAPACITIES

VIII. CONCLUDING CONCLUDING OBSERVATIONS OBSERVATIONS

536 536

537 537

537 537

REFERENCES REFERENCES CHAPTER 22 22

519 519

PSYCHOLOGICAL DEFENSE: CONTEMPORARY THEORY AND AND RESEARCH RESEARCH DELROY DELROY L. L PAULHUS PAULHUS BRAM BRAM FRlDHANDLER FRIDHANDLER SEAN SEAN HAYES HAYES

I. I. CoNTEMPORARY CONTEMPORARY PSYCHOANALYSIS PSYCHOANALYSIS 545 545

II. II. REPRESSION

III. DENIAL III. DENIAL

544 544

546 546

IV. IV. SELF-DECEPTION SELF-DECEPnoN

547 547

V. TRAIT AND TYPE V. TRAIT AND TYPE ApPROACHES APPROACHES

548 548

xiii xiii

CONTENTS CONI'ENTS

VI. HAAN AND COLLEAGUES VI. HAAN

551 551

VII. VII. VAILLANT AND AND COLLEAGUES

553 553

VIII. HOROWITZ AND AND COLLEAGUES

555 555

IX. IX. ERDELYI AND COLLEAGUES

556 556

X. IHILEVICH X. IHILEVICH AND AND GLESER GLESER

557 557

XI. PERCEYf-GENETIC XL PERCEPT-GENETIC ApPROACH APPROACH

558 558

XII. AND COLLEAGUES XII. PLUTCHIK PLUTCHIK AND COLLEAGUES

559 559

XIII. STRESS STRESS AND COPING COPING

560 560

XIV. INFORMATION-PROCESSING INFORMATION-PROCESSING ApPROACHES XIV, APPROACHES XV. SOCIAL PSYCHOLOGY XV.

562 562

563 563

XVI. DEVELOPMENTAL ANALYSES XVI. DEVELOPMENTAL

565 565

THEORETICAL SYSTEMS SYSTEMS XVII. RELATIONS AMONG AMONG THEORETICAL

XVIII. CONCLUSION

567 567

REFERENCES

568 568

23 CHAPTER 23

566 566

iNfflBmoNS AND CONTROLS INTERNAL INmBmoNs E D W I N I. I. MEGARGEE MEGARGEE EDWIN

I. INTRODUCTION I.

581 581

II. DEFINmONAL DEHNITIONAL ISSUES: WHAT Do D O WE W E MEAN MEAN BY INTERNAL CONTROLS? II. III. PHILOSOPHICAL III. PmLOSOPHICAL ISSUES

584 584

IV. ORIGINS OF INTERNAL INHIBmoNs INHIBITIONS IV. V. INHIBITIONS V. OVERCOMING INHIBmoNs

597 597

VI. VI. METHODOLOGICAL PROBLEMS

REFERENCES PARTVD PARTVn CHAPTER 24 CHAPTER 24

586 586

606 606

611 611

PER PERSONALITY AND THE THE SELF SELF SONALITY AND CONGRUENCE OF CONGRUENCE OF OTHERS' OTHERS' AND AND SELF-JUDGMENTS SELF-JUDGMENTS OF PERSONALITY DAVID D A V I D C. C . FUNDER FUNDER C. C. RANDALL R A N D A L L COLVIN COLVIN

I. THE QUESTION OF SELF-OrnER SELF-OTHER AGREEMENT I.

)

II. ANALYSES OF ABSOLUTE ((MEAN) II. MEAN AGREEMENT

617 617 622 622

582 582

xiv XIV

CONTENTS CONTENTS

III. III. CORRELATIONAL CORRELATIONAL ANALYSES ANALYSES OF OF AGREEMENT AGREEMENT IV. CONCEPfUAL IV. CONCEPTUAL ISSUES ISSUES REFERENCES REFERENCES

628 628

639 639

641 641

CHAPTER 25 THE T H E QUEST Q U E S T FOR FOR SELF-INSIGHT: S E L F - I N S I G H T : THEORY T H E O R Y AND AND RESEARCH RESEARCH ON ON CHAPTER 2S

ACCURACY AND BIAS IN SELF.PERCEPTION SELF-PERCEPTION

RICHARD RICHARD W. W . ROBINS ROBINS OLIVER OLIVER P. P. JOHN JOHN I. INTRODUCTION

649 649

II. IS ACCURACY AND How HOW SHOULD IT BE MEASURED? II. WHAT Is

650 650

III. SELF-PERCEFll SELF-PERCEPTION PROCESSES: FOUR METAPHORS OF III. ON PROCESSES: OF THE SELF-PERCEIVER IV. AN AN ApPLICATION APPLICATION OF OF THE FOUR METAPHORS TO IV. TO SELF-ENHANCEMENT BIAS V. V. CONCLUSION

673 673

REFERENCES

674 674

656 656 667 667

CHAFrER26 IDENTITY, A N D SELF-EsTEEM: SELF-ESTEEM: THE T H E SELF SELF LOST LOST CHAPTER26 IDENTITY, SELF-CONCEPT, SELF-CONCEPT, AND AND FOUND FOUND AND ROY F. F . BAUMEISTER BAUMEISTER Roy

I. DEFINmoNs I. DEHNITIONS

681 681

II. FINDING THE THE SELF: OF IDENTITY IDENTITY II. FINDING SELF: CREATION CREATION OF

682 682

III. THE SELF: OF SELF-CoNCEPT AND IDENTITY IDENTITY III. LOSING LOSING THE SELF: PROBLEMS PROBLEMS OF SELF-CONCEPT AND IV. AND IV. SUMMARY SUMMARY A ND CONCLUSION CONCLUSION REFERENCES REFERENCES

CHAPTER 27 CHAPTER17

691 691

701 701

703 703

AND PERSONALITY PERSONALITY MENTAL REPRESENTATIONS OF PERSONS AND JOHN JOHN F. F . KIHLSTROM KIHLSTROM REID REID HASTIE HASTIE

L OF SOCIAL COGNmoN COGNITION I. THE DOMAIN OF

II. VIEWS OF SOCIAL II. ALTERNATE ALTERNATE VIEWS SOCIAL COGNmoN CoGNmoN III. PERSON PERSON PERCEFll ON III. PERCEPTION

717 717

712 712

713 713

XV XV

CONTENIS CONTENTS

IV. PERSON CONCEPTS IV.

720 720

V. PERSON MEMORY V.

722 722

VI. ApPROACH TO SOCIAL COGNITION V I . PROSPECTS PROSPEAS FOR THE INFORMATION PROCESSING APPROACH REFERENCES

CHAPTER 28 28 CHAFfER

728 728

PERSONALITY STRUCTURE: STRUCTURE: THE THE RETURN RETURN OF OF THE THE BIG BIG FIVE FIVE JERRY JERRY S. S. WIGGINS WIGGINS PAUL PAUL D. D . TRAPNELL TRAPNELL

I. INTRODUCTION I.

737 737

II. THEORETICAL PERSPECTIVES ON THE BIG FIVE II. III. CONCLUSION III.

758 758

REFERENCES

758 758

CHAFfER CHAPTER 29 29

744 744

EXTRAVERSION EXTRAVERSION AND A N D ITS I T S POSITIVE EMOTIONAL EMOTIONAL CORE CORE DAVID WATSON WATSON DAVID L E E ANNA A N N A CLARK CLARK LEE

I. INTRODUCTION

767 767

II. EARLY CONCEPTIONS OF EXTRAVERSION II.

768 768

III. III. CONTEMPORARY CONCEPTIONS OF EXTRAVERSION IV. IV. AN A N INTEGRATIVE MODEL OF THE CONSTRUCT V. EXPLORING THE CORE OF EXTRAVERSION V.

VII. VII. IMPLICATIONS AND AND CONCLUSIONS REFERENCES

772 772

774 774 777 777

VI. ORIGINS OF THE HIGHER-ORDER VI. HIGHER-ORDER CONSTRUCT

CHAYI'ER CHAPTER 30 30

727 727

786 786

788 788

790 790

AGREEABLENESS: AGREEABLENESS: A A DIMENSION DIMENSIONOF OFPERSONALITY PERSONALITY WILLIAM WILLIAM G. G . GRAZIANO GRAZIANO NANCY NANCY H. H . EISENBERG EISENBERG

I. CONCEPTUALIZATIONS OF AGREEABLENESS

795 795

II. THEORETICAL PERSPECTIVES ON AGREEABLENESS II.

797 797

xvi xvi

CONTENTS III. THE PROSOCIAL PERSONALfIY PERSONALITY

808 808

IV. SUMMARY AND CONCLUSIONS

815 815

REFERENCES

CHAPI'ER CHAPTER 31 31

817 817

CONCEPTIONS CONCEPTIONS AND AND CORRELATES CORRELATESOF OF OPENNESS OPENNESS TO TO EXPERIENCE EXPERIENCE ROBERT ROBERT R. R . MCCRAE MCCRAE PAUL PAUL T. T . COSTA COSTA,, JR. JR.

I. OPENNESS AS A A BASIC DIMENSION OF OF PERSONALfIY PERSONALITY TRADITIONAL CONCEmONS CoNCEPnoNS OF OF OPENNESS OPENNESS II. TRADmONAL

826 826

828 828

III. ALTERNATIVE CONCEmONS: CoNCEPnoNs: STRUCTURAL AND AND MOTIVATIONAL IV. FURTHER RESEARCH AND ApPLICATIONS APPLICATIONS

REFERENCES

CHAPTER 32 CHAPTER32

835 835

840 840

842 842

CONSQENTIOUSNESS AND AND INTEGRITY INTEGRITY AT AT WORK WORK CONSCIENTIOUSNESS JOYCE JOYCE HOGAN HOGAN DENIZ DENIZ S. S. ONES ONES

I. OVERVIEW

849 849

II. CONCEPl'UALIZING II. CONCEPTUALIZING CONSCIENTIOUSNESS III. ASSESSING CoNSCIENTIOUSNESS CONSCIENTIOUSNESS III.

851 851

854 854

IV. EMPIRICAL EVIDENCE FOR FOR THE VALIDfIY VALIDITY OF OF CoNSCIENTIOUSNESS CONSCIENTIOUSNESS IV.

V. V. MEASUREMENT BREADTH AND THEORETICAL DYNAMICS REFERENCES

PARTVm PART vm CHAPI'ER33 CHAPTER 33

858 858

863 863

866 866

APPLIED ApPLIED PSYCHOLOGY PSYCHOLOGY PERSONALITY, INTERACTIVE INTERACTIVE RELATIONS, AND AND

APPLIED PSYCHOLOGY ApPLIED WILLIAM WILLIAM F. F . CHAPLIN CHAPLIN

I. I. PREDICTION MODELS

873 873

II. DESCRImON II. DESCRIPHON AND EVALUATION OF OF SOME INTERACTIVE MODELS

876 876

xvii xvil

CONTENTS CONTENTS III. IMPLICATIONS FOR FOR ApPLIED APPLIED PSYCHOLOGY III.

IV. IV. CONCLUSION CONCLUSION

886 886

REFERENCES

886 886

CHAPTER 34 CHAFfER34

883 883

PERSONALITY AND HEALTH: PROGRESS PROGRESS AND PROBLEMS PROBLEMS PERSONALITY

IN IN PSYCHOSOMATICS PSYCHOSOMATICS DEBORAH DEBORAH J. J. WIEBE WIEBE TIMOTHY TIMOTHY W. W . SMITH SMITH I. MODELS OF OF ASSOCIATIONS ASSOCIATIONS BETWEEN PERSONALITY AND HEALTH I. AND HEALTII II. II. PERSONALITY CoNSTRuas CONSTRUcrS IN CURRENT RESEARCH

892 892

897 897 906 906

III. LIMITATIONS LIMITATIONS AND CHALLENGES IN PERSONALITY AND III. AND CHALLENGES IN PERSONALITY AND HEALTH RESEARCH IV. CONCLUDING CONCLUDING THOUGHTS IV.

REFERENCES CHAPTER 3S 35 CHAFfER

908 908

908 908

DISORDERS PERSONALITY DIAGNOSIS AND PERSONALITY DISORDERS LESLIE C. C . MOREY MOREY LESLIE

I. INTRODUCTION INTRODUCTION I.

919 919

II. OF PERSONALITY II. THE THE HISTORY HISTORY OF PERSONALITY DISORDER DISORDER

919 919

III. TO PERSONALITY DISORDER III. CONTEMPORARY CONTEMPORARY ApPROACHES APPROACHES TO

922 922

IV. OF PERSONALITY IV. THE THE NATURE NATURE OF PERSONALITY DISORDER: DISORDER: CURRENT CURRENT CONTROVERSIES CONTROVERSIES V. CONCLUSION V. CONCLUSION

941 941

REFERENCES REFERENCES

941 941

CHAPTER 36 36 CHAFfER

932 932

IN PSYCHOTHERAPY OUTCOME TRENDS AND PRACTICES IN PSYCHOTHERAPY ASSESSMENT AND THEIR IMPLICATIONS FOR PSYCHOTHERAPY AND ApPLIED APPLIED PERSONALITY PERSONALITY AND MICHAEL LAMBERT MICHAEL J. J. LAMBERT EDWIN C.. SUPPLEE EDWIN C SUPPLEE

I, HISTORICAL OVERVIEW I.

948 948

II. OF KNOWLEDGE II. PERSONALITY PERSONALITY CHANGE CHANGE MEASUREMENT MEASUREMENT INSTRUMENTS: INSTRUMENTS: CURRENT CURRENT STATUS STATUS OF KNOWLEDGE AND AND PATTERN OF USE 953 PATTERN OF USE 953

xviii xvijl

CONTENTS III. WHAT ASPECTS OF PSYCHOTHERAPY III. OF PERSONALITY ARE ARE BEING MEASURED IN IN PSYCHOTHERAPY OUTCOME 958 OUTCOME RESEARCH? RESEARCH? 958

IV. SUMMARY, AND FUTURE IV. SUMMARY, CONCLUSIONS, CONCLUSIONS, AND FUTURE DIRECTIONS DIRECTIONS REFERENCES REFERENCES

INDEX INDEX

969 969

965 965

963 963

CONTRIBUTORS

authors' contributions begin. Numbers in parentheses indicate the pages on which the authors'

James R. Averill (513) Department of Psychology, University of Massachusetts, Amherst, Massachusetts 01003 Baumeister (681) (681) Department Department of of Psychology, Psychology, Case Case Western Western Reserve Reserve Univer UniverRoy F. Baumeister sity, Cleveland, Ohio 44106 Arnold H. Buss (345) Department of Psychology, University of Texas, Austin, Texas 78712 David M. Buss (317) Department of Psychology, University of Texas, Austin, Texas 78712 William F. Chaplin (873) Department of Psychology, University of Alabama, Tusca Tuscaloosa, Alabama 35487 Lee Anna Clark (767) Department of Psychology, University of Iowa, Iowa City, Iowa 52242 C. Randall Colvin (617) Department of Psychology, Northeastern University, BosBos ton, Massachusetts 02115 Paul T. Costa, Jr. (269, 825) NIA-NIH Gerontology Research Center, Baltimore, Maryland 21224 Laurie Couch (465) Department of Psychology, Fort Hays State University, Hays, Kansas 67601 Rebecca A. Eder (209) St. Louis Children's Hospital, St. Louis, Missouri 63110 Nancy H. Eisenberg (795) Department of Psychology, Arizona State University, Tempe, Arizona Arizona 85287 Tempe, Robert A. Emmons (485) Department of Psychology, University of California, Davis, California California 95616 Davis, 95616 John F. Finch (143) Department of Psychology, Texas A&M University, College Station, Texas Texas 77843 Station, 77843 (543) California California Pacific Pacific Medical Medical Center Center and andUniversity Universityof ofCalifor CaliforBram Fridhandler (543) nia, San Francisco, San Francisco, California 94143 xix

XX

CO NTRIBUTORS CONTRIBUTORS

David C. Funder (617) Department of Psychology, University of California, River Riverside, California 92521 Stella Garcia (167) Department of Psychology, University of Texas at San Antonio, San Antonio, Texas 76010 Russell G. G. Geen (387) (387) Department of Psychology, University of Missouri, Columbia, Missouri 65211 William G. Graziano (795) Department of Psychology, Texas A&M University, College Station, Texas 77843 Charles F. Halverson, Jr. (241) University of Georgia, Athens, Georgia 30602 Reid Hastie (711) University of Colorado, Boulder, Colorado 80309 Sean Hayes (543) Department of Psychology, Pepperdine University, Malibu, CaliCali fornia 90263 Ravenna Helson (291) Institute of Personality and Social Research, University of of California at at Berkeley, Berkeley, Berkeley, California California 94720 California Joyce Hogan (849) University of Tulsa, Tulsa, Oklahoma 74104 Ariington, William Ickes (167) Department of Psychology, University of Texas at Arlington, Arlington, Texas Texas 76010 76010 Arlington, Oliver P. John (649) Department of Psychology, University of California, Berkeley, Berkeley, California 94704 John A. A. Johnson Johnson (73) (73) DuBois DuBois Campus, Campus, Pennsylvania Pennsylvania State State University, DuBois, Pennsylvania 15801 Warren H. Jones (465) Department of Psychology, University of Tennessee, KnoxKnox ville, Tennessee Tennessee 37996 ville, John F. Kihlstrom (711) Department of Psychology, Yale University, New Haven, Connecticut 06520 Michael J. Lambert (947) Psychology Department, Brigham Young University, Provo, Utah 84602 James T. Lamiell (117) Department of Psychology, Georgetown University, WashWash ington, District of Columbia 20057 Jane Loevinger (199) Department of Psychology, Washington University, St. Louis, Louis, Missouri 63130 Dan P. McAdams (3) School of Education and Social Policy, Northwestern Univer University, Evanston, Illinois 60208 Robert R. McCrae (269, 825) NIA-NIH Gerontology Research Center, Baltimore, (269,825) Maryland 21224 Mangelsdorf (209) Department of Psychology, University of Illinois, Sarah C. Mangelsdorf Urbana-Champaign, Champaign, Champaign, Illinois Illinois 61820 Urbana-Champaign, 61820 Edwin I. Megargee (581) Department of Psychology, Florida State University, Tallahassee, Florida 32306 Morey (919) (919) Department Department of of Psychology, Psychology, Vanderbilt Vanderbilt University, University, Nashville, Nashville, Leslie C. Morey Tennessee 37240 Deniz S. Ones (849) University of Minnesota, Minneapolis, Minnesota 55455 Jennifer Pals (291) Institute of Personality and Social Research, University of of California at Berkeley, Berkeley, California 94720

CONTRIBUTORS

xxi xxi

Delroy L. Paulhus Paulhus (543) (543) Department of of Psychology, Psychology, University Universityof of British BritishColumbia, Columbia, Vancouver, British Columbia V6T 1Z4, Canada Thomas F. Pettigrew (417) Stevenson College, University of California, Santa Cruz, California 95064 Richard W. Robins (649) Department of Psychology, University of California, Davis, Davis, California 95616 David C. Rowe (367) School of Family and Consumer Resources, University of Arizona, Tucson, Arizona 85721 William McKinley Runyan (41) School of Social Welfare, University of California at Berkeley, Berkeley, California 94720 Susan Scott (465) Department of Psychology, University of Tulsa, Tulsa, OklaOkla homa 74104 Timothy W. Smith (891) Department of Psychology, University of Utah, Salt Lake City, Utah 84112 Mark Snyder (167) Department of Psychology, University of Minnesota, Minneapo Minneapolis, Minnesota Minnesota 55455 lis, 55455 Marjorie Solomon Solomon (291) (291) Institute Institute of of Personality PersonaHty and and Social Social Research, Research, University University Marjorie of California California at at Berkeley, Berkeley, Berkeley, Berkeley, California Cahfomia 94720 94720 of Edwin C. C. Supplee (947), Medical Center, Honolulu, HI HI 96859 96859 (947) , Tripler TripIer Regional Medical Paul D. Trapnell (737) Department of Psychology, University of British Columbia, Vancouver, British British Columbia Columbia V6T V6T 1Z4, 1Z4, Canada Vancouver, Harry C. C. Triandis Triandis (439) (439) Department Department of of Psychology, Psychology, University University of of Illinois, UrbanaHarry Urbana Champaign, Champaign, Illinois 61820 Karen S. Wampler (241) Texas Tech University, Lubbock, Texas 79409 David Watson (767) Department of Psychology, University of Iowa, Iowa City, Iowa 52242 Iowa Stephen G. West (143) Department of Psychology, Arizona State University, Tempe, Arizona 85287 J. Wiebe (891) Department of Psychology, University of Utah, Salt Lake Deborah 1. City, Utah 84112 Jerry S. Wiggins (95, 737) Department of Psychology, University of British ColumColum (95,737) bia, Vancouver, British Columbia V6T 1Z4, Canada

This Page Intentionally Left Blank

PREFACE PREFACE

This Handbook was conceived in the early 1980s, during a time when personality This scientifically worthless worthless endeavor. endeavor.Research Researchpsychology was under heavy attack as aa scientifically ers had had trouble trouble publishing publishingin inmainstream mainstreamjournals journalsor orgetting gettinggrant grantproposals proposalsfunded, funded, ers and the the discipline discipline seemed seemed in in danger danger of of disappearing disappearing from from the the intellectual intellectual radar radar and scope. scope. Such aastate reached stateof ofaffairs affairsseems seemsvery verystrange strangegiven giventhat: that:(1) (1)the theconclusions conclusions reached by anthropology, criminology, economics, history, political science—indeed, all by anthropology, criminology, economics, history, political science-indeed, all nature; (2) (2) schol scholthe social social sciences-depend sciences—depend heavily on assumptions assumptions about about human the heavily on human nature; ars in inthese thesefields fieldsrarely rarelyexamine examinetheir theirpsychological psychologicalassumptions; assumptions;and and(3) (3)personality personality ars psychology is the one discipline that takes the self-conscious evaluation of human its central central intellectual intellectual task. task.This Thispoint pointhighlights highlightsthe thesignificance significanceof ofperson personnature as its aUty psychology in modern social science. ality It is hard to imagine a more important topic. Despite the overwhelming significance of the topic, personality psychology, as asnoted noted above, above,almost almostdisappeared disappeared in the 1970s and early 1980s. The reasons were varied, but the biggest problem was the enthusiasm of American psychology for behaviorism. Behaviorism is the antithesis of personality psychology; ititargues arguesthat thatwhat whatpeople peopledo dodepends dependsprimarily primarily on their social circumstances. It also denies that there are stable structures inside people that can explain their behavior. Behaviorism is a useful methodology for training animals to perform, but as a model of human nature it is seriously inadequate. The problem problem is that it ignores evolutionary theory, one of the most important developments in the history of science. Evolutionary theory, combined with with human humanbehavior behaviorgenetics, genetics,reveals revealssensi sensible and reproducible evidence for stable tendencies within individuals. Over the past 10 years personality psychology has made a remarkable comecome back. There are probably two reasons for this. First, social psychologists have discovered individual differences and have learned that, by incorporating personperson ality measures in their experiments, they get better results. Second, industrial/ industriall aUty organizational psychology has discovered that that well-constructed measures measures of personxriii xxiii

This Page Intentionally Left Blank

PART I

INTRODUCTION NATURE NATURE OF PERSONALITY PSYCHOLOGY PSYCHOLOGY

This Page Intentionally Left Blank

.

j j j

j j

j

j

j j j

j j

j

CHAPTER!1 CHAPTER

A A CONCEPTUAL HISTORY OF PERSONALITY PSYCHOLOGY DAN P. P . McADAMS MCADAMS DAN NORTHWESTERN UNIVERSITY NORTHWESTERN UNIVERSITY

A history is an an interpretation of the the past in in light of what followed. It is is a story how events events and and changes changes led up up to aa subsequent subsequent state state of of affairs. affairs. If If that that about how about the current state of affairs, then the the historian must subsequent state of affairs is the of how how the the past past may may have have led led up up to to the the present. present. The The prevailing prevailing view view make sense sense of make the present, present, therefore, therefore, influences influences how how the the past past is is to to be be seen. seen. If If recent recent reviews reviews of the of any indication, there appear to be be increasingly increasingly positive positive signs signs concerning concerning the are any are indication, there appear to the present health of of personality psychology (D. (D. M. 1991; Craik, Craik, 1993; 1993;Digman, Digman, present health personality psychology M. Buss, Buss, 1991; 1990; McAdams, McAdams, 1994; 1994; Pervin, Pervin, 1985, 1990; Singer Singer & & Kolligan, KoUigan, 1987; 1987; Wiggins Wiggins & & 1990; 1985, 1990; Pincus, 1992). 1992). This This history, history, therefore, therefore, begins begins with with aa mildly mildly optimistic optimistic perception perception of Pincus, of the current current state state of of affairs affairs in in personality. personality. the When the present is seen in relatively optimistic terms, narratives of the past to manifest manifest themes themes of of either progress progress or or rebirth. rebirth. Both Both are areapparent apparent in in are likely likely to are this history, though they are tempered by themes of stagnation and disorganization in histories of science, for most most people people as well. The theme of progress is a favorite in assume that science moves forward, toward greater understanding and truth. From in the the history history of of personality personality psychology psychology the current current vantage vantage point, point, some some progress progress in the be seen, but the the progress appears uneven andfitful. Sanford (1963) (1963) warns warns that that can be seen, but progress appears uneven and fitful. Sanford can of psychologists psychologists at at any any particular particular time time may may not not accurately accurately the research research activity activity of the main ideas ideasin in the the field. Otherreviewers reviewershave havenoted notedthat thatthe the field fieldof ofpersonality personality reflect main reflect field. Other is especially prone to to conceptual conceptual and and methodological fads (Sechrest, (Sechrest, psychology is psychology especially prone methodological fads 1976). A challenge challenge for for this this history, history, therefore, therefore, is is to to discern discern broad broad conceptual conceptual trends 1976). A trends As such such this this account account is is highly highly selective, selective, passing passing as they they have have evolved evolved over over time. as time. As IIANDBOOK OF HANDBOOK OF PERSONALITY PERSONALITY PSYCHOLOGY PSYCHOLOGY

3 3

COPYRIGHT CI © 1997 BY ACADEMIC PRESS PRESS.. ALL ALL IIGHlli RIGHTS OF OFREPRODUcnON REPRODUCHONININANY ANYFORM FORMRESERVED. RESERVED.

4 4

DAN DANP. P. McADAMS MCADAMS

over many ephemeral fads as well as many substantive contributions to. to the field that simply cannot be included in the short space given for such a daunting exposi exposition. Finally, this is a recent history of personality psychology, concentrating on the twentieth century, and it is heavily weighted toward concept trends and develop developments in America.

1 I. PERSONALITY PSYCHOLOGY'S UNIQUE FEATURES: HOLISM, MomAnON, AND INDIVIDUAL DIFFERENCES DIFFERENCES MOTIVATION, Personality psychology became an identifiable identifiable discipline in the social sciences in the 1930s. During that decade a number of separate lines of inquiry came together, culminating in the highly integrative programs for the field generated by Allport AUport (1937), Murray (1938), and Lewin (1935). The first issue of the journal Character and Personality Personality appeared in 1932. The journal aimed to join and join German studies of character with British and American studies of individual differences differences in persons, incorporating case studies, correlational surveys, experiments, and theoretical disdis cussions. McDougall (1932) wrote the lead article, exploring various meanings of the terms "character" and "personality." Early contributors included Adler, Jung, Spearman, Frenkel-Brunswik, Rosenzweig, and MacKinnon. The publication of Allport's AUport's (1937) Personality: Personality: A Psychological Psychological Interpretation Interpretation marked the formal arrival of personality on the scene of social science. Although textbooks on mental hygiene, abnormal psychology, and character and personality 1908;Jastrow, Jastrow, 1915; 1915;Roback, Roback, had appeared in earlier years (e.g., Bagby, 1928; Bruce, 1908; 1927), Allport's AUport's was the first to articulate a grand vision for the field of personality and to place it within the context of historical and contemporary scholarship in the arts and sciences. (Stagner's [1937] textbook textbook in personality, written from an experimentaVbehaviorist experimental/behaviorist point of view, appeared in the same year, but its historical influence influence has not been nearly as great as that of Allport's.) AUport's.) Allport viewed personal personality psychology as the study of the individual person. He defined personality as "the dynamic organization within the individual of those psychophysical systems that determine his unique adjustments to his environment" -a definition that, shorn of environment"—a its sexist language, is still serviceable today. Allport later (1961) changed "unique adjustments to his environment" to "characteristic behavior and thought." From the beginning, personality psychology was a dissident field in the context of American experimental psychology (Hall & & Lindzey, 1957). Whereas American psychology tended toward the elementaristic, personality was holistic, taking the whole whole person person as the primary unit of study. Whereas American learning theory focused on the relations between external stimuli and publicly observable responses in rats and other animals, personality concerned itself itself with the problem of human motivation, conceived in terms of unobservable dynamics and promptings from within. Whereas experimental psychologists searched for universal laws applicable to all individuals, personality focused on how people were different different from each other as well as how they were alike. In addition, personality psychology invited

CHAPTER CHAPTER 11 CONCEYIlJAL CONCEPTUALHISTORY HISTORY

5

of disciplines lying outside the mainstream of of collaborations with a wide variety of American experimental psychology. These included psychoanalysis and other depth psychologies, German characterology, mental testing, and abnormal psychology. With its triple emphasis on the whole person, motivation, and individual differences, personality psychology psychology has always held a rather tenuous and ambiguous status in American psychology. On the one hand, some have proclaimed that proper personality theory is (or should be) nothing less than integrative psychological psychological theory at the highest level, placing personality at the center of of all of psychology (e.g., G. S. Klein & & Krech, 1951; 1951; McCurdy, 1965). On the other hand, some have suggested that the discipline is so amorphous that it should be abolished, or allowed & Mouton, 1959). Both extremes are problematic in that they to fade away (Blake & fail to affirm a separate identity for the field of of personality psychology: it is either nothing or everything (which is also nothing). This historical account affirms an identity for personality psychology as a discipline that has its own distinguishing distinguishing features. Three of of these are the emphases on (1) the whole person, (2) (2) motivation and dynamics, dynamics, and and (3) differences. and (3) individual individual differences.

D. ROOTS: BEFORE mE n. HISTORICAL ROOTS; THE 1930s 1930s A. The Whole Person Before Allport, a distinguished group of of European scholars championed the whole person as a unit of of scientific scientific study. Relatedly, these scholars tended to conceptualize striving toward toward wholeness, toward unity and purpose in their persons themselves as striving lives. Comte (1852) envisioned a new science, which he named la morale, dedicated of to the examination of the individual person as both ''a "a cause and consequence of society" (Allport, 1954, p. 8). In Comte's view, the person is more than a biological and a cultural being. The person is a uniquely patterned moral agent existing as a unified and directed whole. La La morale morale could be viewed from from either a biological biological or transcend a social/cultural perspective, but the identity of the new science was also to transcend these two views. Writes Allport (1954), "what Comte was seeking was a science of of personality—unfortunately some years before such a science was possible" (p. 8). personality-unfortunately (1900/1976) Dilthey (1900 /1976) argued for a purely human life science, divorced from the objective approaches of the natural sciences. Rejecting virtually all canons of of conventional scientific methodology, his psychology of Verstehen set as its goal the of of the individual life. Somewhat less empathic understanding of the inner unity of radical was the personalistic personalistic psychology of of Stem (1924), premised on the assumption that the person is a "multiform dynamic unity." Closer to the mainstream in acaaca demic psychology was McDougall's (1908) view that whereas personality may be seen in terms terms of of aa number number of of different different instincts instincts and and sentiments, the master master sentiment sentiment seen in sentiments, the is self-regard, which makes for the unity of self, or what McDougall called which makes for the unity of self, or what McDougall called character. is Both James (1890) and Freud (1923/1961) left room in m their writings for the person's strivings toward unity and wholeness. For James the self-as-me self-as-me (the objec-

6

DAN DAN P. P . McADAMS MCADAMS

tive self or self-concept)-in self-concept)—in its material, social, and spiritual aspects-encompasses aspects—encompasses all that the person understands to be "me" and "mine." By contrast, James' selfas-I self-as-I (the subjective self self or ego) is a fleeting state, a person's moment of consciousness, destined to slip instantaneously away into a rushing stream. For Freud, "the I" I" {das Ich, Ich, or or ego) ego) isis aa unifying unifying agent agent of of sorts, sorts, working working to to effect effect compromises compromises among among (das conflicting conflicting forces within the person and between the person and the environment, funcwith the final goal of reducing anxiety and promoting effective, reality-based func tioning. When wholeness wholeness or or unity is is destroyed, destroyed, the the person is is bound to to suffer suffer tioning. problems. problems. B. B. The Problem of Motivation Motivatioii

Hall and Lindzey (1957) write that personality theorists have traditionally assigned processes. More than most other fields fields in the social a crucial role to the motivation processes. sciences, personality psychology concerns itself itself with the internal springs of human action. This orientation is evident even in textbooks written before Allport AUport (1937): "It is surely in the springs of human action, if anywhere, that the key to the problem of personality is to be found" (Garnett, 1928, 1928, p. 14). 14). Conceptualizations of human motivation may be broadly classed into those emphasizing (1) biological constructs such as drives, instincts, and brain rewards, and (2) cognitive processes such as expectancies, values, schemas, and attributions (Geen, Beatty, & & Arkin, 1984). 1984). In the early years of this century, personality theories tended to draw on the former former class, invoking various biological or quasi-biological explanations for for why people ultimately do what they do. There Therefore, Freud (1905/1953, (1905/1953, 1920/1955) argued that biologically anchored drives concerning sexuality and aggression provide the energy and the direction for much of human behavior, though their manifestations manifestations are disguised through the defensive machinations of the ego and the generally antagonistic demands of of the social world. McDougall (1908) presented a more differentiated theory of differentiated of motivation, suggesting that human behavior is energized and guided by 12 instincts and five five "nonspecific "nonspecific innate tendencies." Major Major instincts include those concerned with reproduction, food seeking, construction (building things), and food gregariousness. Whereas Freud's Eros and death instincts are generally viewed to be irrational and at odds with the demands of social life, McDougall saw instinctive activity as complexly patterned, reality-based behavior that is sustained until some natural goal is achieved. For McDougall, instinctive behavior is socially adaptive and situationally malleable. McDougall's (1908) concept of sentiment represents an attempt to move to toward the cognitive pole of motivational theorizing. Similar Similar in in meaning meaning to to the the contem contemporary term "value," a sentiment develops when the mental image of an object or activity in the environment becomes associated with one or more instincts. People form sentiments with regard to concrete objects (one's favorite book), collectivities form (church, gang), and abstractions (qualities of honesty, courage) (Hogan, 1976). 1976). At offered a more thoroughly cognitive apthe most abstract level, Spranger (1928) (1928) offered

CHAPTER CHAPTER 11 CONCEPTUAL CONCEPTUALHISTORY HISTORY

7

Men proach to human motivation in positing six central value types in personality. Men motivated by their allegiance to either theoretical, social, and women are primarily motivated values—a typology captured in Allport AUport and political, artistic, economic, or religious values-a " A Study of Values." Whereas biological drives seem Vernon's (1931) measure, "A efficient-cause manner, values seem to "pull" behavior behavior to "push" behavior in an efficient-cause toward certain goal states, suggesting something more of a teleological or final finaltoward cause explanation of human motivation (Rychlak, 1981). Differences among People C. Differences Western conceptualizations of individual differences differences in personality can be traced back at least 2,000 years to the ancient typologies of Hippocrates and Galen. The characterization of the sanguine, phlegmatic, melancholic, and choleric behavioral characterization has proven an amazingly durable contribution to psychological theorizing, brought forward to modern times by Kant, Wundt, and, in the twentieth century, Eysenck forward reconceptuaUzes the types along the dimensions extraversion and neu neu(1973), who reconceptualizes roticism. Posterity Posterity has has been been less less kind kind to to the the ancient ancient belief belief that that blood, blood, bile, bile, and and roticism. phlegm are are the the physiological physiological underpinnings underpinnings of of these these behavioral behavioral types, types, but but the the phlegm somatotype theories theories of of Kretschmer (1921) and and Sheldon Sheldon (1940) retain aa biological biological somatotype Kretschmer (1921) (1940) retain emphasis, as as have have numerous numerous theories theories of of human human "temperament." "temperament." Much Much of of the the emphasis, groundwork for for differential differential psychology psychology was was laid laid before before the the 1930s, 1930s, extending extending back back groundwork to the the pioneering pioneering work work of of Galton Galton on on mental mental testing testing and and Binet Binet on on intelligence, intelligence, the the to invention of correlation correlation and and factor factor analysis, analysis, and and the the emergence emergence of of formal formal test test theory theory invention of and the the psychometric movement in in the the United States and and Britain (see Anastasi, Anastasi, and psychometric movement United States Britain (see 1958,1976; Jackson & & Paunonen, Paunonen, 1980, 1980, for for reviews). reviews). Spurred Spurred by by the the mobilization mobilization 1958, 1976; Jackson of large large military military forces forces for for World World War War I, I, psychologists psychologists began began to to invent invent self-report, self-report, of multi-item tests tests to to assess assess individual individual differences differences in in personality personality functioning. functioning. A A fore foremulti-item runner to to the the MMPI, MMPI, Woodworth's Woodworth's (1919) (1919) Personal Personal Data Data Sheet Sheet was was used used to to screen screen runner out men men who who were were unfit unfit for for military military service service because because of of personality personality maladjustment. maladjustment. out Bernreuter (1931) developed developed the the first first multitrait multitrait personality personality inventory, inventory, containing containing Bernreuter (1931) scales to assess neuroticism, introversion, dominance, self-sufficiency, confidence, scales to assess neuroticism, introversion, dominance, self-sufficiency, confidence, and sociability. sociability. and

m. in. THE FORMATION OF SYSTEMS: 1930-1950 of the Individual A. Allport's Psychology of

Allport's AUport's (1937) vision for personality psychology was a humanistic alternative alternative to the prevailing mechanistic paradigms of stimulus-response psychology in the determinism and the 1930s. In addition, it was an optimistic antidote to Freudian determinism growing emphasis, in clinical writings, on human pathology. In his autobiography, autobiography, AUport (1968) states that he wished to create a field of study centered on an Allport image of man "that would allow us to test in full whatever democratic and

8

DAN P. McADAMS DANP.MCADAMS

possess" (p. 394). Toward the end of the Great humane potentialities he might possess" Depression in Europe and America and on the eve of World War II, Allport AUport wrote Personality Personality in the spirit of social reform and the hope for a better world. Allport's AUport's text was cosmopolitan, erudite, and steeped in old-world European scholarship. But it was also profoundly profoundly American, in its unabashed optimism and egalitarian tone. Allport AUport presented an eclectic array of concepts and hypotheses, loosely tied to one dominant theme: the person is a unique whole. It is somewhat ironic that in the history of personality psychology the central theme of the seminal textbook in the field has remained the most controversial and hotly disputed aspect of AUport's legacy, as if the field's raison d'etre doubles as its perpetual nemesis. How Allport's can a science of the person assume that each person is unique? If science seeks lawfulness across persons (nomothetics), then how can it make sense of, of, even leave room for, the uniqueness of the individual (idiographics)? Many, if not most, personality psychologists have traditionally objected to Allport's AUport's insistence that personality psychology must focus on the uniqueness of the individual case (e.g., (e.g.. Holt, 1962). But they have been kinder, at least in their rhetoric if not in their AUport's insistence that personality personaUty psychologists con conresearch, when it comes to Allport's cern themselves with the person's wholeness. AUport's own theorizing, the person's wholeness and unity are probably In Allport's best captured in his concept of self, or the proprium. proprium. The "proprium includes all (AUport, 1955, p. 40). Eight aspects of personality that make for inward unity" (Allport, different aspects of the proprium can be identified, each emerging at a particular point in the development of the person. In their developmental order of emergence, selfthese are (1) the sense of bodily self, (2) self-identity, (3) self-esteem, (4) self extension, (5) self-image, (6) self-as-rational-coper, self-as-rational-coper, (7) propriate strivings toward life goals, and (8) a unifying sense of self-as-knower, or the sense of the totality of the person as a process that is continually changing and becoming. In Allport's AUport's view, both human motivation and individual differences differences are ac accounted for by traits. For Allport, AUport, a trait is "a neuropsychic structure having the stimuli functionally equivalent, and to initiate and guide capacity to render many stimuU equivalent (meaningfully consistent) forms of adaptive and expressive behavior" (1961, p. 347). Allport AUport held that traits are real, causal entities that correspond to as yet unknown neurophysiological structures. They are not mere descriptive categories of functionally equivalent behaviors. Rejecting the distinction between AUport insisted that traits have motivational features, serving to motive and trait, Allport energize, direct, and select behavior. Despite popular misconceptions, however, AUport did did not not argue argue that that traits traits make make for for extraordinarily extraordinarily high high cross-situational cross-situational Allport generality human behavior (Zuroff, 1986). 1986). A A single single person person may may be be characterized characterized generality in in human behavior (Zuroff, by contradictory contradictory traits. traits. Furthermore, Furthermore, behavior behavior is is always always aa function function of of the the situation, situation, by in that ''the ever changing nature of traits and their close dependence upon the the in that "the ever changing nature of traits and their close dependence upon fluid conditions of the environment forbid a conception that is over-rigid or overfluid conditions of the environment forbid a conception that is over-rigid or over simple" (AUport, 1937, 1937, p. 312). p. 312). simple" (Allport,

99

CHAPTER CHAPTER 11 CONCEPTUAL CONCEPTUALHISTORY HISTORY

B. Murray's Personology After After the death of Prince in 1928, Murray became the director of the Harvard Harvard commanded a remarkable remarkable intellec intellecPsychological Clinic, and for the next decade he commanded from a wide variety of disciplines under under tual expedition, rallying together scholars from the banner of personology 1987). personology (Robinson, 1992; Shneidman, 1981; 1981; White, 1981, 1981,1987). Like Allport, AUport, Murray championed a science of the whole person. But if Allport's AUport's vision was steeped steeped in the Enlightenment, Enlightenment, Murray's personology personology was born bom of Roman Romanticism (Shweder, 1984). Whereas Allport viewed the human mind as potentially potentially rational and orderly, Murray focused focused his attention on that which is relatively irratio irrational, passionate, and laden with conflict and emotion. Murray sought to bring the center of academic psychology. insights of Freud and (especially) Jung toward the center from McDougall and His eclectic theory blends psychoanalytic ideas with strands from from the study of literature, mythology, and medicine. While Lewin and themes from Murray was was just as committed committed as as Allport Allport to to conceiving conceiving persons as integrated wholes, Murray just as persons as integrated wholes, he was was less less sanguine sanguine about about the the possibility possibility that that personality personality is is aa unified unified and and self selfhe consistent totality. totality. There There is is nothing nothing like like aa proprium proprium in in Murray's Murray's personology. personology. Instead Instead consistent typical personality personality is is aa typical life, conscious and unconscious: a whispering a flow of powerful subjective life, gallery in which voices echo from the distant past; a gulf gulf stream of fantasies fantasies gallery of contending complexes, plots with floating memories of past events, currents of and counterplots, counterplots, hopeful hopeful intimations intimations and and ideals ideals . .. .. . . A A personality personality is is aa and full congress of orators and pressure-groups, of of children, demagogues, commu commufull nists, isolationists, isolationists, war-mongers, war-mongers, mugwumps, mugwumps, grafters, grafters, log-rollers, log-rollers, lobbyists, lobbyists, nists, Promethean revolu revoluCaesars and Christs, Machiavels and Judases, Tories and Promethean tionists. tionists. .

pp. 160-161) 160-161) (Murray, 1940, 1940, pp.

forth the basic principles of personology in Murray and his colleagues set forth the landmark volume, volume. Explorations Explorations in Personality Personality (1938). Among the more more influential unity thema. thema. The primary motiva motivainfluential concepts are need, press, thema, and unity �ogenic needs, such tional constructs in Murray's system are the 20 or so psyc psychogenic as the needs for achievement, affiliation, affiliation, dominance, play, and succorance. Each Each need stands for a force "which organizes perception, apperception, intellection, inteHection, conation and action in such a way as to transform transform in a certain direction an existing, unsatisfying unsatisfying situation" (Murray, 1938, p. 123). A A press is an environmental environmental situation that functions functions as an opportunity for or obstacle to the expression of a particular need. The person's subjective perception of the situation is termed termed the "beta press"; the objective nature of the situation is the "alpha "alpha press." A A thema is a recurrent need-press interaction. A unity thema is a dominant pattern of related needs and press (plural) which organizes or gives meaning to a large from infantile infantile experience, a portion of the individual's life. Ultimately derived from motif of a person's bio biounity thema may be viewed as the central, organizing motif graphy.

10 10

DAN P. McADAMS DANP.MCADAMS

C. Lewin's Field Theory

Both Allport and Murray assumed that behavior is a function function of the interaction of the person and the environment (Ekehammer, 1974; Zuroff, 1986). But Lewin was A Dynamic Theory of of Personality, Personality, Lewin Lewin more explicit about the interaction. In A (1935) conceived both the person and the environment as differentiated differentiated aspects of an integrated integrated life space. space. The The life life space space contains contains the the totality totality of of possible possible facts facts which which an are capable capable of of determining determining the the behavior behavior of of an an individual individual at at aa given given moment. moment. Strongly Strongly are influenced by by the the Gestalt Gestalt theories theories of of Wertheimer Wertheimer and and Kohler, Kohler, Lewin Lewin viewed viewed the the influenced person-in-the-environment as a contemporaneous gestalt—a field of forces that person-in-the-environment as a contemporaneous gestalt-a field of forces that assumes aa characteristic characteristic form form at at aa particular particular moment moment in in time. time. All All of of the determinants assumes the determinants of behavior at a given moment are in the field at the moment. Thus, Lewin's of behavior at a given moment are in the field at the moment. Thus, Lewin's approach, in contrast to Allport and Murray, tends to deemphasize developmental approach, in contrast to Allport and Murray, tends to deemphasize developmental constructs. Whereas Whereas Murray Murray (1938) (1938) said said that that "the "the history history of of the the organism organism is is the the constructs. organism" (p. 39) and Allport (1937) spoke of stages in the development of the organism" (p. 39) and Allport (1937) spoke of stages in the development of the proprium, Lewin advocated an ahistorical analysis of person-situation interactions. proprium, Lewin advocated an ahistorical analysis of person-situation interactions. Lewin viewed viewed human human motivation motivation in in terms terms of of energy energy transformations transformations in in aa Lewin field. Energy is released when the person attempts to return to equilibrium dynamic dynamic field. Energy is released when the person attempts to return to equilibrium after the the onset onset of of aa state state of of tension. tension. The The person person experiences experiences tension tension when when one one part part after of the inner-personal region is thrown out of balance vis-^-vis other parts. This is of the inner-personal region is thrown out of balance vis-it-vis other parts. This is caused by the arousal of a need—generally defined as either (1) a physiological caused by the arousal of a need-generally defined as either (1) a physiological condition (e.g., (e.g., hunger), hunger), (2) (2) aa desire desire for for something, something, or or (3) (3) an an intention intention to to do do condition something. A valence is the value of a particular region of the environment for something. A valence is the value of a particular region of the environment for aa person. A A region region of of positive positive valence valence is is one one that that contains contains aa goal goal object object which which will person. will reduce tension when the person enters the region. Therefore, valences become reduce tension when the person enters the region. Therefore, valences become coordinated with with needs, needs, in in aa way way not not unlike unlike Murray's Murray's characterization characterization of of the the need needcoordinated press thema. thema. Lewin's Lewin's conceptualizations conceptualizations of of energy, energy, tension, tension, need, need, and and valence valence paved paved press the way way for for subsequent subsequent expectancy-value expectancy-value theories theories of of motivation, motivation, as as in in Atkinson Atkinson the (1964) and and Rotter Rotter (1954). (1964) (1954). D. The The Integration Integration of of Psychoanalysis Psychoanalysis and and Leaming Learning Theory Theory D. While Allport and Murray labored on behalf of personality at Harvard and Lewin founded research programs at Cornell (1933-1935) and the University of Iowa (1933-1945), aa group group of of social social scientists scientists at at Yale's Yale's Institute of of Human Human Relations Relations were working to bring about closer collaboration among the fields of psychology, psychiatry, sociology, and anthropology. Hull's (1943) learning theory served as the overarching conceptual framework while psychoanalysis and social anthropology provided data, ideas, and agendas for empirical research and theoretical syntheses. N. E. Miller and Dollard DoUard (1941; Dollard DoUard & & Miller, 1950) sought to reformulate reformulate psychoanalysis in learning-theory terms. They believed that all significant human behavior is learned learned in particular social, cultural, and historical contexts. Learning drives, conceptual conceptualfundamental factors. First, learning is motivated by drives, involves four fundamental ized as strong internal stimuli that propel behavior. Learned drives are social needs

CHAFfER CHAPTER 11 CONCEPTUAL CONCEPTUALHISTORY HISTORY

11U

that are ultimately derived from primary physiological drives, like hunger and thirst. Second, learning is given direction and guidance guidance cues, which which are stimuli stimuli in the environment environment that provide information information concerning what the organism should attend attend to and how the organism should respond. Third, learning involves a response: response: propelled by drive and guided by cue, cue^ the organism acts. Such action leads to a reduction reduction in drive, which in itself is rewarding and thus constitutes reinforcement, reinforcement, the fourth fourth and final component of learning. There can be no reinforcement reinforcement without some kind of drive reduction. reduction. DoUard translated a number number of classic Freudian Freudian ideas into the Miller and Dollard more objective and operational language of drive, cue, response, and reinforcement. reinforcement. "pleasure principle" the principle of For instance, they substituted for Freud's "pleasure reinforcement, understood as the reduction of a primary or learned learned drive. The reinforcement, psychoanalytic concept of "transference" was seen as a special case of stimulus generalization. "Repression" became inhibition. "Anxiety" was viewed as a learned, acquired through repeated experiences of pain (a primary drive secondary drive, acquired development was explained according to itself) and threatened pain. Psychosexual development appUed to the realms of feeding and weaning, principles of learning and conditioning applied socialization of a child's anger. These cleanliness training, early sex training, and the socialization longitudinal investigation investigation of translations became central concepts in the important longitudinal personality launched launched by Sears, Maccoby, and Levin (1957) and child rearing and personality the cross-cultural studies of Whiting and Child (1953).

E. Factor Theories Relying on factor analysis, Cattell (1946, 1950) developed a comprehensive system of (1946,1950) conceptualizations from from McDougall, McDougall, personality that resembles in various features conceptualizations AUport. For Cattell, the central problem in personality Freud, Lewin, Murray, and Allport. prediction of behavior. Indeed, he defined generpsychology is the prediction defined personality quite gener situation" ally as "that which permits a prediction of what a person will do in a given situation" prediction is to be accurate, then the psychologist psychologist must obtain (1950, p. 2). If prediction information on a great many variables at many different different levels, weigh quantitative information information appropriately, and combine the information information into a specifi specifiand scale the information cation equation. The specification specification equation equation is a linear combination combination of quantitative indices of certain traits, roles, and states, each weighted according to its relevance in the present behavior-that behavior present situation. situation. Thus, the interactional nature of behavior—that function of the person interacting with the environment—is mathematical is a function environment-is given mathematical form in Cattell's specification specification equation. Like Allport, AUport, Cattell viewed the trait as a form central personality variable. For Cattell, a trait is a mental structure that may be inferred from from observable behavior to account for regularity and consistency in inferred behavior. Surface Surface traits represent represent clusters of manifest manifest variables that appear appear to go together; source traits are the underlying factors that that determine the multiple surface surface manifestations. Traits may also be divided into three general categories with respect to their content and function: dynamic dynamic traits, which set the individual into action action to accomplish a goal; ability traits, which concern the effectiveness with which the to accomplish a goal; ability traits, which concern the effectiveness with which the

u 12

DAN P. McADAMS DANP.MCADAMS

individual reaches a goal; and and temperament temperamenttraits, traits,which whichconcern concernsuch suchstylistic stylisticaspects aspects of responding as speed, energy, and emotional reactivity. Other factor theories were developed by Guilford Guilford (1959; Guilford Guilford & & Zimmer Zimmerman, 1949) and Eysenck (1952). (1952). Eysenck's conceptualization has become increas increasingly inftuential influential over the past 30 years. Eysenck divides personality personality into three very broad traits, existing as higher order factors in the analysis of responses from from thousands of subjects on hundreds of self-report self-report questionnaire items. items. The three dimensions are extraversion-introversion, neuroticism, and psychoticism. Eysenck patternings hypothesizes that all three are grounded in particular neurophysiological patternings differences in each are substantially inftuenced influenced by one's geand that individual differences ge netic makeup.

F. Dominant Dominant Trends Trends in in the tiie Grand Grand Systems Systems F. Between 1930 1930 and 1950 1950 a number of personality psychologists developed grand systems for understanding the whole person and predicting what the person will In addition to Allport, Murray, Lewin, Miller and DoUard, Guilford, do. In Dollard, Cattell, Guilford, (1947), Angyal Angyal personaUty were proposed by Murphy (1947), and Eysenck, broad systems of personality (1941), Lecky (1945), (1945), and the psychoanalytic ego psychologists such as Erikson (1950) as well well as as the the neo-Freudian neo-Freudian perspectives from from Fromm Fromm (1950) and Hartmann (1939), (1939), as (1941), (1939), Rank (1945), (1945), and M. M. Klein (1948). (1948). Amidst Amidst the therich richdiversity, diversity, (1941), Horney (1939), discerned. a few consistent trends in these conceptual systems may still be discerned. created in the the 1930s 1930sand and1940s 1940swere werebased based First, most of the personality systems created from many different different perspectives on the assumption that the person may be seen from different levels. Most of the systems, therefore, proposed multiple and on many different constructs organized on multiple levels. For Allport, Murray, and Cattell no single trait, need, attitude, or sentiment is to be seen as the "key" to personality. Rather, various constructs are organized in complex hierarchies (Murray, Cattell) or idioidio graphic patterns unique to the individual (Allport). Despite the plethora of variables variables and levels, however, many of the systems make a second important important claim-that claim—that the person may still be viewed as a unified and organized totality. Such constructs as proprium (Allport), unity thema (Murray), and dynamic lattice (Cattell) attempt to account for the potentially integrated and holistic nature of human personality. "Self' and "ego" are parallel constructs proposed by Lecky (1945) and the ego psychologists, respectively. Most personality systems from explic from this era are either explicitly or implicitly organismic in that they emphasize the consistency and coherence of of normal personality and view the individual organism as an organized and complexly structured whole. whole. A third trend involves motivation. motivation. Many Many of of the the systems systems propose propose some somevariation variation of tension reduction as as a theory of human motivation. This is most appparent in Miller and Dollard, but it is also prominent in Cattell's concept of erg, Lewin's view of dynamics in the life space, and Murray's concept of a need as as transforming transforming "in a certain direction an existing, unsatisfying unsatisfying situation." The general view is that equilibrium, drives or needs increase tension, and the organisms seek some sort of eqUilibrium,

CHAPTER 1 CONCEP11J CONCEPTUAL HISTORY CHAmR AL HISTORY

13 13

organism is motivated to act upon drives or needs in order to reduce tension, which is ultimately satisfying or reinforcing. The emphasis on tension reduction is most apparent in the work of the two psychologists who, in the overall, were probably the most influential influential general theorists during this time: Freud and Hull. Finally, many systems conceived of personality development development in terms of learn learning in society, or what might be termed socialization. The individual begins as an unsocialized and assumedly self-centered self-centered creature, but over time he or she learns how to be an effective effective and relatively cooperative member of a complex social world. With the exceptions of Eysenck and Cattell, the personality systems of the time placed a great deal of stock in Lockean environmentalism-the environmentalism—the person is a product of his or her environment; traits, motives, sentiments, and attitudes are learned in the environment. The most important learning occurs in childhood, especially in the family. Development Development is continuous and relatively gradual, a product of basic principles of learning that remain pretty much the same across the entire life Ufe span.

rv. THE ELABORATION OF CONSTRUCTS: 1950-1970 IV. Psychologists returned to their university settings at the end of World War II to face what would become the greatest expansion in higher education in the history of American society. Large numbers of war veterans returned to or entered college, many taking advantage of the GI Bill. Universities scrambled to keep up with burgeoning enrollments, building new laboratories, classrooms, and residence halls and enlarging their departments well beyond their prewar size. Like most other university departments, psychology departments grew in size and diversity. Federal funding for psychological research became much more plentiful, stimulating and promoting a multitude of applied and basic research programs across the country. psycholoThe expansion brought with it increasing specialization. Fewer and fewer psycholo gists saw themselves as "generalists." Rather they were now "developmental "developmental psy psychologists," "social psychologists," or "physiological psychologists," not simply "psychologists." After World War II, psychology psychology expanded with exuberance into many nonacademic settings, as witnessed by the tremendous growth of clinical psychology and other "applied" subdisciplines, subdisciplmes, the boom in psychotherapies and various forms of counseling and behavior change, the expansion of psychology into the schools, and the growing professionalization of a field whose primary roots were in academia. Within academic psychology, certain traditions of scholarship seemed to ride the crest of the general expansion while others risked being washed away. Stimulated by exciting new theories (e.g., Festinger, 1954; Heider, 1958) and bold laboratory simulations (e.g., Asch, 1951; 1951; Milgram, 1963), experimental social psychology en enjoyed something of a golden age through the mid-1960s. By contrast, personality psychology seemed to flounder. As a whole, personality psychology was generalist by nature in an age of specialization, sympathetic to correlational approaches for research in an era that glorified the experimental method, and interested in differ-

14 14

DAN P. MCADAMS DAN P* MC ADAMS

ences among a m o n g people during during a time when when social psychology psychology was suggesting suggesting that, in extent that some ways, everybody everybody is pretty much much the same. And A n d to the the extent that people people might seemed to suggest, in ways might be b e "different," "different," they are likely to be different, different, some some seemed ways related to pathology-a pathology—a province province of of the growing growing discipline discipline of clinical psychology. psychology. related of clinical Between 1950 and and 1970, personality personality psychology psychology witnessed witnessed a gradual gradual erosion erosion of of Between identity within within psychology psychology as a whole. As A s one one reviewer reviewer put put it a few few years years its identity later, personality personality psychology psychology can be spelled spelled in one one of of two ways: c-I-i-n-i-c-a-l c-M-n-i-c-a-1 or or s-o-c-i-a-1 s-o-c-i-a-l (Sechrest, 1976). After the the war, personality personality psychologists psychologists settled settled down down to to do do hypothesis-testing hypothesis-testing After Through conventional conventional nomothetic nomothetic methods, they they sought sought to articulate articulate some some research. Through of the the key personality personality constructs constructs embedded e m b e d d e d in the the grand order to do do of grand theories. In order often had had to disembed disembed those constructs. In focus on on a single single this, they often In order order to focus personality construct, the researcher researcher might might have have to pull it out out of of its theoretical theoretical personality context. Once O n c e the construct construct was out, it was sometimes sometimes difficult difficult to to fit it back back in. A . The Focus on Constructs and Their Measurement Measurement A. A s World World War W a r II was coming coming to an end, the editorial editorial board board of of Character Character and and As Personality announced announced its first first "editorial reorientation" reorientation" since the the journal's journal's inception inception Personality Anticipating the postwar postwar return return of of psychologists psychologists to to universities universities and and the the in 1932. Anticipating coming increase increase in subsmissions for for publication, pubUcation, the board board decided decided that that the the journal journal coming should shift shift from from its rather rather eclectic role-incorporating role—incorporating aa wide range of of articles, should wide range from theoretical theoretical essays to case studies studies to research research reports-to reports—to one one focused focused more more from exclusively on on empirical empirical research. They They wrote, "appropriate "appropriate methodological, methodological, histori historiexclusively cal, and and theoretical theoretical contributions contributions will continue continue to be be accepted, but the the major major empha emphaaccepted, but upon reports of of original, empirical, and, as far far as material material permits, sis will be placed upon significant experimental experimental investigations, without without restriction restriction as to technicality technicality of of pre presignificant sentation" (Zener, (Zener, 1945, p. p . 1). The T h e journal journal was also to change change its name n a m e to the the Journal broader shift Journal of of Personality. Personality. This editorial change was indicative indicative of of a broader shift that that became became very apparent apparent in the years to come: personality personality psychology psychology was becoming becoming more m o r e self-consciously self-consciously empirical. empirical. The T h e shift shift is apparent apparent in one of of the early and and important important postwar postwar textbooks textbooks in the Allport and and Murray, McClelland the field: field: McClelland's McClelland's (1951) Personality. Personality, Like AUport McClelland argued concerned with argued that that the personality personality psychologist psychologist should should be be concerned with the whole whole person. person. As A s if to t o underscore underscore his point, McClelland McClelland made m a d e good good use of of an an extensive extensive case case sstudy-the t u d y — t h e case of of Karl-in Karl—in the text. However, McClelland's McClelland's vision vision for for the the field field of personality psychology of personaHty psychology in 1951 was quite different different from from that that promoted promoted by by the the grand theorists a few grand few years before, as is evident in the the following following passage from from the the b o o k ' s Preface: Preface: book's with concrete lives like this [the case of of Karl], as they they proceed through through Working with prevent students or or anyone else the theoretical discussions in this book, should prevent from gaining the impression impression that I am trying to present present "a system" or "a theory" theory" from of personality. No No one one knows knows enough enough at at present to to build build aa theory. theory. Rather Rather what what of is needed needed and and what what II have have tried tried to to do do is is to to find find aa number number of of constructs constructs in in terms terms is

1515

CHAPTER 11 CONCEPTUAL CONCEPTUALHISTORY HISTORY CHAPTER

of which which we we can can collect collect data about personality, personality, perhaps perhaps with the ultimate ultimate hope of data about of building building aa theory. theory. of

(McClelland, (McClelland,1951, 1951,p. p.xiv) xiv)

The history of personality psychology between the years 1950 and 1970 is aptly foreshadowed in McClelland's words. The time for building theories was over. Rather, personality psychologists were now to identify key constructs in terms of which data might be collected and analyzed. The promise was that construct elaboration would increase psychologists' knowledge of different parts of the person. Once psychologists knew more about the parts, they would be able to put together better grand theories about the whole. Many of the classic contributions to the literature on personality psychology in the 1950s and 1960s concern problems and issues in the measurement of constructs. psycholCronbach and Meehl (1955) struggled with the question of how personality psychol ogists might determine the worth and validity of a given measure designed to assess individual differences on on such such psychologically psychologically meaningful but ultimately ultimately assess individual differences meaningful but hypothetical as "intelligence," "intelligence," "extraversion," "extraversion," "ego and the the hypothetical dimensions dimensions as "ego strength," strength," and like. Such dimensions, dimensions, which which are are indeed indeed the the staple staple of of virtually all personality theories like. Such virtually all personality theories ever invented, be directly observed but but exist as open open concepts concepts ever invented, cannot cannot be directly observed exist instead instead as (Meehl, 1977) 1977) or or "constructs" "constructs" whose whose workings can be known only only by the network network (Meehl, workings can be known by the of laws laws in in which which they they occur occur (Hogan, (Hogan, 1988). 1988). of Along with Loevinger (1957), Cronbach and Meehl presented guidelines for construct validity validity in psychological research. The process of the establishment of construct science—a dynamic construct validation is essentially that of hypothesis testing in science-a process through which constructs become further defined and articulated as new findings and new measures accumulate over time. Campbell and Fiske (1959) zeroed in on two two derivatives derivatives of of construct construct validity validity—convergent and discriminant discriminant validity. validity. in on convergent and Different measures of the same construct should be highly correlated whereas Different measures of the same construct should be highly correlated whereas measures of of constructs constructs that that purport purport to to be be different different should indeed be be uncorrelated. uncorrected. measures should indeed Thus, measures of constructs should measure what they claim to measure, and nothing else. The emphasis on convergent and discriminant validity reflected a general concern that personality psychologists clarify and make more precise the meanings of their constructs. The 1950s and 1960s saw the construction and refinement of a number of omnibus personality inventories designed to measure many different constructs at once. The clinically oriented MMPI, whose scales were derived solely from meaempirical-key coding, remained the most popular self-report inventory. Newer mea sures for assessing individual differences in normal populations, however, employed more eclectic scale construction strategies, drawing explicitly, in some cases, upon personality theory. theory. Popular Popular inventories inventories developed developed during during this this time time include include the the personality California Psychological Psychological Inventory Inventory (CPI; (CPI; Gough, Gough, 1957), 1957), Cattell's Cattell's (1957) (1957) Sixteen Sixteen California Personality Factor Factor Questionnaire Questionnaire (16PF), (16PF), and and two two measures measures of of Murray's Murray's needs: needs: Personality Edwards' (1957) (1957) Personal Personal Preference Preference Schedule Schedule (EPPS) (EPPS) and and Jackson's Jackson's (1970) (1970) Person PersonEdwards' ality Research Research Form Form (PRF). (PRF). ality -

16 16

DAN P. McADAMS DANP.MCADAMS

Amidst Amidst the flurry flurry of of activity in test construction and validation, two measure measurement controversies rose to the fore: clinical versus statistical prediction (Meehl, 1954; Sawyer, 1966) and the problem of of response styles (Christie & & Lindauer, 1963; Edwards, 1957; 1957; Jackson Jackson & & Messick, 1958). The latter latter preoccupied a great number number of of researchers for for many years, producing a voluminous literature in personality personality journals and books. At At stake was the validity validity of of self-report self-report scales designed designed to assess individual differences differences in personality constructs. Do these scales assess the content content instead tap general test-taking styles that that variables they claim to assess or do they instead cut across a wide variety variety of of content content domains? The controversy was never fully fully of the most compelling evidence for for the content integrity of of resolved, but some of personality tests was summoned forth forth by Block (1965), who, for for example, demon demonstrated strated that the factor factor structure of of the MMPI remained remained essentially essentially unchanged unchanged whether or not one controls for for the social desirability desirability of of the items. People primarily whether respond to the content of of the items, regardless of of their rated desirability. Nonethe Nonetherespond of social less, test developers came to pay closer attention to the potential problem of for response bias when possible possible desirability and sought to mitigate or control for 1971; Wiggins, 1973). ((Jackson, Jackson, 1971; B. Popular Constructs of of the 19505 1950s and 1960s Four personality personality constructs that received a tremendous amount amount of of empirical atten attenFour tion during this time are authoritarianism, achievement motivation, anxiety, and of the four four attracted creative and dedicated investigators investigators field independence. Each of measurement who developed ambitious research programs anchored to specific measurement operationprocedures. Thus, the constructs were generally well conceived, well operation marketed to the scientific community at large. Each of of the the alized, and boldly marketed to constructs generated empirical findings and new theoretical ideas that spoke to central issues and problems in personality personality functioning. In at least two of the cases (authoritarianism (authoritarianism and achievement achievement motivation), psychologists extended their in inquiries into the realms of of societal structures, economics, and history. history. The most important important reasons for for the popularity of of these four, however, may may reside in the nature of of American American society in the 1950s and 1960s. 1960s. Each of of the the four four constructs constructs reflects reflects prevalent concerns and preoccupations preoccupations among middle-class middle-class Americans of of the day. Fresh Fresh from from the great victory of of World War II, Americans Americans moved moved optimistically optimistically forward forward as the world's preeminent role models of of economic economic and technological success driven driven by individual know-how and dedicated teamwork. teamwork. Democracy Democracy had triumphed triumphed over authoritarian dictatorships. The community commimity of of free freehad proven stronger, more efficient, efficient, and more flexible flexible than than thinking individualists had for the (short-term) benefit benefit the rigidly hierarchical systems that oppressed the many for of the few. Yet these optimistic assessments of of America's America's role and destiny destiny lived of alongside more pessimistic viewpoints that decried mindless conformity conformity and rigid rigid authoritarianism and warned of of a smoldering cultural uneasiness in the 19508 1950s (Sara (Saraauthoritarianism A central cultural tension was that between between the individual individual and and the the group. group. son, 1988). A In The The Lonely Lonely Crowd, Riesman Riesman (1950) explored the intractable problems of of group

CONCEPTUALHISTORY HISTORY CHAPTER 11 CONCEPTUAL

17 17

life among the "inner of American "inner directed" and the "outer "outer directed" members of American postwar society. Kenniston Kenniston (1963) wrote to youth's alienation alienation from from the traditional traditional postwar of America of the social upheavals of of the America on the eve of groupings and institutions of Erikson (1959) spoke spoke of of identity crises in modem modem industrial societies, late 1960s. Erikson imploring youth to live boldly boldly within a dialectic between conformity conformity to and rejection rejection imploring of the status quo. of Within Within the cultural context of of middle-class America America in the 19508 1950s and early 1960s, the authoritarian authoritarian personality personality (Adorno, (Adomo, Frenkel-Brunswik, Frenkel-Brunswik, Levinson, & & San San19608, ford, 1950) represented represented an idealized idealized type-a type—a character character syndrome syndrome personifying personifying Ger Gerford, which free-thinking free-thinking American American individualists had assumedly tri triman Nazism, over which encounter again again in the guise of of American American bigotry bigotry and racism, umphed, only to encounter portrayed in increasingly stark stark relief portrayed relief as the civil rights movement grew. By contrast, the need need for for achievement achievement (McClelland, 1961; McClelland, Atkinson, Clark, & & Lowell, 1953) enjoyed enjoyed middle-class society's unambivalent blessing in the 19508 1950s as Ameri Americans worked worked hard hard to consolidate their position as the number-one economic power in the world. A personality construct construct that celebrated celebrated entrepreneurship entrepreneurship and and innovation innovation the A personality resonated well with the values and goals of corporate America. resonated well with the values and goals of corporate America. Anxiety Anxiety was the price Americans had to pay for for living in a postwar, newly nuclearized age. Although Although this third third personality personality construct construct is traditionally found found at nuclearized center of of many many different and theories of of personality, both ancient and the center different systems and research topic in the 19508 1950s and and 19608 1960s (e.g., Sarason & & modem, its salience as a research Spielberger, 1966; Taylor, 1953) may have reflected reflected some of of the Mandler, 1952; Spielberger, cultural concems verse of of W. H. Auden, when he christened christened the concerns captured captured in the verse middle part part of of the twentieth twentieth century "the age of of anxiety." Finally, the construct constmct of of middle field independence independence (Witkin, 1950) considers the perceptual problem problem of of individual field figure and embedfigure and common common ground. To what extent can the individual divorce the embed context from from the embedded embedded phenomenon? Those who are able to interpret ding context reaUty in a decontextualized, decontextualized, inner-directed inner-directed manner manner are deemed deemed field independent. independent. reality phenomena in context-those context—those whose perceptions perceptions are By contrast, those who view phenomena more outer outer directed-are directed—are considered considered be field dependent. The polarities of of individual and group, figure and ground, and self self and context reflect reflect a cultural tension that, in America, America, is probably probably as old as de Tocqueville's nineteenth nineteenth century century appraisal of of American life. Yet the tension seemed to grow stronger and more salient after World American Yet seemed grow stronger after World War II, subtly influencing influencing the questions asked and answers sought by American American War personaUty psychologists. personality C. Conceptual Conceptual Trends Three general trends may be discerned discerned in the history of of personality personality psychology between approximately approximately 1950 and 1970. These are (1) the splitting splitting of of the whole whole between person decontextualized dispositional constructs, (2) the downf downfall of tension tension all of person into decontextualized reduction as an organizing idea in human human motivation, and (3) the emergence of of reduction cognitive approaches to understanding the person. With respect to the first, postwar cognitive approaches personality psychologists borrowed personaUty borrowed liberally from from the grand theories of of the previ-

18 18

DAN DAN P. P. McADAMS MCADAMS

ous generation to identify identify important individual-difference individual-difference variables for hypothesis hypothesistesting research. But, with few exceptions, they abandoned the spirit of those earlier integrative attempts. To paraphrase McClelland's (1951) text, "nobody knew enough" yet to conceptualize the whole person within a single meaningful frame framework. By the end of the 1960s, Allport, AUport, Murray, and Lewin were generally viewed as heroic but rather naive pioneers, and their quests to understand the whole person were considered anachronistic in an era of precise measurements, no-nonsense factor analysis, and tough experimental designs (Fiske, 1971). There was reason to believe, furthermore, that the general concept of a "whole person" might itself itself be an anachronism. Sociologists like Goffman (1959) argued that much of of life is mere role playing and impression management in response to situational demands and unifying and unified core of the person need be considered in understanding that no unifying social-learning what people do and think. Similarly, many social psychologists and social-learning of situations, not the person, is theorists were beginning to suggest that the nature of the primary determinant of of how a person will behave. of the 196Os, 1960s, the stage was set for an ideological ideological battle between By the end of former sought to account for the "trait psychologists" and the "situationists." The former of personality constructs, like achievement motivation and field behavior in terms of of the environment. For both independence; the latter focused on the exigencies of camps, however, the whole person was no longer a factor to be considered, for the first group had split him into little pieces and the second had disregarded him Of course, there were important exceptions to this trend. White (1952, completely. Of of Murray in their 1963) and his colleagues carried on the personological tradition of California of lives." Block (1971) and others at the University of California idiographic "studies of sought to to discern individual differences differences in inpatterns patterns (Barron, 1969; MacKinnon, 1965) sought of traits in the same person evolving over time. of of the unifying themes in the grand theories of of the 1930s and 1940s was One of the central role of of tension reduction in human motivation. In the 1950s, however, the concept received a series of of fierce blows from a number of of different directions. While no single knockout punch was ever delivered, by 1970 the referee was about to call the fight. Research on animals began to suggest that motivation often does not involve any detectable decrease in tension or drive. For instance, Sheffield, Sheffield, Wolff, and Backer (1951) reported that male rats would cross an electric grid to before copulate with a receptive female even though they were always interrupted before orgasm so there was no drive reduction. Harlow, Harlow and Meyer (1950) found that rhesus monkeys would work to disentangle a mechanical puzzle even in the of primary drive reduction. Closer to home, home. White (1959) composed composed a absence of devastating critique of tension reduction in human behavior and argued for a of motivation along the lines of of mastery and competence. reconceptualization of Bowlby (1969) substituted substituted cybernetics and modern ethology ethology for oral libidinal dis discharge to explain the development development of mother-infant mother-infant attachment. Psychoanalysts Psychoanalysts began to disregard Freud's "metapsychology" for its outdated emphasis on erotic and aggressive drives, drives, energy energy transfers, transfers, and and cathexis cathexis (Eagle, 1971). and aggressive (Eagle, 1984; 1984; Guntrip, Guntrip, 1971).

CHAPTER CHAPTER 11 CONCEPTIJAL CONCEPTUALHISTORY HISTORY

19 19

As many observers have noted, psychology as a whole was beginning to experience the cognitive revolution. The gradual erosion of of the doctrine of of tension reduction was part of a larger transformation in American psychology from a cognitive model mechanistic, drive-oriented, stimulus-response viewpoint to a more cognitive experience emphasizing information information processing, image of human behavior and experience making, and the subjective construction of meaning (Singer & & Singer, 1972). Kelly's (1955) personal construct psychology was a harbinger of of cognitive things to come. For Kelly, the person is a quasi-scientist seeking to predict and control his or her world. Each person seeks to make sense of reality through the use of bipolar comprecognitive categories, or personal constructs. To know the whole person is to compre hend the vicissitudes and nuances of of his or her construct system; to comprehend individual differences differences is to compare and contrast the structures, and to a lesser extent contents, of different different persons' construct systems. In a somewhat similar vein, G. A. Miller, Galanter, and Pribram (1960) employed the language of of cybernetics to explain how behavior is rather more guided by rational plans and goals than driven by blind instinct. In the 1960s, more and more personality psychologists were beginning to couch their explanations for human behavior in cognitive terms.

CoNnDENCE:1970 1970TO TOTHE THEPRESENT PRESENT V. DOUBT AND A RENEWAL OF CONFIDENCE: In 1951, McClelland claimed that personahty personality psychologists did not "know much" yet, but the tenor of his text was extremely hopeful, and the reader was still able to conclude that greater knowledge might be just around the corner. Sears's (1950) chapter on personality in the first Annual Annual Review Review of of Psychology Psychology is similarly cautious Bronfenbrenner (1953), and Nuttin but optimistic, as is true of MacKinnon (1951), Bronfenbrenner (1955). Signs of discontent, however, began to appear in the mid-to-late mid-to-Iate 1950s as reviewers seemed to become becoine increasingly frustrated about contradictory empirical results, nagging peccadillos in personality measurement, and the field's overall lack of coherence. By the late 1960s, personality psychology was being called "a of "well controlled studies disconcerting sprawl" (Adelson, 1969; Sanford, 1963) of & that are virtually irrelevant to the questions they are supposed to answer" (Rorer & Widiger, 1983), yielding results that are "inconsequential, trivial, and even pointless" (Sechrest, 1976). The rising tide of of discontent culminated in the publication of of a few extraordi extraordinarily influential critiques of the field-Carlson field—Carlson (1971), Fiske (1974), and Mischel (1968, 1973)-and 1973)—and the spread of a general view that personality psychology was experiencing a major crisis. In the 1970s some even suggested that the field was dead. Outside academia, furthermore, certain social and cultural changes seemed to create a less than hospitable scene for personality psychology. Social upheavals in the late 1960s and early 1970s cast serious doubt on the adequacy of of traditional of people and stable individual differences. frameworks for identifying "types" "types" of of normal persons, personality diagnosis and Both in clinical work and in the study of assessment could be viewed as mere "labelling" by an unempathic and out-of-

20

DAN P. McADAMS DANP.MCADAMS

(Goffman, 1961; 1961; Rosenhan, 1973). 1973). The antiwar, civil rights, touch establishment (Goffman, and women's movements all sensitized Americans to the pervasive influence of influence of culture and environment on human behavior and experience-influence experience—influence experi experienced in the contexts of family, class, ethnicity, race, and nation-state. The implicit product—even a victim-of victim—of social context; there theremessage was this: The person is a product-even person—on social influence influence rather fore, one should focus on context rather than person-on than individuality. In addition, some came to see traditional personality psychology as dominated by an Anglo-masculine world view. One could reasonably argue in 1970 that the only whole persons whom personality psychologists ever studied anyway were upper middle-class, white males. A. De The Decade Decade of of Doubt Doubt A. In an article entitled "Where is the Person in Personality Research?" Carlson (1971) suggested that personality psychology had lost its center. Sampling 226 articles published in two major major personality journals in the late 19608, 1960s, Carlson found found not a single study that fulfilled fulfilled the promises of AUport and Murray concerning Allport personality's commitment to the investigation of the whole person. Rather, the prototypical study was a contrived laboratory experiment or a simple correlational investigation of a large group of college men, about whom the researchers collected information and with whom the researchers spent, on the only a few pieces of information organizaaverage, less than an hour. Virtually abandoned were inquiries into (a) the organiza tion of personality, (b) the stability of personality, (c) problems of the mature developindividual, (d) psychosexuality, (e) striving for personal goals, and (f) the develop friendship or love. She concluded: ment and power of friendship

to be be paying paying an an exorbitant exorbitant price price in in potential potential Personality psychology psychology would would seem seem to Personality knowledge for for the the security security afforded afforded by by preserving preserving norms norms of of convenience convenience and and knowledge methodological orthodoxy. orthodoxy. Must Must these these important, important, unanswered unanswered questions questions be be left left methodological to literature literature and and psychiatry? psychiatry? to

{p^207) (p. 207)

Carlson implied that personality psychologists had lost their way during the era of construct elaboration. Research and theorizing had become so narrow that personality psychologists were no longer able to address the central questions of of the field posed by the grand theorists. By contrast, Fiske (1974) suggested that personality psychology had gone about as far as it could go. From Fiske's even more pessimistic outlook, personality psychology had begun to reach the limits of of scientific study of the person could conceivably achieve. The constructs of what a scientific of personality are inevitably linked to the conventions of everyday language, Fiske claimed. Meanings are bound to be ambiguous, like language. No truly cumulative knowledge base can be built on the shifting sands of personality conceptualizations. The most influential Personalinfluential critique, however, was delivered by Mischel. In Personal ity and Assessment (1968), Mischel's highly selective review concluded that personal personality dispositions, typically measured via paper-and-pencil tests and questionnaires, questionnaires.

CHAPTER 11 CONCEPTUAL CONCEPTUALHISTORY HISTORY CHAPTER

2121

of the variance in human behavior. For the most part, there account for very little of is very little cross-situational generality in behavior, Mischel argued. Instead, human action tends to be dictated by situation ally specific situationally specific factors. Individual differences differences effective predictors of behavior than are individual differences differences in situations are more effective in traits. Mischel raised the possibility that the only place that traits may truly exist of the personality psychologist. Thus, personality personaUty psychologists may be is in the mind of guilty of committing a fundamental fundamental attributional error by imposing broad categories concerning internal dispositions to explain (and predict) the behavior of others, when in fact that behavior is better explained by factors specific to the situation. The critiques of of Carlson, Fiske, and Mischel ushered in a decade of doubt in the history of personality psychology. Many personality psychologists began to doubt the credibility of the entire enterprise of studying persons; others seemed to become become highly defensive, hastily dismissing the critiques as overly simplistic or idealistic lamentations. Over the course of of the decade, an increasing number of of confidence. It is important to note, journal articles considered the mounting crisis of however, that there was more than one crisis during this time, for the critiques are different from each other. Neither Fiske nor Mischel seems especially con convery different cerned with the question, "Where is the person?" And Carlson seems to suggest that personality psychologists could recapture the prize of of the whole person if if they of yesteryear and thereby release the creative would only summon up the will of of methodological orthodoxy. energies that lie trapped beneath the norms of It is also interesting to note that only one of of the critiques was ever seriously addressed by personality psychologists during the 1970s and 1980s. Mischel's indict indictment of of trait psychology ultimately met with a barrage of countercriticism, stimulat stimulating a lengthy "debate" about the relative contributions of of traits and situations in the prediction of of behavior (Alker, 1972; D. J. Bern Bem & & Allen, 1974; Block, 1977; & Goldberg, 1984; Cheek, 1982; Ekehammer, 1974; 1974;Endler Endler & & Bowers, 1973; Chaplin & 1983; Hogan, DeSoto, & & Ozer, 1983; & Magnusson, 1976; Epstein, 1979, 1984; Funder & Solano, 1977; Kenrick & 1981; Mis & Stringfield, 1980; Lamiell, 1981; 1981; McClelland, 1981; Mischel & & Peake, 1982; Ozer, 1986; Rushton, Brainerd, & & Pressley, 1983; Snyder, 1983; West, 1983). What is intriguing to note here is that personality personality psychologists sought to settle the debate on primarily empirical empirical grounds, much in the spirit of Mischel's original critique. This is to say that researchers designed new studies and collected new data to determine (a) the extent to which individual differences differences in traits and situations are able to predict behavior and (b) the extent to which people's behavior can be seen to be consistent over time and across different different situations. Like Mischel, they proceeded in a pragmatic and empirical fashion. In an effort effort to improve the predictive power of of traditional trait measures, certain personality personaUty psychologists have championed (a) moderator variables and (b) aggregation. With respect to the first, D. J. Bern Bem and Allen (1974) and others have suggested that predictions of of behavior can be enhanced when assessments of a person's level on a given trait measure are coupled with assessments of the extent extent to which the given trait is relevant, salient, or important for the person. The latter assessment is conceived as a moderator variable. The argument suggests that only assessment

22 22

DAN P. MCADAMS DANP.MCADAMS

differwhen a trait is relevant, salient, or important for the person may individual differ ences in the trait be predictive of of behavior. In other words, personality psychologists can predict some of the people some of the time. A A number of other moderator approaches have been developed, such as Snyder's (1983) position that the personal personality variable of of self-monitoring self-monitoring serves as a general moderator. According to Snyder, individual differences differences in personality traits are especially predictive of behavior among persons who are low in self-monitoring. These are the people who are relatively oblivious to the demands of situations and, therefore, more likely to act in accord with inner dispositions. Epstein (1979, 1984) has championed the judicious (1979,1984) use of of aggregation in personality studies to increase predictive power. In Epstein's view, trait measures are bound to do a poor job in predicting the single act because the single act is not a reliable index of behavioral trends. When functionally similar acts are aggregated over time and across situations, reliability is enhanced and personality trait measures are able to do a better job of predicting behavior. The trait-situation debate appeared to die down in the 1980s as many psychol psychologists settled on a compromise position that most of them suggested they had advocated all along. Though major differences differences in emphasis are still apparent, many personality psychologists now seem to agree that behavior is a function of both traits (or internal dispositional variables in general) and situations: that the person and the environment interact interact to produce behavior. Though Interactionism is nothing new (see Lewin, 1935), the perception among many reviewers is that personality psychologists are now more explicitly interactionist in their thinking and in in their research designs (Kenrick & & Funder, 1988). The less sanguine view, however, is that a lot of of time and energy have been wasted marshalling empirical support for "great deal various ideological positions. Rorer and Widiger (1983) assert that a **great of of nonsense has been written on the trait-situation trait-situation topic, and as far as we can tell all the data that have been collected are irrelevant to solving the problem, which is conceptual" (p. 446). This is not to suggest that the trait-situation controversy has been a mindless exercise in number crunching. Many of the contributions have been well conceived conceived and ingeniously designed. But the controversy controversy has not directly produced the broad conceptual advances in personality psychology psychology that some observers of the field believe are needed (Carlson, 1984; Helson & & Mitchell, 1978). Furthermore, the empirical activity has tended not to speak directly to the concerns raised by Carlson's (1971) and Fiske's (1974) critiques, which were much more conceptual in nature and, it is probably fair to say, more challenging. Nonetheless, personality psychology appeared to move through the 1980s and into the 1990s with a renewed optimism and vigor (Hogan & 1985; Maddi, & Jones, 1985; 1984; West, 1983). Although the serious doubts raised in the previous decade had not been put to rest, researchers and theorists in the field seemed to have found a new confidence confidence and credibility. credibiUty. Signs of renewal are increasingly manifest in in many different places today. With respect to research methodology, personality psycholo psychologists appear to be employing a wider range of approaches, including naturalistic naturaUstic strategies for experience sampling (Hormuth, 1986), behavioral genetic methods

CHAPTER 11 CONCEYfUAL CONCEPTUALHISTORY HISTORY CHAPTER

23 23

(Plomin, 1986), structural equation models ((Judd, Judd, Jessor, lessor, & Donovan, 1986), and various qualitative methodologies (Helson, 1982; Mendelsohn, 1985; 1985; Runyan, 1982; Wrightsman, 1981). With respect to research topics, personality psychologists have broadened their inquiries to incorporate important issues in health psychology Kobasa, 1985; 1985;Suls Suls& &Rittenhouse, Rittenhouse, 1987) 1987)and andlife-span life-span development development ((Jemmott, Jemmott, 1987; Kobasa, (Eichorn, Clausen, Haan, Honzik, & Mussen, 1981; 1981; Wrightsman, 1988; Zucker, Rabin, Aronoff, & Frank, 1992), and they have made important contributions in of personal personal relationships (Duck, 1986; 1986; Hendrick & Hendrick, studies of the quality of 1986), loneliness and shyness (Briggs, 1985; Shaver & Rubenstein, 1980), gender 1981; Cook, 1985; 1985; Franz & Stewart, 1994), optimal and sex roles (S. L. Bern, 1981; experience (Csikszentmihalyi, 1982; Privette, 1983), adaptation to life changes of personality (A. H. Buss & (Stewart & Healy, 1985), and the biological bases of 1991; Revelle, 1995). An increasing number of Plomin, 1984; D. M. Buss, 1984, 1991; personality and social psychology research projects integrate perspectives from both personality differ(Blass, 1984). And within psychology in general today interest in individual differ ences in persons appears to be increasing.

B. The Whole Person A renewed emphasis on the whole person in contemporary personality psychology A is perhaps most evident in the burgeoning theoretical and research literature on self (Honess & Yardley, 1987; Lapsley & Power, 1988; Loevinger & Knoll, the self of self has traditionally served as 1983; Schlener 1985; Shaver, 1985). The concept of persons as wholes and a rallying point for those psychologists inclined to view persons disposed to ask questions about how persons find unity and coherence in their lives. self has been rediscovered rediscovered in a number of of new guises, including In recent years, the self of "schema" (Markus, 1977), "prototype" (Kuiper & & Derry, 1981), "theory" those of & Gergen, 1983; McAdams, (Berzonsky, 1988; Epstein, 1973), and "story" (Gergen & of Kohut's (1977) self-psychology self-psychology 1985). In psychoanalytic circles, the emergence of is an important conceptual development. of the most fruitful theorizing about the self self comes from the interface Some of of cognitive developmental psychology and personality of personaUty (Loevinger, 1987). Blasi (1988), Damon and Hart (1982), and Kegan (1982) have formulated developmental theories of of the self self that draw on the tradition of structural developmentalism as epitomized in the writings of of Piaget and Kohlberg. These stage theories seek to chart the self's development development from a simple and undifferentiated undifferentiated structure to increasing autonomy, differentiation, and integration. Compared to the developmental formu formulations of of the 1930s and 1940s, these tend to place less emphasis on basic principles of learning and the socialization of of the individual in a particular cultural system. The most influential scheme of of this kind for personality psychology is Loeving Loevinger's (1976) conception of of ego development, which has been carefully carefully operationalized framethrough a sentence completion test. For Loevinger, the ego is one's overall frame of meaning for interpreting experience, encompassing aspects of of character work of development and impulse control, interpersonal style, conscious preoccupations,

24

DAN DAN P. P. McADAMS McADAMS

and cognitive complexity. In the earliest (immature) stages of of ego development, the person adopts a simplistic, global, and egocentric framework framework for understanding experience: the impulsive and self-protective self-protective stages. In the middle stages, one's framework framework of of meaning is more differentiated differentiated and integrated and less egocentric, conventional ways: but reality is now apprehended in stereotypic, banal, and highly conventional the conformist and conformist/conscientious conformist/conscientious stages. At the highest stages (conscien (conscientious, autonomous, or integrated), one comes to question the simple dictates of of issues becomes more highly convention, and one's understanding of a range of differentiated and integrated so that contradiction and ambiguity become tolerable differentiated and the individuality of others is accepted, even "cherished." Persons at the highest stages manifest a rich inner life and complex understanding of of self self as an evolving whole in a social and historical context. Few people reach the highest stages; most "stop" "stop" developing somewhere in the middle. In Loevinger's developmental of ego development is the major individual difference difference typology, one's terminal stage of of personality. personaHty. variable of of the self provide integrative frameworks for viewing Although some theories of unifying whole, others suggest a multiplicity in self self and the person as a unified and unifying identity. Horowitz (1979) presents a clinically anchored scheme of of multiple selves or "states of mind." Markus and Nurius (1986) conceptualize the person in terms of a wide assortment of of "possible selves," each functioning functioning as a semiautonomous of structure containing information concerning what the person believes he or she of "self "selfmight be or fears to be. Similarly, Higgins (1987) has developed a theory of discrepancy" in which various "actual selves," "ideal selves," and "ought selves" of me's. Rosenberg and Gara (1985) have underscored the coexist in a confederacy of of personal identity. Borrowing from deconstructionist deconstructionist literary theory, multiplicity of Sampson (1985) suggests that psychologists should consider the possibility that self need not be unified or coherent. Instead, he argues for a "decentralized, the self nonequilibrium ideal, whose very being hinges on continuous becoming" (Sampson, 1985, p. 1203). In a somewhat somewhat similar vein, Shotter and Gergen (1989) have suggested suggested that the self self is to be viewed as a set of dynamic texts that are constructed and negotiated through social interaction, no single text serving as an integrative core. (1985,1993) self in textual terms but argues that, begin begmMcAdams (1985, 1993) also views self ning in late adolescence and young adulthood, a person strives to create unity and of a single, purpose in life through the conscious and unconscious formulation of text—a dynamic and internalized life story, or personal myth, that inte intedominant text-a grates one's reconstructed past, perceived present, and anticipated future while motivational content situating the person in a social niche and in historical time. The motivational of a person's self-defining self-defining life story is organized along the "thematic lines" of of agency of (power/achievement/autonomy) and communion communion (love/intimacy/care) (Bakan, (power/achievement/autonomy) 1966; Wiggins, 1991). The story displays a characteristic "narrative tone" (ranging from comic or romantic optimism to tragic or ironic pessimism), a unique quality of personal "imagery," pivotal scenes (called "nuclear episodes"), main characters of of idealized self-personifications self-personifications (called "imagoes"), and an anticipated in the guise of story ending that serves to "leave something behind" for the next generation

CONCEPTUALHISTORY HISTORY CHAPTER 11 CONCEPTUAL

25 25

(the "generativity script"). McAdams' narrative conception of of the self self draws from from Adler's (1927) concept concept of of the "guiding "guiding fiction" fiction" and Sartre's notion of of the "true "true human living, and it connects to a growing literature literature novel" (Charme, 1984) in human of narrative in human human lives and personality personality (e.g., Baumeister, Baumeister, on the importance of Wotman, 1992; Bruner, Bruner, 1986, 1986,1990; 1991; Hermans, Kempen, & Stilwell, & Wotman, 1990; Gregg, 1991; Howard, 1991; 1991; Polkinghorne, 1988; Rosenwald Rosenwald & Ochberg, 1992; van Loon, 1992; Howard, Sarbin, 1986; Spence, 1982). C. Motivation Recent years have continued the trend away from from tension-reduction tension-reduction theories of of human motivation and toward toward cognitive approaches for for understanding understanding the dynamics human of action. In the 1970s, Weiner reconceptualized reconceptualized achievement motivation in cogni cogniof attributional terms (Weiner, 1980). Depression and learned helplessness have tive attributional interpreted from from the standpoint standpoint of of dysfunctional dysfunctional attributional attributional styles (Abram (Abrambeen interpreted son, Seligman, & Teasdale, 1978) and faulty faulty schemata (Beck, 1976). More recently, 1976). personality psychologists have proposed proposed a host of of cognitive cognitive variables to account account personality for the goal-directed, inner-motivated inner-motivated features features of of human behavior (see Cantor Cantor & for "personal strivings" (Emmons, 1986), and "personal projects projects Kihlstrom, 1985), "personal "cognitive evaluation (Palys & Little, 1983). Deci and Ryan (1985) have developed a "cognitive of intrinsic motivation" positing a basic human desire to feel competent competent and theory of self-determining. Carver Carver and Scheier (1981) have sought to explain motivation motivation in self-determining. of a hierarchy of of control systems and feedback feedback loops. terms of script theory theory represents represents an ambitious attempt to integrate integrate Tomkins' (1987) script certain certain cognitive themes within within a theory theory of of motivation and personality personality that places affect. Tomkins identifies identifies approximately approximately 10 primary primary affects, affects, prime emphasis on affect. such as joy, excitement, sadness, and anger. Izard (1977) has articulated articulated a very very similar view. Each of of these affects affects has served an adaptive function function throughout throughout human evolution, and each is associated with a particular physiological response, including a corresponding corresponding set of of facial facial expressions. In Tomkins' view, affects affects are the primary motivators of human behavior, amplifying biological drives and provid of amplifying providing life's goals with the emotional coloring that makes tl.tem worthy of pursuit. them worthy of Tomkins Tomkins views the person as a playwright who fashions fashions his or her personal drama from the earliest weeks of life. The most basic component from of component of of the drama drama is the "scene," which is viewed as an idealized recollection recollection of of a specific happening or event in one's life life which which contains at least one affect affect and one object object of of that affect. affect. A A "script" is a set of of rules for for interpreting, creating, enhancing, or defending defending against a family of related scenes (Carlson, 1988). The process of connecting scenes into into a meaningful meaningful pattern is called "psychological magnification"-a magnification"—a process that works works differently differently for for different different sorts of of scripts and affect affect patterns. Understanding Understanding the unique patterning patterning of of human human motivation motivation in an individual's life involves an intensive analysis of of the recurrent recurrent affects, critical scenes, scripts, and different different modes of of psychological magnification magnification that the person manifests manifests across the life span.

26 26

DAN P. McADAMS DANP.MCADAMS

theory is developed developed more fully fully in The evolutionary theme in Tomkins' script theory Hogan's (1987) socioanalytic theory ties together socioanalytic theory theory of of personality. Socioanalytic theory strands of of evolutionary biology, psychoanalysis, and sociological role theory. Human Human beings have evolved to live in small groups that are variously organized organized into status hierarchies. In this context, the two central motivational tendencies in human human behav behavior are toward seeking acceptance and seeking status in social groups. As Hogan Hogan puts it, "getting along and getting ahead are the two great problems in life that each person must solve" (Hogan, Jones, & & Cheek, 1985). The two great problems problems are always addressed and resolved resolved in the context context of of ritualized ritualized social interaction. interaction. Following Mead Mead (1934) and Goffman Goffman (1959), Hogan Hogan views social behavior behavior as an Following elaborate game, governed governed by rules and conventions, scripted scripted into roles and routines, and mastered mastered by the most skillful skillful managers of of impressions. Through role playing and impression impression management management the individual finds a part part to play in society, a social identity that specifies specifies a recognized recognized niche in the community. This is not to trivialize social behavior. Rather, the striving for for status and and acceptance acceptance through through ritualized social interaction interaction is an unconscious, central, genetic tendency tendency for for all human human beings: .. .. .. self-presentation self-presentation and and impression impression management management are not trivial trivial party party games. games. They are fundamental fundamental processes, rooted in our history as group-living animals. They are archaic, powerful, powerful, compUlsive compulsive tendencies that are closely tied to our chances for for survival and reproductive success.

(Hogan et al., 1985, p.p.181) al, 1985, 181)

and Hogan's socioanalytic theory are are indicative of of the Tomkins' script theory and of affect affect and growing interest in personality psychology today in the concepts of Zajonc (1980) and Rychlak (1988) have argued argued that instinct. Zajonc that the first step in the apprehension affective judgment-that apprehension of of any event or experience is a basic affective judgment—that emo emotional preferences preferences precede, and are more basic than, cognitive inferences. Other theorists have sought to integrate cognitive and affective affective approaches (Izard, 1977; Singer & & Kolligan, 1987). The concept of of biological instincts has attained attained a new Singer respectability, as expressed in ethological conceptions conceptions like Bowlby's (1969) attach attachment ment theory-which theory—which has been expanded in creative ways to organize research research & Shaver, 1987)-and 1987)—and in Hazan & and theory on human love and adaptation (e.g., Hazan provocative speculations about the application application of of sociobiological sociobiological viewpoints to personality and social psychology (D. M. Buss, 1984; Cunningham, 1981). personality Differences among People D. Differences trait-situation Personality psychologists have come back to traits. Now that the trait-situation controversy controversy has subsided, a steady stream of research findings findings have documented documented impressive longitudinal consistency of important important individual difference consistency in a number number of difference variables (e.g., Conley, 1985; Costa, McCrae, & & Arenberg 1980). New ways of of understanding traits have also been proposed. In their "act-frequency "act-frequency approach" approach" to personality, D. M. Buss and Craik Craik (1984) conceive of of traits as summary categories categories Different act members of of a containing discrete and representative behavioral acts. Different

CHAnER CHAPTER 11 CONCEPTUAL CONCEPTUALHISTORY HISTORY

27 27

trait family differ in their "prototypicality." Those closest to the "center" of of the family are "best examples" of a given trait, as the act "talking to a stranger" might function as an especially prototypical example for extraversion. Those acts on the representative and likely to shade into other adjacent trait cate cateperiphery are less representative gories. In the 1980s, personality psychologists expressed a great deal of interest in formulating a single systematic taxonomy taxonomy for personality traits. Such a framework identify a finite set of central, most salient, saUent, or highest-order personality traits might identify and/or place various traits into a conceptually appealing order. Expanding upon the early work of Leary (1957), Wiggins and Broughton (1985) refined a circumplex model of traits organized according to the orthogonal axes of of strength (e.g., dominant-submissive) and warmth (e.g., agreeable-quarrelsome). Eysenck (1973) extraversion-introversion has proposed his own circumplex, organized according to extraversion-introversion and neuroticism-stability. Covering the same conceptual space as Eysenck, Gray physiologi(1987) suggests that anxiety and impulsivity represent two primary and physiologi cally grounded orthogonal dimensions in personality, each tilted 45° to Eysenck's pair. Another increasingly influential system for conceptualizing differconceptualizing individual differ ences comes from the longitudinal investigations of Block (1971, 1993; Funder, of Parke, Tomlinson-Keasey, Tomlinson-Keasey, & & Widaman, Widaman, 1993) employing the California Q Set. Set. Parke, 1993) employing the California Two major dimensions underlying the various personality types and developmental Two major dimensions underlying the various personality types and developmental trajectories identified identified by by Block Block and and his his colleagues colleagues are are ego ego resiliency resiliency and ego control. control. trajectories and ego of individual differences differences At the current time, the most influential formulation of of Fiske in personality is the "Big Five" trait taxonomy. Building on the early work of of personality (1949), Norman (1963), and Tupes and Christal (1961), a number of of trait dimensions can be reduced psychologists have proposed that the universe of 1981,1993; to approximately five basic bipolar categories (Digman, 1990; Goldberg, 1981, 1993; McCrae & & Costa, Costa, 1987). 1987). Different Different factor-analytic factor-analytic studies studies John, 1990; McCrae, 1992; McCrae different ways, but a representative representative breakdown is that have cut the pie in slightly different of McCrae and Costa (1987), who identify identify the five as (1) extraversion-introversion extraversion-introversion of (E) (2) neuroticism (N), (3) openness to experience (0), (O), (4) agreeableness agreeableness(E) (A), and (5) conscientiousness-undirectedness conscientiousness-undirectedness (C). Goldberg's pains painsantagonism (A), taking lexical analyses suggest that these five dimensions are encoded in language. dimenAt least in the case of English, these five may serve as the grand organizing dimen sions with respect to which virtually all trait labels for describing describing general noncondi noncondiexperience can be construed. tional individual differences in human behavior and experience

VI. VI. CONCLUSIONS: PROGRESS AND STAGNATION In conclusion, the history of personality psychology in the twentieth century may be broadly viewed from the standpoint of conceptual progress and stagnation. The field of of personality has traditionally emphasized the study of of the whole person, the dynamics of human motivation, and the identification identification and measurement of of individ-

28 28

DAN P. McADAMS DANP.MCADAMS

ual differences among persons. How much conceptual progress has been made in each of these three areas? First, significant progress may be seen in the conceptualization of of human motivation. The decline of general drive-reduction theories and the emergence of highly differentiated differentiated cognitive-affective cognitive-affective approaches to understanding the dynamics of action appear to represent a major conceptual advance. The recent formulations formulations of of Tomkins (1987), Izard (1977), Hogan (1987), McClelland (1985), and Bowlby of (1969) draw upon some of the best ideas from modern cognitive psychology while grounding motivational theory in human evolution and emotional dynamics. These less-than-rational emotions and instinc instinctheories do not simply "leave room for" the less-than-rational cognitivetual tendencies. Rather they portray human motivation in complex cognitive affective-instinctual terms and, in the cases of of Tomkins and Hogan, provide a very affective-instinctual of human compelling sociocultural context within which to understand the dynamics of behavior and social interaction. moderate progress may be observed. In the area of individual differences, moderate The empirical elaboration of personality constructs beginning in the 1950s, the personsubsequent trait-situation debate, and the recentflurry flurry of of research on broad person ality dispositions have combined to enrich and broaden psychologists' understanding of key personality traits while underscoring their limitations and their situationally of specific manifestations. Those who complained that the only good way to organize (Lonthe plethora of possible personality traits was that provided by the alphabet (Lon & Exner, 1978) may now take heart in the emergence of of circumplex models don & efforts to order and the Big Five as compelling organizing schemes. Although the efforts trait dimensions deserve resounding applause, one begins to be concerned in this area about creeping conceptual imperialism. Psychologists should not be too quick framework. to assimilate every conceptual scheme under the sun to the Big Five framework. Simply reducing the person to five trait scores will not satisfy satisfy those who seek a more differentiated portrait for comprehending individual differences. Furthermore, the differentiated Big Five dimensions do not directly address many issues with which personality psychologists have traditionally been concerned-issues concerned—issues such as personality dynam dynamics, personality development, life changes, life histories, identity and the develop development of self, and the relation between the person on the one hand and society, culture, and history on the other (McAdams, 1992). Finally, it is disappointing to note that little progress appears to have been made in the conceptualization of the whole person. The reemergence reemergence of of the self as a viable construct in personality psychology is surely a positive development in of ego develop developthis regard. But with the possible exception of Loevinger's theory of self theories have yet to provide the breadth and depth necessary necessary to integrate ment, self comprehending the whole person. The hope of of disparate conceptual strands in comprehending AUport and Murray that personality psychology would someday someday provide provide a coherent Allport way of understanding the whole person has not yet been realized. The grand theories of the 1930s and 1940s have not proven adequate to the task, though they continue of to provide insights and guidelines. More recent theorizing about the person has been more limited in scope, with the possible exception of of Tomkins (1987), whose

CHAPTER 1

CONCEPTUAL HISTORY

29 29

multifaceted theory still needs to be systematized systematized and fleshed fleshed out before before it can expect from the lack of expect to gain wide appeal. The field field of personality personality still still suffers suffers from of a persuasive integrative framework framework for understanding understanding the person person as a differentiated differentiated complex social largely the dynamic whole whole living living in a complex and integrated dynamic social context. It was largely encompassing frameworks frameworks in the 1930s 1930s and 1940s that established generation of such encompassing personality psychology's reputation Ass the grand dissident field. field. A grand theories reputation as that of a dissident field of personality personality seemed seemed to become came to be rejected, the field become more more conventional, dissenting champion champion of of the whole Perhaps any losing its unique status as the dissenting losing whole person. Perhaps integrative conceptual whole person person is doomed conceptual framework framework for comprehending comprehending the whole doomed to be rejected sooner sooner or later. But until the field field of of personality personality begins begins again to generate such candidates for rejection, it will fall fall somewhat somewhat short short of of fulfilling fulfilling the promise of pioneers. of its pioneers.

ACKNOWLEDGMENT ACKNOWLEDGMENT I thank Rae Rae Carlson, Jonathan Cheek, Bob Bob Hogan, John Johnson, Jane Loevinger, Dave Dave and Mac Runyan Runyan for their gracious and extensive comments on McClelland, Bob Nicolay, and early drafts of this chapter. The preparation of the manuscript was aided by a grant from The Spencer Foundation. Correspondence should be addressed to Dan P. McAdams, Graduate and Social Policy, Northwestern University, 2115 2115 North Program in Human Development and Campus Drive, Evanston, IL 60208.

REFERENCES REFERENCES L. Y., Seligman, M. M. E. P., J. D. (1978). Learned helplessness in Abramson, L. P., & Teasdale, J. and reformulation. Journal Journal of of Abnormal Abnormal Psychology, Psychology, 87, 49-74. humans: Critique and Adelson, J. J. (1969). Personality. Annual Annual Review Review of of Psychology, Psychology, 20, 20, 217-252. Adler, A. (1927). The practice practice and theory theory of of individual individual psychology. psychology. New York: Harcourt Brace World. Adorno, T. W., Frenkel-Brunswik, E., Levinson, D. J., J., & Sanford, R. N. (1950). The authoritar authoritarian personality. personality. New York: Harper & Brothers. Alker, H. A. (1972). Is personality personality situationally specific specific or intrapsychically intrapsychically consistent? Journal Journal

of of Personality, Personality, 40, 40, 1-15. Allport, AUport, G. W. (1937). Personality: Personality: A psychological psychological interpretation. interpretation. New York: Henry Holt. Allport, AUport, G. W. (1954). The historical background of modem social psychology. In G. Lind Lind1, pp. 3-56). Cambridge, MA: MA: zey (Ed.), Handbook Handbook of of social social psychology psychology (Vol. 1, Addison-Wesley. Allport, for a psychology W. (1955). Becoming: Basic Basic considerations considerations for psychology of of personality. personality. New AUport, G. W. CT: Yale University Press. Haven, CT: AUport, G. W. W. (1961). Pattern and growth growth in personality. personality. New York: Holt, Rinehart & Allport, Winston. Allport, W. Allport AUport (Ed.), The person person in psychology psychology AUport, G. W. (1968). An autobiography. In G. W. (pp. 376-409). Boston: Beacon Press.

30 30

DAN P. McADAMS DANP.MCADAMS

Allport, G. W., values. Journal AUport, W., & & Vernon, P. P. E. E. (1931). (1931). A A test test for for personal personal values. Journalof ofAbnormal Abnormal and Social Social Psychology, Psychology,25, 25,368-372. 368-372. Anastasi, A A. (1958). Differential Differentialpsychology psychology(3rd (3rded.). ed.).New NewYork: York:Macmillan. Macmillan. Anastasi, A Psychologicaltesting testing(4th (4thed.). ed.).New NewYork: York:Macmillan. Macmillan. A. (1976). Psychological Angyal, A ofof personality. New York: Commonwealth Fund. A. (1941). Foundations Foundationsfor foraascience science personality. New York: Commonwealth Fund. Asch, S. E. (1951). Effects of group pressure upon the modification and distortion of judgmen (pp. 177-190). Pittsburgh, ments. In H. Guetzkow (Ed.), Groups, Groups, leadership, leadership,and and men (pp. 177-190). Pittsburgh, PA: PA: Carnegie Press. Atkinson, J. W. (1964). An introduction introductiontotomotivation. motivation.New NewYork: York:Van VanNostrand-Reinhold. Nostrand-Reinhold. Bagby, E. (1928). The The psychology psychologyof ofpersonality. personality.New NewYork: York:Henry HenryHolt. Holt. Isolation andand communion in Western man. man. Bakan, Bakan, D. (1966). (1966). The Theduality dualityof ofhuman humanexistence: existence: Isolation communion in Western Boston: Beacon Press. Barron, F. process. New York: Holt, Rinehart & Winston. F. (1969). (1969). Creative Creativeperson personand andcreative creative process. New York: Holt, Rinehart & Winston. A, & Baumeister, R. R. F., F., Stilwell, Stilwell, A., & Wotman, Wotman, S. S. R. R. (1990). (1990). Victim Victim and and perpetrator perpetrator accounts accounts of interpersonal conflict: conflict: Autobiographical Autobiographical narratives narratives about about anger. anger.Journal Journalof ofPersonal Personalof Social Psychology, Psychology,59, 59,994-1005. 994-1005. ity and Social Beck, A disorders. New York: International Cognitivetherapy therapyand andthe theemotional emotional disorders. New York: International A. T. (1976). Cognitive Universities Universities Press. Bern, Bem, D. J., & & Allen, A. (1974). On predicting some of the people some of the time: time: The 81, search for cross-situational consistencies in behavior. Psychological Psychological Review, Review, 81, 506-520. 506-520. Bern, Bem, S. S. L. L. (1981). (1981). Gender Gender schema schema theory: theory: A A cognitive cognitive account account of of sex sex typing. typing. Psychological Psychological Review, Review, 88, 88, 354-364. 354-364. Bernreuter, R. G. (1931). The Stanford, CA: Stanford University Press. Bemreuter, Thepersonality personalityinventory. inventory. Stanford, CA: Stanford University Press. Berzonsky, Lap Berzonsky, M. D. D. (1988). (1988). Self-theorists, Self-theorists, identity identity status, status, and and social social cognition. cognition. In In D. D. K. K. LapIntegrative approaches (pp.(pp. 243-262). sley sley & & F. F. C. C.Power Power (Eds.), (Eds.), Self, Self ego, ego,and andidentity: identity: Integrative approaches 243-262). New York: York: Springer-Verlag. Springer-Verlag. New Psychology, 10,10,203-232. 203-232. Blake, R. R., & & Mouton, Mouton, J.J.S.S.(1959). (1959).Personality. Personality.Annual AnnualReview Reviewofof Psychology, Blasi, A. (1988). Identity and the development of the self. In D. K. Lapsley & & F. C. C. Power (Eds.), Self, Self ego, ego, and identity: identity: Integrative Integrative approaches approaches (pp. (pp. 226-242). New York: Springer-Verlag. Blass, T. ofof Person T. (1984). (1984). Social Social psychology psychology and andpersonality: personality:Toward Towardaaconvergence. convergence.Journal Journal Personality and Social Social Psychology, Psychology,47, 47,1013-1027. 1013-1027. Block. sets: Unconfounding meaning, acquiescence, and and Block, J. (1965). The The challenge challengeofofresponse response sets: Unconfounding meaning, acquiescence, social desirability in the MMPI. New York: Appleton-Century-Crofts. desirability in the MMPI. New York: Appleton-Century-Crofts. Block, J. (1971). Lives through throughtime. time.Berkeley, Berkeley,CA: CA:Bancroft BancroftBooks. Books. Block, J. (1977). Advancing the psychology of personality: Paradigmatic shift or improving the quality of research? In D. Magnusson & & N. S. Endler (Eds.), Personality Personalityatatthe the Hillsdale, NJ: crossroads: Current issues in interactional psychology. crossroads: Current issues in interactional psychology. Hillsdale, Erlbaum. NJ: Erlbaum. Block, J. J. (1993). Studying personality the long way. In D. C. C. Funder, R. D. Parke, C. Tomlinson-Keasey, & & K. Widaman (Eds.), Studying Studying lives lives through through time: time:Personality Personality and development development (pp. (pp.9-44). 9-44). Washington, Washington,DC: DC:American American Psychological Psychological Association. Association. Bowers, K. S. (1973). Situationism in psychology: An analysis and a critique. Psychological Psychological Review, Review, 80, 80, 307-336. Attachment. New York: Basic Books. Bowlby, J. (1969). Attachment Attachmentand andloss: loss:Vol. Vol.1. L Attachment. New York: Basic Books. Briggs, S. R. (1985). A trait account of social shyness. Review of of Personality Personality and andSocial Social Psychology, Psychology, 6,6,35-64. 35-64. Bronfenbrenner, U. (1953). Personality. Annual Review Review of of Psychology, Psychology,4,4,157-182. 157-182.

CHAPTER 1

HISTORY CONCEPTUAL HISTORY

31 31

riddle of of personality. personality. New York: Moffat, Yard. Bruce, H. A. (1908). The riddle Bruner, J. (1986). Actual Actual minds, possible possible worlds. Cambridge, MA: Harvard University University Press. Acts of of meaning. Cambridge, MA: Harvard University Press. Bruner, J. (1990). Acts R. (1984). Temperament: Early developing developing personality personality traits. Hillsdale, Buss, A. H., & Plomin, R NJ: Erlbaum. Buss, D. M. (1984). Evolutionary Evolutionary biology biology and personality personality psychology: psychology: Toward a conception conception of of human nature and individual differences. American American Psychologist, Psychologist, 39, 361-377. Buss, D. M. (1991). Evolutionary personality psychology. Annual Annual Review Review of of Psychology, Psychology, 42, 459-491. Buss, D. M., & Craik, K. H. (1984). Acts, dispositions, and personality. Progress Progress in Experimen Experimental Personality Personality Research, 13, 241-301. Campbell, D. T., & Fiske, D. W. (1959). Convergent and discriminant validation by the multitrait-multimethod multitrait-multimethod matrix. Psychological Psychological Bulletin, 56, 81-105. basis of personality. Cantor, N., & Kihlstrom, J. F. (1985). Social intelligence: The cognitive basis Review of of Personality Personality and and Social Social Psychology, Psychology, 6, 15-34. Review Carlson, R R. (1971). Where is the person in personality research? Psychological Psychological Bulletin, 75, 203-219. 75, Carlson, R R. (1984). What's social about social psychology? Where's the person person in personality Journal of of Personality Personality and Social Social Psychology, Psychology, 47, 1304-1309. research? Journal R. (1988). Exemplary lives: The use of of psychobiography for theory development. development. Carlson, R Journal of of Personality, Personality, 56, 105-138. Journal selfCarver, C. S., & Scheier, M. F. (1981). A control systems approach to behavioral self Review of of Personality Personality and Social Social Psychology, Psychology, 2, 107-140. regulation. Review R. B. (1946). Description Description and and measurement measurement of of personality. personality. New York: World Book. Book. Cattell, R Cattell, R R. B. (1950). Personality: Personality: A A systematic, theoretical, and and factual factual study. New York: McGraw-Hill. R. B. (1957). Personality Personality and and motivation motivation structure and measurement. New York: Cattell, R World Book. R. (1984). A failure to replicate the Bern Bem and Allen study Chaplin, W. F., & Goldberg, L. R of individual differences differences in cross-situational consistency. Journal Journal of of Personality Personality and of Social Psychology, Psychology, 47, 1074-1090. Social Charme, S. L. (1984). Meaning and and myth myth in the study study of of lives. Philadelphia: University University of of Pennsylvania Press. Cheek, J. M. (1982). Aggregation, moderator variables, and the validity of of personality tests. Journal Journal of of Personality Personality and and Social Social Psychology, Psychology, 43, 1254-1269. Christie, R, R., & Lindauer, F. (1963). Personality structure. Annual Annual Review Review of of Psychology, Psychology, 14, 201-230. Comte, A. (1852). The positive positive polity. polity. London: Longmans, Green. Conley, J. J. (1985). Longitudinal stability of personality traits: A A multi-trait-multimethod multi-trait-multimethodmultioccasion analysis. Journal Journal of of Personality Personality and Social Social Psychology, Psychology, 19, 1266-1282. Cook, E. P. (1985). Psychological Psychological androgyny. androgyny. New York: Pergamon. Costa, P. T., Jr., McCrae, R R. R, R., & Arenberg, P. (1980). Enduring dispositions in adult males. Journal Journal of of Personality Personality and Social Social Psychology, Psychology, 38, 793-800. Craik, K. H. (1993). The 1937 Allport AUport and Stagner texts in personality psychology. psychology. In K. H. Craik, R R. Hogan, & R R. N. Wolfe (Eds.), Fifty years years of of personality personality psychology psychology (pp. 3-20). New York: Plenum Press. Cronbach, L. J., & Meehl, P. E. (1955). Construct validity in psychological psychological tests. Psychological Psychological Bulletin, 52, 281-302.

32 32

DAN DAN P. P. McADAMS MCADAMS

Csikszentmihalyi, M. M. (1982). (1982).Toward Towardaapsychology psychologyof ofoptimal optimalexperience. experience.Review ReviewofofPersonal Personal-

ity and Social Social Psychology, Psychology,3,3,13-36. 13-36.

Cunningham, M. R. (1981). Sociobiology as aa supplementary supplementary paradigm paradigm for for social social psychologi psychological research. Review 2, 2, 69-106. Review of of Personality Personalityand andSocial SocialPsychology, Psychology, 69-106. Damon, W., & «&Hart, Hart,D.D.(1982). (1982).The Thedevelopment developmentofofself-understanding self-understandingfrom frominfancy infancythrough through Damon, W., adolescence. Child Child Development, Development,53, 53,841-864. 841-864. Deci, E. L., in in human L., & & Ryan, R. M. (1985). Intrinsic Intrinsicmotivation motivationand andself-determination self-determination human behavior. behavior.New NewYork: York:Plenum PlenumPress. Press. M. (1990). Personality Digman, J. J. M. Personahty structure: structure: Emergence Emergence of of the the five-factor five-factor model. model. Annual Annual

Review Review of of Psychology, Psychology, 41 41,417-440. 417-440. ,

Dilthey, Dilthey, W. W. (1976). The The development of of hermeneutics. In In H. H. P. P. Rickman Rickman (Ed.), (Ed.), W. W. Dilthey: Dilthey: Selected Selected writings. writings,Cambridge, Cambridge,England: England:Cambridge CambridgeUniversity UniversityPress. Press.(Original (Originalwork work published 1900). published Dollard, psychotherapy. New York: McGraw-Hill. DoUard, J., J., & & Miller, Miller,N. N.E. E.(1950). (1950).Personality Personalityand and psychotherapy. New York: McGraw-Hill. Duck, S. (1986). Human Human relationships: relationships: introduction to social psychology, London: Duck, AnAn introduction to social psychology. London: Sage.Sage. Eagle, M. M. N. (1984). Recent Recentdevelopments developmentsinin psychoanalysis: A critical evaluation. York: Eagle, psychoanalysis: A critical evaluation. NewNew York: McGraw-Hill. McGraw-Hill. Edwards, A A. L. L. (1957). (1957).The TheEdwards Edwardspersonal personal preference schedule. New York: PsychologiEdwards, preference schedule. New York: TheThe Psychologi cal Corporation. cal Corporation. Eichom, J. A., Haan, N., N., Honzik, Honzik, M. M. P., & Mussen, Mussen, P. P. H. H. (1981). Present Present Eichorn, D. D. H., H., Clausen, Clausen, J. A, Haan, P., & in middle middle life. life. New NewYork: York:Academic AcademicPress. Press. and past in Ekehammer, Interactionism in in personality personality from from an an historical historical perspective. perspective. Psycho PsychoEkehammer, B. B. (1974). Interactionism

logical Bulletin, Bulletin,81, 81,1026-1048. 1026-1048. logical

Emmons, R. A A. (1986). Personal Personal strivings: strivings: An An approach approach to to personality personality and and subjective subjective well wellEmmons, Journalof ofPersonality Personalityand andSocial Social Psychology, 1058-1068. being. Psychology, 51, 51, 1058-1068. being. Journal Endler, Endler, N. N. S., & & Magnusson, Magnusson, D. (1976). Toward Toward an an interactional interactional psychology psychology of of personality. personality. Psychological 956-974. PsychologicalBulletin, Bulletin,83,83, 956-974.

Epstein, S. The self-concept self-concept revisited: revisited: Or Oraatheory theoryofofaatheory. theory.American American Psychologist, Epstein, S. (1973). The Psychologist, 28, 28, 404-416.

Epstein, S. (1979). The stability of behavior: behavior: 1. 1. On predicting most of the people much of of Psychology, 77, 77, 1097-1126. the time. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 1097-1126. Epstein, S. (1984). The stability of behavior across time and situations. In R. A A. Zucker, J. prediction ofbehavior (pp.(pp. 209-268). Aronoff, & &A A. I.I.Rabin Rabin(Eds.), (Eds.),Personality Personalityand andthe the prediction of behavior 209-268). New York: Academic Press. Erikson, E. H. (1950). Childhood Childhood and andsociety. society.New NewYork: York:Norton. Norton. Erikson, E. H. (1959). Identity and the life cycle: Selected papers. Psychological PsychologicalIssues, Issues,

1(1), i(l), 5-165.

Eysenck, H. J. (1952). The personality. London: Routledge && Kegan Paul. Thescientific scientificstudy studyofof personality, London: Routledge Kegan Paul. Wiley. Eysenck, H. J. (1973). Eysenck Eysenck on onextraversion. extraversion.New NewYork: York: Wiley. Festinger, L. (1954). A theory of social comparison processes. Human 1 17-140. HumanRelations, Relations,7, 7,117-140. Fiske, D. W. (1949). Consistency of the factorial structures of personality ratings from 329-344. different sources. Journal Journalof ofAbnormal Abnormaland andSocial SocialPsychology, Psychology,44,44, 329-344. Fiske, D. W. (1971). Measuring personality. Chicago: Aldine. Measuringthe theconcepts conceptsofof personality, Chicago: Aldine. Fiske, D. W. ofof Personal W. (1974). (1974).The Thelimits limitsofofthe theconventional conventionalscience scienceofofpersonality. personality.Journal Journal Personal-

ity, ity, 42, 42, 1-11.

Franz, c., lives. Boulder, CO:CO: Westview Press. C, & &Stewart, Stewart,A A.J.J.(Eds.). (Eds.).(1994). (1994).Women Womencreating creating lives, Boulder, Westview Press. Freud, S. (1953). Three essays on the theory of sexuality. In J. Strachey (Ed. (Ed. and Trans.), The psychological works of Sigmund Freud (Vol.(Vol. 7, 7, The standard standardedition editionofofthe thecomplete complete psychological works of Sigmund Freud pp. 126-243). London: London: Hogarth Press. (Original work published 1905).

CHAPTER 1

CONCEPTUAL CONCEPTUAL HISTORY HISTORY

33 33

Freud, S. (1955). Beyond the pleasure principle. In J. Strachey (Ed. and Trans.), The standard standard edition of Sigmund Freud of the complete complete psychological psychological works works of of Sigmund Freud (Vol. 18, pp. 3-64). London: Hogarth Press. (Original work published 1920). Freud, S. (1961). (1%1). The ego and the id. In J. Strachey (Ed. and Trans.), The standard standard edition edition of of the complete complete psychological psychological works works of of Sigmund Sigmund Freud (Vol. 19, pp. 12-66). London: Hogarth Press. (Original work published pubhshed 1923). Fromm, E. (1941). Escape Escape from from freedom. freedom. New York: Farrar & & Rinehart. Funder, D. C., C, & & Ozer, D. J. (1983). Behavior as a function of of the situation. Journal Journal of of Personality Personality and Social Social Psychology, Psychology, 44, 107-112. Funder, D. C., C , Parke, R R. D., Tomlinson-Keasey, c., C, & & Widaman, K. (Eds.). (1993). Studying Studying lives through time: Personality Personality and development. development. Washington, DC: American Psycho Psychological Association. Gamett, A. C. (1928). Instinct Instinct and and personality. personality. New York: Dodd, Mead. Garnett, R., Beatty, W. W., & & Arkin, R R. M. (1984). Human Human motivation: motivation: Physiological, Physiological, behavioral, behavioral, Geen, R, and social social approaches. approaches. Boston: Allyn AUyn & Bacon. and R. Sarbin & & K. E. & Gergen, M. M. (1983). Narratives of of the self. In T. R Gergen, K. J., & social identity. New York: Praeger. Scheibe (Eds.), Studies in social presentation of of self self in everyday everyday life. Garden City, NY: Doubleday. Goffman, E. (1959). The presentation Asylums. Garden City, NY: Doubleday. Goffman, E. (1961). Asylums. R. (1981). Language and individual differences: The search for universals in Goldberg, L. R Review of of Personality Personality and Social Social Psychology, personality lexicons. Review Psychology, 2, 141-166. American Psychologist, Psychologist, R. (1993). The structure of phenotypic personality traits. American Goldberg, L. R 48, 26-34. Gough, H. G. (1957). California psychological psychological inventory: inventory: Manual. Palo Alto, CA: Consulting Psychologists Press. Gray, J. A. (1987). Perspectives on anxiety and impulsivity: A A commentary. Journal Journal of of Research in Personality, Personality, 21, 493-509. studies in identity Gregg, G. S. (1991). Self-representation: Life Life narrative narrative studies identity and and ideology. ideology. New York: Greenwood Press. Guilford, J. P. (1959). Personality. New York: McGraw-Hill. Guilford, J. P., & & Zimmerman, W. S. (1949). The Guilford-Zimmerman Guilford-Zimmerman temperament temperament survey. survey. Beverly Hills, CA: Sheridan Supply Co. self. New York: Basic Books. Guntrip, H. (1971). Psychoanalytic Psychoanalytic theory, therapy, and and the self Hall, C. S., & & Lindzey, G. (1957). Theories of of personality. personality. New York: Wiley. Harlow, H. F., Harlow, M. K., & & Meyer, D. R R. (1950). Learning motivated by a manipulation drive. Journal Journal of of Experimental Experimental Psychology, Psychology, 40, 228-234. Hartmann, H. (1939). Ego psychology psychology and and theproblem problem of of adaptation. adaptation. New York: International Universities Press. Hazan, c., C, & & Shaver, P. (1987). Romantic love conceptualized conceptualized as an attachment attachment process. Journal Journal of of Personality Personality and and Social Psychology, Psychology, 52, 511-524. Heider, F. (1958). The psychology psychology of of interpersonal interpersonal relations. New York: Wiley. Helson, R R. (1982). Critics and their texts: An approach to Jung's theory of of cognition and personality. Journal Journal of of Personality Personality and Social Social Psychology, Psychology, 43, 409-418. Helson, R, Review of Psychology, 29, R., & Mitchell, V. (1978). Personality. Personahty. Annual Annual Review of Psychology, 29, 555-586. Hendrick, c., C , & Hendrick, S. (1986). A theory and method of of love. Journal Journal of of Personality Personality and Social Social Psychology, Psychology, 50, 392-402. Hermans, J. J. M., Kempen, H. J. G., & & van Loon, R R. J. P. (1992). The dialogical self: Beyond individualism and rationalism. American American Psychologist, Psychologist, 47, 23-33.

34 34

DAN DAN P. P. MCADAMS MCADAMS

Higgins, E. E. T. T. (1987). (1987). Self-discrepancy: Self-discrepancy: A A theory theory relating relating self self and and affect. affect. Psychological Psychological Review, Review, 94, 94, 319-340. 319-340. Englewood Cliffs, NJ:NJ: Hogan, R R. (1976). Personality Personality theory: theory:The Thepersonological personologicaltradition. tradition, Englewood Cliffs, Prentice-Hall. & Hogan, R R. (1987). Personality psychology: Back to basics. In J. Aronoff, A. I. Rabin, & R. A. A. Zucker Zucker (Eds.), (Eds.), The The emergence emergenceofofpersonality personality(pp. (pp.79-104). 79-104).New NewYork: York:Springer. Springer. R (1988). The Themeaning meaningof ofpersonality personalitytest testscores. scores.American AmericanPsychologist, Psychologist, Hogan, R R. (1988). 43,43,621-626. 621-626. Hogan, R, R., DeSoto, DeSoto, C. C. B., & & Solano, Solano, C. C. (1977). (1977). Traits, Traits, tests, tests, and and personality personality research. Hogan, American American Psychologist, Psychologist,32, 32,255-264. 255-264. Hogan, R, R., & & Jones, Jones, W. H. H. (Eds.). (Eds.). (1985). (1985). Perspectives Perspectivesininpersonality personality(Vol. (Vol.1).1). Greenwich, Greenwich, Hogan, CT: JAI JAI Press. CT: Hogan, R., & Cheek, J. M. (1985). Socioanalytic theory: An alternative to R, Jones, W. H., & armadillo psychology. In B. R Theself selfand andsocial sociallife life(pp. (pp.175-198). 175-198). R. Schlenker (Ed.), The New York: York: McGraw-HilI. McGraw-Hill. New Holt, R R. R R. (1962). (1962). Individuality Individuality and and generalization generalization in in the the psychology psychology of of personality: personality: An An evaluation. Journal Journalof ofPersonality, Personality,30,30,377-402. 377-402. evaluation. Honess, T., T., & & Yardley, Yardley, K. K. (Eds.). (Eds.). (1987). (1987). Self and identity: identity:Perspectives Perspectivesacross across lifespan, thethe lifespan. Honess, London: Routledge Routledge & & Kegan Kegan Paul. London: Personality^ Hormuth, S. S. E. E. (1986). (1986). The The sampling sampling of of experiences situ. Journal Journal of of Personality, Hormuth, experiences in situ. 54,54, 262-293. 262-293. Horney, K. K. (1939). (1939). New ways in in psychoanalysis. psychoanalysis.New NewYork: York:Norton. Norton. Horney, Horowitz, Statesof ofmind. mind.New NewYork: York:Plenum PlenumPress. Press. Horowitz, M. M. J. J. (1979). (1979). States Howard. G. G. S. S. (1991). (1991). Culture Culture tales: tales: A A narrative narrative approach approach to to thinking, thinking, cross-cultural cross-cultural psycholpsycholHoward. ogy, and psychotherapy. American American Psychologist, Psychologist,46,46,187-197. 187-197. Hull, Hull, C. C. L. L. (1943). (1943). Principles Principlesof ofbehavior. behavior.New NewYork: York:Appleton-Century-Crofts. Appleton-Century-Crofts. Izard, C. Human emotions. emotions.New NewYork: York:Plenum PlenumPress. Press. C. E. (1977). Human Jackson, D. Jackson, D. N. (1970). (1970). A sequential system for personality scale development. In C. D. Spielberger (Ed.), Current psychology (Vol. 2, 2, pp.pp. Current topics topicsininclinical clinicaland andcommunity community psychology (Vol. 61-96). New York: Academic Press. Jackson, D. N. (1971). The dynamics of structured personality tests. Psychological PsychologicalReview, Review, 78, 78, 229-248. 229-248. Jackson, D. D. N., N., & &Messick, Messick,S.S.(1958). (1958).Content Contentand andstyle styleininpersonality personalityassessment. assessment.Psychologi Psychological Bulletin, Bulletin,55, 55,243-252. 243-252. Annual Jackson, D. N., & & Paunonen, S. V. (1980). Personality Personality structure and assessment. Annual Review of of Psychology, Psychology,31, 31,503-552. 503-552. (1890). Principles James, W. (1890). Principlesof ofpsychology. psychology.New NewYork: York:Holt. Holt. Jastrow, J. New York: Appleton. J. (1915). Character Characterand andtemperament. temperament. New York: Appleton. Jemmott, J. B., III. III. (1987). Social motives and susceptibility susceptibility to disease: Stalking individual differences 267-298. Journalof ofPersonality, Personality,55,55, 267-298. differences in health risks. risks. Journal John, O. P. (1990). The "Big Five" factor taxonomy: Dimensions P. (1990). The "Big Five" factor taxonomy: Dimensionsofofpersonality personalityininthe thenatural natural language and in questionnaires. In L. L. Pervin (Ed.), Handbook of of personality personality theory theory and research research(pp. (pp.66-100). 66-100).New NewYork: York:Guilford GuilfordPress. Press. & Donovan, Donovan,J.J.E. E. (1986). (1986). Structural Structural equation equation models models and and personality personality Judd, C. M., Jessor, R, R., & research. Journal Journalof ofPersonality, Personality,54,54,149-198. 149-198. Kegan, R process inin human development. Cambridge, R. (1982). The Theevolving evolvingself: selfProblem Problemand and process human development, Cambridge, MA: MA: Harvard University Press. Kelly, G. G. (1955). (1955). The Thepsychology psychologyof ofpersonal personalconstructs. constructs.New NewYork: York:Norton. Norton. Kelly,

CHAPTER 1

HISTORY CONCEPTUAL HISTORY

3S 35

R. W. White (Ed.), The study study of of Kenniston, K. (1963). Inburn: An American Ishmael. In R lives (pp. 40-71). New York: Atherton Press. Kenrick, D. T., & & Funder, D. C. (1988). Profiting from controversy: Lessons from the person personAmerican Psychologist, Psychologist, 43, 23-34. situation debate. American & Stringfield, D. O. (1980). Personality traits and the eye of the beholder: Kenrick, D. T., & Crossing some traditional philosophical boundaries in the search for consistency in all Psychological Review, Review, 87, 88-104. the people. Psychological & Krech, D. (1951). The problem of personality and its theory. Journal of of Klein, G. S., & Personality, Personality, 20, 2-23. Klein, M. (1948). Contributions Contributions to psychoanalysis: psychoanalysis: 1921-1945. 1921-1945. London: Hogarth Press. Kobasa, S. C. (1985). Personality and health: Specifying Specifying and strengthening the conceptual links. Review Review of of Personality Personality and Social Social Psychology, Psychology, 6, 291-311. Kohut, H. (1977). The restoration of of the self. self New York: International Universities Press. Physique and character. Berlin: Springer. Kretschmer, E. (1921). Physique A., & & Derry, P. A. (1981). The self as cognitive prototype: An application to Kuiper, N. A, person perception and depression. In N. Cantor & & J. F. Kihlstrom (Eds.), Personality Personality cognition, and social social interaction (pp. 214-232). Hillsdale, NJ: Erlbaum. of personality. American American Psychologist, Psychologist, Lamiell, J. T. (1981). Toward an idiothetic psychology of 36, 276-289. R., & & Power, F. C. (Eds.). (1988). Self, Self ego, and and identity: Integrative approaches. approaches. Lapsley, D. R, New York: Springer-Verlag. Interpersonal diagnosis diagnosis of of personality. personality. New Leary, T. (1957). Interpersonal New York: Ronald Press. Lecky, P. (1945). Self-consistency. New York: Island Press. dynamic theory theory of of personality. personality. New York: McGraw-Hili. McGraw-Hill. Lewin, E. (1935). A dynamic of psychological theory. Psychological Psychological Loevinger, J. (1957). Objective tests as instruments of Reports, 3, 635-694. development. San Francisco: Jossey-Bass. Loevinger, J. (1976). Ego development. Paradigms of of personality. personality. San Francisco: Jossey-Bass. Loevinger, J. (1987). Paradigms & Knoll, E. (1983). Personality: Stages, traits, and the self. Annual Annual Review Review of of Loevinger, J., & Psychology, Psychology, 34, 195-222. London, H., & & Exner, J. E., Jr. (Eds.). (1978). Dimensions Dimensions of of personality. personality. New York: Wiley. MacKinnon, D. W. (1951). Personality. Annual Annual Review Review of of Psychology, Psychology, 2, 113-136. MacKinnon, D. W. (1965). Personality and the realization of creative potential. American American Psychologist, Psychologist, 20, 273-281. Maddi, S. R R. (1984). Personology for the 19805. 1980s. In R R. A A. Zucker, J. Aronoff, & &A A. I. Rabin behavior (pp. 7-41). New York: Academic (Eds.), Personality Personality and and the prediction prediction of of behavior Press. Markus, H. (1977). Self-schemata and processing information about the self. Journal of of Personality Personality and Social Psychology, Psychology, 35, 63-78. Markus, H., & & Nurius, P. (1986). Possible selves. American American Psychologist, Psychologist, 41, 954-969. McAdams, D. P. (1985). Power, intimacy, and the life story: story: Personological Personological inquiries into identity. New York: Guilford Press. McAdams, D. P. (1992). The five-factor model in personality: A A critical appraisal. Journal of Personality, Personality, 60, 329-361. of live by: Personal Personal myths myths and and the the making McAdams, D. P. (1993). The stories we live making of of the self self. New York: Morrow. McAdams, D. P. (1994). The person: psychology (2nd ed.). person: An An introduction introduction to personality personality psychology Fort Worth, TX: TX: Harcourt Brace.

36 36

DAN P. McADAMS DANP.MCADAMS

McClelland, D. C. C. (1951). Personality. Personality.New NewYork: York:Holt, Holt,Rinehart Rinehart&&Winston. Winston. McClelland, D. D. C. C. (1961). The The achieving achievingsociety. society.New NewYork: York:Van VanNostrand. Nostrand. McClelland, D. C. C. (1981). Is personality consistent? In A. I. Rabin, J. Aronoff, A. M. in in personality (pp.(pp. 87-113). New &R R. A. A. Zucker Zucker (Eds.), (Eds.),Further Furtherexplorations explorations personality 87-113). New Barclay, & York: York: Wiley. Wiley. McClelland, D. D. C. C. (1985). (1985). Human Human motivation. motivation,Glenview, Glenview,IL: IL:Scott, Scott,Foresman. Foresman. McClelland, McClelland, C., Atkinson, J. McClelland, D. D. C, J. W., W., Clark, Clark, R R. A., & & Lowell, Lowell, E. E. L. L. (1953). (1953). The The achievement achievement motive. motive. New New York: York: Appleton-Century-Crofts. Appleton-Century-Crofts. McCrae, R R.R R.(Ed.). (Ed.).(1992). (1992).The Thefive-factor five-factormodel: model:AAspecial special issue ofthe Journal ofPersonality, McCrae, issue ofthe Journal ofPersonality. Durham, NC: NC: Duke University Press. McCrae, R R. R, R., & & Costa, Costa, P. P. T., T., Jr. Jr. (1987). (1987). Validation Validation of of the the five factor factor model model of of personality personality McCrae, across instruments and observers. Journal 52,52, Journal of of Personality Personalityand andSocial SocialPsychology, Psychology, 81-90. 81-90. McCurdy, H. H. G. G. (1965). (1965). Personality Personalityand andscience: science:AAsearch search self-awareness, Princeton, McCurdy, forfor self-awareness. Princeton, NJ: NJ: Van Van Nostrand. Nostrand. McDougall, W. W. (1908). (1908). Social Socialpsychology. psychology,London: London: Methuen. Methuen. McDougall, McDougall, W. W. (1932). (1932). Of Of the the words words character character and and personality. personality. Character Characterand andPersonality, Personality, McDougall, 1, 3-16. 3-16. 1, Mead, G. G. H. H. (1934). (1934). Mind, Mind, self, self and andsociety. society,Chicago: Chicago:University Universityof ofChicago ChicagoPress. Press. Mead, Clinicalvs. vs,statistical statisticalprediction: prediction: theoretical analysis a review Meehl, P. P. E. E. (1954). (1954). Clinical A Atheoretical analysis and and a review of of Meehl, evidence,Minneapolis: Minneapolis:University UniversityofofMinnesota MinnesotaPress. Press. the evidence. Meehl, P. E. E. (1977). (1977). Specific Specific etiology etiology and and other other forms forms of of strong strong influence: Some quantitative quantitative Meehl, P. influence: Some meanings. Journal Journalof ofMedicine Medicineand andPhilosophy, Philosophy,2, 2, 33-53. 33-53. Mendelsohn, G. G. (1985). (1985). La La Dame Dame aux aux Camelias Camelias and and La La Traviata: Traviata: A A study study of of dramatic dramatic Mendelsohn, transformations transformations in the light of biography. In R R. Hogan & & W. W. H. H. Jones Jones (Eds-.), (Eds^.),Perspec Perspectives in in personality personality(Vol. (Vol.1,1,pp. pp.271-303). 271-303).Greenwich, Greenwich,CT: CT:JAI JAIPress. Press. Milgram, S. S. (1963). (1963). Behavioral Behavioral study study of of obedience. obedience. Journal Journalof ofAbnormal Abnormaland andSocial SocialPsychol PsycholMilgram, ogy, 67, 67, 371-378. Miller, G. A., Galanter, E., & & Pribram, K. H. (1960). Plans Plans and and the thestructure structureofofbehavior. behavior. New York: York: Holt, Rinehart & & Winston. Miller, N. E., E., & Yale learningand andimitation. imitation.New NewHaven, Haven,CT: CT: Yale & Dollard, DoUard, J. (1941). Social Social learning University Press. New York: Wiley. Mischel, W. (1968). Personality Personalityand andassessment. assessment. New York: Wiley. Mischel, W. (1973). Toward a cognitive social leaming learning reconceptualization of personality. Psychological PsychologicalReview, Review,80,80,252-283. 252-283. Mischel, W., & & Peake, P. K. (1982). Beyond Deja D6ja Vu in the search for cross-situational consistency. Psychological PsychologicalReview, Review,89,89,730-755. 730-755. Murphy, G. (1947). Personality: origins andand structure. NewNew York: Personality:AAbiosocial biosocialapproach approachto to origins structure. York: Harper and Brothers. York: Oxford University Press. Murray, H. A. (1938). Explorations Explorationsininpersonality. personality.New New York: Oxford University Press. ofof Abnor Murray, Murray, H. H.A. A.(1940). (1940).What Whatshould shouldpsychologists psychologistsdo doabout aboutpsychoanalysis? psychoanalysis?Journal Journal Abnormal and Social Social Psychology, Psychology,35, 35,150-175. 150-175. Norman, W. W. T. (1963). Toward an adequate taxonomy of personality attributes: attributes: Replicated factor structure in peer peer nomination nomination personality personality ratings. ratings.Journal Journalof ofAbnormal Abnormaland andSocial Social Psychology, Psychology, 66, 66,574-583. 574-583. Nuttin, J. (1955). Personality. Annual Review of of Psychology, Psychology,6,6,161-186. 161-186. A Amethodological framework. New York: Ozer, D. J. (1986). Consistency Consistencyininpersonality: persoruility: methodological framework. New York: Springer-Verlag.

CHAPTER 11

CONCEPTUALHIsTORY HISTORY CONCIPJUAL

37 37

Palys, T. S., & & Little, B. R. (1983). Perceived life satisfaction and the organization of of personal project systems. Journal Journal of of Personality Personality and and Social Social Psychology, Psychology, 44, 44, 1221-1230. Pervin, L. A. (1985). Personality: Current controversies, issues, and directions. Annual Annual Review Review of of Psychology, Psychology, 36, 83-114. Pervin, L. A. (1990). Personality theory and research: Prospects for the future. In L. A. Pervin (Ed.), Handbook Handbook of of personality personality theory theory and and research (pp. 723-727). New New York: Guilford Press. Plomin, R. (1986). Behavioral genetic methods. Journal Journal of of Personality, Personality, 54, 226-261. Polkinghome, D. (1988). Narrative Narrative knowing knowing and the human sciences. Albany, NY: SUNY SUNY Press. Privette, G. (1983). Peak experience, peak performance, and flow: A A comparative analysis of and Social Psychology, 45, 1361of positive human experience. Journal Journal of of Personality Personality and Social Psychology, 45,13611368. Rank, O. (1945). Will therapy therapy and and truth and and reality. New York: Knopf. Revelle, W. (1995). Personality processes. In L. W. Porter and M. R. Rosenzweig (Eds.), Annual Annual review of of psychology psychology (Vol. 46, pp. 295-328). Palo Alto, CA: Annual Annual Re Reviews, Inc. Riesman, D. (1950). The lonely lonely crowd. New Haven, CT: CT: Yale University Press. Roback, A. A. (1927). The psychology psychology of of character. character. London: Kegan Paul. Love*s story story told: A life of of Henry Henry A. Murray. Cambridge, MA: Harvard Harvard Robinson, F. (1992). Love's University Press. Rorer, L. G., & Annual Review & Widiger, T. A. (1983). Personality PersonaUty structure and assessment. Annual Review of of Psychology Psychology 34, 431-463. Rosenberg, S., & & Gara, M. A. (1985). The multiplicity of of personal identity. Review Review of of Personality Personality and Social Social Psychology, Psychology, 6, 87-114. Rosenhan, D. L. (1973). On being sane in insane places. Science, 179, 250-258. Rosenwald, G. C., self C, & & Ochberg, R. L. (Eds.). (1992). Storied Storied lives: the cultural politics politics of of selfunderstanding. New Haven, CT: CT: Yale University Press. Rotter, J. B. (1954). Social Cliffs, NJ: Social learning and and clinical clinical psychology. psychology. Englewood . Prentice-Hall. fe histories and psychobiography: Explorations Explorations in theory Runyan, W. M. (1982). Li Life andpsychphiography: theory and method. method. New York: Oxford University Press. Rushton, J. P., Brainerd, C. J., & & Pressley, M. (1983). Behavioral development and construct validity: The principle of of aggregation. Psychological Psychological Bulletin, 94, 18-38. Rychlak, J. E. (1981). Personality Personality and and psychotherapy psychotherapy (2nd ed.). Boston: Houghton Houghton Miftin. Miflin. Rychlak, J. E. (1988). The psychology psychology of of rigorous rigorous humanism humanism (2nd ed.). New York: New York University Press. Sampson, E. E. (1985). The decentralization of identity: Toward a revised concept of of personal personal and social order. American Psychologist, 40, 1203-1211 . American Psychologist, 1203-1211. Sanford, R n psychology. IIn nS R.. N N.. (1963). Personality: Its place iin S.. Koch (Ed.), Psychology: Psychology: A A study study of of a science (Vol. 5, pp. 488-592). New York: McGraw-Hill. Sarason, S. B. (1988). The making autobiography. San making of of an American American psychologist: psychologist: An An autobiography. Francisco: Jossey-Bass. Sarason, S. B., & & Mandler, G. (1952). Some correlates of of test anxiety. Journal Journal of of Abnormal Abnormal and and Social Social Psychology, Psychology, 47, 810-817. of human human conduct. Sarbin, T. (Ed.). (1986). Narrative Narrative psychology: psychology: The storied storied nature of conduct New York: Praeger.

38 38

DAN P. McADAMS DANP.MCADAMS

Bulletin, Sawyer, J. J. (1966). (1966). Measurement Measurement and andprediction, prediction,clinical clinicaland anrfstatistical. statistical.Psychological Psychological Bulletin, Sawyer, 66, 66, 178-200.

Schlenker, B. R. (Ed.). (1985). The The self selfand andsocial sociallife. life.New NewYork: York:McGraw-Hili. McGraw-Hill. Sears, R. R. (1950). Personality. Annual Annual Review Review of of Psychology, Psychology,1,1,105-118. 105-118. Sears, R. R., Maccoby, E. E., & IL:IL: & Levin, H. (1957). Patterns Patternsof ofchild childrearing. rearing.Evanston, Evanston, Row Row Peterson. Peterson. Psychology,27, 27,1-27. Sechrest, L. L. (1976). Personality. Personality. Annual Review Review of of Psychology, Sechrest, 1-27. Shaver, P. P. (Ed.). Self situations, situations,and andsocial socialbehavior: behavior: Review ofpersonality social Shaver, (Ed.). (1985). Self, Review ofpersonality andand social psychology (Vol. (Vol. 6). Beverly Beverly Hills, Hills, CA: CA: Sage. Shaver, P., P., & & Rubenstein, Rubenstein, C. C. (1980). Childhood Childhood attachment attachment experience experience and and adult adult loneliness. loneliness. Shaver, Review of 42-73. of Personality Personalityand andSocial SocialPsychology, Psychology,1, 1, 42-73. Sheffield, F. F. D., Wolff, Wolff, J. J. J., J., & & Backer, Backer, R. R. (1951). Reward Reward value value of of copulation copulation without without sex sex Sheffield, drive Psychology, 44, 44, 3-8.3-8. drive reduction. reduction. Journal Journalof of Comparative Comparativeand andPhysiological Physiological Psychology, Sheldon, W. W. H. H. (1940). The The varieties varietiesof ofhuman humanphysique: physique: introduction to constitutional Sheldon, AnAn introduction to constitutional psychology. New New York: Harper. York: Harper. psychology. Shneidman, E. S. S. (Ed.). (Ed.). (1981). Endeavors Endeavors in inpsychology: psychology:Selections Selections from personology from thethe personology Shneidman, E. of Henry Henry A. A. Murray. Murray.New NewYork: York:Harper Harper&&Row. Row. of Textsof ofidentity. identity.London: London:Sage. Sage. Shotter, J., J., & & Gergen, Gergen, K. K. (Eds.). (Eds.). (1989). Texts Shotter, Shweder, R. R. A. A. (1984). Anthropology's Anthropology's romantic romantic rebellion rebellion against against the the enlightenment, enlightenment, or or Shweder, there's there's more more to to thinking thinking than than reason reason and and evidence. evidence. In In R. R. A. A. Shweder Shweder & & R. R. A. A. Levine Levine (Eds.), Culture Culture theory: theory:Essays Essaysononmind, mind,self, selfand andemotion emotion(pp. (pp.27-66). 27-66).Cambridge, Cambridge, (Eds.), England: Cambridge Cambridge University University Press. England: Press. Kolligan, J., J., Jr. Singer, J. J. L., & & KoUigan, Jr. (1987). Personality: Personality: Developments Developments in in the the study study of of private private experience. experience. Annual Annual Review Reviewof ofPsychology, Psychology,38,38,533-574. 533-574. Singer, J. 23,23, 375-412. J. L., & & Singer, Singer, D. D. G. G. (1972). (1972).Personality. Personality.Annual AnnualReview ReviewofofPsychology, Psychology, 375-412. Snyder, M. (1983). The influence of individuals on situations: situations: Implications Implications for for understanding understanding the 497-516. Journalof ofPersonality, Personality,51,51, 497-516. the links links between between personality personality and and social social behavior. behavior. Journal Spence, D. (1982). Narrative truth. New York: Norton. Narrativetruth truthand andhistorical historical truth. New York: Norton. Spielberger, C. C. D. (1966). The effects of anxiety on complex learning and academic achievement. In C. D. Spielberger (Ed.), Anxiety and behavior. behavior.New NewYork: York:Academic AcademicPress. Press. Spranger, E. (1928). Types Types of of men. men.Halle: Halle: Niemeyer. Niemeyer. Stagner, R. (1937). Psychology Psychologyof ofpersonality. personality.New NewYork: York:McGraw-Hili. McGraw-Hill. Leipzig, Germany: J. A. Barth. Stern, menschlichePersiJnlichkeit. Persdnlichkeit. Leipzig, Germany: J. A. Barth. Stem, W. (1924). Die menschliche Stewart, A. J., adaptation toto change. In In R. R. Hogan & & J.,& &Healy, Healy,J.J.M., M.,Jr.Jr.(1985). (1985).Personality Personalityand and adaptation change. Hogan pp. 117-144). CI': W. H. Jones (Eds.), Perspectives Perspectivesininpersonality personality(Vol. (Vol.1, 1, pp. 117-144).Greenwich, Greenwich, CT: JAI Press. AA special issue Suls, J., & & Rittenhouse, J. D. (Eds.). (1987). Personality Personalityand andphysical physicalhealth: health: special issue of NC: Duke University Press. of the the Journal Journalof ofPersonality. Personality.Durham, Durham, NC: Duke University Press. Taylor, J. (1953). A personality scale of manifest anxiety. Journal Journalof ofAbnormal Abnormaland andSocial Social

Psychology, Psychology,48, 48,285-290. 285-290.

Tomkins, S. S. (1987). Script theory. In J. Aronoff, A. I. Rabin, & (Eds.), The & R. A. Zucker (Eds.), The emergence emergenceof ofpersonality personality(pp. (pp.147-216). 147-216).New NewYork: York:Springer. Springer. Tupes, E. C, C., & & Christal, R. E. (1961). Recurrent Recurrentpersonality personalityfactors factorsbased basedonontrait traitratings. ratings.

U.S. Command, Aeronautical Systems Division, Technical Report,Report, U.S.Air AirForce ForceSystems Systems Command, Aeronautical Systems Division, Technical No. 61-97. Weiner, B. (1980). Human Human motivation. motivation.New NewYork: York:Holt, Holt,Rinehart Rinehart&&Winston. Winston. West, S. G. (1983). Personality and prediction: prediction: An introduction. Journal Journal of of Personality, Personality, 51, 51, 275-285.

CHAPTER 11

CONCEPTUAL IbsTORY HISTORY CONCEPTUAL

39 39

White, R. W. (1952). Lives progress (1st ed.). New York: Holt, Rinehart & Lives in progress & Winston. White, R. W. (1959). Motivation Motivation reconsidered: The concept of competence. Psychological Psychological Review, Review, 66, 297-333. study of of lives. New York: Atherton Press. White, R. W. (Ed.). (1963). The study of lives. In A. I. Rabin, White, R. W. (1981). Exploring personality the long way: The study of Aronoff, A. M. Barclay, & & R. A. Zucker (Eds.), Further explorations explorations in personality personality J. Aronoff, (pp. 3-19). New York: Wiley. A memoir: memoir: Seeking Seeking the shape shape of of personality. personality. Marlborough, NH: The White, R. W. (1987). A Homestead Press. & Child, I. L. (1953). Child training and and personality: personality: A cross-cultural cross-cultural Whiting, J. W. M., & study. study. New Haven, CT: CT: Yale University Press. Wiggins, J. S. (1973). Personality Personality and prediction: prediction: Principles Principles of ofpersonality personality assessment. assessment. Reading, Reading, MA: Addison. Agency and communion as conceptual conceptual coordinates coordinates for for the understanding understanding Wiggins, J. S. (1991). Agency and measurement of of interpersonal behavior. In D. Cicchetti & & W. Grove (Eds.), Thinking clearly about about psychology: psychology: Essays Essays in honor honor of of Paul Paul Everett Everett Meehl Meehl (pp. 89-113). Thinking of Minnesota Press. Minneapolis: University of & Broughton, R. (1985). The interpersonal circle: A A structural model for the Wiggins, J. S., & integration of of personality research. In R. Hogan & & W. H. Jones (Eds.), Perspectives Perspectives personality (Vol. 11,, pp. 1-47). Greenwich, CT: CT: JAI Press. in personality & Pincus, A. L. (1992). Personality: Structure and assessment. Annual Annual Review Review Wiggins, J. S., & of Psychology, Psychology, 43, 473-504. of differences in ease of of perception of of embedded figures. Witkin, H. A. (1950). Individual differences Journal of of Personality, Personality, 19, 11-15. Journal -15. Personal data sheet. Chicago: Stoelting. Stocking. Woodworth, R. S. (1919). Personal personality Wrightsman, L. S. (1981). Personal documents as data in conceptualizing adult personality development. Personality Personality and and Social Psychology Psychology Bulletin, 7, 367-385. Personality development development in adulthood. adulthood. Newbury Park, CA: Sage. Wrightsman, L. S. (1988). Personality American Psy PsyZajonc, R. B. (1980). Feeling and thinking: Preferences need no inferences. American chologist, 35, 151-175. Zener, K. K. (1945). A A note note concerning concerning editorial editorial reorientation. reorientation. Journal Journal of of Personality, Personality, 14, 11-2. Zener, -2. Aronoff, J., & & Frank, S. (Eds.). (1992). Personality Personality structure structure in Zucker, R. A., Rabin, A. I., Aronoff, the life course. New York: Springer. AUport a trait theorist? Jorunal Jorunal of of Personality Personality and Social Social Zuroff, D. C. (1986). Was Gordon Allport Psychology, 51, 993-1000 993-1000.. Psychology,

This Page Intentionally Left Blank

CHAPTER CHAPTER 22

STUDYING LIVES AND THE THE CONCEPTUAL STRUCTURE PSYCHOBIOGRAPHY AND OF OF PERSONALITY PERSONALITY PSYCHOLOGY PSYCHOLOGY

WILLIAM WILLIAM MCKINLEY MCKINLEY RUNYAN RUNYAN UNIVERSITY OF OF CALIFORNIA CALIFORNIA AT AT BERKELEY BERKELEY UNIVERSITY

I. INTRODUCTION INTRODUCTION L

and There is a puzzling history in the relationships between personality psychology and the study of individual lives. Most simply, the the study of individual persons and lives the central concerns and and motivating agendas for founders of the field field was one of the AUport (1937) and and Henry Murray (1938), but was then lost sight such as Gordon Allport the 1950s and and 1960s (with some exceptions, as as in the work of Robert White of in the as far far greater attention was was given to psychometric [1952] or Erik Erikson [1958]), as [1952] concerns and and the the experimental study of particular processes. in the constitution and changing defini definiturn of events in This was a remarkable tum tion of the field. field. Major texts of the period, such as Hall and and Lindzey's Lindzey's Theories Theoriesofof Personality (1957), (1957),which whichasasthe themost mostwidely widelyused usedtext texteventually eventually sold soldmore morethan than Personality copies, and and Walter WalterMischel's Mischel'sPersonality Personalityand andAssessment Assessment (1968), gave almost 700,000 copies, (1968), gave almost the study of individual persons or lives. Hall and and Lindzey argued no attention to the the fruitfulness of personality theories "is "is to be judged primarily by how that the how as a spur to research" research" (p. (p. 27), the effectively they serve as 27), while Mischel argued for the superiority of of experimentally experimentally based based social social learning learning and and cognitive theories theories over over trait superiority for the the prediction prediction and modification of of behavior. and psychodynamic psychodynamic approaches approaches for and and modification behavior. an improved improved understanding understanding of of individual individual persons persons was significant Note that that an Note was not not aa significant in either either of of these these influential influential formulations formulations of field. criterion in criterion of the the field. 41 HANDBOOK OF PERSONAUTY PSYCHOLOGY

JlANDBOOK OF PERSONALITY PSYCHOLOGY

41

COmlGHT COPYRIGHT C © 1997 1997 IY BY ACADEMIC ACADEMIC PRESS PRESS.. ALL liGHTS Of IEPIODUcnON IN ANY fORM IESElVED.

ALL RIGHTS OF REPRODUCHON IN ANY FORM RESERVED.

42 42

WILL IAM MCKINLEY RUNYAN WILLIAM MCKINLEYRUNYAN

How of individual persons and lives fall by the wayside? It H o w did the study of seems to have have been been due to a combination combination of of changing changing intellectual intellectual fashions fashions about what it means to "be scientific," personal and temperamental temperamental preferences preferences for partic particular kinds of of research, the kinds of of graduate students attracted to the field field in the growing competition World War competition with clinical psychology psychology after after World War II, patterns of of funding funding and grant support, and institutional institutional processes determining determining who who was or was not other major universi hired and promoted promoted at Harvard, Stanford, Yale, Berkeley, and other universities around the country. Precisely how however, remains somewhat how all this unfolded, unfolded, however, somewhat obscure and needs to be illuminated through more detailed detailed research on the intellec intellectual and institutional institutional history history of of personality personality psychology. psychology. In any case, there there has been been a remarkable resurgence of interest in recent years in the study of of individual lives lives and in psychobiography psychobiography among personality personality psychologists psychologists (e.g., Alexander, 1990; Anderson, 1981a, 1988; Cohler, 1988; Elms, 1988a, 1988b, 1994; McAdams, 1990; 1988a, 1988b, 1994). McAdams McAdams & & Ochberg, 1988a; Runyan, 1982, 1982,1988a, This chapter is intended to help reintegrate reintegrate or reweave reweave the study of of lives lives back into the fabric of personality psychology. The objective is not to stomp out other of personality objective forms of of research, or to argue that personality personality psychology psychology should should be nothing nothing but forms of lives, but rather, to argue that the study of of persons or lives is one one of the study of the central central objectives of of the field. field. A A discipline discipline of of psychology psychology which which does does not not contrib contribute to a better understanding understanding of of persons is pretty pathetic. Particularly Particularly pathetic pathetic is a personality psychology which fails to do so. My hope is to contribute to an an personality psychology, psychology, so that the study of of persons and emerging "gestalt shift" in personality lives is again seen as one of of its central objectives. objectives. of this chapter are to reexamine, perhaps even even help rehabili rehabiliThe objectives of tate, the study of of lives in the field field of of personality personality psychology psychology and to examine examine how the study of of lives is related to quantitative and experimental experimental research tradi traditions. The second section section will explore explore the conceptual conceptual structure structure of of personality personality psychology psychology and argue for a conception conception of of the field field in terms of of the four four objectives objectives of individual and group of developing developing general theories of of personality, analyzing analyzing individual differences, studying specific processes or classes of of behavior, and, finally, finally, studying individual lives. The third section discusses the concept of progress as a way concept of of addressing methodological and epistemological problems in the study of of addressing methodological epistemological problems individual lives. The fourth foiwth section section will explore explore relationships relationships between between the study of of lives and quantitative and experimental experimental traditions of of research in personality. personality. The fifth fifth section examines the uses of of a "hard" to "soft" continuum continuum for for understanding relationships relationships among among a variety of of traditions, objectives, and methods methods understanding personality psychology. I will argue argue that thriving thriving interdisciplinary syntheses interdisciplinary syntheses in personality of psychology, psychology, such as as cognitive cognitive science and neuroscience, might might at the hard end of supplemented by a synthesis at the soft soft end end of of psychology, with the well be supplemented of lives in social, cultural, and historical contexts as one one leading leading candidate candidate study of for such a soft soft synthesis. for

CHAPTER 2

STUDYING LIVES STUDYING

43 43

n. THE CONCEPTUAL STRUCTURE OF PERSONALITY PSYCHOLOGY n.

It field of It is is very difficult difficult to get a conceptual grasp on on the thestructure structure of of the the field of personality personality psychology due to the enormous range of things going on within it. What common ground, if any, is there between theories of psychosexual stages, factor-analytic factor-analytic studies of questionnaire responses, experimental studies of aggression, assessment and prediction studies, behavior genetics, genetics, and and psychobiography? psychobiography? How How to to understand understand and the structure of a field field which includes such apparently bewildering diversity? first common way of organizing the field field of personality psychology is in A first as psychoanalysis, behav behavterms of major theoretical orientations or traditions, such as iorism or learning theory, trait and psychometric approaches, and humanistic psypsy which are are almost almost always always discussed, discussed, while while other other chology. These are the "Big Four," Four," which traditions sometimes discussed are culture and personality, behavior genetics and sociobiology, the study of lives, and cognitive approaches to personality. This apap proach is frequently used in undergraduate personality courses and is adopted in as Liebert and Spiegler's Personality: Personality:Strategies Strategies andIssues Issues and undergraduate texts such as Personality:Theory, Theory,Assessment, Assessment, Research (1980), Peter(1987), Pervin's Personality: andand Research (1980), andand Peter Personality(1988). (1988).AAvariant variantof ofthis thisisisaabiographical biographicalanalysis analysisof ofthe thework workof of son's Personality Theoriesof ofPersonality Personality major theorists in thefield, field, such as with Hall and Lindzey's Theories Beneath the theMask: Mask:An AnIntroduction Introduction (1957, and subsequent editions) and Monte's Beneath to Theories Theoriesof ofPersonality Personality(1987). (1987). to A second way of dividing up the field is in terms of core conceptual issues which cut across theoretical orientations, including topics such as the structure of personality, the dynamics of personality, development of personality, personaUty, assessment of personality, and the change of personaHty. personality. This strategy was used by Gordon Personality:AAPsychological PsychologicalInterpretation Interpretation (1937), (1937), Allport in his foundational text Personality: and is used in the preface to the more than 40 volumes of the Wiley Series on Personality Processes. A third way of dividing up the field is in terms of different methodological traditions or techniques. The most influential version of this is probably by Lee Cronbach, who in "The Two Disciplines of Scientific Psychology" (1957) and "Be"Be yond the Two Disciplines of Scientific Psychology" (1975) argued that the field of of psychology could be seen as developing along an experimental tradition and a quantitative-correlational tradition, which needed to be integrated into a more comprehensive interactional tradition, examining the interaction of individual difdif ferences with responses to experimental and situational conditions. These correla correlational, experimental, and interactional research designs have all been important in personality psychology, along with longitudinal, cross-cultural, archival, and case study methods. Another approach to conceptualizing the field of personality from a methodological perspective is provided in Kenneth Craik's (1986) analysis of the history of personality psychology in terms of the the rise, fall, and resurrection of seven biographical/archival. different methodological traditions, which he identifies as biographical/archival,

44 44

WILLIAM RUNYAN WILLIAMMCKINLEY MCKINLEYRUNYAN

field studies, laboratory laboratory methods, naturalistic observational observational assessment, observer observer judgments, judgments, personality personality scales and inventories, and projective projective techniques. A fourth way way of of dividing up the field field of of personality personality is in terms of of empirical empirical A substantive processes and classes classes of of behavior, such as aggression, sexual sexual research on substantive behavior, behavior, creativity, altruism, anxiety, anxiety, psychopathology, psychopathology, locus locus of of control, delay delay of gratification, gratification, achievement achievement motivation, and stress and coping. coping. Many textbooks textbooks of combine a discussion discussion of of four four or five five major theoretical theoretical traditions with with a number number of of combine specific personality processes (e.g., Mischel, 1981; Phares, 1988; Wiggins, Renner, specific personality Mischel, Renner, Clore, & & Rose, 1976). Textbooks Textbooks are, of of course, also organized organized according according to various combinations of of these four four principles, such such as a discussion discussion of of four four theories followed followed combinations by a set of of substantive processes, or a review review of of four four or five five major major theoretical theoretical orientations with research and applications discussed discussed under under each; or theory, empiri empirical research, and applications may may be placed in separate separate sections; and so on. propose a fifth fifth way way of In this chapter I want to propose of conceptualizing conceptualizing the structure of personality personality psychology; one one which cuts across the prior conceptual conceptual frameworks frameworks of interrelatedness of of the and which raises intriguing questions about the degree degree of of interrelatedness disparate intellectual intellectual enterprises which constitute the field. field. If If successful, framesuccessful, this frame work may bring into view view aspects of of the structure structure of of the field field not not previously work previously visible. central idea is that the field field of of personality personality psychology concerned with with The central psychology is concerned four major tasks or objectives: (1) developing developing general general theories theories of of personality, personality, four differences, (3) analyzing analyzing specific specific processes processes (2) studying individual and group differences, of behavior, and (4) understanding understanding individual individual persons and classes of persons and lives. between these four four objectives, and the development development of of each each The relationships between of them them over time, are outlined in Figure 1. Starting with the top row, row. General General of development of of a number number of of the major major theoretical theoretical programs programs theory, we can trace the development in personality psychology, beginning beginning with psychoanalysis psychoanalysis around around 1900 with the of Freud's Interpretation Interpretation of of Dreams Dreams (1900/1958); behaviorism behaviorism around around publication of 1913 with John B. Watson's "Psychology "Psychology as the Behaviorist Behaviorist Views Views It"; culture and personality in the 1930s with Margaret Mead, Ruth Benedict, Benedict, Edward Edward Sapir, and others; the psychometric psychometric approach in the 1940s with the early of Hans Hans early publications publications of Eysenck and R. B. Cattell; the humanistic-phenomenological of Carl Eysenck humanistic-phenomenological approach of Rogers work of Rogers and Abraham Maslow; cognitive approaches approaches with the work of George George Kelley extensively with cognitive-experimentalists in 1955, but then more extensively cognitive-experimentalists such as Mischel, work in behavior behavior genetics genetics and sociobiology sociobiology Bandura, Cantor, and KihlstrOm; and work becoming more prominent prominent in the 1970s and 1980s with E. O. Wilson, Arnold Arnold Buss, emphasized that these these historical datings are David Buss, and others. It should be emphasized approximate, but the primary point is to outline outline the the historical emergence emergence of highly approximate, each theoretical theoretical tradition and then be able to raise questions questions about about its relations to to developments in studies studies of of individual individual and group group differences, differences, of of specific specific processes processes developments of behavior, and of of individual individual persons and lives, as represented represented in the and classes of bottom three rows of of Figure Figure 1. bottom of studying individual individual and group differences, repreThe second row, that of differences, is repre sented with a sample of of relatively influential influential programs programs of of this type, such as studies of intelligence by Binet, Terman, Wechsler, Eysenck, Eysenck, Howard Howard Gardner, and others; of

CHAPTER! CHAPTER 2

45 4S

STUDYING LIvEs LIVES STUDYING Psychoanalytic Psychoanalytic Behavioral Behavioral

General General Theory Theory

Culture & & personality personality CuHure

Biological

Terman

Binet Intelligence: Psychopathology. Kraepelin Kraepelin Psychopathology: Personality traits. Personality dimensions. dimensions. types:

Individual Individual & & group group differences differences

&

Self·understanding: Self-understanding: Clinical patients: patients:

Research subjects: subjects: Research

figures: Biographical figures:

Eysenck

DSM-I Menninger OSM-I

Introversion IntroverslonExtraversion Extraversion

Murray Mun^ay

MMPI

Gardner H. Garmer OSM·III DSM-III

CPI CatteD Cattell Wiggins 'Big Meehl CPI "Big 5' 5" a·sort Q-sort Block

C.rOflbach Qropbach Eysenck authoritanan authontanan personality personality act frequency dreams slips jokes anxiety

processes & & processes classes of classes behavior behavior

Individual Individual persons persons and and lives lives

Wechsler

Grou Group differences: differences: (gen r, age, race, class, (gender, dass, cuHure, culture, historical period)

Specific Specific

Psychometric Psychometric Humanistic Cognitive Cognitive Behavior Behavior genetic genetic & & SociObiological Sociobiological

honesty

phobias

achievement social cognition cognitk)n motivation . stress & & coping frustr & & frustration aliC?n self.m aHruism self-monitoring onitoring aggress creativity goal.seeking goal-seeking ual dmguse drug use . , ^®'^y of ^ t)ehavior delay behavior SUICIde suidde cation anti-Semitism gratificaticj emitism gratifi anti.S symptoms symptoms

:�

Freud's case studies: studies: Dora, Little Hans, Rat Man, Dr. Schreber, et al. Dr.Schreber,etal. "Eamst" 'Earnst'

Case Studies in Behavior Modification

DSM-III OSM·III caseboo Casebookk

Cocaine Lives in Letters Lives Progress.� from Jenny Users Progres Leonardo da Vinci Leonardo da Vinci oung �an Luther Y Dostoevsky Ea�y Mental Hitler Dostoevsky The The Eariy Hitler Young Man Luther Gandhi es Moses Trails Moses Traits of of 300 300 Wllson^^"^*^ cd?^^^ Wilson i III ^ ^ Geniuses GenKises George George'11 f ^ ^ E. Dickinson Melville Van Gogh Melville Gogh, etc etc..

� J�

1900 1900

1940 1940

FIGURE FIGURE 11 The The history of of four interrelated types of of inquiry in personality psychology.

1980 1980

studies Studies in psychodiagnostic categories, as with Kraepelin, Karl Menninger, and the Diagnostic (DSM-I, DSM-III); studies ofof personality traits, Diagnosticand andStatistical StatisticalManual Manual (DSM-I, DSM-III); studies personality traits, dimensions, and types, as with studies of extroversion-introversion by Jung and others, Henry Murray's Murray's studies studies of of needs, needs, development development of of the the MMPI, MMPI,Meehl's Meehl'sanaly analysis of clinical versus statistical prediction, Gough's California Psychological InvenInven tory (CPI); the work of Cattell, Eysenck, Block, Cronbach, Wiggins, and others on personality personaUty measurement, and recent discussion of the "Big Five" dimensions of personality. personahty. The study of group differences according to gender, age, race, social class, culture, culture, and and historical historicalperiod periodmay may also alsobe beincluded includedin inthis thisrow, row,but butfor forpurposes purposes of simplicity in the diagram, I have focused on studies of individual differences. differences. The third row is concerned with studies of specific processes and classes of behavior, as with aa set of phenomena phenomena examined examined in in psychoanalysis, psychoanalysis, including including dreams, dreams, slips, jokes, and anxiety; the study of phobias studied by Watson and other later behaviorists; the famous study of "honesty" by Hartshorne Hartshome and May (1928); (1928); the study of frustration and aggression at Yale by DoUard, Dollard, Doob, Miller, Mowrer, and (1948,1953); Sears (1939); the study of sexual behavior by Kinsey and colleagues (1948, 1953);

46

WILLIAM WILLIAM MCKINLEY McKiNLEY RUNYAN RuNYAN

personthe study of anti-Semitism in conjunction with research on the authoritarian person ality; the study of achievement motivation by McClelland, Atkinson, Clark, and Lowell (1953); studies of creativity at the Institute of Personality Assessment and Research, U.c. U.C. Berkeley, by Donald MacKinnon (1978), Frank Barron (1969), studies of of delay delay of of gratification gratification by by Walter Walter Ravenna Helson (1984-1985), and others; studies Mischel in the 1960s and later (Mischel, 1966); 1966); and a variety of other more recent studies of different classes of behavior. The items noted on the chart are obviously only selections from a much larger set. The bottom row deals with studies of individual persons and lives, which have been divided into the four subgroups of studies in self-understanding, of clinical patients, of research subjects, subjects, and and of ofbiographical biographicalfigures. Afew fewof ofthe theitems itemsincluded included figures. A in the diagram are Freud's famous clinical case studies of Dora, Little Hans, the Rat Man, and Dr. Schreber et al.; a1.; Freud's psychobiographical analyses of Leonardo and Moses; the study study of of intelligence intelligence in in 300 300 historical historical geniuses geniuses da Vinci, Dostoevsky, and Moses; the Terman; the case study of by Catherine Cox (1926) in association with Lewis Terman; ExplorationsininPersonality Personality "Earnst" written by Robert White in Henry Murray's Explorations Adolf Hitler by the O.S.S. in World World II, and many other (1938), the study of Adolf subsequent psychobiographies; Henry Murray's studies of Herman Melville; the Progress (White, (White, 1952); 1952);Gordon GordonAllport's Allport's in Progress study of three normal lives in Lives in Jenny(1965); (1965);the theinfluential influentialedited editedcollection collectionofofCase CaseStudies Studies Lettersfrom fromJenny analysis of Letters (Ullmann&&Krasner, Krasner,1965); 1965);Erik ErikErikson's Erikson'spsychobio psychobioin Behavior Modification Modification (Ullmann Man Luther Luther(1958) (1958)and andGandhi's GandhVsTruth Truth (1969); a persongraphical studies of Young Man (1969); a person Cocaine Users Users(Spotts (Spotts&&Shontz, Shontz,1980); 1980);case casestudies studiesinin ality research study of nine Cocaine the DSM-III casebook; and psychobiographical studies of Henry James, James, Joseph Stalin, Emily Dickinson, Vincent Van Gogh, and many others (see Gilmore, 1984; 1982,1988a, Runyan, 1982, 1988a, 1988b). Although methodological approaches are not explicitly included in this diadia gram, it may be noted that different methodological approaches tend to be associ associated with particular objectives, and thus with particular rows in the chart. Most simply, the bottom row of studying individual persons and lives tends to rely on case study, archival, historical, and interpretive methods; methods; the third row, of studying specific processes and and classes of behavior, behavior, tends tendsto to rely relymore moreheavily heavilyon onexperimental experimental methods; while the second row, of studying individual and group differences, tends methods; to rely more on psychometric, correlational, and factor-analytic methods. The top row, general theory, may draw on varying combinations of methodological apap proaches. This conceptual partitioning of the field field of personality psychology into four different perspecdifferent objectives or streams of work is useful from a number of different perspec tives. First, it makes clear some of the very different kinds of objectives pursued different investigators. It is across these lines that different individuals and by different groups in personality psychology sometimes have little interest in, respect for, or even knowledge of each other's research. These differences have led at times to severe criticisms of each other's work, such as experimentalists being unhappy with the relatively grand or untestable claims of general theorists or, conversely, of

CHAPTER 22 STUDYING STUDYINGLIVES LIVES CHAPTER

47 47

of spe spegeneral theorists being uninterested in the more microexperimental studies of cific behaviors or the quantitative measurements of of particular dimensions. Some such analysis of the internal lines of division and criticisms within the field is necessary for understanding its intellectual and interpersonal structure. A second issue suggested by this diagram is that of of the fascinating epistemolog epistemologA of the very possibility of knowledge in each of of these four enterprises. ical question of of knowledge are and are not possible in each of these four lines of What kinds of inquiry, from developing general theory down to studying individual lives? The of this paper will focus on the issue of intellectual progress in the third section of of individual lives, but similar questions may be raised about each of of the study of other levels. Third, laying out the historical evolution of work within each of these four tasks raises interesting empirical and historical questions about what connections of general there are between them over time. What influence has the development of of specific individual and group differences, upon specific specific theory had upon the study of of behavior, or upon the study of of individual lives? How, for processes and classes of example, has psychoanalysis as a general theory had an influence upon diagnostic classifications, upon the study of of specific classes of of behavior such as dreams, jokes, or psychiatric symptoms, or upon the study of individual lives such as Leonardo da Vinci or Martin Luther? In turn, what influence has research at each of of these of psychoanalytic theory? Or, to take an three levels had upon the development of difference level, how have the diagnostic example from the individual and group difference categories in the Diagnostic Diagnostic and Statistical Manual (DSM-IV) (DSM-IV) been influenced by by been influenced categories in the advances in in psychoanalytic psychoanalytic theory, theory, behavioral behavioral theory, theory, cognitive cognitive theory, theory, and and biologi biologiadvances cal theory theory (Millon (Millon & & Klerman, Klerman, 1986)? 1986)? How How have have diagnostic categories such such as as cal diagnostic categories schizophrenia and and borderline borderline disorders disorders been been related to the the study of specific specific sympschizophrenia related to study of symp toms or or clusters of symptoms symptoms and and to to the the interpretation interpretation of of individual individual clinical or toms clusters of clinical or historical figures? In short, a whole research agenda is opened up by examining historical figures? In short, a whole research agenda is opened up by examining the existence existence or or degree degree of of interconnectedness or not not between research within within each each the interconnectedness or between research of these partially independent of work. work. of these four four partially independent streams streams of of personality psychol psycholAt a minimum, this conceptualization of the structure of of four distinct tasks or objectives and their relationships to each other ogy in terms of of the useful ways of of conceptualizing the structure of of the field of personality, is one of orienalong with the previously discussed conceptualizations in terms of theoretical orien conceptual issues, methods, and classes of substantive phenomena. tations, core conceptual fundamental about the My hunch is that this conceptualization reveals something fundamental of the field, although I will only begin to be able to argue that here. structure of

m n i . PROGRESS IN PSYCHOBIOGRAPmCAL PsYCHOBioGRAPmcAL INQUIRY INQUIRY psychobiography has developed not only within psychology, but also within Work in psychobiography psychoanalysis and psychiatry, history, political science, literature, and an assort assortment of of other fields including religion, the history of of science, and so on. The field

48 48

WILLIAM RUNYAN WILLIAM McKINLEY MCKINLEYRUNYAN

of of psychobiography psychobiography is traditionally defined defined as beginning beginning with Freud's Leonardo Leonardo da da Vinci Vinci and and a Memory Memory of of His His Childhood Childhood (1910/1957), with a number of of the earliest earliest psychobiographies psychobiographies summarized summarized in Dooley's Dooley's "Psychoanalytic "Psychoanalytic Studies of of Genius" Genius" (1916). I have reviewed the the intellectual intellectual and institutional growth of of work in psychobiog psychobiography and psychohistory psychohistory elsewhere elsewhere (Runyan, 1988a), including a quantitative quantitative analy analysis of of the growth growth of of publications publications in the field and a review review of of professional professional organiza organizations, specialty journals, conferences, conferences, academic courses, and dissertations in the the field, so I will not repeat that broader survey survey here. I will, however, discuss several several developments developments within personality personality psychology. The study of of individual lives was championed championed by Henry Murray (1938, 1981), Gordon Allport AUport (1937, 1942, 1965), Robert Robert White (1952, 1963, 1972), and others at Harvard beginning beginning in the the 1930s. Although Although there there were were earlier earlier academic academic psychologists involved involved in psychobiography, psychobiography, such as Morton Morton Prince's Psychology Psychology of of the the Kaiser: Kaiser: A A Study Study of of His His Sentiments Sentiments and and His Obsession Obsession (1915) and "Roosevelt **Roosevelt as Analyzed Analyzed by the New New Psychology" (1912), His and G. Stanley Hall's Jesus, Jesus, the the Christ, Christ, in the the Light Light of of Psychology Psychology (1917), these these and fell on infertile infertile soil, leaving no continuous legacy. In were isolated works that fell contrast, contrast, the personological personological and study study of of lives tradition tradition begun begun by Murray, Murray, Allport, AUport, and White has had an enduring enduring impact. Later Later personality personality psychologists psychologists working working of lives tradition include Alan Elms with psycho psychowithin the personological personological and study of biographical of Allport Allport (1972), Freud (1980), Skinner (1981), Murray (1987), biographical studies of Anderson with methodological methodological writings on psychobiography psychobiography and others; James Anderson Robert Stolorow Stolorow (1981b) and on William James (1981a) and Henry Murray (1988); Robert George Atwood Atwood on personality theorists (1979); Peter Newton Newton on Samuel Samuel and George Johnson (1984); Ravenna Ravenna Helson Helson on E. Nisbet (1984-1985); Gerald Mendelsohn Mendelsohn Johnson Verdi (1985); Rae Rae Carlson Carlson in applying Silvan Tomkins' script theory (1981); on Verdi myself on conceptual conceptual and and methodological methodological issues (1982, (1982,1983,1988a,b); Alexmyself 1983, 1988a,b); Irving Alex a n McAdams McAdams on on ander on Freud, Jung, and Harry Stack Sullivan (1990); and D Dan Mishima (1985). Yukio Mishima In Life Life Histories Histories and and Psychobiography: Psychobiography: Explorations Explorations in Theory Theory and and Method Method (Runyan, (Runyan, 1982), I attempted attempted to provide a critical review of of basic methodological methodological and conceptual conceptual problems encountered encountered in the intensive study of of individual lives, whether in the form form of of biographies, psychobiographies, psychobiographies, or clinical clinical case studies. This This whether included a review review of of psychological psychological literature on the case study method, idiographic idiographic included psychobiography. The discussion of of psychobiography psychobiography analyzed issues issues methods, and psychobiography. of evidence evidence needed needed for a psychobiographical psychobiographical interpretation, the such as the kinds of of alternative psychobiographical psychobiographical explanations, the dangers of critical evaluation of psychological reductionism, reductionism, the extent extent to which which adult personality personality and behavior behavior can can psychological cannot be explained explained by childhood childhood experience, the problems problems of of attempting attempting to to or cannot early life events, the trans-historical trans-historical and cross-cultural cross-cultural applicability applicability "reconstruct" early of psychological psychological theory, and finally, the relative contributions contributions of of psychoanalytic psychoanalytic and and of non-psychoanalytic theory theory to psychobiography. psychobiography. The The relations relations of of psychobiography psychobiography non-psychoanalytic to the wider field field of of psychohistory, psychohistory, as as well as to other hybrid disciplines such as

CHAPTER 22 STUDYING STUDYINGLIVES LIVES CHAPTER

49 49

political psychology, historical psychology, and psychological anthropology, are analyzed in Psychology Psychology and Historical Historical Interpretation Interpretation (Runyan, 1988a). The recent resurgence of interest in the study of lives among personality psychologists is well represented in a special issue of the Journal Journal of of Personality Personality on "Psychobiography *Tsychobiography and Life Narratives" (McAdams & & Ochberg, 1988b), also published as a book with the same title (McAdams & & Ochberg, 1988a). As stated in the Introduction, Today, personality psychologists seem less ashamed than they did 20 years ago to to admit admit that the the subject subject of of their their study study is is human human lives. lives. .. .. .. Once Once again, again, it it is is okay okay to to study study the "whole "whole person." person." Better, Better, contemporary contemporary personologists insist, insist, as did did pioneers pioneers like like Gordon Gordon Allport AUport and and Henry Henry Murray, Murray, that that such such an an endeavor endeavor as is the the personologist's personologist's raison raison d'�tre. d*itre,(McAdams, (McAdams,1988a, 1988a,p.p.1)1) is

of lives back into person personTwo further indicators of the integration of the study of psychology are a general personality textbook, The Person: An An Introduction Introduction ality psychology Personality Psychology Psychology (McAdams, 1990), which gives substantial attention to to Personality individual life stories and psychobiography, and a recent book on Personology: Personology: Content and Method in Personality Assessment and Psychobiography Content and Method Personality Assessment Psychobiography (Alexander, 1990), which contains psychobiographical psychobiographical interpretations of of Freud, Jung, and Harry psychobiographical interpretation, and out outStack Sullivan, suggests principles for psychobiographical lines a teaching format for integrating personality assessment with the study of individual lives. An An important recent book on psychobiography is Uncovering Uncovering Lives: Lives: The Uneasy Alliance of Biography and Psychology (Elms, 1994), which provides Alliance of Biography and Psychology practical methodological advice and contains fascinating psychobiographical por portraits of of psychologists such as B. F. Skinner, Freud, Jung, and Gordon Allport, AUport, as well as of of selected science fiction writers and political figures. Either explicitly or implicitly, many psychologists still have a number of of objec objections to the detailed study of individuals. We have been trained to think about social science in a way which makes the study of individuals seem somehow trivial, irrelevant, or misguided. Typical concerns are that the study of of individuals is not rigorous enough, is too subjective, is not generalizable generalizable enough, or is not sufficiently sufficiently scientific. I will briefly review a number of of these criticisms and respond to them. Perhaps the most widespread criticism of of studies of of particular lives is that it is difficult difficult to generalize from from them. Staub (1980) suggests that "if we focus focus on the uniqueness of of every human being, we cannot generalize from one person to another. Since the aim of of science is to discover laws or principles-applicable principles—applicable at least to some, if not to all people—what people-what we will learn will not contribute to a science of AUport's summary of of such criticisms is that "We'd have to psychology" (p. 3). Allport's of any scientific value" (Allport, (AUport, generalize to other people or else we'd have nothing of 1962, p. 406). These criticisms seem to be based on the unwarranted assumption that the of personaUty goal of personality psychology is solely to produce generalizations at the highest possible level of of abstraction, preferably preferably universal generalizations. As argued earlier, psychology needs to attend to at least four different different kinds of of objectives, personality psychology

so 50

WIWAM McKINLEY RUNYAN WILUAMMCKINLEYRUNYAN

ranging from general theory through the study of individual differences differences and specific classes of of behavior to the study of individual persons. Although there is some transfer between these four levels, they are at least partially independent of each other. To the cry of, "How can you generalize from that idiographic study?," study?,'* the equally appropriate response is, "How can you particularize from that group or population study?" study?'* Work on all four tasks is necessary, and the fact that inquiry at one level does not automatically answer questions at the other levels is not a telling criticism. A A second objection is that interpretations of of individual cases are seen as of most people's people*s lives are too arbitrary or subjective. For example, "The events of sufficiently sufficiently variegated and multifarious that virtually any theoretical template can investigator freedom to locate the be validated. The case study simply allows the investigator preformulated convictions" (Gergen, 1977, p. facts lending support to his or her preformulated 142). Is interpretation of the single case little more than an arbitrary application of one's theoretical prejudices? It may be possible to interpret any life with any of theory, but often only at the cost of distortion or selective presentation of of the evidence. Any explanatory conjecture can be made, but not all of of them stand up under rigorous cross-examination. A third objection is that it is not only impractical, but literally impossible impossible to A idiographic study of of every individual. If individuals individuals are largely largely dissimilar, conduct an idiographic then "every sparrow would have to be separately identified, named and intuitively understood" (Murray, 1938, p. 715). If all individuals are unique, unique. then it would be necessary to formulate "as many theories as there are persons in the universe" (Levy, 1970, p. 76). This criticism raises an important question about the costs and (Levy. 1970. of detailed studies of of individuals. Granted that there are not sufficient sufficient benefits of resources for studying every individual in the universe, it is still entirely feasible to conduct detailed idiographic of particular interest idiographic studies of individuals of interest to us, includ includAdolf Hitler, Sigmund Freud, and Virginia Woolf, ing historical figures such as Adolf particular clinical patients, and other individuals of of interest. We do not have the time and money to study all individuals, but neither do we have the resources to test all possible theories. It is necessary to be selective, both in theoretical inquiries and in studies of specific individuals. A A fourth objection, and the final one to be discussed here, is that there is nothing wrong with the idiographic study of of individuals, but it is not science. Levy (1970), for example, argues that the meaning of of data about individual cases "can found within the context of of laws that hold for for all individuals. . . . It is not only be found possible to go beyond this and remain within the confines confines of of science" (p. 76). of The suggestion that science as a whole is not concerned with the study of particulars is clearly untenable, as this criterion would rule out significant significant portions of geology, astronomy and cosmology, and evolutionary evolutionary biology. These sciences of are concerned not solely with general principles and processes but also with topics, respectively, such as the structure and evolution of of this particular earth, earth. the structure of our solar system, and the particular sequence of of species leading to and origins of evolution of humans. There is, in short, a whole set of the evolution of "historical sciences"

CHAPfER 2 STIJDYING CHAPTER! STUDYINGLIVES LIVES

51 51

(Gould, 1989) concerned with the study of particular historical processes as well as with theoretical generalizations. Personality Personality psychology, in order to study persons persons and lives, must be (in part) a historical science as well as a nomothetic science. A useful way of looking at methodological and epistemological problems in the study of individual lives is to toconsider consider the the extent extent to to which which such such research research programs programs are "progressive" or not. What constitutes progress in our knowledge and under understanding of an individual life? To the extent that progress occurs, what processes bring it about? And finally, how do advances in other areas of personality psychology relate to progress in the study of individual lives? For example, what progress, if any, has there been in our psychological under understanding of Adolf Adolf Hitler during the course of research on his life, from from the Office Office of Strategic Services study in World War II (Langer, 1972), to to Alan Bullock's classic biography in 1952, to Robert Waite's The Psychopathic Psychopathic God: Adolf Adolf Hitler Hitler (1977)? What progress, if any, has there been in our knowledge and understanding of of Sigmund Freud, from to Ernest Ernest Jones' Jones' standard standard from an early biography by Wittels (1924), to three-volume biography (1953-1957), to more recent studies by Roazen (1975), SuUoway (1979), Gay (1988), and others? Finally, in the clinical realm, what ad adSulloway vances, if any, have there been in our knowledge and understanding of Freud's Wolf Man, the Rat Man, Dora, and others classic case studies of Little Hans, the Wolf through decades of reanalysis and reinterpretation (e.g., EUenberger, Ellenberger, 1970; Kanzer & Glenn, 1980)? Kanzer

A. Conceptualizing Conceptualizing Progress Progress in in Biography Biography A. To respond to questions about whether certain sequences of life history studies are progressive or not requires a clarification and definition definition of the concept of progress. The literature on the concept of progress is surprisingly extensive, from from studies of of the history of the idea of progress, to analyses of progress in physics, biology, the social sciences, history, and other disciplines, to progress in technology and material benefits, to economic progress, to progress in morals, and, finally, to progress in human welfare as a whole (cf. Almond, Chodorow, & Pearce, 1982). Underlying these many uses of the concept of progress, the idea may be defined defined most simply as change over time in a direction perceived as desirable or preferable. Thus, it involves a temporal or historical component and a valuative component. A third possible component of the idea of progress, which is sometimes but not necessarily implied, is that of progress as inevitable. inevitable. Let Letme memake makeclear clearthat that I am not claiming that a sequence of biographical studies is necessarily progressive. Some are, and some are not. Rather, the concept of progress is introduced as a way of addressing epistemological issues in the study of lives, as a way of comparing life history studies not to some absolute standard of truth, which can be impossibly difficult to specify, but rather of comparing a given study with prior studies in terms difficult of aa variety of specifiable criteria. of variety of specifiable criteria. of How should we look at progress in our knowledge and understanding of psychobiographical biographical studies can be individual lives? It seems to me that progress in psycho

52 52

WaUAM McKlNu;y McKlNLEY RUNYAN WILIJAM

meaningfully assessed in terms of criteria such as (1) the comprehensiveness of the evidential base, (2) the insightfulness and persuasiveness of interpretation, and (3) the literary or aesthetic appeal of the narrative account. This discussion focuses on thefirst first two of these criteria, criteria, the quality of evidence and of interpretation, while other works have focused on literary appeal and other criteria (Novarr, 1986). Advances in understanding can can occur through a variety of of processes, such as collect collecting additional evidence, developing more powerful background theory and research to draw on, and proposing and testing new interpretations. The processes involved in advancing our knowledge and understanding of individual lives can, for the sake of simplicity, be divided into eight steps or compo components, as in Figure 2. This set of processes is related to the specific criteria I am proposing. Other criteria of progress, such as moral rectitude, metaphoric expresexpres siveness, and political correctness, would suggest a somewhat different set of propro cesses. The components in Figure 2 have been numbered from 11 to 8 for purposes of identification, rather than to identify any rigidly fixed sequence of steps. The top left-hand box, Evidence, and Processes of Data Collection, includes activities such as finding finding additional letters or diaries, conducting further interviews, and findingadditional additionalarchival archivalrecords recordsororphysical physicalevidence. evidence.InInresearch researchon onHitler, Hitler,for for finding Kampf, interviews interviewswith withand and Mein Kampf, example, this would include material such as Mein documents by those who knew him, the discovery of his burned corpse, records of his personal physician, and the alleged discovery of previously unknown "Hitler Diaries." The second step is the Critical Examination of Evidence and Sources, Diaries." including activities such as detecting forgeries or falsifications in the evidential base and learning how much weight to give to the testimony of different witnesses. In the case of Hitler research, dental records supported the claim that the partly 11 Evidence, Evidence, and and processes of data collection

/

2

ICrltical Critical examination of levldence evidence and sources

L

.,,,, roUnd ... ry Background theory and knowledge

..

\

AI

4

Proposed Interpreexplanations! tations, explanation� 5 Critical testing testing of of Critical Interpretations & Interpretations & explanations

•

I

A "\

6 6 Life histoi hlsto ry accounts 1.2,3... ,2,3 ... n

J

�

I�

7 Critical evaluation of the account

88 Social, psychological and Social, pOlitical. political, psychological and historical historical factors factors

nGURE 22 FtGVRI

of individual lives: a simplified model of Advances in our knowledge and understanding of component processes.

CHAPTER 2 S1lJDYlNG CHAPTER! STUDYINGLIvEs LIVES

S3 53

burned body found by Russian soldiers outside the Fuhrer's air raid shelter in was that of Hitler (Waite, 1977), while a study of the paper used in the Berlin was was produced after after his death. alleged Hitler diaries revealed that it was The third component is Background Theory and Knowledge, which is drawn upon in interpreting the individual case, and would include theories of personality development, an understanding understanding of the relevant cultural and historical background, background, In particular, particular, and knowledge of relevant medical conditions and biological processes. In background theory and knowledge would include advances in all other areas this background of personality, such as general theories of personality, research on individual and group differences, and studies of specific processes and classes of behavior. The individfourth step is the generation of new interpretations and explanations of the individ falsification of of ual case, while the fifth step is the critical evaluation and attempted falsification proposed interpretations (cf. Runyan, 1981). the production of of aa narrative narrative account account of of the the life, which The sixth component is the incorporates a number of specific interpretations and explanations, organizes a substantial amount of data about the life, and draws on theoretical and background knowledge. The numbered subscripts indicate a whole sequence of accounts which 1944, one one might be produced about the same life, such as a biography of Hitler in 1944, and another another in in 1977. 1977. All Allof ofthe thepreceding precedingfive fiveprocesses processesused usedin inconstructing constructing in 1952, and the account can be repeated in an iterative cycle. The seventh step is the critical evaluation of the narrative account, as in the form of a book review for a biography biography or a case case conference conference for foraaclinical clinicalpresentation, presentation, factors such as the adequacy of the evidence, the appropriateness of considering factors of the background theory, and the credibility of the proposed interpretations. The is Social, Social, Political, Political,Psychological, Psychological, and and Historical Historical Factors Factors eighth and final final component is which influence each of the other processes. They influence what data are collected influence and seen as relevant, and how critically they are scrutinized. These factors influence influence the kinds of background theory and knowledge which are drawn upon, influence proposed and and how how critically critically they they are are evaluated, evaluated, affect affect the interpretations which are proposed and influence the critical the shape and structure of the finished narrative account, and reception that the finished work receives. developments The impetus for a new psychobiographical study can come from developments in any one of the eight components in Figure 2, 2, such as the discovery of new sources of evidence, advances in theoretical knowledge which make possible the interpretation of previously inexplicable events, or the critique and dismissal of of earlier interpretations and the proposal of new new ones.

RELATiONSfflPS BETWEEN BETWEEN THE THESTUDY STUDYOF OFLIVES LIVESAND ANDOTHER OTHER IV. RELATIONSIDPS AREAS OF OF PERSONALITY PERSONALITY PSYCHOLOGY PSYCHOLOGY AREAS What relations are there between progress in the psychological analysis of individual lives and advances in other branches of personality psychology? What relations, if any, are there between progress in the study of individual lives and progress in

54 54

WILLIAM RUNYAN WILLIAM MCKINLEY MCKINLEYRUNYAN

general general personality personality theory, theory, research research on individual individual and group group differences, differences, or research research on specific processes and classes of behavior? In terms of Figure 1, this is equivalent specific processes of of equivalent to asking what connections connections there are between between the study of of individual persons in the fourth row with developments in the top three rows. fourth row developments Studies of of individual individual lives can have implications for inquiry at each of of the the three levels. Methodology texts often state that the study of single cases can other other Methodology often of provide hypotheses, hypotheses, which then need to be tested with quantitative or experimental experimental methods (as in the second second and third rows). Or, general general theories of of personality personality can can methods personal or subjective origins in the lives of of individual individual influenced by their personal be influenced personality Stolorow and Atwood Atwood (1979) have argued is true in the personality theorists, as Stolorow work of of Sigmund Freud, Carl Jung, Wilhelm Wilhelm Reich, and Otto Rank. Abraham Abraham work Maslow reported that his ideas about self-actualization self-actualization came initially from from Maslow has reported reflections about about the similarities similarities of of two of of his mentors, mentors. Max Wertheimer Wertheimer and Ruth Ruth reflections Benedict (Maslow, (Maslow, 1971). Rae Carlson (1988) has discussed discussed how howpsychobiographical Benedict psychobiographical contribute to the development development of of general general theories of of personality. personality. research can contribute There unquestionably unquestionably are many influences influences of of the study of of individual lives on on There other lines of of research in personality, but for the present present discussion, I will focus focus influences going in the other direction-the direction—the ways in which progress in the study on influences of individual individual lives has been influenced influenced by advances in other areas of of personality personality of psychology. Most simply, how how is the course course of of research on individual individual lives influenced influenced psychology. general theories theories of of personality, by studies of of individual individual and group group by advances in general differences, and/or by research on specific specific processes and classes of of behavior? behavior? differences, of Figure Figure 2, which focuses focuses on processes processes contributing contributing to progress progress in In terms of of lives, all of of these developments in personality personality theory and research research the study of would be channeled channeled through through the third component, component, Background Theory Theory and Knowl Knowledge. This background background theory theory and knowledge knowledge would would include developments developments in all personality psychology, other areas of of personality psychology, but is not restricted restricted to advances in personality personality psychology, psychology, and also includes advances advances in other areas of of psychology psychology and contribu contributions from from disciplines such as history, sociology, anthropology, anthropology, political political science, science, and biology. biology. The The contributions contributions of of other areas of of personality personality psychology psychology to the study study of of lives lives will be be reviewed reviewed under under four four different different headings: (1) influences influences of of general personality personality theory theory upon upon the study of of lives, including including psychoanalytic, psychoanalytic, neoanalytic, and nonana nonanalytic theories theories of of personality; (2) influences influences of of research on individual individual and group group differences differences upon upon the study of of lives, such as research on intelligence intelligence or categories categories of psychopathology; (3) influences of psychopathology; influences of of research on specific specific processes processes and classes of behavior, such such as altruistic behavior; and (4) influences influences of of all different different kinds of of personality personality theory theory and research as they are funneled funneled into the study of of a single life, life, such as that of of Adolf Adolf Hitler. First, what influences influences have there been from advances advances in in general general theories of personality personality upon psychobiography psychobiography and the study study of of individual individual lives? The most most extensive extensive influence influence upon the study of of individual individual lives is certainly from from psychoanaly psychoanalysis, beginning beginning with Freud and his Viennese Viennese followers, spreading spreading through Europe Europe and the United United States, and taking another another step forward forward with Erik Erikson's work work

CHAPTER CHAPTER 22 STUDYING STUDYINGLIVES LIVES

55 55

in Young Young Man Luther Luther (1958) and Gandhi's Gandhi's Truth (1969). Recent discussion and reviews of of the influence of of psychoanalytic theory upon biography and psychobiogra psychobiography are contained in Edel (1984), Mack (1971), and Moraitis and Pollock (1987). different schools or traditions of of psy psyWithin psychoanalysis, the influence of different choanalytic theory upon biography and psychobiography psychobiography can be analyzed in more detail, such as moving from traditional Freudian drive theories to ego psychology, object relations theory, and self-psychology. Developments in ego psychology and object relations theory approaches to psychobiography are reviewed by Loewenberg Adolf Hitler to illustrate (1988), who uses examples such as Richard Nixon and Adolf how changing theoretical frameworks frameworks within psychoanalysis have led to revised interpretations of of the same historical figures. In the area of self-psychology, self-psychology, Strozier (1985) covers applications of self-psychology to biography and history, of Kohutian self-psychology and Strozier and Offer Offer (1985) examine a number of of applications of of self-psychology self-psychology to political leaders such as Abraham Lincoln, Woodrow Wilson, Kaiser Wilhelm II, and Mahatma Gandhi. At a clinical level, the implications of changes in psychoan psychoanof Freud's case studies of of Dora, Little Hans, alytic theory for the reinterpretation of the Rat Man, the Schreber Schreber Case, and the Wolf Wolf Man are explored in a volume edited by Kanzer and Glenn (1980). What, though, about the uses of of personality personaUty theories other than psychoanalysis of personality could be drawn on in in psychobiography? In principle, any theory of psychobiography could interpreting an individual life history, so psychoanalytic psychobiography behavioral psychobiography, phenomenological-humanistic phenomenological-humanistic be supplemented with behavioral cognitive psychobiography, and so on. What, though, has actually psychobiography, cognitive frameworks in psychobiography? been done in using these other theoretical frameworks Perhaps the most extensively developed behavioral interpretation of of a life is in the three volumes of of B. F. Skinner's autobiography (1976, 1979, 1983), in which (1976,1979,1983), of changes in the external environment he attempts to describe his own life in terms of and their effect overt behavior, without reference to inner experiences effect on his overt or feelings. One rare attempt to apply social learning theory in a book-length psychobiography is a study of of Elizabeth Cady Stanton in terms of of Bandura's social psychobiography learning theory (Griffith, 1984). In the clinical realm, a far greater number of of individual individual cases have been interpreted in terms of of behavioral behavioral or learning theory (e.g., Turkat, 1985; 1985; UUmann Ullmann & Krasner, 1965; Wolpe & Rachman, 1960). Uses of phenomenological-humanistic of phenomenological-humanistic personality personality theory may also occasion occasionally be found, such as Carl Roger's reinterpretation of of the case of of Ellen West (1980), who suffered suffered from anorexia nervosa and eventually committed suicide; a study of of Clarence Darrow in terms of of Charlotte Buhler's theory of of stages of goal seeking (Horner, 1968); or Sartre's use of of existential theory as well as psychoanalysis in his study of Flaubert (1981). of The The uses of of alternative theoretical perspectives in the study of of lives are also of Personality on "Psychobiography indicated in a special issue of of the Journal Journal of Personality "Psychobiography and Narratives" (McAdams & Ochberg, 1988b), which illustrates the application Life Life of Silvan Tomkins' script theory to the lives of Nathaniel Hawthorne and Eleanor of Marx (Carlson, 1988) and to the analysis of of personal documents and clinical material

56

WIWAM WILLIAMMCKINLEY MCKINLEYRUNYAN RUNYAN

(Alexander, 1988); the quantitative analysis of Eriksonian themes of of identity, inti intimacy, and generativity in the writings of Vera Brittain (Stewart, Franz, & & Layton, 1988); and a study of of achievement affiliation, and power motives in Richard Nixon (Winter & & Carlson, 1988). In spite of the growing number of of applications of of other personality theories in psychobiography, it still seems fair to say that far more psychobiographical studies have been influenced by some versions of of psychody psychodypsychobiographical combinanamic theory than any other personality theory, or even all the others in combina tion. Elsewhere I have speculated on different possible explanations for the relative contributions of psychoanalytic and nonanalytic theory in psychobiography (Runyan, 1988a). Looking at the second row of Figure 1, we may ask what developments in the study of individual and group differences have been used in advancing our of individual lives? To mention just a few few examples, let us start with understanding of LQ.'s of the example of intelligence. Lewis Terman began trying to estimate the I.O.'s historical figures such as Francis Galton (Terman, 1917), and Terman's student LQ.'s 's of of 301 famous men and women Catherine Cox (1926) estimated the childhood 1.0. Early Mental Mental Traits of of Three Hundred Hundred Geniuses Geniuses (1926). For example, on in The Early LQ. ratings for recorded behavior up to age 17, the estimate for John Stuart Mill 1.0. was 190, for Goethe, 185, Pascal, 180, Voltaire, 170, David Hume, 155, Hegel, 150, Descartes, 150, Thomas Jefferson, 145, Napoleon, 135, and so on. Cox and Terman are explicit that these estimates are only for early recorded recorded behavior, and may or may not accurately reflect reflect the individual's actual I.Q. LQ. I am not trying to vouch for the adequacy of these analyses, but rather to indicate how advances in the study of individual differences, in this case, I.Q., LQ., were used in the study of of particular of historical figures. Terman points out how biographer's ignorance of age norms in in historical figures. Terman points out how biographer's ignorance of age norms the development of intelligence led them to misinterpret the behavior of their the development of intelligence led them to misinterpret the behavior of their subjects. For For example, example, Karl Karl Pearson's Pearson's biography biography of of Francis Francis Galton Galton reports reports data data on on subjects. Galton's childhood performance and then says that it gives no significant significant indication of his future genius, where Terman assesses the same material and says it indicates of childhood 1.0. LQ. of near 200. aa childhood Staying at the level of of individual differences, differences, consider the implications of of advances in psychodiagnostic categories for our understanding of of individual lives. of England during the American American Revolution Revolution and whose reign George III, King of of physical and lasted from 1760 to 1820, suffered from a perplexing combination of psychological disorders periodically throughout his later life, including symptoms such as delirium, excitement, sleeplessness, painful weakness of of the arms and legs, visual and auditory disturbances, delusions, and agitated talking and hyperactivity. How was such a puzzling array of of symptoms to be explained? To simplify, the history of of different explanations of the king's disorders can be roughly divided into five stages: (1) contemporaneous explanations, which fell back on the theory of of an imbalance between the four humours of black bile, yellow yellow bile, phlegm, and choler; (2) classification classification according to 19th century descriptive psychiatry, in which he was diagnosed as having having "ordinary ''ordinary acute acute mania"; psychodynamic explanation explanation in in diagnosed as mania"; (3) (3) aa psychodynamic 1941, in which his breakdowns were understood as breakdowns of his vulnerable 1941, in which his breakdowns were understood as breakdowns of his vulnerable

STUDYINGLIvEs LIVES CHAPTER 22 STUDYING

57 57

defenses under pressure of of stressful political and domestic events; (4) explanations defenses in the 1960s in terms of of the metabolic disease porphyria, which leads to physical and psychological disorders similar to those of of George III; and (5) finally to criticisms criticisms of of the porphyria hypothesis because the genetic evidence seems inconsistent with it, and a search for alternative explanations consistent with both George Ill's symptoms and with the genetic evidence. Details of of these symptoms and the chang changing classifications and diagnostic diagnostic assessments of the King's disorders are presented elsewhere (Runyan, 1988b), but the point for present purposes is to show how understanding of of an individual case is dependent upon changing background knowl knowledge in the form of of available diagnostic systems and categories. As new diagnostic categories emerge, such as borderline disorders or narcissistic disorders, they are applied to a host of then freshly freshly appHed of historical and clinical cases, such as Adolf Adolf Hitler, Thomas Wolfe, or Pablo Picasso. of Figure 1, how has research on specific processes Looking at the third row of of behavior affected affected our knowledge and understanding of of individual and classes of lives? What contributions to biographical and case study analyses have been made through research on such classes of of behavior as specific psychiatric symptoms, sexual behavior, anti-Semitism, creativity, obedience obedience to authority, bystander intervention, altruism, stress and coping, drug use, or suicide? To mention one example, Samuel Altruistic Personality: Personality: Rescuers Rescuers of of Jews in Nazi Nazi Europe Europe and Pearl Oliner in The Altruistic reason(1988) draw on the existing literature on altruism, prosocial behavior, moral reason interpersonal attachment, and empathy in their study of more than 400 400 rescuers ing, interpersonal of Jews in Nazi-occupied Europe. of As a second example, consider that Woodrow Wilson did not learn his letters until age 9, and could not read until he was 11. Alexander George and Juliette George (1964) originally suggested that Wilson as a boy was filled with rage at his perfectionist father father which he could not openly acknowledge acknowledge or demanding and perfectionist express, and that his failure to learn was motivated by unconscious resentment of his father. In response, Weinstein, Anderson, and Link (1978) argued that Wilson's difficulties, but to developmental developmental dyslexia, delay in reading was not due to emotional difficulties, estabUshment of dominance of one hemisphere, which is caused by a delay in the establishment usually the left, for language. In rebuttal, the Georges (1981-1982) drew on details of recent research on dyslexia. In particular, they argued that it is not established that the absence of cerebral dominance is responsible for dyslexia, that many specialists continue to believe beUeve that emotional emotional factors are responsible for some reading disorders, and that life—such as the amount of his reading, the neatness of of his details of Wilson's life-such excellent spelling-are spelling—are all inconsistent inconsistent with a diagnosis of handwriting, and his excellent developmental dyslexia. In this debate they draw on a specialized body of of psycholog psychologdevelopmental of behavior in order to critique an ical theory and research on a particular class of alternative explanation and to argue that the bulk of the evidence is consistent with their original interpretation. (This debate continued in subsequent years, with references cited in Link et al, references aI., 1986.)

58 58

WILLIAM RVNYAN WILLIAM McKINLEy MCKINLEYRUNYAN

A. A. Adolf Adolf Hitler Hitler Thus far, I've indicated how advances in the areas of general theory, the study of of individual differences, and research on specific processes and classes of behavior have been used in the study of individual lives. The array of uses of other areas of of from the bottom personality psychology in the study of lives can also be illustrated from up, as it were, by examining the course of research on a specific life and seeing up, affected by theory and research from from different different areas of personality how it has been affected fulcrum or lens from from which to view psychology. The individual life can be used as a fulcrum the range of uses of psychology. This array of uses of psychological theory and Godstudy. The Psychopathic Psychopathic God: research will be illustrated with a psychobiographical study, Adolf Hitler (1977) by Robert Waite. Adolf At the level of general theory, Waite draws most heavily on psychoanalytic theory, with discussions of the anal stage in Hitler's development (pp. 148-149), of a "primal scene trauma" and its consequences (pp. 162-168), of Hitler's Oedipus of as displace displacecomplex (pp. (pp. 162-165), and of the operation of defense mechanisms such as (p. 190). 190). Waite also draws on the psychoso psychosoment and projection in his anti-Semitism (p. cial theory of Erik Erikson in discussions of trust and mistrust in Hitler's childhood, with pervasive feelings of mistrust remaining consequential throughout his life (pp. 383-386), and with discussions of identity development development and identity crises in (pp. Hitler's adolescence and early adulthood (pp. 184-205). differences, research on a number of At the level of individual and group differences, of drawn upon in interpreting his groups to which Hitler could be assigned were drawn personaUties, monorchids monorchids (males with one testicle), anal life, such as borderline personalities, characters, and anti-Semites. For example, on the basis of Soviet autopsy reports on Hitler's partially burned body, Waite believes there is convincing evidence that Hitler's left testicle was missing. Hitler also had a wide variety of psychological characteristics which match those of studies of other patients with this characteristic, such as feelings of social inadequacy, concerns with bowel movements and feces, belief in themselves as special persons, and passive tendencies with a reaction belief formation formation against them in an insistence on hardness, toughness, and ruthlessness. Another study suggests that monorchid patients often often have an intense concern with redesigning and reconstructing buildings, as if to quell anxieties about defects in which is consistent with Hitler's preoccupation with with designing and their own bodies, which Vienna, and and Berlin Berlin (pp. (pp. 150-162). 150-162). redesigning elaborate architectural plans for Linz, Vienna, classification has been a subject of Hitler's psychodiagnostic and medical classification of different diagnostic possibilities, extensive debate, and Waite reviews at least six different effects, medical poisoning (from (from including Parkinson's disease with psychiatric side effects, an incompetent physician), rapidly progressive coronary arteriosclerosis, syphilis, damage to his left cerebral hemisphere, and borderline personality (pp. (pp. 349-359). Waite finds the last of these diagnoses most persuasive, as it best fits many (though not all) of Hitler's behavior patterns. behavior and experience, Waite At the level of research on specific classes of behavior draws on studies in areas such as anti-Semitism (pp. 359-373), survivor guilt (over

CHAPIER CHAPTER 1 2

S9 59

STUDYING STUDYING LIVES LIVES

the death of his brother) (pp. 171-172), sexual perversion, in that Waite argues that Hitler had women urinate or defecate on his head (pp. (pp. 237-243), masochistic self-destructive behavior, which Waite argues was partially responsible for some or self-destructive of Hitler's disastrous military decisions such as invasion of the Soviet Union or declaring war on the United States (pp. 391-411), andfinally in an an attempt finally suicide, in to understand Hitler's end (pp. 411-426). This discussion is not an exhaustive analysis of Waite's use of psychology in interpreting Hitler, but does illustrate how our understanding of an individual life can be informed by theory and research at the three levels of general theories of personality, studies of individual and group differences, and research on specific processes and classes of behavior. An intriguing question for future inquiry is what additional advances in personality and other branches of psychology can or will be drawn upon in further illuminating Hitler's behavior, personality, and career. V. THE STUDY OF LIVES AS A "SOFT SYNTHESIS" IN PSYCHOLOGY V.

to soft soft dimension dimension seems seems useful useful in in thinking thinking about about the the array arrayof oftheoreti theoretiA simple hard to and research methods that constitute thefield personcal orientations, objectives, and field of person aUty psychology (see Fig. 3). The "hard" ality "hard" end of the continuum is concerned with quantitative measurement, experimental control, objectivity, and being scientific in natural science science sense, while the "soft" "soft" end of the continuum is concerned with a natural subjective experience, meaning and interpretation, social-historical context, and being scientific in a human science tradition (see Fig. 3). "Hard" "Hard"

Traditions

Objectives

"Soft" "Sor

Behaviorist Behaviorist

Psychometric Psychometric

Psychoanalytic Psychoanalytic

PhenomenologicalPhenomenological-

Cognitive

Biological Biological

Study of lives

humanistic humanistic Culture & & personality personality

(1) (1) Developing Developing general theories of personality personality (2) (4) Understanding lives in (2) Analyzing individual and group differences differences

(3) Studying specific processes processes and classes of behavior behavior (3)

context social-historical context

Experimental (laboratory,

Quantitative

longitudinal Longitudinal

(psychometric,

(prospective,

Interpretive (language,

field, multi-variate,

correlational,

retrospective,

symbolization, symbolization.

taxonomic, epidemiological,

sequential)

intentionality,

Methods, Methods, quasi-experimental) Disciplines

probabilistic)

meaning)

HistOrical-Contextual Historical-Contextual (case (case study, narrative, particular context, randomness, complex sequences,

psychobiographical, psychoblographlcal, cross-cultural, transhistorical)

FIGURE FIGURE 33 "Hard" "Hard" and and "soft" **soft" traditions, traditions, objectives, objectives, and and methods methods in in personality personality psychology. psychology.

60 60

WILLIAM RVNYAN WILLIAMMcKINLEY MCKINLEYRUNYAN

With With a hard to soft soft continuum continuum on a horizontal horizontal axis, we can roughly roughly place a number number of of personality personality psychology's theoretical traditions, objectives, and methods methods along this dimension. In terms of of traditions or theoretical orientations, as shown in the top row of of Figure 3, the behavioral, cognitive, psychometric, and biological traditions would be at the hard end of the continuum, while the psychoanalytic, psychoanalytic, study of of lives, phenomenological-humanistic, phenomenological-humanistic, and culture and personality traditions traditions would be at the soft soft end. In the middle row, where the four four major major objectives of of personality psychology discussed earlier are considered, the objective of of studying specific processes and classes of of behavior is at the hard end, understanding understanding lives in their social-historical context context is at the soft soft end, and developing general theories of of their differences are in the middle. Meth Methpersonality and analyzing individual and group differences of research research are in the bottom bottom row, with experimental experimental and quantitative quantitative methods methods ods of interpretive and historical-contextual historical-contextual methods at the soft soft end, and at the hard end, interpretive longitudinal methods in the middle. There There is a tendency tendency for for theoretical traditions, objectives, and research research methods methods at the "hard" end of of the continuum to go together. Behavioral and cognitive traditions are, for for example, likely to be concerned concerned with studying specific processes of behavior, and to use primarily experimental and quantitative methods, and classes of rather than interpretive interpretive or historical historical methods. At At the other end of of the continuum, continuum, rather "soft" traditions, objectives, and methods also tend to cluster together. For For the "soft" example, the psychoanalytic tradition is concerned concerned primarily with developing a general theory theory of of the mind and with understanding particular particular lives, and relies general rather than quantitaquantitaprimarily on interpretive and historical or case study methods rather tive or experimental methods. , This is, certainly, only an approximation, as theoretical traditions may pursue pursue a variety of objectives and utilize a variety of research methods. Psychoanalysis, Psychoanalysis, for for example, may be tested with quantitative and experimental methods (Fisher & & Greenberg, 1977), although it is more closely associated with clinical case study and interpretive methods. The point of the diagram is that the relative emphasis on different different objectives and research research methods is not randomly distributed across different different theoretical traditions, but rather that there is a rough clustering into hard and soft soft traditions, objectives, and methods. ^^Soft Synthesis" in in PersonaHty Personality Psychology Psyciiology A. Toward 8a "Soft of psychology as a whole, one can identify In the history of identify roughly three stages in the way the field has been divided. The first stage was the division division of of psychology psychology into major major systems or schools, such as structuralism, functionalism, functionalism, psychoanalysis, behaviorism, gestalt, and purposive or hormic psychology, up until roughly 1935 (cf. (cf. Heidbreder's Heidbreder's classic Seven Seven Psychologies, Psychologies, 1933; 1933; Hilgard, 1987). A A second stage, from approximately approximately 1935 to recent recent years, divided the discipline into the familiar familiar from of social, personality, developmental, clinical, experimental, physiological, "areas" of A third period, which has emerged emerged in recent years, and comparative psychology. A integration of of different different areas of of psychology with other disciplines into synthetic is the integration

CHAPTER22 SruDYJNG STUDYINGLIVES LIVES CHAYlER

61 61

fields such as "cognitive science" and "neuroscience," as well as "sensory science" and behavior genetics. Productive as these new syntheses may be, an important limitation is that they are restricted to the "hard" end of psychology and leave out a number of objectives and methods of of the "soft" traditions such as psychoanalysis, humanistic psychology, the study of of lives, and culture and personality. Their methods are typically experimental and quantitative, with relatively little attention given to historical-contextual methods. These hard the use of interpretive, case study, and historical-contextual syntheses in cognitive science and neuroscience raise an important set of questions, of questions about the understand understandbut they also leave unaddressed an important set of ing of persons and lives, of inner subjective experience, of texts and their meanings, and of relations with the social-historical context as traditionally pursued in the soft end of psychology. One intriguing possibility is that there might also be an emerging synthesis at the "soft" end of psychology around the study of of lives in their social-historical social-historical context, a synthesis which could fruitfully integrate a number of common concerns of personality, social, developmental, and clinical psychology with those in adjoining of social and historical sciences (Runyan, 1988a). At its best, such a synthesis might identify shared issues at the soft soft end of psychology, clarify clarify research methods and identify criteria, provide fresh energy and optimism for "soft" psychology, and lead to integrative theoretical and empirical work cutting across separate areas. One important component of of this soft synthesis is that it emphasizes the overlap of of psychology with the social sciences and humanities, in contrast to the biological sciences and computer hard syntheses, which emphasize overlaps with the biological science. Howard Gardner's (1985) extremely useful survey of the history of cognitive science emphasizes the overlap of cognitive psychology with artificial intelligence, linguistics, philosophy, anthropology, and neuroscience, but explicitly deemphasizes of affective affective factors or emotions, the contribution of historical and "the influence of of the background context in which particular actions cultural factors, and the role of or thoughts occur" (p. 6). It is just such neglected factors which may be central to a "soft synthesis" around the study of persons and lives. In one of of the two plenary addresses at the first annual convention of the American Psychological Society Society in 1989, George Miller said that he would give a young person starting out in psychology two pieces of of advice: Learn all you can about biology, and learn all you can about computers. This may be responsible advice for the hard end of psychology, but not for the field as a whole, and certainly not for the soft end of of psychology. The complementary advice for those going into the softer be, Learn all you can about the softer human science end of psychology might be. social sciences and humanities which overlap with psychology, and learn all you can about people and lives, including yourself. These are different different pieces of advice, different perspectives on the discipline. Together, they point toward coming from different psychology. a more adequate and comprehensive vision of the whole field of psychology. Although there is a lot of recent work on the study of persons and lives (Alexander, 1990; Elms, 1994; Gardner, 1993; 1993; McAdams, 1990; 1990; McAdams & & Och-

62 62

WILLIAM RUNYAN WILLIAM MCKINLEY MCKINLEYRUNYAN

berg, 1988a; Rabin, Zucker, Emmons, & 1988a, 1988b), & Frank, 1990; Runyan, 1992, 1992,1988a, it may be too early to talk about about this soft soft synthesis synthesis as something something already constructed constructed or completed. completed. Rather, there are diverse lines of of work in progress which might well be integrated integrated into such a wider synthesis, or have only partially been integrated integrated into such a synthesis. My own own current view view is that a soft soft synthesis would include include at least the following following My five five components: (1) theory theory and conceptualization, conceptualization, (2) quantitative quantitative empirical empirical re research, (3) clinical and practical-applied issues, (4) subjectivity, and (5) historical historicalinterpretive work, such as case studies and psychobiography. psychobiography. Hard syntheses focus focus on the first two of of these issues, the relationships relationships between between theory theory and research, and in more more liberal versions on the relation to practical-applied practical-applied concerns concerns as well. A A soft soft synthesis would would include attention attention to all five five components, components, subjectivity and historical-interpretive historical-interpretive work being of of substantial substantial importance. importance. with subjectivity This conceptualization conceptualization of of the internal internal structure of of a soft soft synthesis raises ques quesabout the relationships relationships between between all five five of of these components. Consider, Consider, for tions about relationships of of subjectivity subjectivity to the other elements. What are the example, the relationships relationships of of subjectivity subjectivity to theory development, development, as in the subjective origins of relationships personality theory? What are the relations of of subjectivity subjectivity to clinical practice, as in personality recent literature on "countertransference" "countertransference" and the use of of self self in the extensive recent clinical work? What What are the relations of of subjectivity subjectivity to historical-interpretive historical-interpretive work, work, clinical Erikson's (1975) discussion discussion of of the use of of "disciplined "disciplined subjectivity" in psychohis psychohisas in Erikson's These are samples of of the kinds of of issues left left out of of the hard syntheses, but tory? These newly developing soft synthesis. central for a newly developing soft Additional Additional questions questions suggested by these five five components components include, What are the relative relative contributions contributions of of theory, research, clinical clinical case experience, and subjective subjective self-awareness self-awareness to clinical practice? How How are theory and research related to historical historicalinterpretive work? To To what what extent extent is the value of of a theory (such as psychoanalytic psychoanalytic theory) determined determined by its relations to clinical practice, to quantitative-empirical quantitative-empirical work, to subjective subjective self-understanding, self-understanding, and/or to historical-interpretive historical-interpretive work? There There is, in short, a whole whole set of of important important questions about the relationships between between theory, research, applied work, subjectivity, and historical-interpretive historical-interpretive work which which can only be suggested suggested here, but which have long preoccupied preoccupied "soft" psychologists, psychologists, are peripheral to most "hard" psychologists, and may be freshly freshly illuminated illuminated with developments developments coming coming out of of a soft soft synthesis between psychology and the human humansocial-historical social-historical sciences. A A soft soft synthesis would draw on a variety of of methods, including hermeneutic hermeneuticinterpretive, case study, idiographic, psychobiographical, psychobiographical, historical, and narrative narrative methods (Bromley, 1986; Meehl, 1973; Polkinghorne, Polkinghome, 1988; Runyan, 1982). The The integrative integrative synthesis around this set of of problems and methods methods could also draw on on developments developments in related fields fields such as philosophy philosophy of of the social sciences, philosophy philosophy of of mind and consciousness, consciousness, principles of of other historical sciences, psychohistory, psychohistory, and the philosophy philosophy of of history (Campbell, 1988; Fiske & & Shweder, 1986; 1986; Searle, 1983). 1983). One One form of of this soft soft synthesis may may well well be around the study of of lives in social socialhistorical historical contexts. This, however, however, need not be the only soft soft synthesis. Just as there there

CHAPTER 2 STUDYING CHAPTER! STUDYING LIVES LIVES

63 63

are a variety of syntheses at the hard end of psychology, such as cognitive science, different syntheses syntheses neuroscience, and sensory science, so may there be several different soft end of psychology. emerging at the soft At their best, such syntheses can raise new issues, apply novel methods, and have important practical applications, as well as casting fresh light on old problems and pointing out similarities and integrations between previously disparate and Unes of of work. Many have spoken in recent years about a renaissance unconnected lines of personality psychology psychology (cf. Cantor & & Kihlstrom, Kihlstr5m, 1987; Craik, or revitalization of 1986; A 1986; Hogan, 1985; Maddi, 1984; 1984; Millon, 1984; 1984; Pervin, 1990; Tomkins, 1981). A of lives in social-historical context can be an important synthesis around the study of socialcontributor both to personality psychology and to the whole softer social developmental-clinical side of psychology. developmental-clinical

CONCLUSION VI. CONCLUSION VI. To summarize, this chapter argued that the structure of personaUty personality psychology can usefully be seen as involving four central tasks or objectives, namely, developing usefully of personality, analyzing individual and group differences, studying general theories of of behavior, and developing a better understanding of of individual specific classes of of raising intriguing questions, persons. This conceptual framework has the virtue of both theoretical and historical, about the possible and actual relationships between of "progress" each of these four enterprises. I then briefly sketched the dynamics of of lives. The next section explored relationships between the study of in the study of lives and other lines Hues of of work within personality psychology. Finally, I outlined a "hard" "hard" to "soft" array array of theoretical theoretical orientations, objectives, and methods methods in personality psychology, and suggested the possibiUty of a soft synthesis around the the personality psychology, and suggested the possibility of a soft synthesis around study of persons and lives. study of persons and lives. personahty measurement, in the causal analysis of Advances in theory, in personality personality processes, and in the understanding of individual lives are all important objectives for personality psychology. The challenge is to help clarify relationships between the possibilities and the limitations of theoretical, psychometric, experi experimental, and life-historical Ufe-historical forms of inquiry. Such an understanding of the relation relationships between these four enterprises can contribute to a better understanding of the structure and history of of the field. One way of of viewing the history of personality psychology is in its changing emphases on these four different different tracks of of inquiry. When personality psychology crystallized as a field in the 1930s with the seminal books by Allport AUport (1937) and Murray (1938), it could be seen as reacting against the sterile formalisms formaUsms of of academic psychophysics, and as an effort to integrate the rigor of academic methods with of persons and lives. In the tum turn away from the study of of lives in the understanding of 1960s, there was far greater emphasis on the "hard" the 1950s and 19608, "hard" traditions of measurement and experimentation, with a decline of interest in the "softer" issues of studying whole persons and in developing general theories of personality. With of

64 64

WOLUAMMCKINLEYRUNYAN WILUAM McKuuv RUNYAN

the current resurgence resurgence of of interest in the study of of persons and lives, there is a need to reexamine reexamine the relationships between between the array of of hard to soft soft traditions, objectives, and methods. Such reexamination reexamination can lead to a more coherent coherent and broadly integrative vision of of personality psychology psychology and contribute to the intellec intellecrevitalization of of the field. While much work work in recent years years tual and institutional revitalization has focused focused on syntheses at the hard end of of psychology, such as cognitive science science exciting possibilities for a synthesis at the soft end of and neuroscience, there are now exciting psychology around around the study of of lives in their social, cultural, and historical historical contexts. psychology

REFERENCES Alexander, I. E. E. (1988). (1988). Personality, Personality, psychological psychological assessment, assessment, and and psychobiography. psychobiography. Journal Journal Alexander, L of Personality, Personality, 56(1), 5(5(1), 265-294. 265-294. of Alexander, I. E. (1990). Alexander, I. E. (1990). Personology: Personology: Content Content and and method method in personality personality assessment assessment and psychobiography. psychobiography. Durham, Durham, NC: NC: Duke Duke University University Press. Press. Allport, G. G. W. W. (1937). (1937). Personality: Personality: A A psychological psychological interpretation. New New York: York: Holt. Holt. Allport, Allport, G. G. W. W. (1942). (1942). The use of of personal personal documents documents in psychological psychological science. New New York: York: Social Social Science Science Research Research Council. Council. Allport, G. Allport, G. W. W. (1962). (1962). The The general general and and the the unique unique in in psychological psychological science. science. Journal Journal of of Personality, Personality, 30, 405-422. 405-422. Allport, G. Allport, G. W. W. (1965). (1965). Letters Letters from from Jenny. New New York: York: Harcourt, Harcourt, Brace Brace & & World. World. Almond, G., Chodorow, Chodorow, M., & Pearce, Pearce, R. R. (Eds.). (Eds.). (1982). (1982). Progress Progress and and its discontents. discontents. Berke BerkeM., & Almond, G., ley: University University of of California California Press. Press. ley: Anderson, J. J. W. W. (1981a). (1981a). The The methodology methodology of of psychological psychological biography. biography. Journal of of Interdisci Interdisciplinary History, History, 11, 11, 455-475. 455-475. plinary Anderson, Anderson, J. J. W. W. (1981b). (1981b). Psychobiographical Psychobiographical methodology: methodology: The The case case of of William William James. James. In In L. L. Wheeler Wheeler (Ed.), Review Review of of personality personality and and social social psychology psychology (Vol. (Vol. 2, 2, pp. pp. 245-272). 245-272). Beverly Hills, CA: Sage. Anderson, J. W. (1988). Henry A. Murray's early career: A A psychobiographical psychobiographical exploration. Journal Journal of of Personality, Personality, 56(1), 139-172. process. New New York: Holt, Rinehart & Barron, F. (1969). Creative & Winston. Winston. Creative person person and and creative process. Bromley, D. B. (1986). The case-study case-study method method in psychology psychology and related disciplines. Chiches Chichester, England: Wiley. Bullock, A. (1952). Hitler,' Hitler; A A study study in tyranny. New York: Harper. Campbell, D. T. T. (1988). Methodology Methodology and epistemology epistemology for for social social science. Chicago: University of of Chicago Chicago Press. Cantor, N., N., & & KihlstrOm, J. J. F. F. (1987). (1987). Personality Personality and social social intelligence. Englewood Englewood Cliffs, Cliffe, NJ: NJ: Prentice-Hall. Prentice-Hall. Carlson, R. (1981). Studies Studies in in script script theory: theory: I. I. Adult Adult analysis analysis of of aa childhood childhood nuclear nuclear scene. scene. Carlson, R. (1981). Journal of of Personality Personality and and Social Social Psychology, Psychology, 40, 40, 501-510. 501-510. Journal Carlson, Carlson, R. R. (1988). (1988). Exemplary Exemplary lives: lives: The The uses uses of of psychobiography psychobiography for for theory theory development. development. Journal of of Personality, Personality, 56(1), 5(5(1), 105-138. 105-138. Journal Cohler, B. B. J. J. (1988). (1988). The The human human studies studies and and the the life life history: history: The The Social Social Service Review Review Cobler, Lecture. Lecture. Social Social Service Review, Review, 62(4), 62(4), 552-575. 552-575. Cox, C. C. (1926). (1926). The early mental mental traits of of three hundred hundred geniuses. Stanford, Stanford, CA: CA: Stanford Stanford Cox, University Press. Press. University

CHAnER CHAPTER22 STUDYING STUDYINGLIVES LIVES

65 65

Craik, K. K. H. H. (1986). (1986). Personality Personality research research methods: methods: An An historical historical perspective. perspective. Journal Journal of of Craik, Personality, Personality,54(1), 54(1),18-51. 18-51. Cronbach, L. J. (1957). The two disciplines discipUnes of scientific psychology. American American Psychologist, Psychologist, 12, 12, 671-684. 671-684. Cronbach, Psychol Cronbach, L. L. J.J.(1975). (1975).Beyond Beyondthe thetwo twodisciplines disciplinesofofscientific scientificpsychology. psychology.American American Psychologist, ogist, 30, 30,116-127. 116-127. Dollard, J., J., Doob, Doob, L. L. W., W., Miller, Miller, N. N. E., E., Mowrer, Mowrer, O. O. H., H., & & Sears, Sears, R. R. R. R. (1939). (1939). Frustration Frustration Dollard, and aggression. aggression.New NewHaven, Haven,CT: CT:Yale YaleUniversity UniversityPress. Press. Dooley, L. (1916). (1916). Psychoanalytic Psychoanalytic studies studies of of genius. genius. American American Journal JournalofofPsychology, Psychology,27,27, Dooley, L. 363-416. 363-416. biographica. New York: Norton. Edel, Edel, L. (1984). (1984). Writing Writinglives: lives:Principia Principia biographica. New York: Norton. New York: Basic Books. Ellenberger, The discovery discoveryofofthe theunconscious. unconscious. New York: Basic Books. EUenberger, H. H. F. F. (1970). (1970). The Elms, Elms, A. A. C. C. (1972). (1972). Allport, Allport, Freud, Freud, and and the the clean clean little little boy. boy. Psychoanalytic PsychoanalyticReview, Review,59,59, 627-632. 627-632. Elms, A. A. C. C. (1980). (1980).Freud, Freud,Irma, Irma,Martha: Martha:Sex Sexand and marriage "Dream Irma's injection." marriage in in thethe "Dream of of Irma's injection." Elms, PsychoanalyticReview, Review,67,67, 83-109. Psychoanalytic 83-109. Elms, A. A. C. C. (1981). (1981).Skinner's Skinner'sdark darkyear yearand andWalden WaldenTwo. Two. American Psychologist, 36,470-479. American Psychologist, 36, 470-479. Elms, Elms, A. A. C. C. (1987). (1987). The The personalities personalities of of Henry Henry A. A. Murray. Murray. In In R. R. Hogan Hogan & & W. W.Jones Jones (Eds.), (Eds.), Elms, Perspectivesininpersonality personality(Vol. (Vol. 1-14). Greenwich, CT: JAI Press. Perspectives 2, 2, pp.pp. 1-14). Greenwich, CT: JAI Press. Elms, A. A. C. C. (1988a). (1988a). Freud Freud as as Leonardo: Leonardo: Why Whythe thefirst psychobiographywent wentwrong. wrong.Journal Journal first psychobiography Elms, of Personality, Personality,56(1), 56(1),19-40. 19-40. of Elms, A. A. C. C. (1988b, (1988b, August). August). The The psychologist psychologistas as biographer. biographer.Henry HenryA.A.Murray MurrayAward Award Elms, Lecture, American American Psychological Psychological Association Association annual convention, Atlanta, Atlanta, GA. GA. Lecture, annual convention, Elms, A. A. C. C. (1994). (1994). Uncovering Uncoveringlives: lives:The The uneasy alliance of biography and psychology. Elms, uneasy alliance of biography and psychology. New New York: Oxford Oxford University University Press. York: Press. Erikson, E. E. H. H. (1958). (1958). Young Young man man Luther: Luther: AA study studyininpsychoanalysis psychoanalysisand andhistory. history.New New York: Norton. York: Norton. Erikson, E. E. H. H. (1969). (1969). Gandhi's Gandhi'struth. truth.New NewYork: York:Norton. Norton. Erikson, history and andthe thehistorical historicalmoment. moment. New York: Norton. Erikson, E. E. H. H. (1975). (1975). Life history Erikson, New York: Norton. & Greenberg, Greenberg, R. R. P. P. (1977). (1977). The Thescientific scientific credibility of Freud's theories therapy. Fisher, S., S., & credibility ofFreud's theories and and therapy. Fisher, New York: Basic Basic Books. New York: Books. social science: Pluralisms and subjectivi Fiske, Fiske, D., D., & &Shweder, Shweder,R. R.(Eds.) (Eds.)(1986). (1986).Metatheory Metatheoryinin social science: Pluralisms and subjectivities. ties. Chicago: Chicago:University Universityof of Chicago ChicagoPress. Press. In In J. Strachey (Ed. and Freud, Freud, S. S. (1957). (1957). Leonardo Leonardo da daVinci Vinciand anda amemory memoryofofhis hischildhood. childhood. J. Strachey (Ed. and complete psychological works of Sigmund Freud Trans.), The standard standardedition editionofofthethe complete psychological works of Sigmund Freud Trans.), The 12, pp. 3-82). London: (Vol. 12, London: Hogarth Press. (Original work published 1910) dreams. In J. (Ed. andand Trans.), TheThe standard Freud, Freud, S. S. (1958). (1958). The The interpretation interpretationofof dreams. InStrachey J. Strachey (Ed. Trans.), standard (Vols. 4-5,4-5, pp. pp. 1-621). edition works ofof Sigmund Freud edition of ofthe thecomplete completepsychological psychological works Sigmund Freud (Vols. 1-621). London: London: Hogarth Press. (Original work published 1900 1900)) Gardner, A history ofthe cognitive revolution. New New York:York: Themind's mind'snew newscience: science: A history of the cognitive revolution. Gardner, H. H. (1985). (1985). The Basic Books. Books. Gardner, H. H. (1993). (1993). Creating Creatingminds: minds:An Ananatomy anatomy creativity seen through the lives of Freud, ofof creativity seen through the lives ofFreud, Gardner, Einstein,Picasso, Picasso,Stravinsky, Stravinsky, Eliot, Graham, Gandhi York: Basic Books. Einstein, Eliot, Graham, andand Gandhi. NewNew York: Basic Books. Gay, P. P. (1988). (1988). Freud: Freud:AAlife lifefor forour ourtime. time.New New York: Norton. Gay, York: Norton. George, A. A. L., & George, George, J. J. (1964). (1964). Woodrow Wilson Wilson and and Colonel ColonelHouse: House:AApersonality personality George, L., & study. New New York: Dover. study. York: Dover. George, J. J. L., L., & & George, George, A. A. L. L. (1981-1982). (1981-1982). Woodrow Woodrow Wilson Wilson and and Colonel Colonel House: House: A A reply reply George, to Weinstein, Anderson, and Link. Political 96,96, 641-665. Political Science ScienceQuarterly, Quarterly, 641-665.

66 66

WILLIAM MCKINLEY MCKINLEYRUNYAN RUNYAN WILUAM

Gergen, K. J. (1977). Stability, change, and chance in understanding human development. development. & H. Reese (Eds.), Life-span Lifespan developmental developmental psychology: psychology: Dialectical Dialectical per perIn N. Datan & spectives spectives on experimental experimental research. New York: Academic Press. Gilmore, W. J. (1984). Psychohistorical Psychohistorical inquiry: A A comprehensive comprehensive research bibliography. bibliography. New York: Garland. Wonderful life. New New York: Norton. Gould, S. J. (1989). Wonderful right: The life of of Elizabeth Elizabeth Cady Cady Stanton. New York: Oxford Griffith, E. (1984). In her own own right: Griffith, University Press. Hall, C. S., & & Lindzey, G. (1957). Theories of of personality. personality. New York: Wiley. Hall, G. S. (1917). Jesus, the Christ, in the light of of psychology. psychology. Garden City, NY: Doubleday. & May, M. A. (1928). Studies in deceit. New York: Macmillan. Hartshorne, H., & psychologies. New York: Century. Heidbreder, E. (1933). Seven psychologies. Helson, R. (1984-1985). E. Nisbet's forty-first forty-first year: Her life, times and symbolizations of personality growth. Imagination, Imagination, Personality, and Cognition, 4, 53-68. Personality, and Psychology in America: America: A A historical historical survey. San Diego, CA: Harcourt Hilgard, E. (1987). Psychology Brace Jovanovich. Hogan, R. W. (1985). Introduction. In R. Hogan & &,W. W.Jones Jones(Eds.), (Eds.),Perspectives Perspectives in inpersonality personality CT: JAI Press. (Vol. 1). Greenwich, CT: Horner, A. (1968). The evolution of Homer, of goals in the life of of Clarence Darrow. In C. Buhler & & Massarik (Eds.), The course course of of human life. New York: Springer. F. Massarik Jones, E. (1953-1957). The lif lifee and work work of of Sigmund Sigmund Freud (Vols. 1-3). New York: Basic Books. & Glenn, J. (Eds.). (1980). Freud and his patients. New York: Aronson. Kanzer, M., & Kinsey, A., Pomeroy, W., & & Martin, C. (1948). Sexual Sexual behavior behavior in the human human male. Philadel Philadelphia: Saunders. Kinsey, A., Pomeroy, W., & Sexual behavior behavior in the human female. Philadel& Martin, C. (1953). Sexual phia: Saunders. Langer, W. C. (1972). The mind mind of of Adolf Adolf Hitler. New York: Basic Books. personality. New York: Random House. Levy, L. (1970). Conceptions Conceptions of of personality. Liebert, R. M., & & Spiegler, M. D. (1987). Personality: Personality: Strategies and and issues (5th ed.). Chicago: Dorsey Press. Link, A., et al. (Eds.). (1986). Introduction. In The papers papers of of Woodrow Woodrow Wilson (Vol. 54). Princeton, NJ: Princeton University Press. Loewenberg, P. (1988). Psychoanalytic models of history: Freud and after. In W. M. Runyan (Ed.), Psychology Psychology and and historical historical interpretation (pp. 126-156). New York: Oxford Uni University versity Press. Mack, J. E. (1971). Psychoanalysis and historical biography. Journal Journal of of the American American Psycho Psychoanalytic Association, Association, 19, 143-179. analytic MacKinnon, D. W. (1978). In search of of human effectiveness: Identifying Identifying and and developing developing creativity. Buffalo, NY: Creative Education Foundation. Maddi, S. (1984). Personology for the 19808. 1980s. In R. Zucker, J. Aronoff, Aronoff, & & A. Rabin (Eds.), Personality and and the prediction prediction of of behavior behavior (pp. 7-41), New New York: Academic Press. Personality Maslow, A. (1971). The farther farther reaches of of human nature. New York: Viking Press. McAdams, D. P. (1985). Fantasy and reality in the death of of Yukio Mishima. Biography, Biography, 8, 292-317. person: An An introduction introduction to personality personality psychology. psychology. San Diego, McAdams, D. P. (1990). The person: CA: Harcourt Brace Jovanovich.

CHAPTER 22 STUDYING STUDYINGLIVES LIVES CHAPTER

67 67

McAdams, D., D., & & Ochberg, Ochberg, R R. (Eds.). (Eds.). (1988a). (1988a).Psycho Psychobiography andlife lifenarratives. narratives. Durham, McAdams, biography and Durham, NC: NC: Duke University Press. McAdams, D., & & Ochberg, R R. (Eds.). (1988b). Psychobiography and life narratives [Special issue]. issue]. Journal Journalof ofPersonality, Personality,56(1). 56(1). McClelland, D. C., C, Atkinson, J. W., Clark, R R. A, A., & Lowell, E. L. (1953). The The achievement achievement motive. motive. New New York: York: Appleton-Century-Crofts. Appleton-Century-Crofts. Meehl, Meehl, P. P. E. E. (1973). (1973). Why Why II do do not not attend attend case case conferences. conferences. In In Meehl, Meehl, P. P. E. E. (Ed.), (Ed.), Psychodiag Psychodiagnosis: nosis:Selected Selectedpapers. papers.Minneapolis: Minneapolis:University UniversityofofMinnesota MinnesotaPress. Press. Mendelsohn, G. A. (1985). (1985). La Dame aux aux Camelias Cameliasand andLa La Traviata: Traviata:AAstudy studyofofdramatic dramatic Mendelsohn, G. A transformations in in the the light light of of biography. biography. In In R R. W. W. Hogan Hogan & & W. W.Jones Jones(Eds.), (Eds.),Perspec Perspectransformations tives tives in in personality personality(Vol. (Vol.1). 1).Greenwich, Greenwich,Cf: CT:JAI JAIPress. Press. Millon, T. (1984). (1984). On On the the renaissance renaissance of of personality personality assessment assessment and and personality personality theory. theory. Millon, T. Journalof ofPersonality PersonalityAssessment, Assessment, 450-466. Journal 48,48, 450-466. Millon, T., T., & & Klerman, Klerman, G. G. (Eds.). (Eds.). (1986). (1986). Contemporary Contemporarydirections directions psychopathology. New in in psychopathology. New Millon, York: York: Guilford Guilford Press. Mischel, W. W. (1966). (1966). Theory Theory and and research research on on the the antecedents antecedents of of self-imposed self-imposed delay delay of of reward. reward. Mischel, personality research (Vol.(Vol. 3, pp.3,85-132). In B. B. A A. Maher Maher (Ed.), (Ed.), Progress Progressininexperimental experimental personality research pp. 85-132). New New York: York: Academic Academic Press. New York: Wiley. Personalityand andassessment. assessment. New York: Wiley. Mischel, W. W. (1968). (1968). Personality Mischel, Introductiontotopersonality personality(3rd (3rded.). ed.).New NewYork: York:Holt, Holt,Rinehart Rinehart&& Mischel, W. W. (1981). (1981). Introduction Winston. Winston. Monte, C. C. (1987). (1987). Beneath Beneaththe themask: mask:An Anintroduction introduction theories of personality. Monte, to to theories of personality. (3rd (3rd ed.).ed.). New New York: York: Holt, Holt, Rinehart Rinehart & & Winston. Winston. Moraitis, G., & Pollock, Pollock, G. G. H. H. (Eds.). (Eds.).(1987). (1987).Psychoanalytic Psychoanalyticstudies studies biography. New York: Moraitis, G., & ofof biography. New York: International Universities Universities Press. Murray, H. H. A, A., et et al. al. (1938). (1938). Explorations Explorationsininpersonality. personality. New York: Oxford University Press. New York: Oxford University Press. Murray, Murray, H. A. (1981). (1981). Endeavors Endeavors ininpsychology: psychology:Selections Selections from personology of Henry from thethe personology of Henry Murray, H. A A. Murray Murray (E. A. (E. Shneidman, Shneidman, Ed.). Ed.). Harper Harper & &Row. Row. Newton, P. P. M. M. (1984). (1984). Samuel breakdown and and recovery recovery in in middle middle age: age: A A life life span span Newton, Samuel Johnson's Johnson's breakdown and its itscure. cure.International InternationalReview Review Psychodevelopmental approach approach to to mental mental illness illness and ofof Psycho developmental Analysis, 11, 11,93-1 93-118. 18. Analysis, of of biography, 1880-1970. WestWest Lafayette, Ind.: Ind.: Novarr, Novarr, D. D. (1986). (1986). The The lines linesof oflife: life:Theories Theories biography, 1880-1970. Lafayette, Purdue University Press. Oliner, S. P., & OHner, P. M. (1988). The of of Jews in in Nazi altruisticpersonality: personality:Rescuers Rescuers Jews Nazi & Oliner, The altruistic Europe. New NewYork: York:Free FreePress. Press. Europe. Personality:Theory, Theory,assessment, assessment, research. York: Wiley. Pervin, L. L. A A. (1980). (1980). Personality: Pervin, andand research. NewNew York: Wiley. Pervin, L. A. A. (Ed.). (Ed.). (1990). (1990). Handbook Handbook of of personality: personality: Theory Theoryand andresearch. research.New NewYork: York: Pervin, L. Guilford Press. Guilford Press. Peterson, C. (1988). Personality. Personality.San SanDiego, Diego,CA: CA:Harcourt HarcourtBrace BraceJovanovich. Jovanovich. Scott, Phares, Phares, E. E. J. J. (1988). (1988). Introduction Introductiontotopersonality personality(2nd (2nded.). ed.).Glenview, Glenview,IL:IL: Scott,Foresman. Foresman. Polkinghorne, D. (1988). (1988). Narrative Narrativeknowing knowingand and human sciences. Albany: University Polkinghorne, D. thethe human sciences. Albany: StateState University of New New York York Press. of Press. Prince, M. M. (1912, (1912, March March 24). 24). [Theodore] [Theodore] Roosevelt Roosevelt as as analyzed analyzed by by the the new new psychology. psychology. New New Prince, Times, pp. pp.1-2. 1-2. York Times, Prince, M. M. (1915). Thepsychology psychologyof ofthe theKaiser: Kaiser: study sentiments his obsession. AA study ofof hishis sentiments andand his obsession. Prince, (1915). The Boston: Badger. Boston: Badger. Rabin, A A. I., Zucker, R, R., Emmons, Emmons, R, R., & & Frank, Frank, S. S. (Eds.). (Eds.). (1990). (1990). Studying Studyingpersons personsand and Rabin, I., Zucker, lives. New NewYork: York:Springer. Springer. lives.

68 68

WIWAM WILLIAM McKINLEy MCKINLEY RUNYAN RUNYAN

Roazen, P. (1975). Freud and and his followers. followers. New York: Knopf. Rogers, C. R. (1980). Ellen West-and West—and loneliness. In C. R. Rogers, A A way way of of being. Boston: Houghton Mifflin. Mifflin. off his ear? The problem problem of of alternative Runyan, W. M. (1981). Why did Van Gogh cut off explanations in psychobiography. Journal Social Psychology, 40, Journal of of Personality Personality and and Social Psychology, 40, 1070-1077. Life histories histories andpsychobiography: method. psychobiography: Explorations Explorations in theory theory and and method. Runyan, W. M. (1982). Life New York: Oxford University Press. of lives. Journal Journal of of Runyan, W. M. (1983). Idiographic goals and methods in the study of Personality, Personality, 51, 413-437. Runyan, W. M. (Ed.). (1988a). Psychology Psychology and and historical historical interpretation. interpretation. New York: Oxford University Press. Runyan, W. M. (1988b). Progress in psychobiography. Journal of of Personality, Personality, 56(1), 295-326. work of of Henry A. Murray. Runyan, W. M. (1994). Coming to terms with the life, loves, and work Contemporary Psychology, Psychology, 39(7), 39(1), 701-704. Contemporary family Gustave Flaubert, 1821-1857 amily idiot: Gustave 1821-1857 (Vol. 1) (C. Cosman, Sartre, J.-P. (1981). The f Trans.). Chicago: University of Chicago Press. Intentionality. Cambridge, England: Cambridge Cambridge University Press. Searle, J. (1983). Intentionality. Particulars of of my my life. New York: Knopf. Skinner, B. F. (1976). Particulars shaping of of a behaviorist. New New York: Knopf. Skinner, B. F. (1979). The shaping A matter matter of of consequences. New York: Knopf. Skinner, B. F. (1983). A v., & & Shontz, F. C. (1980). Cocaine Cocaine users: A representative representative case approach. approach. New Spotts, J. V., York: Free Press. Personality: Basic Basic aspects and current current research. Englewood Cliffs, NJ: Staub, E. (1980). Personality: Prentice-Hall. Stewart, A. J., Franz, C C.,, & & Layton, L. (1988). The changing self: Using personal documents documents to study lives. Journal Journal of of Personality, Persormlity, 56(1), 41-74. Subjectivity in personality Stolorow, R. D., & Atwood, G. E. (1979). Faces in a cloud: Subjectivity personality theory. theory. New York: Aronson. Strozier, C. (Ed.). (1985). Self-psychology Self-psychology and and the humanities. New York: Norton. Strozier, c., C, & & Offer, Offer, D. (Eds.). (1985). The leader: Psychohistorical Psychohistorical essays. New York: Plenum Press. Sulloway, F. J. (1979). (1979), Freud, biologist biologist of of the mind: Beyond Beyond the psychoanalytic psychoanalytic legend. New York: Basic Books. lton in childhood. American Terman, L. (1917). The intelligence intelligence quotient of Francis Francis Ga Galton American Journal Journal of of Psychology, Psychology, 28, 209-215. Tomkins, S. S. (1981). The rise, fall fall and resurrection of of the study of of personality. Journal Journal of of Mind Mind and and Behavior, Behavior, 2, 443-452. Turkat, I. D. (Ed.). (1985). Behavioral Behavioral case formulation. formulation. New York: Plenum Press. Ullmann, L. P., & & Krasner, L. I. (Eds.). (1965). Case studies studies in behavior behavior modification. modification. New York: Holt, Rinehart & & Winston. psychopathic god: god: Adolf Adolf Hitler. New York: Basic Books. Waite, R. G. L. (1977). The psychopathic Psychological Review, Review, 20, Watson, J. B. (1913). Psychology as the behaviorist views it. Psychological 158-177. Weinstein, E., Anderson, J., & & Link, A. (1978). Woodrow Wilson's political personality: A A Political Science Quarterly, 93, 585-598. reappraisal. Political Lives in progress. progress. New York: Holt, Rinehart Rinehart & Winston. White, R. W. (1952). Lives study of of lives. New York: Atherton. White, R. W. (Ed.). (1963). The study

CHAPTER CHAPTER 22 SllIDYlNG STUDYINGLIvEs LIVES

69 69

White, R. (1972). The The enterprise enterpriseofofliving. living. New York: Holt, Rinehart Winston. White, R. W. W. (1972). New York: Holt, Rinehart && Winston. Wiggins, J. J. S., Renner, Renner, K. E., Clore, Clore, G. G. L., L., & & Rose, Rose, R. R. J. J. (1976). (1976). Principles Principlesof ofpersonality. personality. Wiggins, Reading, MA: MA: Addison-Wesley. Winter, D. D. G., & & Carlson, Carlson, L. L. (1988). (1988). Using Using motive motive scores scores in in the the psychobiographical psychobiographical study study Winter, of Journalof of Personality, Personality,56(1), 56(1),75-104. 75-104. of an an individual: individual: The The case case of of Richard Richard Nixon. Nixon. Journal Wittels, personality, his his teaching, and and his school. New New York:York: Wittels, F. F. (1924). (1924). Sigmund SigmundFreud: Freud:His His personality, teaching, his school. Dodd, Mead. Dodd, Mead. Wolpe, Wolpe, J., J., & & Rachman, Rachman, S. S. (1960). (1960). Psychoanalytic Psychoanalytic "evidence": "evidence": A A critique critique based based on on Freud's Freud's 131, 135-148. case case of of Little Little Hans. Hans. Journal Journalof ofNervous Nervousand andMental MentalDisease, Disease, 131, 135-148.

This Page Intentionally Left Blank

PART II PARTH

CONCEPTUAL AND MEASUREMENT ISSUES IN PERSONALITY

This Page Intentionally Left Blank

j j j j j j j j j j j j j j

CHAPTER CHAPTER 33

UNITS OF ANALYSIS FOR THE THE DESCRIPTION AND EXPLANATION OF PERSONALITY A. JOHNSON JOHN A. PENNSYLVANIA PENNSYLVANIA STATE STATE UNIVERSITY, UNIVERSITY, DuBOIS DUBOIS CAMPUS CAMPUS

1 I. THE T H E NEED N E E D FOR FOR UNITS U N I T S OF OF ANALYSIS ANALYSIS Every science has has a nomenclature that describes and and defines defines its domain of study. Nuclear physicists talk about subatomic particles; chemists analyze molecules and and compounds; and evolutionary biologists ponder genes, populations, and species. In what sort of language do and explain do psychologists describe and explain personality? personality? The list of the the units of analysis used by personality personality psychologists psychologists is wildly diverse: cognitive styles, complexes, current concerns, dispositions, folk folk concepts, goals, instincts, interests, motives, needs, personal projects, plans, personal con constructs, strivings, sentiments, themas, types, and values are are aa few few (cf. (cf. D. D. M. M.Buss Buss& & Cantor, 1989; Emmons, 1989, this volume, chap. 20). The length and complexity of of this list suggest that our problem problem concerns organizing and simplifying simplifying our potpourri of of conceptual units rather than choosing among them. This chapter suggests that the of the apparent diversity of the proposed units of analysis masks a unity underlying them and that this unity is captured by the the term trait (Allport, and suggest that the meaning of traits and (AUport, 1937). I begin by examining the understanding of personality. the trait notion is virtually required for a systematic understanding the trait concept and and suggest that these criticisms are I then review criticisms of the not always well founded. HANDBOOK OFPERSONALITY PERSONAUTY PSYCHOLOGY H ANDBOOK OF PSYCHOLOGY

7373

COPYRIGHT © 1997 1997 IY BY ACADEMIC ACADEMIC PlEss PRESS.. CoPYII GHT C A LRIGImi L RIGHTS REPRODUOION IN ANY FORM RESERVED. ALL OF OF REnODucnON IN ANY FORM RESERVED.

74 74

JOHN JOHN A. A. JOHNSON JOHNSON

Next I distinguish distinguish between phenotypic and genotypic traits, and discuss the implication implication of of this distinction distinction for for the twin tasks of of description description versus explanation explanation and for the validity of self-assessments versus observer assessments. This leads to for of self-assessments the view that, in the process of of assessment, the genotypic/phenotypic genotypic/phenotypic distinction distinction disappears. Finally, I consider whether a special unit of consider whether of analysis is necessary to of individuals, and whether type language might be as describe the uniqueness of useful as trait language. useful

II. II. TRAITS AS UNITS OF ANALYSIS of the Trait Concept A. The Meaning of

Traits are consistent patterns of of thoughts, feelings, or actions that distinguish distinguish people from one another. The reader reader should note three features features of of the foregoing foregoing definition. definition. from refer to thoughts, feelings, or behavior. This point is often often overlooked overlooked First, traits can refer define personality personality only in terms of of consistent behaviors. Sec Secby psychologists who define ascription invariably invariably involves comparisons between people. If If I say that ond, trait ascription obsessive-compulsive, I mean that he seems to have more intrusive thoughts thoughts Jones is obsessive-compulsive, feelings and demonstrates more ritualistic behavior behavior than people in general. and guilt feelings for traits to distinguish distinguish people from from one another, they must must display some Third, for consistency. If If Jones's obsessesive obsessesive experiences experiences and compulsive compulsive activities distinctive consistency. frequent than those of of the general diminish to the point that they are no more frequent from people in general. population, then they would no longer distinguish Jones from required for for science of of personality, because any science Traits seem to be required consistent patterns (Hanson, 1958). Imagine trying involves detecting and explaining consistent construct a science of of chemistry if elements and compounds did not possess to construct properties—if sodium chloride were only sometimes water soluable. If If people stable properties-if scientifically. had no stable properties (i.e., traits), they could not be studied scientifically. B* De The Situationist Challenge to the Trait Concept B.

Although Although a science of of personality personaUty seems to depend depend on studying consistencies, the of traits has been been questioned questioned from from World War I to the present (see existence of Kenrick & & Funder, 1988, for for a review). It is primarily writers with a behaviorist behaviorist Kenrick orientation who doubt doubt the existence of of traits. They claim that behavior behavior depends depends orientation on social contexts and that consistencies consistencies are an artifact artifact of of a person being in similar Stimulus-response behaviorism behaviorism (Hendrick, 1977) seeks to situations (Ross, 1977). Stimulus-response identify stable S -» R laws rather than stable properties of of persons. --+ R rather identify behavior depends on situational situational cues seems congruent congruent with The claim that behavior differently at wedding receptions and funerals. We We common sense. We behave differently differently in the different different roles we play, for for example, as employee, spouse, behave differently from everyday experience experience are consistent with the view or parent. These examples from confirmed this that social situations determine our behavior. Behaviorists have confirmed

CHAPTER 3 CHAPTERS

UNITS UNITS OF OFANALYSIS ANALYSIS

7S 75

common sense observation observation countless times in their laboratories, where they show that people's behavior behavior responds to experimental treatments or manipulations manipulations further evidence for for the power of of situations, trait critics cite (A. H. Buss, 1989). 1989). As further presumably reflect reflect traits) Mischel's (1968) claim that personality test scores (which presumably correlate higher than .30 with behavioral criteria. seldom correlate

C. Responses to the Situationist Challenge Despite the intuitive appeal of of the behaviorist argument, their demonstrations fail fail to disprove the existence of traits for for five reasons. 1. / . Reliable Reliable Situational Situational lnJIuence Influence Requires Requires an Enduring Enduring Capacity fluenced Capacity to to Be Be In Influenced

reUably control behavior, then people must have a capacity to First, if situations reliably respond to situational cues, that is, the trait trait of of being responsive to situations. This respond point has been recognized recognized even by the most vocal critics of of traits (e.g., Ross, 1977, "For instance, in accounting for for Jack's purchase of of a house house the 'situational' p. 176): "For explanation (i.e., 'because it was so secluded') implies a disposition on the part of of explanation particular actor actor to favor favor seclusion." this particular 2. Individuals Individuals Respond Respond Diff Differently to the Same Same Situation Situation 2. erently to

Second, even in the most (allegedly) powerful powerful situations, such as the Asch perceptual perceptual conformity paradigm, people's responses to the situation will differ differ as a function function conformity of of personality (Barron, 1953). This point has also been acknowledged acknowledged by trait critics (Ross & Nisbett, 1992). Having a Trait Trait Means Means Reacting Reacting Consistently Consistently to to the the Same Same 3. Having Situation, Not Not Different Different Situations Situations Situation, Third, the fact fact that a person is inconsistent across different different situations is completely completely irrelevant to the validity of of the trait concept. This point is often often overlooked overlooked even by trait trait defenders defenders (e.g., Kenrick Kenrick & Funder, 1988), who state that traits imply "cross ''crosssituational situational consistency." I regard the issue of cross-situational consistency as a red herring. Salt need not dissolve in benzene before we describe it as water soluable; need not exhibit identical behaviors in different different environments environments likewise, persons need before we can say they have traits. The trait concept implies consistent reactions reactions before to similar similar situations over time, not consistent reactions across across different different situations. characterized by a trait automatically automatically implies the relevant type of of situation; Being characterized for example, cooperative cooperative means consistently complying with with reasonable reasonable requests requests for indiscriminately complying with others' wishes on every oc oc(Alston, 1975), not indiscriminately casion. 4. Having Having a Trait Trait Does Does Not Not Mean Mean Your Your Reactions Reactions Are Are Absolutely Consistent Consistent Absolutely

for behavior behavior to be consistent consistent across time (temporal continuity) it need not Fourth, for be identical in every every relevant relevant situation. For example, a lascivious lascivious person person takes takes

76 76

JOHN JOHN A. A. JOHNSON JOHNSON

advantage of opportunities for for sex more often often than the average person, but does not attempt to copulate with every person he or she meets (Johnson, 1993). This point point is missed by those who criticize trait measures for for not predicting behavior behavior in situation.^ Proponents of traits never claimed that their mea meaa single experimental situation.1 predict trends trends in behavior behavior over over time sures could do this. Rather, trait measures predict A trait score is like a batting average. Knowing that a baseball (Epstein, 1983). A predict what what he or she will do in a particular particular player hits .300 does not allow you to predict performance over over the course course of of a season. at-bat, but does predict his or her performance S. 5. Behavioral Behavioral Inconsistency Inconsistency Does Does Not Not Rule Rule Out Out

Inner Inner Consistency Consistency

Fifth, the lack of consistency in behavior behavior over time does not rule out the existence of emotional or cognitive traits. An emotional trait, for for example, may be rarely of expressed because of of conflicts with other emotional traits. A consisA person might consis attend parties, but might do so infrequently infrequently due to a stronger tently desire to attend of opportunity prevent the consistent desire to work (Alston, 1975). Lack of opportunity can also prevent of emotional and cognitive traits. A courageous courageous person—that someone expression of person-that is, someone certainly overcome overcome fear fear and act decisively decisively in an emergency-may emergency—may rarely who could certainly faced with emergencies in which he or she could actually be heroic. Cognitive be faced not be de depersonality traits (e.g., persistently attending to details) likewise might not tected unless a person is given the opportunity to demonstrate the trait. Finally, Alston (1975) also points out that needs may be (like hunger) cyclical. D. D* Pbenotypic Phenotypic versus Genotypic Traits Two forms forms of of traits have been traditionally traditionally distinguished distinguished (Allport, (AUport, 1937; MacKin MacKinnon, 1941): outer outer traits that can be directly observed observed (behavioral (behavioral traits) and inner traits (emotional and cognitive traits) that must be inferred. inferred. Meehl (1956) borrowed borrowed from from biology the terms phenotypic phenotypic and genotypic genotypic to distinguish distinguish the two types of of traits. Others (Weimer, 1974; Wiggins, this volume, chap. 4) use a linguistic meta metaphor. Behavioral traits are analagous to the surface surface structure of of a sentence sentence whose whose meaning is indeterminate, whereas inner traits represent represent the intention intention or deep structure that disambiguates (explains) the meaning meaning of of behavior. Two common assumptions of traits are worth assumptions about the two types of worth noting. The first first concerns the view that outer traits are descriptions of of behavior behavior that that need explanation, explanation, whereas inner inner traits are the causes or reasons reasons that that explain the outer 1990a; Wiggins & & Trapnell, this volume, chap. 28). traits (Alston, 1975; Johnson, 199Oa; privileged access to outer outer traits, and actors Second, observers are assumed to have privileged turn, is assumed to affect affect the validity of of observer to inner traits. This access, in tum, Traitmeasures measurescan canalso alsofail failtotopredict predictbehavior behaviorbecause becausethey theylack lackreliability reliabilityor orvalidity validity(Block, (Block, 1*Trait 1977). When When reliable, reliable, valid valid measures measures are are used, used, Mischel's MischePs alleged alleged .30 .30 ceiling ceiling is is easily easily broken broken (Hogan, (Hogan, DeSoto, & & Solano, 1977).

CHAPTER CHAPTER 33

UNITSOF OFANALYSIS ANALYSIS UNITS

7777

ratings versus self-ratings self-ratings of of the two types of of traits. Although Although both assumptions assumptions have merit, they are only partially correct, as I argue below. E. Traits as Descriptions and as Explanations

To explain behavior in terms of of traits-e.g., traits—e.g., Joe hit Fred because because Joe is aggres aggressive-is sive—is sometimes criticized as description rather than a genuine explanation explanation 1984). Trait critics often (Weimer, 1984). often note the apparent apparent circularity in statements such as "He acted aggressively because he is [behaviorally] aggressive" (Rholes, New Newman, & & Ruble, 1990, p. 371). However, to explain a single single behavioral act as an exemplification exemplification of of a behavioral trait is a valid account of of an act (Wiggins, this 4)^. When When II say say that that Joe Joe hit hit Fred Fred because because Joe Joe is is an an aggressive aggressive person, person, volume, chap. 4)2. impUes that Joe's behavior behavior is not unusual for for him and therefore therefore requires no this implies further explanation.3 explanation.^ Many nonscientists nonscientists are satisfied satisfied with explanations such as further "that's typical" or "that's his or her nature" (Young, 1975). People generally seek deeper explanations only for for behaviors that are out of of character or unusual. deeper Psychologists, on the other hand, also want to explain normal behavioral traits. One approach to this is to hypothesize hypothesize inner emotional and cognitive traits that generate the behavioral behavioral traits. In many accounts, these inner traits form form part of of the generate basic level of of theoretical explanation, as illustrated in the following (see Weimer, Weimer, 1984): behavioral act, "Joe hit Fred" Fred" it behavioral behavioral trait, aggressiveness iT inner inner trait, aggressive feelings feelings

fact fact it law iT theory

explained nonexplainer nonexplainer iT explained explained explainer iT nonexplained nonexplained explainer

In the Weimer Weimer (1984) model, "explained nonexplainers" are single events events that require an explanation but themselves explain nothing. Isolated behavioral behavioral acts fit this category. "Explained "Explained explainers" are empirical regularities or laws used to explain single events, but require an explanation explanation themselves. Common Common behavioral behavioral traits fall into this category. "Nonexplained "Nonexplained explainers" are metatheoretical metatheoretical assump assumptions about nonverifiable, structural structural entities that explain empirical laws but are are themselves beyond explanation. Unobservable, psychic structures represent a type of of metatheoretical metatheoretical primitive. The notion that emotional emotional and cognitive traits underlie and explain behavioral behavioral common sensical and "familiar "familiar to all of of us since childhood" (Alston, traits is actually common Behavioraltraits traitscan canalso alsobe beinvoked invokedto tohelp helpexplain explainthe thereactions reactionsof ofan anaudience audienceto tothe theactor, actor, 2^Behavioral that is, is, as as part part of of aa social-psychological social-psychological explanation. explanation. For For example, example, the the poor poor morale morale and and unproductivity unproductivity that of aa team team might might be be attributed attributed to to the the ineffective meffective behavioral behavioral traits traits of of the the leader. leader. of ^Alternatively, Alternatively,what whatIImight mightreally reallymean meanwhen whenIIsay, say,"Joe "Joehit hitFred Fredbecause becauseJoe Joeisisan anaggressive aggressive 3 person," isthat thatJoe Joehas hashostile, hostile,aggressive aggressiveurges urgesororfeelings. feelings.IfIfthis thisisisthe thecase, case,I am I am invoking "aggressive" person," is invoking "aggressive" as aa motivational motivational concept concept rather rather than than aa behavioral behavioral trait. as trait.

78 78

JOHN JOHN A. A. JOHNSON JOHNSON

1975, P p 24). An example from Johnson (1990a) illustrates this. Suppose that Mary, a therapist, consistently treats her clients in an empathic manner. She uses reflective reflective listening to promote accurate understanding, appears attentive and interested in her clients' problems, and often often reassures and supports them. The trait empathic empathic aptly describes her counseling style. Why is Mary consistently empathic? A A common sense explanation would refer to Mary's desires (emotional traits) and beliefs (cognitive traits). Presumably, Mary wants to promote her clients' psy psychological well-being. There are many ways to say this: Mary desires their well wellbeing; she values their well-being; she has established their well-being as a goaL goal. The precise term is unimportant-what unimportant—what counts is that Mary is motivated to promote her clients' well-being. Being motivated to achieve goals involves emotional pro processes (see Averill, this volume, chap. 21). Promoting the well-being of of her clients makes Mary feel good whereas failing to achieve this goal would make her feel bad. Mary's motives, desires, or goals provide only half of the explanation, however. We also need to know about her beliefs concerning how she can achieve her goals. empathically will have positive outcomes for Mary believes that treating her clients empathically them. In short, Mary behaves empathically because (a) she wants to promote the of her clients, cHents, and (b) believes that behaving empathically will help her well-being of enhance their well-being. Should this seem too simple, I might note that behavioral traits often serve several goals simultaneously. Mary's empathic behavior may satisfy not only her need to promote her clients' well-being, but also her need to be liked and admired by her clients. She might also believe that her empathic style will lead to a good performance evaluation and a larger paycheck. Furthermore, some behavioral traits may be habits that once promoted goals but now are no longer useful. Or, if the depth psychologists are correct, behavioral traits may express unconscious desires. Depth psychologists also suggest that con conflicting motives can become fused and lead to a compromise activity not clearly linked to any one goal. Ethologists talk about displacement displacement activities, (e.g., eating, self-grooming) self-grooming) that serve to relieve tension when conflicting conflicting instincts arise. Thus, there is not always a simple one-to-one correspondence between behavioral traits and underlying motivational or cognitive traits. Even if one could identify identify all of the relevant motives and beliefs underlying a behavioral trait, some would regard this explanation as incomplete. One might further inquire into the origin of of the motives and beliefs. Why Why does Mary want to be a therapist rather than a truck driver? How did she conclude conclude that Rogerian therapy is more effective effective than Freudian therapy? The answers to these questions can be found in personality development (e.g., Eder & & Mangelsdorf, Mangelsdorf, this volume, chap. 9) and in evolutionary psychology (e.g., D. Buss, this volume, chap. 13; A. Buss, this volume, chap. 14). Developmental and evolutionary explanations explanations complement intentional explanations (Wakefield, 1989). Some psychologists feel feel that it is insufficient insufficient to explain behavioral traits in of goals and beliefs for yet another reason: "Scientific" ''Scientific" explanations should terms of transcend common sense (McCrae, Costa, & & Piedmont, 1993). Theoretical physics

CHAPTER 3 CHAPTERS

79 79

UNITS OF OFANALYSIS ANALYSIS UNITS

is often often presented presented as a science that contradicts our common common sense assumptions assumptions of objects objects and the absolute nature of of time and space. From From this about the solidity of important concepts in personality psychology are not well perspective, the most important represented in ordinary ordinary language. McCrae (1990) proposes openness openness to experience experience represented of such a concept. as an example of Hofstee, de Raad, and Goldberg (1992) warn us, however, about the dangers of of stepping out of of the area of of common language: "There is nothing against this advice, as long as the pertinent pertinent outcomes do not have to be communicated communicated in words" (p. 162) 162). Cattell transcended transcended ordinary language with terms such as Harria, Harria, Presemsia, Alaxia, and Protension; ultimately these terms had to be translated into common trait language to be useful useful to practitioners (IPAT Staff, 1979). In summary, most most psychologists regard "outer" "outer" (behavioral) traits as descrip descriptions that need need explanation, and and they they assume that "inner" (emotional and cognitive) traits generate and therefore therefore explain outer traits. Behavioral traits or consistencies consistencies determined by the interaction of several emotional and cognitive traits. may be determined personality psychologists, in order order to provide deeper deeper explanations explanations of of inner Some personality genetic and physiological bases, developmental developmental histories, traits, inquire into their genetic evolution of of the human species. Whether Whether a scientific scientific conception conception and roots in the evolution of traits needs to go beyond beyond ordinary trait language language is a matter of of current debate. of .

m i n . ISSUES ISSUES IN IN mE THE ASSESSMENT ASSESSMENT OF OF TRAITS TRAITS A. Genotypic and Phenotypic Traits from the Perspective of Self and Others the Self Self-ratings of of traits correlate substantially substantially (rs in the .4-.6 range-see range—see Johnson, Johnson, Self-ratings of the same trait made by others. Nonetheless, in individual 1994) with ratings of self-ratings sometimes disagree with ratings by other persons. This raises an cases, self-ratings When self-ratings self-ratings disagree with ratings by others, whom are we to to interesting issue: When The chapters in this handbook handbook by Funder Funder and Colvin (chap. 24) and and Robins believe? The John (chap. 25) discuss this issue in some detail. This section addresses the and John of ratings of of genotypic genotypic versus phenotypic traits made by the self self and others. accuracy of / . Actors Actors Can Can Directly Directly Experience Experience Their Their Own Own Inner Inner Traits, Traits, 1. but Observers Observers Must Infer Others' Others^ Inner Inner Traits Traits but Must Infer often assumed to affect affect the validity of of personality personality The outer/inner trait distinction is often self-ratings versus ratings by other persons. Self-assessment Self-assessment of of geno genoassessment by self-ratings other-assessment of of these traits because typic traits is potentially more valid than other-assessment infer people may directly experience their own inner traits whereas observers must infer from verbal reports and nonverbal behavior. For example, anxious individuals them from of their anxiety level, whereas observers must infer infer their anxiety should be aware of from signs such as a furrowed furrowed brow, sweaty palms, tremors, and hesitant speech speech from Kenrick and Stringfield Stringfield (1980) report that personality personahty scores are more validvalid— style. Kenrick

80 80

JOHN A. A. JOHNSON JOHNSON

i.e., self-ratings correlate more highly with other other ratings-for ratings—for individuals who openly express their traits (see (see also Funder Funder & & Colvin, this volume, chap. 24). Funder Funder and Dobroth (1987) found found interjudge interjudge agreement agreement to be higher for traits related related to be Dobroth highly visible. 2. 2. Observers Observers Can Can Directly Directly Experience Experience Others' Others^ Outer Outer Traits, Traits,

but Actors Outer Traits Traits but Actors Must Must Infer Infer Their Their Own Outer

Individuals may but they cannot directly may have direct access to their inner traits, but observe their own own behavior. This implies that external observers may may provide a more valid assessment of a person's phenotypic phenotypic traits. For example, people are poor judges of how how charming they are, are, but this outer outer trait is obvious to typically poor s e l f suggests that our an observer. Cooley's (1902) concept of the "looking glass self" understanding of our own personality is determined determined by the feedback feedback reflected reflected from from understanding around us, the perspective of others to others around us, at least until we are able to take the appear to them (Mead, 1934). Whether Whether we listen to a real or an imagine how how we appear our outer traits is indirect and and inferential. inferential. imagined audience, our knowledge of our Defensiveness Hinders Accurate Accurate Self-Assessment Self-Assessment 3. Def ensiveness Hinders Although the relationship between outer/inner outer/inner traits and the validity of self- and other-assessment intuitively compelling, it is incomplete. Individuals may other-assessment may seem intuitively self-deceived about their inner traits (Averill, this vol volbe unaware, mistaken, or self-deceived ume, chap. 21; 21; Paulhus Faulhus & & Reid, 1991; 1991; Paulhus, Fridhandler, & & Hayes, this volume, chap. 22; 22; Robins & & John, this volume, chap. 25). 25). Aspects of personality that are unknown to the individual but known to others are found found in the "blind area" of unknown of the Johari window (named after after Joe Joe Luft Luft and Harry Ingram; see see Luft, Luft, 1966). The Johari window is illustrated illustrated in Figure 1. provide 1. In the case of blind spots, observers provide more accurate assessments of inner inner personality than the person observed. observed. Individuals may may also be reluctant to describe all their inner traits as they actually perceive them, preferring preferring instead to describe traits that they would like others to believe they have. Paulhus and Reid (1991) call this process impression management. Inner traits that are not described to others are the "secret area" of of the Johari window.

Known to to Self Self Known Known to to Others Unknown to Others Others

Public PubUc

Area Area

1Secret Secret Area Area

FIGURE FIGURE 11 The The lohari Johari window. window.

Unknown to to Self Self Unknown Blind Area Area Unconscious Area Area

CHAPTER 3 CHAPTERS

UNITS UNITS OF OFANALYSIS ANALYSIS

81 81

The Johari window and Paulhus' research on social desirability desirabiUty responding responding suggest that self-reports self-reports of inner traits may may be more accurate than than observer observer assess assesssuggest of self-deception self-deception the converse converse is true. We may also have ments, but in the case of "secret areas" that we choose not to reveal. 4. Self-Observer Self'Observer Agreement Agreement Requires Requires a Common Common Understanding f Language Understanding o of Language

Accurate knowledge of one's inner traits and a willingness wilUngness to share this with others insufficient to ensure agreement agreement between between self-description self-description and description description are still insufficient of the self self provided provided by others. Full agreement agreement also depends depends on the actor and of for communication. For observer using the same semantic and pragmatic rules for person rating himself himself or herself herself for for thoughtfulness thoughtfulness will provide an inap inapexample, a person rating if he or she believes thoughtful thoughtful to mean mean considerate but the rater considerate but propriate rating contemplative. Misunderstandings Misunderstandings of of the meaning of of trait believes the term means contemplative. often than researchers may realize (Goldberg terms occur more often (Goldberg & Kilkowski, 1985). questionnaires rather When we assess personality with questionnaires rather than rating scales, we encounter pragmatic as well as semantic misunderstandings. Pragmatic rules rules are are encounter about meaning that that can can vary across subcultures who who implicit social conventions about share the same language. The impact of pragmatics on measurement measurement validity is virtually unexplored, but I can cite two illustrations here. that item 77 on the California InvenFirst, Johnson (1997) notes that California Psychological Inven tory (CPI; Gough, 1987), "When I get bored bored I like to stir up some some excitement," is tory reUeve their boredom boredom by doing something illegal. a phrase used by delinquents who relieve For this reason, Johnson (1997) interprets interprets the item as reflecting reflecting (lack (lack of) of) conscien conscienFor apparently prefered prefered a more literal interpretation interpretation tiousness. McCrae et al. (1993) apparently reflect extraversion. We do know how a typical and they judged the item to reflect do not not know respondent interprets this item. respondent narrow, literal meaning meaning of of words We do know that people who focus on the narrow, often and therefore therefore respond often miss the broader broader social implications of of personality items and inappropriately person who answers inappropriately (Johnson, 1993). A A punctual and conscientious conscientious person "false" to the item, "I am never late to appointments," because he or she thinks never means literally never never has missed the point of of the item. Pragmatic rules-i.e., rules—i.e., our our social conventions of of language use-allow use—allow punctual people to say, "I am never late," because we know this really means, "I am a conscientious person person who is rarely late." Similarly, pragmatic rules suggest that an inteUectual intellectual person should answer "true" to CPI item 152, "I read at least ten books a year," even if he or she reads only three books a year. The point of of these examples of of pragmatic ambiguity ambiguity is as follows. People perfectly aware of their actual dispositions to be delinquent, extraverted, extraverted, might be perfectly conscientious, intellectual, and so forth, and also be quite willing to acknowledge these dispositions through the items discussed above. Nonetheless, their responses information only when the test taker and test to these items will convey valid information constructor interpret the item response in the same way. constructor

82 82

JOHN JOHN A. A. JOHNSON JOHNSON

5. 5.

Observer Ratings Ratings oof Outer Traits Traits Are Are Valid Valid Almost Almost Observer f Outer by Definition Definition

Whereas observer ratings of another person's inner traits are only rarely more valid than self-ratings, observer ratings of outer traits are usually more accurate than self-ratings of outer traits. An exception might be a case where an observer is prejudiced prejudiced against the person he or she is rating. But on the whole, outer traits are—almost by definition-whatever definition—whatever impressions an actor makes on observers. If are-almost If an individual is perceived by others as a loudmouth, then by definition that person is a loudmouth. This implies that observer ratings constitute an "ultimate criterion" of sorts for for validating self-reports self-reports of outer traits (Hofstee, 1994). Outer traits are social constructions of reality (Berger & & Luckmann, 1966). "actual'* traits that are assigned to a From a social constructivist perspective, the "actual" majority of observers believe should be assigned. Hogan person are whatever the majority reputation. AA person's person's refer to the social consensus as a person's reputation. and Briggs (1986) refer view of his/her own reputation may be as correct but not more correct than his/ her reputation as constructed by the social group. To summarize, self-ratings of one's inner traits tend to be more valid than observer ratings of those traits except in cases of blind areas (self-deception) or secret areas (impression management). Conversely, observer ratings of one's outer traits (reputation) are almost always more valid than self-ratings of outer traits. People must have highly developed perspective-taking skills to describe accurately & Hogan, 1978). the way they appear to others (Mills & The fact that individuals may not provide valid self-ratings in the blind and secret areas presents problems for those of us who wish to assess personality with questionnaires and self-rating self-rating scales. How does one address this problem? One possible solution is to identify identify subtle items, that is, items with less-than less-thanobvious psychological significance or implications. Different Different responses to subtle items covary empirically with individual differences differences along a trait dimension for reasons unknown to the test taker and sometimes even the test constructor. Subtle items can be found found through brute, dust-bowl empiricism. Unfortunately, research has demonstrated demonstrated that subtle items are almost invariably not valid (Johnson, 1993). Valid self-assessment self-assessment of of blind and secret areas remains a challenge for for personality researchers who are attempting to build better mousetraps. 6. ment ofof Surroundsthe theAssess Assessment 6. Controversy Controversy Surrounds

UnconsciousTraits Traits Unconscious

The Johari window contains one more pane we have not discussed: the unconscious area. Unconscious traits are the foundation foundation of psychoanalytic theories, but are often often ignored by mainstream personality psychologists working within the cognitive 1979; Weinberger & Zeitgeist (Hogan, 1979; «fc McClelland, 1990). 1990). Some might argue that unconscious traits, because they are unseen by the self or others, are not amenable to scientific study. Others would counter that unconscious traits are like nuclear particles. We cannot see these particles, but they leave traces in cloud chambers

CHAPTER CHAPTER 33 UNITS UNITSOF OFANALYSIS ANALYSIS

8383

and certainly certainly have a palpable impact on us. Unconscious traits similarly cannot cannot be of their activity activity and have an impact on us. The The directly observed, but leave traces of question is whether whether we can devise the equivalent equivalent of of a cloud cloud chamber chamber or Geiger Geiger question counter to assess unconscious traits. counter Although clinicians often often rely on their their own intuition intuition to access unconscious unconscious Although traits (Reik, 1948), some modern researchers claim that that projective projective tests constitute a cloud chamber for the unconscious. Projective tests are simply stimuli (inkblots, photographs, sentence fragments) fragments) with open-ended open-ended response options. Rather Rather than respondents can say or or write write as little or as much as they responding true or false, respondents of rules to yield an evaluation evaluation like. Their responses are then scored according to a set of of the respondent's level on various unconscious unconscious needs or motives. of An influential influential review paper paper by Entwistle (1972) cast serious doubts on the reliability and validity of of projective projective tests. Undaunted, McClelland McClelland and his colleagues reliability (McClelland, 1980; Weinberger Weinberger & McClelland, 1990) marshalled marshalled further further evidence for for the reliability, validity, and utility of of projective projective measures. McClelland McClelland also replaced the usual psychoanalytic framework framework for for projective projective testing with an ethologi ethologiframework. Summarizing research on projective projective measures in applied settings, cal framework. about as valid as objective measures. Hogan (1991) concluded that these tests are about Scores on projective projective tests tend not to correlate correlate with scores on objective tests measuring the same construct (Weinberger (Weinberger & McClelland, 1990). This finding led McClelland McClelland (1980) to assert assert that projective and objective objective tests measure measure two different different of traits. Specifically, he suggested that projective tests tap a more primitive, kinds of affect-laden type of of trait, whereas whereas questionnaires assess a more more biologically based, affect-laden of trait. The Weinberger Weinberger and McClelland McClelland (1990) chapter cognitive, symbolic type of projective and objective objective measures predict different different reviews studies indicating that projective types of of activities. activities. types Is McClelland correct to argue that qualitatively qualitatively different different kinds of of traits exist different kinds of of measures to assess these types of of traits? That and that we need different is the question addressed next. Questionnaires B. Trait Measurement through Questionnaires / . Do Do Diff Different Questionnaires Measure Measure Diff Different Kinds 1. erent Questionnaires erent Kinds off Traits? Traits? o Preference Schedule (EPPS; Edwards, 1959) purports to The Edwards Personal Preference "folk concepts," the measure Murray's needs, Gough's (1987) CPI allegedly assesses "folk Guilford-Zimmerman Temperament Temperament Inventory (GZTS; Guilford, Guilford, Zimmerman, & Guilford-Zimmerman Guilford, 1976) obviously aims to measure measure temperaments, and and the Myers-Briggs Myers-Briggs Guilford, Indicator (MBTI; Myers & McCaulley, 1985) is supposed capture the supposed to capture Type Indicator functions in Jung's psychological types. Do Do these four four tests actually mea meacognitive functions four different different kinds of of traits? sure four Costa (1989; McCrae, 1989; McCrae et al., According to McCrae and Costa aI., 1993; question is clearly no. McCrae Piedmont, McCrae, & Costa, 1992) the answer to this question

84 84

JOHN JOHN A. A. JOHNSON JOHNSON

and Costa have demonstrated demonstrated that the scales on these inventories assess the same same five traits measured measured by their own NEO-PI (Costa & & McCrae, 1992): 1992): Neuroticism, Neuroticism, Extraversion, Openness Openness to Experience, Agreeableness, and Conscientiousness. In fact, fact, McCrae McCrae and Costa Costa have ingeniously and systematically demonstrated demonstrated that virtually "Big Five" or Five virtually every every major major inventory inventory assesses some or all of the *'Big FiveFactor Factor Model (FFM) traits. A A look at actual items on these inventories also suggests that they are not not different kinds of of traits. Consider Consider the following following items: "I like to plan measuring different details of of any work work that I have to undertake." "I always see to to and organize the details that my work work is carefully carefully planned planned and and organized." "You "You like work that that requires requires it that attention to details." "When "When you start a big project project that is due in a considerable attention of week, do you (a) take time to list the separate things to be done and the order of plunge in?" "I like to follow follow a strict routine routine in my work." Can doing them, or (b) plunge you tell which item measures a need, which measures a folk folk concept, and which measures a cognitive style? planful and serious-minded serious-minded approach to work. But All the items indicate a planful anyone unfamiliar unfamiUar with these these inventories would likely be unable to tell that that they from the EPPS, CPI, GZTS, MBTI, and NEO-PI, respectively. The empirical empirical came from and semantic overlap overlap in the items across these instruments does not imply that and folk concepts, temperaments, and cognitive styles are identical concepts. concepts. needs, folk questionnaires seem to be measuring similar, if not identical, Nonetheless, the questionnaires constructs.^ 1t It the items reflect reflect patterns of of thoughts, feelings, or actions, one might constructs.4 as well use the generic term trait to describe what they measure. trait I think it is pointless to worry about conceptual distinctions between trait measured in identical ways or if scores from from different different measures measures constructs if they are measured criteria equally well). Kilkowski Kilkowski behave in similar ways (e.g., predict the same criteria (1975), for for example, provides an interesting six-page analysis of of the conceptual conceptual differences differences between Allport's traits and Murrays' needs. But he does not describe describe different different methods for for measuring measuring traits and needs. 2. Do Do Questionnaires Questionnaires Measure Measure Phenotypic Phenotypic or or 2. Genotypic Genotypic Traits? Traits?

A A closer closer look at the five items listed Usted above shows that two refer refer to actual planful planful behavior refer to a liking liking for for organization. Might it be important important to to behavior and three refer (outer behavioral) items from from genotypic (inner (inner cognitive or distinguish phenotypic (outer LOhr (1986) and Werner Werner and Pervin Pervin (1986) emotional) items? Angleitner, John, and L6hr report that different different inventories contain different different proportions of of phenotypic and report affect test validity, validity. genotypic items. They then assert that item characteristics may affect

^Emmons Emmons(this (thisvolume, volume,chap. chap.20) 20)also alsohas hashinted hintedatatthe thefutility futilityof oftrying tryingto todistinguish distinguishallegedly allegedly 4 different goal goal constructs constructs from from each each other. other. As As he he points points out, out, the the statements statements in in Table Table 22 of of his his chapter chapter different representing representing four four purportedly purportedly different different goal goal constructs-current constructs—current concerns, concerns, personal personal projects, projects, life life tasks, and personal strivings-are strivings—are very similar.

CHAPTER 3 UNITS CHAPTERS UNITS OF OFANALYSIS ANALYSIS

8S 85

but neither research team examined whether phenotypic phenotypic or genotypic items were superior for predicting non test criteria. nontest Johnson (1993a) examined the ability abiUty of of phenotypic and genotypic items on the CPI to predict acquaintance ratings. He found that phenotypic items predicted extraversion ratings better than genotypic items. In domains other than extraversion, however, reference to outer or inner traits was not related to validity. I think genotypic versus phenotypic wording is unrelated to validity because genotypic tendencies normally find phenotypic expression. For example, people who want to eventually act in ambitious ways; conversely, people who act ambitiously get ahead eventually normally have ambitious motives. Thus, to endorse the genotypic item, "I have a is tantamount tantamount to to endorsing endorsing the the phenotypic phenotypic strong desire to be a success in the world," is item, "I do whatever I can to get ahead" (and vice versa). geBecause personality questionnaires simultaneously assess phenotypic and ge found it useful to think about personality in terms of a trait notypic traits, I have found self-presentational style. Self Selfconstruct that incorporates both levels of personality: self-presentational presentations are any behaviors (including responses to questionnaire items) guided by inner traits that create impressions in others. I believe that all noncognitive self-presentational styles (Johnson, 1981; 1981; Mills & Hogan, questionnaires assess self-presentational 1978). It does not matter whether the test is intended to measure moral reasoning & Hogan, 1981a), vocational interests (Johnson & Hogan, 1981b), atti atti(Johnson & & Rogolsky, 1981), or philosophical philosophical tudes (Johnson, Hogan, Zonderman, Callens, & & Overton, 1988); 1988); responses to these various world views (Johnson, Germer, Efran, & inventories create a distinctive impression on those who see the responses. Self-presentation of of traits that are already well known to everyone (the public Self-presentation of the Johari window) is direct and straightforward (Wolfe, 1993). In the public area of area of personality one can take item responses at face value. If someone endorses the item, "I am rarely late for appointments," we can accept that this person is punctual. I agree with Wolfe (1993) that personality assessment via questionnaires straightforward fashion in most cases, even in contexts such as person personproceeds in a straightforward nel selection (Hogan, 1991). self-presentation on questionnaires-like questionnaires—like social behavior in Nonetheless, self-presentation everyday life-contains life—contains both conscious, intentional and unconscious, unintentional elements. This implies that we cannot always always take item responses at their face value; ultimately we must determine, in an empirical fashion, what an item response means (Meehl, 1945). In particular, we cannot trust item content when blind, uncon unconscious, or secret aspects of of personality are being assessed. Clearly, persons cannot disclose blind or unconscious traits by endorsing items whose content describes the trait. An overly critical person who is unaware that he or she is overly critical cannot validly respond to an item such as "I am overly A critical." What is needed is an item that allows an observer to infer the trait. A more oblique item, such as "Spare the rod and spoil the child," might be endorsed by overly critical but unaware individuals. In everyday interactions, perceptive observers can make inferences about blind or unconscious areas from another of the tongue or body language; in questionnaires we depend upon person's slips of

86 86

JOHN JOHN A. A. JOHNSON JOHNSON

nonobvious empirical correlates of of personality personality items to reveal information information beyond beyond the manifest manifest content content of of the item.s item.^ During normal normal interaction interaction we know that people do not always tell the truth. We may, therefore, watch for for signs of of dissembling such as laughing nervously, averting the eyes, and touching one's face. Questionnaire items do not give us the dissembling, but over the years researchers have developed nonverbal cues to detect dissembhng, developed various techniques for for detecting intentional misrepresentation. Items on dissembling dissembling keys often often contain contain "unlikely '*unUkely virtues" (see Gough, 1987; Tellegen, in press)-they press)—they describe describe behaviors that are socially desirable desirable but unlikely to be literally true (e.g., "I have never told a lie"). Interpreting these dissembling scales is problematic, everyday life as well as on however, because people exaggerate their virtues in everyday questionnaires, and unlikely virtue scales predict nontest behavior (Johnson, 1990b). argued that that all personality personality questionnaires questionnaires measure measure a trait This section has argued self-presentational style. Do other assessment modes such as cognitive cognitive I call self-presentational projective tests also measure self selftests (Emmons, this volume, chap. 20) and projective presentational style? I beUeve format of of cognitive tests differs differs presentational believe so. Although the format from questionnaires, I believe that persons who endorse statements such as those from found in Table 2 of of Emmons's (this volume, chap. 20) chapter will create a distinctive found impression on others. Whether Whether scores from from the measures of of cognitive style and impression needs are empirically distinguishable from from each other and from from ordinary ordinary personality personaUty questionnaire scores remains to be seen. questionnaire projective tests. Gough Gough (1948) long ago argued that that I am less certain about projective responding to projective projective tests involves self-presentation. self-presentation. More recently, however, however, responding Weinberger and McClelland McClelland (1990) have argued argued that scores on projective projective tests are Weinberger uncorrelated objective tests because responses to the two types of of uncorrelated with scores on objective generated from from two different different parts of of the brain. I think think it is important important to tests are generated Unk units of of personality think much much more link personality analysis to neurophysiology, but I also think data will be required required to forge forge this link. Until then I will stand by my view that personality assessment involve self-presentation. responses to all forms forms of of personaUty self-presentation. IV. ALTERNATIVES TO TRAITS IV.

A. Units for Capturing Uniqueness Because each person is obviously unique, psychologists occasionally suggest that of analysis designed to capture the uniqueness of of we should use special units of individual personality. This position, called the idiographic idiographic approach, contrasts with individual nomothetic view that we should compare individuals with a common set of of units. the nomothetic idiographic-nomothetic issue concerns how detailed our our I believe that the idiographic-nomothetic rather than what kind of of units we use to describe personality. personality. descriptions are rather ^Sadly, Sadly,however, however,the thetrack trackrecord recordfor forsubtle subtleitems itemsisisvery verypoor poor(Johnson. (Johnson,1993b). 1993b).We Wesimply simply 5 subtle but but valid valid items. items. have not been very successful at designing subtle

CHAPTERS CHAPTER 3

UNITS OF OFANALYSIS ANALYSIS UNITS

87 87

Ordinary traits are perfectly perfectly capable of describing what is unique about about us. Consider Consider Ordinary of traits as patterns of of consistent thoughts, feelings, or actions that definition of the definition distinguish persons persons from from one one another. another. Traits, by by definition, definition, describe how we differ differ distinguish from one another, and the sum of of these differences differences defines our uniqueness. from objection of of idiographic psychologists to nomothetic trait trait I think the real objection description is that a limited Umited number number of of trait dimensions (e.g., the Big Five; see description III.B.l) fail to capture the richness and complexity of of a unique person. Section III.B.1) & McCrae, Indeed, the Big Five, even broken down into six facets each (Costa & everything about someone's personality. personaUty. But to think that 1992), cannot describe everything Five-Factor Model or any other other model of of personality personality can completely describe the Five-Factor 1962; Rosenblueth & & Wiener, Wiener, a person is to misunderstand models in science (Holt 1962; A useful useful model is, by definition, a simplification: simpUfication: it retains only the important important 1945). A features of of the infinitely infinitely complex domain it represents (Eckhardt, 1979). The The precise precise features number of important traits is still a matter of debate, but we cannot expect any of of number of personality. our limited models to capture every detail about personality.

B. Types versus Traits One final possible unit of of analysis in the study of personality is the type type construct. construct. of personality types is nearly as complex as personality traits, and I will The notion of of these complexities. Grant Dahlstrom (1972) has written a defini defininot examine all of tive monograph monograph on the meaning of type. I also recommend recommend articles by Gangestad Gangestad and Snyder (1985, 1991) and by Paul Meehl (1992). Rather than reviewing these of types that are most often often cited as distinguishing works, I will discuss two properties of from traits: their holistic character character and their discrete character. To anticipate types from my conclusion, I believe that, in practice, the trait and type concepts are actually almost indistinguishable. 1. Are Are Types Types Holistic? Holistic?

of the type concept (Hogan & & In previous writings, I have stressed the holistic nature of Johnson, 1981; Johnson Johnson & & Ostendorf, 1993). 1993). I conceptualized conceptualized types as constellations constellations of traits that naturally co-occured co-occured in persons. My metaphor metaphor for a type or patterns of compound composed of of simpler elements. Types, like compounds, was a chemical compound possess emergent emergent properties, that is, properties not found found in the traits (elements) taken by themselves-e.g., themselves—e.g., hydrogen hydrogen and oxygen oxygen do not resemble water. If types have emergent emergent properties, then the holistic addage, "the whole is more than the sum of of its parts," applies. In an important important paper, Mendelsohn, Weiss, and Feimer (1982; see also Weiss, Mendelsohn, & & Feimer, 1982) provide a persuasive empirical and conceptual argue arguement against the holistic conception of of types. If If their argument argument is correct, and it seems to be, properties of of types can be derived from from an understanding understanding of of the properties of of the traits that make up the type, not from from an emergent configuration configuration of trait properties. For For example, the type notion "extravert" can be broken broken down of of gregariousness, impulsivity, and excitement-seeking, and all that into the traits of

JOHN A A. JOHNSON

88

is predictable about extraverts can be traced to these traits. And so on for all the other type concepts. If type concepts are not holistic and we reconceptualize them as collections (rather than configurations) of traits, types become nearly synonymous with broad traits. Consider Considerthe the"Big "BigFive" Five"traits traitsassessed assessedby byCosta Costaand andMcCrae's McCrae's(1992) (1992)NEO NEOtraits. PI; each trait is decomposed into narrower facets. These facets can be further PI; decomposed into more specific thoughts, feelings, and behaviors described by indiindi metaphor that contrasts contrasts types types as as compounds with traits vidual items. So perhaps my metaphor as elements is misleading. It may be more accurate to say that types and traits vary on a continuum of breadth.

2. Are Types Discrete? Discrete? 2. The second alleged difference between types and traits is that types are discrete or discontinuous, whereas traits exist along a continuum of values. This may be true at a conceptual level, but at the level of assessment and application, this & de Raad, 1992). To apply typologies in a continuous distinction vanishes (Hofstee & manner, one need only describe a person's degree of resemblance (on as finely a graded a scale as one desires) to as many type constructs as one desires. 3. 3.

But Are Are Traits Traits Continuous? Continuous? But

At a conceptual level, traits exist along a continuum of values. However, at the level of measurement, a person's score on a trait questionnaire does not necessarily indicate the degree or amount of the trait possessed. In the words of Meehl and Hathaway (1946), "simply counting how many responses . . . have been made seems to be very crude; . . .. [our mathematical scaling procedures] should not us into supposing that we we are doing anything very close close to to what what the the physicist physicist mislead us does when he cumulates centimeters" (p. 557). In actuality, the number of keyed responses endorsed by a person yields a probability statement about whether whether the the positive positive or negative pole of the the trait trait concept concept applies to him him or or her. her. This point is most clearly seen in the case of scales constructed constructed empirically by by contrasting the responses of two criterion groups (e.g., schizophrenics versus normals). If one has a very high score, it becomes more probable that we should apply the (type!) label "schizophrenic." But a score at the midpoint does not necessarily mean that the person is moderately schizophrenic. This may be true, but technically an average score means that the probability of correctly labeling the person (as normal or schizophrenic) approaches zero. Average scores on scales constructed by rational means or internal consistency are also ambiguous. An An average score on a rationally constructed scale of, of, say, sociability may indicate or (b) uncertainty about the applicability either (a) a moderate amount of sociability or & Tice, 1988). of the labels "unsociable" or "sociable" (see Baumeister & In In practice, personality test test users users often often convert converttrait traitscores scoresinto intotype typecategories. categories. And, contrary to popular belief, one can do this without losing much information.

CHAFfER 3 UNITS CHAPTERS UNITS OF OFANALYSIS ANALYSIS

89 89

Hofstee and de Raad (1992) explain: Hofstee An optimistic estimate of the proportion of true trait variance of a personality An scale, scale, after after subtraction subtraction of of both both unreliable unreliable and and method method variance, variance, is is 0.5, giving giving a standard standard error error of of measurement measurement of of 0.7. So, So, aa standard standard score score would would have have to to be be below 1.4 or over + 1.4 to be significantly (p = .05) different from O. -1.4 +1.4 0. To trichotomize a population into extraverts, introverts, and neither accordingly would be be quite quite realistic realistic in in view view of of the the large large error error of of measurement, (p. 62) would measurement. (p. =

Real-world decisions about persons are almost always binary (yes-no) or categorical (friend-foe). This means that trait scores are usually transformed into categorical terms. Consider a programmer's task of deciding whether a particular personality description should be triggered in a computer-generated personality report (Johnson, 1996). Either the score is high (or low) enough to print the narrative of personality paragraph or it is not. Consider an employer hiring people on the basis of test scores. Either their scores are sufficiently sufficiently auspicious to hire the person or they are not. The same situation occurs when counselors decide what careers to recommend or which therapies to administer to clients. I conclude, then, that the trait concept, interpreted as a facet of a person's person's self-presentational style, serves as the best unit of analysis for personality research. self-presentational In applied settings, however, the real world forces trait continua to be treated as discrete types.

ACKNOWLEDGMENTS Part of of this this chapter chapter was was written writtenwhile whileIIwas wasaavisiting visitingresearch researchfellow fellowatatUniversitlit UniversitStBielefeld, Bielefeld, Part supported by by aafellowship fellowshipfrom fromthe theAlexander Alexandervon vonHumboldt-Stiftung. Humboldt-Stiftung.IIexpress expressmy mygratitude gratitude supported to Alois Alois Angleitner Angleitner and andRobert RobertWicklund Wicklundfor forhelping helpingtotoarrange arrangemy mystay stayand andtoto the psychology to the psychology department at at Bielefeld and and the the von von Humboldt-Stiftung for for their support support during during that that time. time. also thank thank Robert Robert Hogan Hogan and and Jerry Jerry Wiggins Wiggins for for their their feedback feedback on on an an earlier earlier draft draft of of II also this chapter. this chapter.

REFERENCES Allport, G. W. W. (1937). Personality: Personality:AApsychological psychologicalinterpretation. interpretation. New York: Holt. New York: Holt. Alston, Alston, W. W. P. P. (1975). Traits, Traits, consistency, consistency, and and conceptual conceptual alternatives alternatives for for personality personality theory. theory. Journal 5, 17-48. Journalfor for the theTheory TheoryofofSocial SocialBehaviour, Behaviour, 5, 17-48. Angleitner, A., John, O. P., & you ask ask and and how how you you ask ask it: it:An An & LOhr, Lohr, F.-J. (1986). It's what you itemmetric analysis analysis of of personality personality questionnaires. questionnaires. In In A. A. Angleitner Angleitner & & J. J. S. S. Wiggins Wiggins itemmetric (Eds.), Personality (pp.(pp. 61-107). NY: Springer-Verlag. Personalityassessment assessmentviaviaquestionnaires questionnaires 61-107). NY: Springer-Verlag. Journal of of Barron, F. F. (1953). Some Some personality personality correlates correlates of of independence independence of of judgment. judgment. Journal Barron, Personality,21, 21,287-297. 287-297. Personality, Journalof ofPersonality, Personality,56, 56,571-598. 571-598. Baumeister, R. R. F., & Tice, Tice, D. M. (1988). Metatraits. Journal Baumeister, F., & D. M. (1988). Metatraits. The social socialconstruction constructionofofreality. reality. New York: DouBerger, P. P. L., L., & & Luckmann, Luckmann, T. T. (1966). The Berger, New York: Dou bleday. bleday.

90 90

JOHN JOHN A. A. JOHNSON JOHNSON

Block, J. J. (1977). Advancing Advancing the the psychology of of personality: personality: Paradigmatic Paradigmatic shift shift or or improving improving the quality of research? In D. Magnusson & & N. S. Endler (Eds.), Personality Personalityatatthe the crossroads crossroads (pp. (pp.37-63). 37~63).Hillsdale, Hillsdale,NJ: NJ:Erlbaum. Erlbaum. Buss, A. H. (1989). Personality as traits. American 1378-1388. American Psychologist, Psychologist,44,44, 1378-1388. Cantor (Eds.), Personality Buss, Buss, D. D. M., M., & &Cantor, Cantor,N. N.(1989). (1989).Introduction. Introduction.In InD. D.M. M.Buss Buss&&N.N. Cantor (Eds.), Personality psychology: Recent Recent trends trends and emerging emergingdirections directions(pp. (pp.1-12). 1-12).New NewYork: York:Springer Springerpsychology: Verlag. Verlag. Cooley, C. C. H. H. (1902). Human Human nature natureand andthe thesocial socialorder. order. New York: Scribner's. Cooley, New York: Scribner's. Costa, Costa, P. P. T., T., Jr., Jr., & & McCrae, McCrae, R. R. R. R. (1992). Revised NEO Personality PersonalityInventory Inventory(NEO (NEOPI PIOdessa, FL:FL: RTM) professional manual. R^^) and NEO Five-Factor Five-FactorInventory Inventory(NEO-FFI) (NEO-FFI) professional manual Odessa, Psychological Psychological Assessment Resources. Personalitysystematics systematicsand andthe the problem types. Morristown, Dahlstrom, W. G. G. (1972). Personality problem ofof types. Morristown, NJ:NJ: Dahlstrom, W. General General Learning Press. Eckhardt, B. (1979). The York: McGraw-Hill. Eckhardt, R. R. B. The study studyof of human humanevolution. evolution.New New York: McGraw-Hill. Edwards, A. A. L. L. (1959). (1959).Edwards EdwardsPersonal PersonalPreference Preference Schedule manual. York: PsychologiSchedule manual. NewNew York: Psychologi Edwards, cal Corporation. Corporation. cal Emmons, R. R. A. A. (1989). (1989). Exploring Exploring the the relations relations between between motives motives and and traits: traits: The The case case of Emmons, of trends narcissism. In In D. M. M. Buss & & N. N. Cantor Cantor (Eds.), (Eds.), Personality Personalitypsychology: psychology:Recent Recent trends emergingdirections directions(pp. (pp.32-44). 32-44).New NewYork: York:Springer-Verlag. Springer-Verlag. and emerging Entwistle, D. D. R. R. (1972). To To dispel dispel fantasies fantasies about about fantasy-based fantasy-based measures measures of of achievement achievement Entwistle, 377-391. motivation. Psychological PsychologicalBulletin, Bulletin,77,77, 377-391. Epstein, S. S. (1983). Aggregation Aggregation and and beyond: beyond:Some Somebasic basicissues issueson onthe theproduction productionofofbehavior. behavior. Epstein, Journal Journalof ofPersonality, Personality,51,51,360-392. 360-392. Funder, Funder, D. D. c., C, & & Dobroth, Dobroth, K. K. M. M. (1987). Differences Differences between between traits: traits: Properties Properties associated associated Psychology, 52, 52, 409-418. with with interjudge interjudge agreement. agreement. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 409-418. Gangestad, S., & & Snyder, M. (1985). "To *To carve nature at its joints": joints": On the existence of of discrete discrete classes classes in in personality. personality. Psychological PsychologicalReview, Review,92,92,317-349. 317-349. Gangestad, S., Some statistical considerations S., & &Snyder, Snyder,M. M.(1991). (1991).Taxonomic Taxonomicanalysis analysisredux: redux: Some statistical considerations for for testing testing aa latent latent class class model. model. Journal Journal of of Personality Personalityand andSocial SocialPsychology, Psychology,61,61, 141-146. Goldberg, L. L. R., & & Kilkowski, J. M. (1985). The prediction of semantic consistency in self selfdescriptions: descriptions: Characteristics Characteristics of of persons persons and and of of terms terms that that affect affect the the consistency of of and Social Psychology, response pairs. Journal Journalof ofPersonality Personality and Social Psychology, response to to synonym synonym and and antonym antonym pairs. 48, 48, 82-98. Gough, H. H. G. G. (1948). The The frame frame of of reference reference of of the the Thematic Thematic Apperception Apperception Test. Test. Journal Journal Gough, of of Clinical ClinicalPsychology, Psychology,4,4,90-92. 90-92. Gough, H. H. G. G. (1987). CPI administrator's administrator'sguide. guide.Palo PaloAlto, Alto,CA: CA:Consulting ConsultingPsychologists Psychologists Gough, Press. Press. Guilford, J. J. S., Zimmerman, W. W. S., & Guilford, Guilford, J. J. P. P. (1976). The Guilford-Zimmerman Guilford-Zimmerman Guilford, S., Zimmerman, S., & (1976). The TemperamentSurvey Surveyhandbook: handbook:Twenty-fi Twenty-five years research application. Temperament ve years ofof research andand application. San San Diego, CA: CA: EdITS. Hanson, N. N. R. R. (1958). Patterns Patternsof of discovery. discovery.Cambridge, Cambridge,England: England:Cambridge Cambridge University University Hanson, Press. Hendrick, C. C. (1977). Social Social psychology psychology as as an an experimental experimental science. science. In In C. C. Hendrick Hendrick (Ed.), (Ed.), Hendrick, Perspectiveson onsocial socialpsychology psychology(pp. (pp.1-74). 1-74).Hillsdale, Hillsdale,NJ: NJ:Erlbaum. Erlbaum. Perspectives Hofstee, W. W. K. K. B. Who should should own own the the definition definition of of personality? personality? European EuropeanJournal Journal Hofstee, B. (1994). Who of of Personality, Personality,8,8,149-162. 149-162.

CHAPTER CHAPTER 33 UNITS UNITSOF OFANALYSIS ANALYSIS

9191

Hofstee, W. W. K. K. B., B., & & de de Raad, Raad, B B.. (1992). (1992). Personality Personality structure through traits. traits. In In G.-V. G.-V. Hofstee, Caprara Caprara & & G. G. L. L. Van Van Heck Heck (Eds.), (Eds.), Modern Modern personality personalitypsychology: psychology:Critical Criticalreviews reviews new directions directions (pp. (pp.56-72). 56-72).New NewYork: York:Harvester HarvesterWheatsheaf. Wheatsheaf. and new Hofstee, W. W. K. K. B., B., de de Raad, Raad, B., B., & & Goldberg, Goldberg, L. L. R. R, (1992). (1992). Integration Integration of of the the big big five Hofstee, five and and circumplex approaches approaches to to trait trait structure. structure. Journal Journalof ofPersonality Personalityand and Social Psychology, Social Psychology, circumplex 63, 63, 146-163. Hogan, R. R. (1979, (1979, September). September). Role theory theory as as a depth depth psychology. psychology. Paper Paper presented presented at at the the Hogan, 87th annual annual convention convention of of the the American American Psychological Psychological Association, Association, New New York. York. 87th Hogan, R. R. (1991). (1991). Personality Personality and and personality personality measurement. Dunnette & & L. L. M. Hogan, measurement. In In M. M. D. D. Dunnette M. Hough Hough (Eds.), (Eds.), Handbook of of industrial industrialand andorganizational organizationalpsychology psychology(2nd (2nded., ed.,Vol. Vol. 2, 2, pp. pp. 873-919). 873-919). Palo Palo Alto, Alto, CA: CA: ConSUlting Consulting Psychologists Psychologists Press. Hogan, R., R., & & Briggs, Briggs, S. S. R. R. (1986). (1986). A A socioanalytic socioanalytic interpretation interpretation of of the the public public and and private private Hogan, selves. In R. F. Baumeister (Ed.), Public self and private self self (pp. 179-188). NY: Springer-Verlag. Springer-Verlag. Hogan, R., DeSoto, C. C. B., B., & & Solano, Solano, C. C. (1977). (1977). Traits, Traits, tests, tests, and and personality personality research. Hogan, American Psychologist, Psychologist,32,32, 255-264. American 255-264. Hogan, R., R., & & Johnson, Johnson,J.J.A. A.(1981, (1981,September). September).The Thestructure structure personality. Paper presented ofof personality. Paper presented Hogan, at the the 89th 89thannual annualconvention conventionof ofthe theAmerican AmericanPsychological PsychologicalAssociation, Association,Los LosAngeles. Angeles. at Holt, R. R. (1962). Individuality and generalization in the psychology of personality. Journal Journal of Personality, Personality,30, 30,377-404. 377-404. of IPAT Staff. Staff. (1979). (1979). Admininstrator's Admininstrator'smanual manualfor forthe the1616PF. Champaign, Institute PF. Champaign, IL:IL: Institute forfor IPAT Personality Personality and and Ability Ability Testing. Testing. Johnson, Johnson, J. J. A. A. (1981). (1981). The The "self-disclosure" "self-disclosure" and and "self-presentation" "self-presentation" views views of of item item response response Psychology, Journalof ofPersonality Personalityand andSocial Social Psychology, dynamics and and personality personality scale scale validity. validity. Journal dynamics 40, 761-769. 761-769. 40, Johnson, J. J. A. A. (1990a). (1990a). Empathy Empathy as as aa personality personaHty disposition. disposition. In In R. C. MacKay, MacKay, J. Johnson, R. C. J. R. R. (pp. 49-64). New Hughes, & & E. E. J. J. Carver Carver (Eds.), (Eds.), Empathy Empathyin inthe thehelping helpingrelationship relationship (pp. 49-64). New Hughes, York: Springer. York: Springer. Johnson, J. A. A. (1990b, (1990b, June). Unlikely virtues virtuesprovide providemultivariate multivariate substantive information June). Unlikely substantive information Johnson, J. about personality. personality.Paper Paperpresented presentedatatthe the2nd 2ndannual annualmeeting meetingofofthe theAmerican AmericanPsycho Psychological logical Society, Dallas, Dallas, TX. on item andand scalescale validity. Manuscript Johnson, J. J. A. A.(1993a). (1993a).The Theimpact impactofofitem itemcharacteristics characteristics on item validity. Manuscript submitted for publication. Johnson, J. J. A. (1993b, August). Sex Sex roles roles do not lead to gender gender differences differencesininpersonality. personality. Paper Paper presented presented at at the the annual annual conference conference of of the the Human Human Behavior and Evolution Evolution Society, Society, Binghamton, Binghamton, NY. (BARS). Johnson, J. J. A. A. (1994). (1994). 1nterpreter's Interpreter'sguide guidetotothe theBipolar BipolarAdjective AdjectiveRating RatingScales Scales (BARS). Unpublished Unpublished manuscript, manuscript, Pennsylvania Pennsylvania State University. University. of of individual profiles. Manuscript Johnson, Johnson, J. J. A. A. (1996). (1996). Computer Computernarrative narrativeinterpretations interpretations individual profiles. Manuscript submitted for publication. publication. submitted for Johnson, J. A. A. (1997). (1997). Seven Seven social social performance performance scales scales for for the the California California Psychological Psychological Johnson, J. Inventory. Human Human Performance, Performance,10,10, 1-31. Inventory. 1-31. Johnson, J. J. A., A., Germer, Germer, C. C K., K., Efran, Efran, J. J. S., S., & & Overton, Overton, W. W. F. F. (1988). (1988). Personality Personality as asthe the basis basis Johnson, Social Psychology, 55, 55,824-835. 824-835. for theoretical predilections. Journal Journalof ofPersonality Personalityand and Social Psychology, Johnson, J. J. A., & Hogan, Hogan, R. R. (1981a). (1981a). Moral Moral judgments judgments and and self-presentations. self-presentations. Journal Journalof of Johnson, A., & ResearchininPersonality, Personality,15,15, 57-63. Research 57-63. Johnson, J. J. A., & Hogan, Hogan, R. R. (1981b). (1981b). Vocational Vocational interests, interests, personality, personality, and and effective effective police police A., & Johnson, performance. Personnel PersonnelPsychology, Psychology,34,34,49-53. 49-53. performance.

92

JOHN A. A. JOHNSON JOHNSON JOHN

Johnson, J. J. A., & & Hogan, Hogan, R., Zonderman, Zonderman, A. A. B., B., Callens, c., C, & & Rogolsky, S. S. (1981). Moral and Social judgment, judgment, personality, personality, and and attitudes attitudes toward toward authority. authority.Journal Journalof ofPersonality Personality and Social Psychology, Psychology,40, 40,370-373. 370-373. Johnson, Johnson, J. J. A., A.,& &Ostendorf, Ostendorf,F.F.(1993). (1993).Clarification Clarificationofthe of thefive fivefactor factormodel modelwith withthe theAbridged Abridged Psychology, 65,65, Big Five-Dimensional Five-Dimensional Circumplex. Circumplex. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, Big 563-576. Kenrick, D. D. T., T., & & Funder, Funder, D. D. C. C. (1988). Profitting Profitting from from controversy: controversy: Lessons Lessons from from the Kenrick, person-situation debate. debate. American American Psychologist, Psychologist,43,43, 23-34. 23-34. person-situation Kenrick, (1980). Personality Kenrick, D. D. T., T., & & Stringfield, Stringfield, D. D. O. O. (1980). Personality traits traits and and the the eye eye of of the the beholder: beholder: Crossing Crossing some some traditional traditional philosophical philosophical boundaries boundaries in in the search search for for consistency consistency in in all all of of the the people. people. Psychological PsychologicalReview, Review,87, 87,88-104. 88-104. Kilkowski, Kilkowski, J. J. M. M. (1975). Traits, Traits, situations, situations, and and the the logic logic of of explanation explanation in in psychology: psychology: A A Research Mono selective selective review review of of the the issues issues and and evidence. evidence. Oregon OregonResearch ResearchInstitute Institute Research Monograph 15(3). 15(3). graph Luft, J. J. (1966). Group Group processes: processes:An An introduction introductiontotogroup groupdynamics. dynamics, Palo Alto, CA: NaLuft, Palo Alto, CA: Na tional Press. tional Press. MacKinnon, D. W. W. (1944). The The structure structure of of personality. personality. In In J. J. McV. McV. Hunt (Ed.), Personality Personality MacKinnon, D. Hunt (Ed.), the behavior behaviordisorders disorders(pp. (pp.3-48). 3-48).New NewYork: York:Plenum PlenumPress. Press. and the McClelland, D. C. C. (1980). Motive Motive dispositions: dispositions: The The merits merits of of operant versus versus respondent respondent McClelland, D. personality and andsocial socialpsychology psychology(Vol. (Vol.1,1, measures. In In L. L. Wheeler Wheeler (Ed.), (Ed.), Review Review of of personality measures. pp. 11-41). Beverly Beverly Hills, Hills, CA: CA: Sage. pp. Sage. McCrae, R. R. R. R. (1989). Why Why II advocate advocate the thefive-factor model: Joint Joint factor factor analyses analyses of of the the McCrae, five-factor model: with other other instruments. instruments.In InD. D.M. M.Buss Buss&&N.N.Cantor Cantor(Eds.), (Eds.),Personality Personality psycholNEO-PI with psychol NEO-PI Recenttrends trendsand andemerging emerging directions 237-245). New York: Springer-Verlag. directions (pp.(pp. 237-245). New York: Springer-Verlag. ogy: Recent McCrae, R. R. (1990). Traits and trait names: names: How well is Openness represented in natural 4, 4, 119-229. EuropeanJournal JournalofofPersonality, Personality, 119-229. languages? European McCrae, R. R., & & Costa, P. T., Jr. Jr. (1989). Reinterpreting the Myers-Briggs Type Indicator five-factor model of personality. Journal from the perspective of thefive-factor Journalof ofPersonality, Personality, 57, 57, 17-40. McCrae, R. R., Costa, P. T., Jr., & & Piedmont, R. (1993). Folk concepts, natural language, and psychological constructs: five constructs: The California Psychological Inventory and the fivefactor model. Journal Journalof ofPersonality, Personality,61,61,1-26. 1-26. Mead, O. (1934). Mind, G. H. (1934). Mind, self, self and andsociety. society.Chicago: Chicago:University UniversityofofChicago ChicagoPress. Press. Clinical Meehl, P. E. (1945). The "dynamics" "dynamics" of structured personality tests. Journal Journal of of Clinical Psychology, Psychology, 1,1,296-303. 296-303. Meehl, Meehl, P. P. E. E. (1956). Problems Problems in in the the actuarial actuarial characterization characterization of of aa person. person. In In H. Feigl Feigl & M.Scriven Scriven (Eds.), (Eds.),Minnesota Minnesotastudies studies philosophy ofscience: I. foundations The foundations M. inin thethe philosophy ofscience: VoLVol I. The of science andthe theconcepts concepts psychology and psychoanalysis 205-222). Minneaposcience and ofof psychology and psychoanalysis (pp.(pp. 205-222). Minneapo of lis: University University of of Minnesota Minnesota Press. lis: Press. Meehl, P. E. E. (1992). Factors Factors and and taxa, taxa, traits traits and and types, types, differences differences of of degree degree and and differences differences Meehl, P. in kind. Journalof ofPersonality, Personality,60,60,117-174. 117-174. in kind. Journal Meehl, P. P. E., E., & & Hathaway, Hathaway, S. S. R. R. (1946). The The K K factor factor as as aa suppressor suppressor variable variable in in the the Meehl, Minnesota Multiphasic Multiphasic Personality Personality Inventory. Inventory. Journal Journal of of Applied Psychology, Psychology, 30, Minnesota 525-564. Mendelsohn, Mendelsohn, O. G. A., A., Weiss, Weiss, D. D. S., S., & Feimer, Feimer, N. N. R. R. (1982). Conceptual Conceptual and and empirical empirical analysis analysis of the the typological typological implications implications of of patterns patterns of of socialization socialization and and femininity. femininity. Journal Journalof of of Personalityand andSocial SocialPsychology, Psychology,42,42, 1157-1170. Personality 1157-1170.

CHAPTERS CHAPTER 3

UNITS OF OFANALYSIS ANALYSIS UNITS

93 93

Mills, C, & Hogan, Hogan, R. R. (1978). (1978). A A role role theoretical interpretation of of personality scale item Mills, c., & theoretical interpretation personality scale item 46,46, 778-785. responses. Journal Journalof ofPersonality, Personalityy 778-785. Mischel, W. W. (1968). (196S). Personality Personalityand andassessment. assessment. New York: Wiley. New York: Wiley. Mischel, Myers, I. I. B., & McCaulley, McCaulley, M. M. H. H. (1985). (1985). Manual: Manual:AAguide guidetotothe thedevelopment developmentand and use use ofof Myers, B., & MyerS'BriggsType TypeIndicator. Indicator. Palo Alto, CA: Consulting Psychologists Press. the Myers-Briggs the Palo Alto, CA: Consulting Psychologists Press. Paulhus, D. D. L., L., & &Reid, Reid,D. D.B.B.(1991). (1991).Enhancement Enhancementand anddenial denialinin sociallydesirable desirableresponding. responding. Paulhus, socially Journalof ofPersonality Personalityand andSocial Social Psychologyj 307-317. Journal Psychology, 60, 60, 307-317. Piedmont, R. L., McCrae, R. R., & & Costa, P. T., Jr. Jr. (1992). An assesement of the Edwards Personal Preference Preference Schedule Schedule from from the the perspective perspective of of the thefive-factor model. Journal Journal Personal five-factor model. of of Personality PersonalityAssessment Assessment58,58,67-78. 67-78. Reik, T. T. (1948). (1948). Listening Listeningwith withthe thethird thirdear: ear: inner experiences a psychoanalyst. TheThe inner experiences of a of psychoanalyst. New New Reik, York: Farrar, Straus. Rholes, W. W. S., S., Newman, Newman, L. L.S., S.,& &Ruble, Ruble,D. D.N. N.(1990). (1990).Understanding Understandingself selfand andother: other:Develop DevelopRholes, and motivational motivational aspects aspectsof ofperceiving perceivingpersons personsin interms termsofofinvariant invariantdispositions. dispositions. mental and mental In E. E. T. T. Higgins Higgins & & R. R. M. M. Sorrentino Sorrentino (Eds.), (Eds.), Handbook of of motivation motivation and andcognition cognition In (Vol. pp. 369-407). York: Guilford (Vol. 2, 2, pp. 369-407). New New York: Guilford Press. Rosenblueth, A., A., & & Wiener, Wiener, N. N. (1945). (1945). The The role role of of models models in inscience. science. Philosophy Philosophyof ofScience, Science, Rosenblueth, 12, 12, 316-321. 316-321. Ross, L. L. (1977). (1977).The The intuitive intuitive psychologist psychologist and andhis hisshortcomings: shortcomings:Distortions Distortionsininthe theattribution attribution Ross, process. In In L. L. Berkowitz Berkowitz (Ed.), (Ed.), Advances Advances in inexperimental experimentalsocial social psychology (Vol. process. psychology (Vol. 10,10, pp. pp. 174-221). 174-221). New New York: York: Academic Academic Press. Ross, L., L., & & Nisbett, Nisbett, R. R. E. E. (1992). (1992). Perspectives Perspectives on on personality personality and and social social psychology: psychology: Books Books Ross, waiting waiting to to be be written. written. Psychological PsychologicalInquiry, Inquiry,3, 3,99-102. 99-102. Tellegen, Questionnaire. Minneapolis: University Tellegen, A. A. (in (in press). press). Multidimensional MultidimensionalPersonality Personality Questionnaire. Minneapolis: University of Minnesota of Minnesota Press. Press. Wakefield, C. (1989). (1989). Levels Levels of of explanation explanation in in personality personality theory. theory. In In D. M. Buss Buss & & N. Wakefield, J. J. C. D. M. N. (pp. Cantor Cantor (Eds.), (Eds.), Personality Personalitypsychology: psychology:Recent Recenttrends trendsand and emerging emergingdirections directions (pp. 333-346). 333-346). New New York: York: Springer-Verlag. Springer-Verlag. Weimer, W. B. (1974). Overview Overview of of aa cognitive cognitive conspiracy: conspiracy: Reflections on the the volume. volume. In Weimer, W. B. (1974). Reflections on In W. W. B. Weimer & & D. S. S. Palermo Palermo (Eds.), (Eds.), Cognition Cognitionand andthe thesymbolic symbolicprocesses processes(pp. (pp. 415-442). 415-442). Englewood Cliffs, Cliffs, NJ: NJ: Erlbaum. Erlbaum. W. B. B. (1984). (1984). Limitations Limitations of of the the dispositional dispositional analysis analysis of of behavior. behavior. In InJ.J.R. R.Royce Royce&& Weimer, W. L. L. P. P. Mos Mos (Eds.), Annals Annals of of theoretical theoreticalpsychology psychology(Vol. (Vol.1,1,pp. pp.161-198). 161-198).New NewYork: York: Plenum Press. Weinberger, J., J., & &McClelland, McClelland,D. D.C. C.(1990). (1990).Cognitive Cognitiveversus versustraditional traditionalmotivational motivationalmodels: models: Weinberger, Irreconcilable or or complementary? complementary? In In E. E. T. T. Higgins Higgins & & R. R. M. M. Sorrentino Sorrentino (Eds.), (Eds.), Hand HandIrreconcilable of motivation motivation and andcognition cognition(Vol. (Vol.2,2,pp. pp.562-597). 562-597).New NewYork: York:Guilford GuilfordPress. Press. book of Weiss, D. Mendelsohn, G. G. A., A., & & Feimer, Feimer, N. N. R. R. (1982). (1982). Reply Reply to to the the comments comments of of Block Block Weiss, D. S., S., Mendelsohn, Psychology, 42,42, 1182-1189. and and Ozer. Ozer. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 1182-1189. Journal Werner, P. P. D., D., & & Pervin, Pervin, I. I. A. A. (1986). (1986). The The content content of of personality personality inventory inventory items. items. Journal Werner, 622-628. of 51,51, of Personality Personalityand andSocial SocialPsychology, Psychology, 622-628. Wolfe, R. R. N. N. (1993). (1993). A A commonsense commonsense approach approach to to personality personality measurement. measurement. In In K. K. H. H. Craik, Craik, Wolfe, R. R. Hogan, & & R. R. N. N. Wolfe Wolfe (Eds.), (Eds.), Fifty Fiftyyears yearsof ofpersonality personalitypsychology psychology(pp. (pp.269-290). 269-290). New York: York: Plenum Plenum Press. New Press. Young, N. P. P. (1975). (1975). The The psychology psychology of of explanation and the the measurement of explanatory explanatory Young, N. explanation and measurement of satisfaction dissertation, Pennsylvania Pennsylvania State State University). University). Dissertation DissertationAb Absatisfaction (Doctoral (Doctoral dissertation, stractsInternational, International, 5775B-5776B. stracts 36,36, 5775B-5776B.

This Page Intentionally Left Blank

CHAPTER 44

-

- ----- -------- --

---

-- --

-- --

IN DEFENSE OF TRAITSl TRAITS' JERRY JERRY S. S. WIGGINS WIGGINS UNIVERSITY OF B BRITISH U NIVERSITY OF RmSH COLUMBIA COLUMBIA

In a recent review of the field field generally designated "personality ''personality assessment," I rendered the optimistic albeit highly qualified opinion that visible signs of progress could be discerned in this still youthful science (Wiggins, 1973). Such a conclusion by Vernon (1964), is considerably more sanguine than that reached a decade earlier by or more recently by Mischel (1968). My disagreement with Vernon involves little more than a tendency on my part to view glasses as half-full, rather than halfhalf empty. My differences differences with with Mischel Mischel are are more more profound. In aa classic empty. My profound. In classic RashomonRashomon type example example of of individual differences in in perception, Mischel and and II observed observed the the type individual differences perception, Mischel same events, but provided quite different narratives. This disagreement is all same events, but provided quite different narratives. This disagreement is all the the more striking striking when when it it is is realized the two two observers share many many of more realized that that the observers appear appear to to share of the same same conceptual conceptual biases biases (e.g., (e.g., the the importance importance of of generalizability, generalizability, utility utility analysis, the analysis, and explicit explicit theoretical theoretical bases bases for for assessment). assessment). and as broad as as personality assessment frequently frequently Different perspectives onfields fields as differences in taxonomic behaviors. Categories may be broad arise from stylistic differences or narrow, concrete or overinclusive. In this respect, Vernon (1964) exhibited a width when when he he lumped together together decision decision making, preference for categories of broad width psychoanalytic theory, psychotherapy, descriptive psychiatry, and all things "clini"clini cal" within a single and particularly unwholesome bin (Wiggins, 1964). Mischel (1968) erred even more in the direction of overinclusion when he categorized such

I^ Invited address to the Ninth Annual Symposium Symposium on Recent Recent Developments Developments in the Use of of the February 28, 1974. An earlier earlier version version of of this paper paper was presented presented as MMPI, held in Los Angeles on February Assessment and Research Research in Berkeley Berkeley on November November 27, 27, a seminar at the Institute of Personality Assessment

1973. 1973.

H ANDBOOK OF HANDBOOK OFPERSONALfTY PERSONAUTY PSYCHOLOGY PSYCHOLOGY

9S 95

COPYRIGHT ACADEMIC PlESS COPYRIGHT C © 1997 1997 IY BY ACADEMIC PRESS.. ALL ALL IIGHIS RIGHTS OF OF REPRODUcnON REPRODUOION IN IN ANY ANY FORM FORM RESERVED. RESERVED.

96 96

JERRY JERRY S. S.WIGGINS WIGGINS

diverse offenses against human nature as psychodynamic theory, factor analysis, and the medical model as all belonging to a monolithic **trait" "trait" conspiracy. Quibbling over taxonomic niceties should generally be left to textbook writers and others who continue the scholastic tradition. But there is more at stake in the present instance. Conclusions stemming from Mischel's broadband view of of the "trait construct" have had an extraordinary impact on our field. Behaviorally inclined clinicians appear to be celebrating a decisive victory, a blitzkrieg, so to speak, that defined and destroyed destroyed the enemy almost simultaneously. Psychodynamically oriented clinicians, long used to being "sold up the river," to use Holt's (1958) phrase, must now bear the added humiliation of of having multivariate-trait psycholo psychologists as traveling companions. Mischel's (1968) textbook, and his subsequent writings along similar lines (Mischel, 1969, 1971, 1973a, 1973b), have had a considerable considerable impact on the field of of personality in general, and personality assessment assessment in particular. His views have not gone unchallenged, and there is a still-growing literature of of criticisms directed at one or another facet of his arguments (Alker, 1972; Averill, 1973; 1973; Bowers, 1973; Craik, 1969; Wachtel, 1973a, 1973b; 1973b;Wallach Wallach & &Leggett, Leggett, 1972). 1972). Yet Yet Mischel's Mischel'swritings writings have tended to polarize his readership into a relatively homogeneous group of of satisfied satisfied social behaviorists and social psychologists on the one hand, and a highly heterogeneous and most dissatisfied group of clinicians, psychometricians, and per personality sonaUty theorists on the other. The reason that Mischel's writings have had such a diffuse diffuse effect effect is that his arguments themselves themselves are diffuse diffuse and multipronged. At one level, he is challenging the field of personality as traditionally defined and the field of of personality assessment as it has traditionally been implemented. At another level, he is extolling the virtues of of social-behavioral social-behavioral conceptions, as opposed to psychodynamic and other of certain nonbehavioral views. And at still another level, he is arguing for the utility of methods of behavior modification and control for both practical and theoretical purposes. When considered one at a time, these are each complex issues worthy of of debate. However, when considered in toto, they appear as a shifting shifting myriad of of targets against which it is difficult, if not impossible, to take aim. For example, if if one must assume that the recent triumphs of of some clinicians in reducing or eliminat eliminating fears of snakes in their clients refiect reflect unfavorably on the psychometric adequacy of of multivariate personality inventories, then one is doomed to argue from a position of of weakness and bewilderment. To bring the issues into clearer focus, I would like to consider the trait concept: (1) as it is used, or could be used, in personality measurement and assessment, and (2) as it is used, or should be used, in a theory of of personality. The two sets of considerations are not unrelated. The most modest attempts to quantify quantify personal theoretical considerations; the grandest characteristics should be guided by explicit theoretical reference to concrete measurement proce procetheories must eventually be assessed with reference dures. With this in mind, I will first attempt to discredit the growing philosophical of traits, and then suggest skepticism that has been expressed regarding the existence of of personality. the place that trait measures may have within a theory of

CHAPTER 4

IN DEFENSE OF TRAITS TRAITS

97 97

I. THEORIES THEORIES AND AND VIEWPOINTS VIEWPOINTS L

For a number of reasons, it is convenient to consider (the (the field of) "personality" differences (Wiggins, (Wiggins,Renner, Renner,Clore, Clore,&& as the general psychology of individual individual differences Rose, 1971). In attempting to account for (the fact of) individual differences, a variety of viewpoints must be brought to bear on a common subject matter. At the least, it would seem that the methods and and concepts of biological, experimental, and psychometric-trait approaches are are necessary for providing a complete social, and can result in account of human differences. Pitting one approach against another can such fluid controversies as the ancient "heredity versus environment" issue or the the fluid situation" debate. recently revived "trait versus situation" not When one approach to knowledge is compared with another, the game is not zero sum. Whether from genuine conviction or from the observance of good form, it is common to concede that other approaches have virtues as well as limitations, as well well as as failures. failures. Nevertheless, Nevertheless, while while the the biological, biological, experimental, experimental, and triumphs as triumphs social approaches approaches have have all all recently recently had had their their days days in in the the sun, sun, the the psychometric-trait psychometric-trait social the meth methapproach has has fallen fallen upon upon extremely extremely bad bad times. times. The The possibility possibility exists exists that that the approach and distinctions distinctions of of the the trait trait approach approach have have simply simply outlived outlived their their usefulness usefulness in ods and in ods comparison with with recent recent advances advances in in other other viewpoints. viewpoints. II personally personally do do not not believe believe comparison to be be the the case. case. Instead, Instead, II believe believe that that aa widespread widespread discontent discontent with with certain certain this to this of personality personality has has resulted resulted in in an an attempted attempted purge purge of of concepts concepts essential essential theories of theories the psychometric-trait psychometric-trait viewpoint viewpoint in in personality personaUty study. study. The The distinction distinction between between to the to and theory theory is is critical critical here. viewpoint and viewpoint here. the empirical study of personality that is based A viewpoint is an approach to the and that on assumptions concerning the importance of certain kinds of constructs and and measurement (Wiggins advocates the use of certain methods of observation and in place of "method" "method" to emphasize that et al., aI., 1971). The term "viewpoint" is used in A methods involve constructs and that they impose constraints upon observations. A an extended construct system of broad range and and scope that typically theory is an an integration integration of of constructs constructs from from several several viewpoints.2 viewpoints.^ attempts an I believe beUeve it fair to say say that the the viewpoints of personality study are are reasonably reasonably "established" in their own right, because they represent traditional and respectable psycholareas of psychological investigation (biological psychology, experimental psychol ogy, social psychology, and psychometrics). Theories of personality are, of course, another story. And it is important to note at the outset that although a theory of personality may may achieve achieve aa certain certain prestige prestige by by emphasizing particular viewpoint, viewpoint, personality emphasizing aa particular of the the viewpoint viewpoint cannot cannot be be substituted substituted for for the the propositions propositions of of the the methods methods of the the theory. Bowers Bowers (1973) (1973) has has made made this this point point in in reference reference to to the the misidentification misidentification of theory. of theory with with the the experimental experimental method: S-R theory S-R method: This distinction distinction between between viewpoint viewpoint and andtheory theory isissimilar similarto toFiske's Fiske's (1971) (1971) distinction distinction between between 2^This mode and perspective. perspective. mode

98

JERRY JERRY S. S.WIGGINS WIGGINS

As it it happens, happens, the the experimental experimental method method as as generally generally employed employed is is differentially differentially As sensitive to the impact of situational variables, and correspondingly insensitive to organismic variables variables.. .. .. .. However, the experimental method does not, so to speak, speak, comment on this differential sensitivity; sensitivity; it it is is simply simply aa procedure procedure for for to comment on this differential acquiring aa controlled controlled observation. observation. Thus, Thus, independent-dependent independent-dependent variable variable rela relaacquiring tionships are are metaphysically metaphysically neutral. neutral. This This is is not not the the case case for their S-R S-R counter countertionships for their parts, which which do do carry carry aa great great load load of of metaphysical metaphysical freight. freight, (p. (p. 309) parts, 309) justify a theory, The converse is also true. Just as a method cannot be used to justify A a theory can be discredited without discrediting the method that it espouses. A discreditation of S-R theory is no reflection reflection on the experimental method. However, I believe it to be the case that the psychometric-trait viewpoint has recently been judged guilty in virtue of its association with certain personality theories. The fact that trait-like or dispositional concepts are so ubiquitous in personality theory should not make one particular interpretation interpretation of of traits subject to the criticisms of other interpretations. Traits are many things to many theorists, and it is precisely this conceptual plurality that has provided a composite straw man for those who have criticized trait measurement. One of the chief chief concerns of the psychometric approach to personality personality study of human is the development of quantitative procedures for the measurement of tendencies (pronenesses, proclivities, propensities, dispositions, inclinations) to act "postuor not to act in certain ways on certain occasions. These tendencies are not "postu lated," they are accepted from common sense as expressed in ordinary language usage. If persons are not more or less prone to behave in certain ways on certain occasions, then the psychometric approach is out of of business at the outset, as are all approaches to personality study. Because of its concern with human tendencies as expressed in ordinary lan language, a case could be made for the psychometric-trait viewpoint being propaedeutic psychometric-trait viewpoint to other approaches to personality study. However, as I hope to make clear later, the closeness of of the psychometric-trait psychometric-trait approach to the obvious obvious subject matter of of personality study need not imply, and in fact should not imply, that trait concepts will figure prominently in our eventual systematic accounts or explanations of of that subject matter. It seems more likely that theoretical explanations of of human tendencies will emerge from the social, experimental, and biological biological viewpoints. In the material that follows, I hope to illuminate the nature of of the trait concept by considering the ways in which trait terms are employed in everyday discourse. of traits is provided by But to show that a consistent and meaningful meaningful account of ordinary language usage is not to show that a scientific scientific account of traits is easily of which has to do achieved. A A number of of additional steps are required, the first of of measurement procedures. However, although the current with the specification of of psychometrics may be primitive in comparison with measurement in the state of physical sciences, it is clear from reading Fiske's (1971) recent book book that the primary obstacle to measuring the concepts of personality has been conceptual rather than mensurational. Within the field of personality, there appears to be gre�ter greater agreement concerning how concepts should be measured (e.g., Fiske, 1971) than on what

CHAPTER 44 IN INDEFENSE DEFENSEOF OFTRAITS TRAITS CHAPTER

9999

of a working concepts should be measured (e.g., Levy, 1970). Thus, achievement of consensus on the nature of of the trait concept would be a large first step toward a psychology of of personality. of the trait concept in psychology have focused focused on Most previous discussions of traits as attributes of of persons. persons. Thus, the first task of of psychometrics has been consid considered that of of developing scales and inventories to measure the "tendencies" of persons to act in certain ways on certain occasions (e.g., Edwards, 1970). But this logically presupposes a clear conception of which particular actions are to be accounted for on what particular occasions. To say that a person is "aggressive" is to say that the person has behaved or is likely to behave "aggressively" on certain occasions. But what is an aggressive action? And how would we quantify quantify the of an action? "aggressiveness" of quaUties of ot actions ("John ("John I intend to distinguish among statements expressing: qualities persons ("John is aggressive"), pushed the boy aggressively"), (2) properties of persons of future occurrences ("If ("If frustrated, John is likely to behave aggres aggresand (3) aspects of sively"). I will argue that the first statement conveys an institutional fact; the second, categorical summary of the general trend of a person's conduct to date; and the a categorical third, a hypothetical proposition that is inferred, but not deduced, from statements of the second type. In the specialized terminology terminology of of the psychometric-trait approach of to personality, these distinctions correspond roughly to those sometimes made among "observation," "assessment," and "prediction." n. TRAITS AS ATIRIBUTES ll. VIOR ATTRIBUTES OF HERA BEHAVIOR

"John pushed the boy" describes an action or sequence of of behavior. "John pushed the boy hard, repeatedly, and for a long time" qualifies the description in terms of qualities which may be thought of as attributes of the action, rather than of John, or of the observer. The observer is making judgments here (How hard is "hard"? How often is "repeatedly"? How long is "long"?), but the normative basis for such How judgments can easily be made explicit. The important point is that the descriptive qualities tell how John pushed the boy, not why John pushed, nor why the observer described John's actions in these terms. "John pushed the boy aggressively" qualifies the description of the action, different way. Its meaning is not synonymous with the description yielded but in a different of primary attributes (amplitude, frequency, duration). Nor can the by the use of of aggression be recorded by mechanical devices in the absence of a human quality of observer. Nevertheless, the qualifier ("aggressively") should not automatically be of emotive responses (evaluations) simply because an relegated to the domain of affectionately, observer is involved. One can describe John pushing aggressively, affectionately, playfully in a thoroughly dispassionate manner. or playfully manner. That, of course, is what is meant by an an "objective" "objective" observer. observer. by of John's pushing as "aggressive" does not, in ordinary usage, The description of refer to John's intentions. As Anscombe (1963) put it, "We do not add anything

100 100

JERRYS. S.WIGGINS WIGGINS JERRY

attaching to the action at the time it is done by describing it as intentional" (p. 28). The description "aggressively" tells how John pushed the boy, not why John pushed, nor whether the action was "voluntary" or "involuntary." The latter distinction typically arises in connection with actions that result in improprietous outcomes (Ryle, 1949, pp. 69-74). We may determine whether a person meant or intended to do something (wrong) by inquiring whether he was competent to do it right (and failed to do so) or by establishing whether or not external factors prevented him performing properly. But it is the description of of the present action as "aggres "aggresfrom performing sive" that establishes its character, and subsequent attempts to excuse or condemn of action represent a distinctly different different line of of inquiry. John for taking that kind of of John's pushing as "aggressive" does not refer to John's The characterization of disposition to perform aggressive acts in this or other circumstances. The word "aggressively" modifies the verb (pushed) and not the subject (John). The sentence "Quite uncharacteristically uncharacteristically John gave the boy an aggressive push" makes good sense and conveys both that the act was aggressive and that John is not. True, if John continues to shove people around, we might wish to revise our original appraisal of his aggressiveness. But the direction of inferences is from act to disposition and not the converse. Could it be that the action under consideration is described as "aggressive" of its relation to certain antecedent conditions? It is true, for example, that because of aggression may follow frustration and in that sense be "provoked." Thus, it might of actions which are likely to follow follow be the case that "aggressive" refers to a class of actions or circumstances that are harmful, insulting, or frustrating to the actor. But the lawfulness of of the implied R-R relation does not seem strong enough to enable us to avoid frequent miscategorizations. Provocations may be (and often are) met with a smile, ignored, or submitted to. More damaging to the "antecedent" account of aggression, however, is the fact that the term "aggressive" is applied to topograph topographof ically dissimilar actions that follow submissive, generous, dominant, affectionate, affectionate, or almost any conceivable kind of of action. It is, of course, possible to distinguish "provoked" and "unprovoked" aggression. But such an inquiry into the reasons for an action does not illuminate the qualities that made that action "aggressive" in the first place. If an action is not classified classified as "aggressive" on the basis of of antecedent events, perhaps it is classified classified on the basis of consequent events or outcomes. Let us try: An act is described as "aggressive" if and only if it results in (is followed by) the harm, injury, discomfort, or ridicule of of another. This basis for discriminating the attribute of of aggressiveness has one clear advantage over the antecedent event account: topographically topographically dissimilar actions (pushing, hitting, swearing) are encom encompassed by a single term, "aggressive," which connotes a common property. But the requirement that a specific outcome must occur is too strong. Not all aggressive actions result in harm or injury. If John takes a swing at the boy with a meat axe and misses, the action is still unambiguously "aggressive." of the attribute of of "aggressiveness" is Thus far, I have argued that the sense of found in: (1) primary qualities of the act, (2) evaluative responses of of the not to be found

CHAPTER 44 IN INDEFENSE DEFENSEOF OFTRAITS TRAITS CHAPTER

101 101

of the actor, (4) tendencies of of the actor, (5) the conditions observer, (3) intentions of left? antecedent to the act, or (6) the immediate consequences of the act. What is left? Clearly, we have not exhausted all possibilities; but rather than pursue additional false leads, it seems appropriate to state what I believe is meant when a trait quality is attributed to an action: the action belongs to a class of of actions actions that are likely likely to lead to a particular particular outcome. outcome. The "outcomes" at issue here are social social in nature. They may be characterized by such phrases as "being harmed, injured, discomforted, or ridiculed," "being praised, admired, revered, or lauded," and "being influenced, directed, persuaded, or restrained," considering the likely effects of of aggressive, deferential, and dominant actions, respectively. But how do we know these things, and what is it that we know? Searle (1969, pp. 50-53) has proposed a useful useful epistemological distinction between "brute facts" and "institutional facts." Brute facts are, roughly, those objects, relations, and primary qualities that lend themselves to direct observation and with which the natural sciences are concerned. In contrast, institutional facts do not stand on their own, but presuppose the existence of certain human institu institutions. These institutions are systems of constitutive rules of of the form "X counts as Y in context C". Consider the following institutional facts reported in a newspaper: "In the closing minutes of of the game, pass interference was called in the Redskin's end zone and the Dolphins won by a score of 6 to 0." Clearly, the existence of of the facts recorded in this statement presupposes the existence of of the institution of of football, furnishes the rules that impart a special meaning to the various an institution that furnishes of the game. One rule is of the following form: brute facts of "An action (pushing) that is likely to prevent a receiver from catching the ball (X) counts as interference (Y) in the context of of the rules of of football football (C)." Similarly, the statement "John pushed the boy aggressively" contains an insti institutional fact defined by the following: "An action (pushing) that is likely to harm or injure another (X) counts as aggressive (Y) in the context of of the rules for classifying the consequences of social actions (C)." To designate trait qualities as institutional is not to imply that that they are (merely) (merely) conventions instead of (actual) facts. Trait attributions are conventions about report of reporting facts. As psychologists, we are perhaps less interested in the ontological status of institutional facts than we are in the origins and functions of of the rules which, in ordinary usage, provide definitions of those facts. It seems likely that trait terms of were coined to express the law-like relations that have been observed between certain kinds of of human actions and particular classes of social outcomes. Given the variety and complexity of of actions that may result in the same social outcome, it is not possible for single individual to learn, by direct experience, even a fragment of of these regularities. Trait attributions convey the shared folk wisdom concerning actions and outcomes in an extraordinarily efficient efficient manner. Because the truth of a trait attribution is not defined by a particular immediate outcome, the sense of

102 102

S. WIGGINS WIGGINS JERRY S.

the law-like relation conveyed is "probabilistic" ^'probabilistic" rather than deterministic. Indeed, Indeed, it would be a surprise-if surprise—if not an embarrassment-if embarrassment—if the laws of of common common sense psychology were stronger than those of of scientific psychology. One final final point regarding traits as attributes of of behavior: most psychologists psychologists like to think of of actions that share a common attribute as constituting a "response ''response class." On an observational observational level, we have defined defined trait attributes as response classes that have common environmental effects. Hence, our definition definition should be acceptable acceptable to groups as divergent as ordinary ordinary people, Skinnerians, and trait theorists. As we move from from behavior behavior observation observation to conceptualization, however, we find that these groups diverge in their views. Both Skinnerians and trait theorists impose additional additional requirements beyond the level of of a single observation. In addition to requiring requiring that members of of a response class share the attribute of of a common environmental environmental effect, effect, Skinnerians Skinnerians require that the actions enter enter into the same functional functional relationships relationships requirement imposed by with "controlling" stimulus conditions. The additional requirement trait theorists is that members of of a response class exhibit significant significant covariation covariation of individuals. Ordinary people could be talked into either conceptu conceptuwithin a group of of discourse is that of of psychological theory and not that of of alization since the level of common sense. But agreement agreement as to what trait attributes are on an observational observational common Hghtly. It suggests that we all know what trait attributes attributes level is not to be treated lightly. are, and that they "really" exist.

m. ATTRIBUTES OF PERSONS nit TRAITS AS ATI'RIBUTES Having identified the conditions under which trait qualities are ascribed to behavior, behavior, the manner in which trait terms are used to describe persons should be fairly aggressive," is to say that in certain circumstances, circumstances, John John evident. To say, "John is aggressive," has behaved, or is likely to behave, behave, in a manner likely to result in harm, injury, or discomfort to others. But which is it: "has behaved" or "is likely to behave"? The "is likely to behave" account is the classical argument of Gilbert Ryle, who maintained that tickets. The "has be betraits are dispositions which function as law-like inference tickets. haved" account is a refutation of the classical argument which maintains that trait attributions are summarizing statements that do not commit the speaker to conditional predictions (Hampshire, (Hampshire, 1953). 1953).The The two two views viewsare are sufficiently sufficiently disparate disparate to warrant separate consideration.

A. Traits as as Causal Dispositions Ryle (1949) maintained that trait attributions to persons function in the same aggressive" isis manner as dispositional statements in physics. Thus, to say "John is aggressive" soluble." In this sense, akin to saying "The glass is brittle" or "The sugar is soluble." sense, all trait statements are hypothetical propositions that convey law-like relationships. concepts, as applied to objects, is reasonably clear: "To The status of dispositional concepts,

CHAFfER CHAPTER 4

IN IN DEFENSE DEFENSE OF OF TRAITS TRAITS

103 103

be brittle is just to be bound or likely to fly into fragments fragments in such and such conditions" (Ryle, 1949, 1949, p. 43). 43). Thus, the statement "The glass is brittle" can be verified verified with reference to a bilateral reduction sentence (Carnap, 1936) of the following form: If a case x (the glass) satisfies the test condition S (being struck with a stone), then xX is an instance of of C (brittle) if and only if xx shows the response R (shatters). According to the dispositional view, when we say that John is aggressive, we certain circumstances, circumstances^John Johnwill willbehave behave are asserting that it is a good bet that in certain aggressively.But Butthe theform formofofthe thesubjunctive subjunctiveconditional conditionaland andthe theconditions conditionsfor foritsits aggressively. verification are different, different, in several respects, from those involved in the meaning verification verification of dispositional statements applied to physical objects: and verification

1. The "good bet" (Ryle's words) indicates that the subjunctive subjunctive conditional "if and only if." if." This is not damaging to the is probabilistic in form, rather than "if position since it could be argued that all predictions are probabilistic, or certainly all predictions of human behavior. There is, of course, the question of how good for any the bet has to be before we will make it, but that question is problematic for account of the prediction of behavior. condi2. The "in certain circumstances" clause embraces a much larger set of condi tions than is the case with physical objects. Whereas brittleness is a single-tract disposition that can be defined in terms of a reduction sentence involving a single, sufficient occurrence (shattering), traits convey an indefinite series of hypothetical sufficient propositions. It is not clear which of the many possible circumstances should serve disconfirmation as the test condition for John's aggressiveness, nor is it clear whether disconfirmation of the hypothetical proposition in a specific instance could be discounted on the grounds that the wrong circumstance had been selected. by which we we could decide that John had "behaved aggressively" 3. The criteria by are not as evident as shattering in the case of brittleness. Ryle insists that all trait words are dispositional words, and thus makes no provision for actions which may be described as "aggressive actions" (Powell, 1959). In his view, actions may be described as exercises of John's aggressive disposition in the same way that speech acts may be described as exercises of John's knowledge of French. But just as there are no brittle occurrences, there are no aggressive actions. Although the concept specifying a manageable number of of a physical disposition avoids circularity by specifying of criterion responses, this is not so easily achieved with reference to the concept of of a trait disposition. B. aries B. Traits Traits as as Categorical Categorical Summ Summaries Hampshire (1953) contends that trait attributions do not involve hypothetical or statequasi-hypothetical statements. Instead, trait attributions are summarizing state ments that describe the general trend of a person's conduct to date. The claim of of the dispositional statement is of the form, "So far, the word aggressive is the right word to summarize the general trend of John's conduct."

104 104

JERRYS. S.WIGGINS WIGGINS JERRY

Hampshire's arguments rely on a distinction between statements expressing causal properties and categorical statements that merely summarize. For Ryle, statements about human dispositions are statements expressing causal properties properties that can be restated in terms of of reduction sentences of the "if "if.. . . then" form. For statements about human dispositions are summarizing statements that Hampshire, statements do not comment comment on causality. Such statements may also be applied to material ob objects: .. .. .. one one may may often often choose, choose, or or may may be be compelled compelled by by ignorance, ignorance, to to summarize summarize the the general general character character of of some some physical physical things, things, rather rather than than to to describe describe their behavior behavior in in terms terms of of their their physical physical constitution constitution and and of of the the laws laws which which govern govern the the behavior of of objects objects so so constituted. (Hampshire, (Hampshire, 1952, 1952, p. p. 7) 7)

of such statements would be "It tends to rain in Vancouver" and "This Examples of overflow its banks." river tends to overflow Hampshire emphasizes three bases for distinguishing summarizing statements manifestations of from causal statements: (1) Summarizing statements imply that manifestations the disposition have occurred in the past; causal statements do not: sugar may be soluble without ever having dissolved. (2) Summarizing statements imply that a disposition manifests itself itself more or less continuously over some period of of time; disposition causal statements do not: "being electrically charged is a property which may be switched on and off" o f f (p. 8). (3) The manifestations of dispositions described by summarizing statements are various and indeterminate (John's aggressiveness may be manifested in a multitude of of behaviors). Manifestations of causal dispositions are specific and determinate (shattering and dissolving). If, If, in ordinary usage, the statement "John is aggressive" conveys that John has been observed observed to engage in topographically topographically dissimilar dissimilar aggressive aggressive actions over over a period period of of time, then the advantages of Hampshire's summarizing statement account over Ryle's causal statement account are evident. Therefore, it seems worthwhile to consider at least the more obvious objections that might be raised to the implica implications just stated. That the manifestations manifestations of of a trait are heterogeneous and indeterminate is already conceded in Ryle's account. But could we, or would we, describe John as conceded aggressive on the basis of his repeatedly performing a single act in a given situation? of The disposition to perform a single act in a single situation is ordinarily referred perform to as a habit rather than as a trait. In the case of aggressive behavior, such an action might even be referred to as a tic or mannerism. Would it make sense to assert that John is aggressive if John had never been observed to perform perform an aggressive act? Brandt (1970) does not see such an assertion If we knew a person had lived a sheltered life and as a contradiction. He argues: If had never been required to act courageously, we would not infer that the person cannot be courageous. Further, "there are conceivable psychological tests such that, cannot probably a given a certain result on these tests, we would say that the person is probably courageous person" (p. 26). The fact that such an assertion is probabilistic "shows meaning of of courageous (or other trait-names of of interest something, not about the meaning

CHAPTER TRAITS CHAPTER 44 IN INDEFENSE DEFENSEOF OF TRAITS

105 lOS

to us), but about our convictions on what is adequate evidence for trait-ascriptions" (p. 26). It is readily conceded that if we never observed a person in a situation in which a manifestation of of a trait would have been expected or likely, we cannot conclude that the person does not not have the trait. Nor, of course, can we conclude that the person has the trait. We simply do not know. The "conceivable psychological psychological tests" argument is more likely to appeal to philosophers than to those of us familiar with the grim realities reaUties of psychological psychological testing. But to show that such probabilistic inferences inferences might be made, in principle, is not to give an account of the ordinary usage of of trait terms. The major argument that Brandt (1970) puts forth to show that the summary view is "simply wrong" is based on the fact that trait inferences may be made on of a single act: the basis of But how could couldwe wedraw drawsuch suchan aninference, inference,with withhigh highconfidence, confidence,from fromany any amount But how amount of information information about about aa single single situation situation ifif trait-affirmations trait-affirmations were were assertions assertions about about of the frequency frequency of of behavior behavior in in the the past? past? (The (The present present behavior behavior is, is, of of course, course, one one the case; but but to to say say that that aa person person is is courageous courageous is is surely surely not not to to say say merely merely that that he he case; has acted acted courageously courageously once). once), (p. (p. 26) 26) has

There are circumstances in which a trait inference may be drawn on the basis dispositional of a single action. At issue are the relations that may exist between the dispositional (Ai), and past actions that are heteroge heterogeassertion (D), the single action just observed (At), manifestations of the trait (An). (A„). Table I outlines the assertions and negations neous manifestations different sets of of conditions. For simplicity, the single action that may be made under different just observed (At) (Ai) is not considered a subset of of past actions (A„), (An), although both manifestations (At). are members of the larger set of all possible trait manifestations The first statement in Table I asserts that John is aggressive (D), that John pushed the boy aggressively on this occasion (Ai), (At), and that other instances of John's TABLE II Possible Assertions about a Disposition (D), an (Ai), and Past Actions Action Just Observed (At), (A„) n) That Are Manifestations of the Disposition (A

1. 1. Typical assertion 2. 2. Typical negation

3. 4. 5. 5. 6. 6. 7. 8. 8.

Implausible Contrary disposition Special circumstances Out of character Change of character Single occurrence

D& n & At Ai & &A A„ D D& & At Ai & & AnG An'' D D& & At Ai & & An A^ D D& & At Ai & & A„ An & A, Ai & & A„ D& An D D& & At Ai & & An A„ D& & A, Ai & & An A„ D D& &A Ai, & & An A„

G**Overbar Overbardenotes denotesthe theopposite oppositeconnotation; connotation;for forex example, D lacks the disposition under discussion.

106 106

JERRY JERRY S. S.WIGGINS WIGGINS

aggressiveness have been observed in the past (An). (A„). The second statement asserts that none of of these is true. The normalcy of both these statements is recognized by Brandt (1970, pp. 25-26). The third statement is considered by Brandt to be "not contradictory," but, on the evidence evidence he offers, is labeled implausible. At best, one could argue that John is "potentially" aggressive, with all the attendant difficulties difficulties in disproving such an assertion. In the fourth statement John is said not to be aggressive, even though he just pushed the boy, and even though he has acted aggressively in the past. Such a statement can be true when the speaker is able to cite a variety of incidents which serve as evidence of a contrary contrary disposition: disposition: "the final and conclusive argument must be a balancing of of one set of actual incidents against another set of actual incidents" (Hampshire, 1953, p. 6). In this example, it is assumed that accounts of of John's loving, cooperative, and pacifistic behaviors in the past are weighed more heavily than accounts of of his aggressive actions. In the fifth statement John is said to be aggressive, in light of his past history of aggressive actions, even though he did not push the boy on this occasion. Such of an apparent exception to trait attribution rules does not pose a problem for either causal-dispositional or summary view. Ryle (1949) distinguishes tendencies from the causal-dispositional capacities: "tends to" implies "can" but is not implied by it (p. 131). Hampshire (1953) considers the possibility of of an exception "part of of the force of of calling caUing state stateof disposition summarizing statements" (p. 7). But neither position provides ments of an explanation of of this apparent exception. If there is good reason to believe that aggressively, John is aggressive (e.g., his past actions) and he does not push the boy aggressively, special circumstances circumstances such that John "would have" pushed we could appeal to some special the boy aggressively were it not for those circumstances. However, in the case of desirable or socially socially sanctioned sanctioned actions (not pushing), we are unlikely to seek "excuses," "justifications," "extenuations," and the like (Austin, 1957). In Statement 6, John is said to be not aggressive, in light of of the lack of aggressive incidents in his past, even though he pushed the boy on this occasion. Since the action in this case may be considered reprehensible (pushing), excuses or justifications are likely to be sought. Although special circumstances could be cited (e.g., extreme extreme provocation), the best explanation (defense) would seem to be John's record. In fight light of of the lack of of aggressive actions in his past, it could be argued that John's pushing was an action out of of character. This is not to infer, as have some moral philosophers, that character "causes" actions (Pitcher, 1961). Rather, it is to recognize the possibility possibiHty of of uncharacteristic actions in virtue of the argument that although D is dependent on Am A^, At Ai is independent of of D (Powell, 1959). That is, although our characterization of of John as "aggressive" requires past incidents of aggressive actions, we may classify John's pushing the boy as "aggressive" indepen independently of of John's disposition to be or not to be aggressive. It is not clear how an aggressive action could be performed by a nonaggressive person within Ryle's Kyle's account (Powell, 1959). In Statement 7, John does not push the boy nor is he said to be aggressive, of aggression in his past. When an aggressive person even though there are incidents of

CHAPTER 44 IN INDEFENSE DEFENSEOF OF TRAITS CHAPTER TRAITS

107 107

fails to perform an aggressive action, special circumstances may be invoked. But, as in the present present case, when a person with a history of aggressive actions fails to *'not aggressive," the possibility of a change of of behave aggressively and is called "not character may be entertained. Although the "Contrary disposition" account would be more compelling, people do change, or at least we revise our opinions about them. However, this explanation cannot be invoked too often: "Character may change suddenly; but it must not change suddenly too often, or it ceases to be character" (Hampshire, 1953, p. 6). The eighth statement is the one that Brandt considers damaging to the sum sumof a single action (pushing), and in the absence of previous mary view. On the basis of aggressive actions, the disposition to be aggressive is attributed to John. Although such inferences may not be drawn "often," as Brandt puts it, they may, on some occasions, be drawn. The issue is whether or not the inference is consistent with the claim of of a summarizing statement: "So far, the word aggressive is the right word to summarize the general trend of John's conduct." Clearly, a dispositional inference could be in the form of a summarizing statement if the act in question (e.g., pushing) was the first and only action ever presumobserved. But Brandt is concerned with the situation in which previous (and presum ably extensive) observations were not occasions of trait manifestations. He appears disposito be arguing that John was (latently?) aggressive all this time, but that the disposi tion had not been previously brought into play. This might be analogous to a glass that has always been brittle, though never struck, and therefore never shattered. If we were to unpack all, or some, of the hypothetical propositions implied in the statement that John is aggressive, we would have to conclude that the of the subjunctive conditionals (e.g., "If John is frustrated . . . "")) circumstances of had never before been satisfied. That one condition was satisfied on this recent occasion, and that John did perform an aggressive act (however that might be causal-dispositional framework), is apparently sufficient sufficient to determined within the causal-dispositional such an attribution is indeed inconsis inconsisjustify a trait attribution, in Brandt's view. And such of dispositions since no hypothetical propositions are tent with a summary view of implied by that view, and since the attribution would not provide an accurate of conduct to date. summary of Uncontrived Uncontrived examples of of the borderline case considered by Brandt do not immediately immediately come to mind. The research of of Megargee and his associates (Megargee, 1966; Megargee & Mendelsohn, 1962; Megargee & Menzies, 1971) may provide one such example. Briefly, Megargee has studied the histories and personality characteristics of of prisoners who committed extremly assaultive crimes (e.g., murder), moderately assaultive crimes (e.g., battery), or nonassaultive crimes (e.g., robbery). proporBoth case histories and psychological test data suggested that a substantial propor tion of of prisoners who committed extremely assaultive crimes were less aggressive, more controlled, and less likely to have committed previous offenses than other of extremely assaultive crimes also tended to corrob corrobprisoners. Newspaper accounts of offender proves orate these findings: "In case after case the extremely assaultive offender of aggression" (Megargee to be a rather passive person with no previous history of 1966, p. 2).

108 108

JERRY S. WIGGINS WIGGINS

Both Megargee's research and newspaper newspaper accounts of of such "puzzling," "sense "senseless," and "shocking" crimes are fascinating, or at least, extremely interesting. It is informative informative to ask why this is the case. Would an adherent adherent of of the causal causaldispositional view answer, "It "It is amazing that these aggressive persons should should have been so sheltered sheltered from from circumstances that would provide occasions for for the expression of of their their violence"? Perhaps. But newspaper newspaper readers, the psychologists psychologists who conducted conducted this research, and adherents of the summary view would answer, "It is amazing that such heinous crimes could be committed by such nonaggres nonaggressive persons." It is important important to note note that that a "change of character" does not seem to have been been involved involved since psychological testing and behavior behavior observations observations made after after incarceration still revealed a picture of a nonaggressive personality. Murder is, of of course, an aggressive action. But murder murder can be committed by nonaggressive per persons, at least according to the summary view. for the categorical-summary categorical-summary view over the causal-dispositional account To opt for of Ryle's original distinction between dispositions dispositions is not to deny the importance of hand and episodes and occurrences on the other. When When and tendencies on the one hand disposition we say that John is aggressive, we are asserting that the general trend or disposition of his conduct, to date, has been to engage in a variety of of aggressive actions over over of and a period of time. John's "aggressiveness" is not something that occurs over and under and below, his aggressive actions. It is not something that occurs above, or under for short or long periods of of time, in a real world or in a transpatial world. Nor, for from Ryle, is John's aggressiveness the cause of, of, or and here is where we depart from for, his aggressive actions. In addition addition to appearing closer to ordinary ordinary usage, reason for, categorical-summary account carries less metaphysical freight. freight. the categorical-summary

r v . TRAITS T R A I T S AS AS PREDICTORS PREDICTORS OF OF DEHA BEHAV IOR IV. VIOR

A. Prediction in Everyday Life We can, and on occasion do, use our knowledge of persons' past actions as a basis for for predicting their future future behavior. But the extent to which such predictions are made in everyday social transactions has probably been exaggerated. According According of The Human as Scientist, persons are almost to George Kelly's (1955) model of continuously engaged engaged in gathering gathering data, erecting erecting hypotheses, and subjecting subjecting hypoth hypothcontinuously Although this model has provided a heuristic metaphor metaphor eses to test by prediction. Although for psychological research, it should not be interpreted interpreted literally. As Little (1972) for of academico academicohas noted, to say that all persons are scientists (or predictors) smacks of mimesis: "Everyone is just like me." Some individuals may spend significant significant portions portions mimesis: of their lives predicting the future future behavior of others, but others may be mainly of concerned with reflecting reflecting on the past, or pondering the pres present. �nt. The philosophical view view of of persons as in a constant constant state of of readiness to unpack unpack The the conditional predictions entailed by their trait attributions seems equally academ-

CHAPTER CHAPTER 44 IN INDEFENSE DEFENSEOF OFTRAITS TRAITS

109 109

icomimetic. If you were to assert that John is aggressive, and I were to ask you what you meant meant by that, you would would probably cite corroborative corroborative incidents incidents of aggres aggresfrom John's past. It seems unlikely you would reply that you intended sive actions from John were in such and such a situation, then John to convey the proposition proposition that that if 1/John situation, then would act in such and such a way. The equating equating of the meaning of of a statement statement with would verification is a philosophical language game, not an ordinary the method of its verification discourse game. If you and I were psychologists, we might might conspire to contrive a ordinary people seldom engage in such situational test of John's aggressiveness, but ordinary ordinary people be bound bound to agree that practices. Nor would ordinary that the original attribution experiment yielded yielded negative findings. findings. was false if the experiment

B. Prediction in Personality Assessment unresolved questions concerning the extent There may be unresolved extent to which persons engage in predictive predictive behaviors in in ordinary ordinary life, Hfe, but there is little question prediction question that prediction major professional professional activity of the applied personality personality assessment is the major assessment psychologist (Wiggins, 1973). Since the principal charge leveled against the trait construct construct in recent times has been that of a lack of predictive utility (Mischel, 1968; Peterson, 1968), it is instructive to consider what kinds of behaviors, and in what circumstances, trait measures would be reasonably expected to predict. room for refinement refinement of current current psychometric trait Although there is still room Although behavioral measures, the relatively poor showing of such measures in predicting behavioral reflect inappropriateness inappropriateness of the criteria criteria rather rather than shortcom shortcomcriteria may very well reflect ings of the predictors. This line of reasoning has recently been pursued pursued by Martin 1973; Fishbein & Ajzen, 1972, 1972, 1974) 1974) in the Fishbein and his associates (Fishbein, 1973; context of attitude-behavior attitude-behavior relationships. Fishbein Fishbein and Ajzen Ajzen (1974) have demon demonstrated that whereas dispositional measures are relatively poor poor predictors of single multiple acts in varied acts, they are substantially related to criteria based on mUltiple circumstances. evaluation Jaccard (1974) has recently extended this line of reasoning to the evaluation of traditional trait measures. He assembled a set of multiple-act multiple-act criterion measures measures of of of dominance that included behaviors likely to be pedormed performed by dominant dominant persons subjects were also administered administered Gough's (1957) in a variety of circumstances. His subjects dominance scale, and a single, self selfCPI dominance scale, Jackson's (1967) PRF dominance rated rated scale of dominant tendencies. The average correlations of the trait scales with individual dominant dominant behaviors were barely .20 and were not statistically statistically significant. significant. In contrast, the correlations of the trait measures with the sum of the multiple acts were close to .60 and were highly significant. If we choose, as did Mischel (1968), "personality coefficient" coefficient" to express the relation between trait and behaviors as a "personality based on the average relation between between a trait and single behaviors, then then we are based likely to obtain obtain coefficients coefficients of disappointing size. But if we choose to express the relation as the correlation between a trait measure measure and a criterion of mUltiple multiple acts in multiple situations, there is good reason to believe that the magnitude of validity validity in multiple situations, there is good reason to believe that the magnitude of

110 no

JERRY S. WIGGINS

coefficients will be similar to those found by Jaccard (1974). And such a choice coefficients would be consistent with a summary view of traits. A psychometric measure of a trait should reflect reflect the the general general trend trendof ofaaperson's person's conduct to date. In assessing a single individual, the trait measure should be based on items that provide a broad and representative sampling of relevant acts in multiple situations. Most prediction in personality assessment is nomothetic, howhow ever, and thus we must devise measures that also reflect the relative relativetendencies tendenciesof of persons to behave in certain ways on certain occasions. Fishbein's work in attitude measurement is also relevant here, since he has demonstrated the importance of of several previously neglected item properties that should increase differentiation differentiation among persons. These include: (1) the probability of a trait given an act, p(T I| A), (2) the probability of a trait given the act is not performed, p(T I| A), and the base act in in the the situation situationp(A). The difference between between the thefirst twoprobabilities probabilities rate of the act p(A). The first two & Ajzen, 1974). is a powerful index of item validity (Fishbein & The predictive utility of a trait measure is a direct function of the ability of of the measure to postdict the general trend of an individual's past behavior. Once this is realized, it becomes clear that attempts to assess underlying motives, latent tendencies, and the like are quite beside the point. Prediction from trait measures is based on the logic of the old adage that the best indication of what a person will do in the future is what that person has done in the past. When this is not the case, other measures may be called for, but they are not properly called trait measures. V. VIOR V. TRAITS AS EXPLANATIONS OF REnA BEHAVIOR The summary view of traits that I have advanced thus far does not comment on the causal properties of traits or on the use of traits as explanations of behavior. It is true, of course, that a summary of a person's conduct to date does not provide an account or explanation of his or her conduct. At issue is whether ordinary trait attributions are intended intended to be explanations. Attribution theorists think they are so intended and they view trait attributions as "naive causal inferences" based for the most part on insufficient insufficient data (Heider, 1958; 1958; Jones & & Nisbett, 1971; 1971; Kelley, 1967). But the subjects of attribution studies are required required to provide "causal" explanations on the basis of insufficient insufficient data, leaving open the possibility that naivete naivet6 may be attributed more justly to those who design such experiments. According to the summary view, trait attributions are made in just those circumstances in which the speaker is ignorant of the true causes of the behavior in" for explicit pattern at issue. It may be that such attributions are meant to "stand in" explanations (Harre (Harr6 & & Secord, 1972, 1972, p. 270), but it seems farfetched to regard prescientific attributions of traits as proffered explanations of either a scientific or a prescientific sort. When I say that it tends to rain in Vancouver (or that Vancouver is rainy), I do not presume to be offering an explanation of that tendency, nor to have intimate the nature of whatever meteorological forces may may be operative. Were knowledge of the you to to ask ask me me why it it rains rains in in Vancouver, Vancouver, which which is is aa quite quite different different issue, issue, you you would would you

CHAPTER 44 IN INDEFENSE DEFENSEOF OF TRAITS TRAITS CHAPTER

111 111

expose my "naive inferences" based on my lack of knowledge of meteorology. When I say that John is aggressive (or tends to be aggressive), I would also not be providing anything anything resembling an explanation explanation so presumptuous as to think that I was providing of John's acts in terms of of a well-substantiated theory of of human nature. of There are circumstances in which the ascription of a trait to a person serves of that person's behavior. If you are not acquainted with as a partial explanation of John and if you ask me why John pushed the boy on a certain occasion, I might reply that John is aggressive. In effect, I am saying that such behavior is not unusual or unexpected for John, and such an "explanation" might serve as an answer to teUing you that your question. However, if you and I both know John well, my telling John is aggressive does not answer your question. Were I to inform you that the boy had pushed John yesterday, you might very well feel that I had provided a of the incident (see Averill, 1973, p. 280). These two different different satisfactory account of "causal attributions" correspond, roughly, to the person versus situation dichotomy & Nisbett, 1971). But the first is simply discussed by attribution theorists (e.g., Jones & a statement to the effect that the behavior is not unusual and the second is primarily a justification of that behavior. Neither statement specifies the efficient cause of the behavior in question. Although laymen use trait terms to "stand in" for explanations, psychologists have used them as explanatory constructs. In moving from the level of ordinary language description to the level of theoretical explanation, virtually all "trait theorists" (Allport, 1937; Cattell, 1957; 1957; Eysenck, 1953; Guilford, 1959; 1959; Murray, 1938) consider traits to be causal entities rather than categorical summaries. Patterns energized and granted explanatory status as efficient efficient causes of of past conduct are energized future behavior. To avoid charges of circularity of reasoning, trait theorists speak of traits as "hypothetical constructs," inferred from patterns of past conduct and suffer from from used to predict future behavior. But hypothetical trait constructs often suffer an intrinsic conceptual fuzziness that blurs distinctions between reasons and causes in the explanation of of social behavior. Persons are seen as having certain dispositions of having certain hypothetical constructs, but these constructs are an in virtue of of institutional facts and efficient efficient natural causes. ambiguous blend of The tradition of using traits as causal explanations of behavior has a long history that stems from the early faculty psychologists to the present day. Behavior is explained by reference reference to generative mechanisms (traits, dispositions, needs, instincts, motives, etc.) which are structurally isomorphic with the behavior pattern requiring explanation. Thus a person behaves aggressively because he or she "has" an aggressive trait, need, or whatever that causes him or her to so behave. The behavior is described as phenotypic, manifest, or surface, while the trait is genotypic, latent, or source. The general pattern of the person's social behavior is mirrored in an underlying "structure" of personality traits. This type of theorizing seems, to me, implausible. Recall, first of all, that the criteria for classifying classifying behaviors as instances of a trait are institutional (social) in nature. These criteria would be expected to vary, not only cross culturally, but within cultural subgroups (what is "aggressive" behavior for one socioeconomic

ttl 112

JERRY JERRY S. S.WIGGINS WIGGINS

group may not be for another). Similarly, the situations or "occasions" on which these behaviors occur will be defined quite differently differently in different different cultural settings. Hence, the subject of an efficient efficient cause or of a physical basis for a set of institutional institutional meanconventions must be approached with great caution. Institutional rules impart mean of ing to topographically dissimilar actions in varied circumstances. The existence of isomorphic with institutional rules generative mechanisms which are structurally isomorphic seems most unlikely. It conjures up the image of a cultural homunculus within each of us. of The organization or "structure" of traits may well reside within a pattern of interrelated institutional rules, rather than within individuals. This strong form of the "biosocial" position is implicit in the writings of ethnomethodologists (e.g., ethoGarfinkel, 1967; Goffman, 1959), and is quite explicit in the formulations of etho Harr6 & & Secord, 1972) who would locate "personas" "personas" genic personality theorists (e.g., Harre in the shared perceptions of others and who reject substantive interpretations of traits. of Within the realm of personality assessment, one of of the most highly "struc "structured," replicable, and theoretically meaningful meaningful systems of of behavioral classification classification of interpersonal behavior (Wiggins, 1968, 1973). has been found in the domain of Here I refer to the work of Leary (1957) and his associates and its subsequent refinement by Lorr and McNair (1963), Stern (1970), and others. development and refinement observational and self-report data, it is clear that the relation among From both observational the major categories of interpersonal behavior may be represented structurally as structure? a two-dimensional circumplex of trait vectors. But what is reflected in that structure? The traditional view of the interpersonal circumplex is that it reflects the of needs or traits within individuals. Descriptively this organization organization of refers to summaries of past conduct; but theoretically, the organization is held to mirror the arrangement of of internal needs, dynamisms, or whatever. I think it more plausible that the structure of interpersonal behavior mirrors a set of interrelated social rules for classifying classifying behavior in terms of of its likely interpersonal consequences. The remarkable structural convergences that have been found among diverse theo theoretical systems of of interpersonal interpersonal behavior (Foa, 1961) do not stem from the similarity of of generative mechanisms postulated (needs, traits, dynamisms, etc.). Instead, the convergences reflect the common system of institutional rules that classifies interper interpersonal behavior measured by different different techniques. Further, the promising typologies developed by Leary (1957) and Lorr, Bishop, and McNair (1965) for identifying identifying central interpersonal dispositions (e.g., managerial-autocratic) are more properly viewed as institutionally defined roles than as internal generative mechanisms. Finally, the sequential patterns of of interpersonal interpersonal transactions that have been de described by Leary and others (e.g., power provokes obedience) seem to represent normative rules of conduct rather than mechanistic chain reactions. Once we have clearly separated what is to be explained (patterns of of past conduct) from plausible explanatory constructs (generative mechanisms that cause persons to be rule-following agents), the enormity of the task of personality theory becomes evident. In our present state of ignorance of the nature of human nature,

CHAPTER CHAPTER 44 IN INDEFENSE DEFENSEOF OF TRAITS TRAITS

113 113

we have lapsed into the layman's tendency to allow trait terms to "stand in" in" for genuine explanations. As a consequence, we have neglected the promising leads of those trait theorists who have recognized, at least implicitly, imphcitly, the importance of a distinction between traits and generative mechanisms. Cattell (1946) has, for many years, distinguished distinguished surface traits (ordinary lan language descriptions of person attributes) from source source traits (underlying (underlying generative mechanisms responsible for behavior classified by ordinary language). Perhaps the conceptual significance of this distinction has been overlooked because: (1) despite the use of neologisms, "source" traits appear to represent familiar "surface" themes, (2) it does not seem intuitively obvious that source mechanisms can be detected by the multivariate analysis of ordinary language attributions, and (3) identified "source" patterns have been interpreted, rather casually, by reference to psychoana psychoanalytic mechanisms. But the conceptual conceptual distinction between surface surface traits and source mechanisms resembles, or is at least compatible with, Chomsky's (1965) distinction between the surface and deep structures of of language, a model that may be especially useful in stimulating thought about the plausible origins of of personality traits 1972; Stagner, 1973). (Harre & Secord, 1972;

ACKNOWLEDGMENTS am grateful grateful to to Kenneth Kenneth H. H. Craik Craik and and Brian Brian R. R. Little Little for for their their advice advice and and encouragement. encouragement. II am

REFERENCES Alker, H. H. A. A. (1972). (1972).IsIspersonality personalitysituationally situationallyspecific specificor orintrapsychically intrapsychicallyconsistent? consistent?Journal Journal Alker, of Personality, Personality,40, 40,1-16. 1-16. of AUport, G. G. W. W. (1937). (1937). Personality: Personality:AApsychological psychologicalinterpretation. interpretation. New York: Holt. New York: Holt. Allport, Anscombe, G. E. M. M. (1963). Intention Intention(2nd (2nded.). ed.).Oxford, Oxford,England: England:Basil BasilBlackwell. Blackwell. Austin, J. L. L. (1957). (1957). A A plea plea for for excuses. excuses. Proceedings Proceedingsof ofthe theAristotelian AristotelianSociety, Society, 1-30. 57, 57, 1-30. Averill, J. R. (1973). The dis-position of psychological psychological dispositions. Journal Journalof ofExperimental Experimental Research 6, 6, 275-282. ResearchininPersonality, Personality, 275-282. Bowers, K. B. (1973). Situationism in psychology: An analysis and critique. Psychological Psychological Review, 80, Review, 80,307-336. 307-336. Brandt, R. B. (1970). Traits of character: character: A conceptual analysis. American American Philosophical Philosophical Quarterly, Quarterly,7,7,23-37. 23-37. Camap, R. (1936). Testability and meaning. Philosophy Philosophy of of Science, Science,3,3,419-471. 419-471. Cattell, R. R. B. B. (1946). (1946). The The description descriptionand andmeasurement measurement personality. Yonkers-on-Hudson, Cattell, ofof personality. Yonkers-on-Hudson, NY: World Book. NY: Cattell, R. R. B. B. (1957). (1957). Personality Personalityand andmotivation motivationstructure structure and measurement. Yonkers-onCattell, and measurement. Yonkers-onHudson, NY: NY: World Book. Chomsky, N. N. (1965). (1965). Aspects of of the the theory theoryof ofsyntax. syntax.Cambridge, Cambridge,MA: MA:MIT MITPress. Press. Chomsky, Craik, K. K. H. H. (1969). (1969). Personality Personality unvanquished. unvanquished. Contemporary ContemporaryPsychology, Psychology,14,14,147-148. 147-148. Craik, Edwards, A. A. L. L. (1970). (1970). The Themeasurement measurementofof personality traits scales inventories. personality traits byby scales andand inventories. NewNew Edwards, York: Holt, Rinehart & & Winston.

114 114

JERRY S. WIGGINS JERRYS. WIGGINS

Eysenck, H. J. J. (1953). (1953). The The structure structureofofhuman human personality. London: Methuen. Eysenck, H. personality. London: Methuen. Fishbein, Mor Fishbein, M. M. (1973). (1973). The The prediction prediction of of behaviors behaviors from from attitudinal attitudinal variables. variables. In In C. C. D. D. Mor3-31). New teen & & K. K. Sereno (Eds.), Advances Advances in in communication communicationresearch research(pp. (pp. 3-31). New York: Harper & & Row. Psychology, Fishbein, Fishbein, M., M., & & Ajzen, Ajzen, I. (1972). (1972). Attitudes and and opinions. opinions. Annual Annual Review Review of of Psychology, 23, 23, 487-544. 487-544. Fishbein, M., & & Ajzen, Ajzen, I.I. (1974). (1974). Attitudes Attitudestoward towardobjects objects as aspredictors predictorsof ofsingle singleand andmultiple multiple behavioral criteria. Psychological PsychologicalReview, Review,81,81,59-74. 59-74. Chicago: Aldine. Fiske, D. W. (1971). Measuring Measuringthe theconcepts conceptsofofpersonality. personality. Chicago: Aldine. Foa, Foa, U. G. (1961). Convergences in the analysis of the structure of interpersonal behavior. Psychological PsychologicalReview, Review,68, 68,341-353. 341-353. NJ: Prentice-Hall. Garfinkel, H. (1967). Studies Studiesin inethnomethodology. ethnomethodology.Englewood EnglewoodCliffs, Cliffs, NJ: Prentice-Hall. Goffman, Goffman, E. E. (1959). (1959). The The presentation presentationofofself selfinineveryday everydaylife. life.New NewYork: York:Doubleday DoubledayAn Anchor Books. chor CaliforniaPsychological Psychological Inventory Manual. Alto, Consulting Gough, H. G. (1957). (1957). California Gough, H. G. Inventory Manual. PaloPalo Alto, CA:CA: Consulting Psychologists Psychologists Press. Guilford, J. J. P. P. (1959). (1959). Personality. Personality.New NewYork: York:McGraw-Hill. McGraw-Hill. Guilford, Hampshire, S. S. (1953). (1953). Dispositions. Dispositions. Analysis, Analysis, 14, 14,5-1 5-11. 1. Hampshire, explanationof ofsocial socialbehaviour. behaviour.Oxford, Oxford, England: Harr6, R., R., & & Secord, Secord, P. P. F. F. (1972). (1972). The The explanation England: Harre, Basil Blackwell. Blackwell. Heider, F. F. (1958). (1958). The Thepsychology psychology of of interpersonal interpersonalrelations. relations. New York: Wiley. Heider, New York: Wiley. Holt, R. R. R. (1958). (1958). Clinical Clinical and statistical statistical prediction: prediction: A A reformulation reformulation and and some some new new data. Holt, R. data. Journal 1-12. Journalof ofAbnormal Abnormaland andSocial SocialPsychology, Psychology,56,56, 1-12. Jaccard, Jaccard, J. J. J. J. (1974). (1974). Predicting Predicting social social behavior behavior from from personality personality traits. traits. Journal Journalof ofResearch Research in Personality, Personality,7,7,358-367. 358-367. in Jackson, D. D. N. N. (1967). (1967). Personality PersonalityResearch ResearchForm Form Manual. Goshen, Research PsycholoJackson, Manual. Goshen, NY:NY: Research Psycholo gists Press. gists Press. Jones, E. E. E., & Nisbett, Nisbett, R. R. E. (1971). (1971). The The actor actorand andthe theobserver: observer:Divergent Divergent perceptions perceptions of of E., & the causesof ofbehavior. behavior.New NewYork: York:General GeneralLearning LearningCorporation. Corporation. the causes Kelley, H. H. (1967). Attribution theory in social psychology. Nebraska Nebraska Symposium Symposium on Motivation, Motivation, 14, 14,192-241. 192-241. Kelly, G. A. (1955). The (Vols. l and 2).2). New York: Norton. Thepsychology psychologyof ofpersonal personalconstructs constructs (Vols. 1 and New York: Norton. personality. New York: Ronald Press. Leary, T. (1957). Interpersonal Interpersonaldiagnosis diagnosisofof personality. New York: Ronald Press. andand research. NewNew York: Ran Levy, L. H. (1970). Conceptions Conceptions of ofpersonality: personality:Theories Theories research. York: Random House. Little, B. R. (1972). Psychological man as scientist, humanist and specialist. Journal Journal of of 6, 95-118. Experimental Personality, ExperimentalResearch Researchinin Personality, 6, 95-118. Lorr, M., Bishop, P. F., & & McNair, D. M. M. (1965). Interpersonal types among psychiatric patients. Journal of Abnormal Psychology, 468-472. patients. Journal of Abnormal Psychology,70,70,468-472. Lorr, M., & & McNair, D. M. M. (1963). An interpersonal behavior circle. Journal Journalof ofAbnormal Abnormal and Social Social Psychology, Psychology,67, 67,68-75. 68-75. Megargee, E. I. (1966). Undercontrolled and overcontrolled overcontroUed personality types in extreme antisocial aggression. Psychological (No. Whole 611). PsychologicalMonographs, Monographs,8080 (No.3, 3, WholeNo. No. 611). Megargee, E. I., & & Mendelsohn, G. A. (1962). A cross-validation of twelve MMPI indices of hostility and control. Journal 431-430. Journalof ofAbnormal Abnormaland andSocial SocialPsychology, Psychology,65,65, 431-430. Megargee, E. I., & & Menzies, E. S. (1971). The assessment and dynamics of aggression. In P. McReynolds (Ed.), Advances in in psychological psychological assessment assessment(Vol. (Vol.2,2,pp. pp.133-156). 133-156). Palo Alto, CA: Science and Behavior Books.

INDEFENSE DEFENSEOF OF TRAITS CHAPTER 44 IN TRAITS

115 115

Mischel, W. W. (1968). (1968). Personality Personalityand andassessment. assessment.New New York: Wiley. Mischel, York: Wiley. Mischel, Mischel, W. W. (1969). (1969). Continuity Continuity and and change change in in personality. personality. American American Psychologist, Psychologist,24,24,101210121018. York: Holt, Winston. Mischel, (1971). Introduction Mischel, W. W. (1971). Introductiontotopersonality. personality.New New York: Holt,Rinehart Rinehart& & Winston. Mischel, W. (1973a). On the empirical dilemmas of psychodynamic approaches: Issues and alternatives. Journal Journalof ofAbnormal AbnormalPsychology, Psychology,82, 82,335-344. 335-344. Mischel, W. W. (1973b). (1973b). Toward Toward aa cognitive cognitive social social learning learning reconceptualization reconceptualization of of personality. personality. Mischel, Psychological PsychologicalReview, Review,80, 80,252-283. 252-283. Murray, H. H. A. A. (1938). (1938). Explorations Explorationsininpersonality. personality.New New York: OxfordUniversity University Press. York: Oxford Press. Murray, Peterson, R. (1968). (1968). The The clinical clinical study of of social behavior. York: AppletonPeterson, D. R. behavior. New New York: AppletonCentury-Crofts. Century-Crofts. PhilosophicalQuarterly, Quarterly,11,11, 201-212. Pitcher, G. (1961). (1961). Necessitarianism. Necessitarianism. Philosophical Pitcher, G. 201-212. Powell, B. B. (1959). (1959). Uncharacteristic Uncharacteristic actions. actions. Mind, Mind, 68, 68,492-509. 492-509. Powell, Ryle, G. G. (1949). (1949). The The concept conceptof ofmind. mind.New NewYork: York:Barnes Barnes&&Noble. Noble. Ryle, Searle, J. J. R. R.(1969). (1969).Speech Speechacts: acts: essay in the philosophy of language. Cambridge, England: AnAn essay in the philosophy oflanguage. Cambridge, England: Searle, Cambridge University University Press. Cambridge Press. Stagner, R. R. (1973, (1973, August). August). Traits Traitsare arerelevant: relevant:Theoretical Theoretical analysis empirical evidence. analysis andand empirical evidence. Stagner, Paper presented at at the the annual annual meetings meetings of of the the American American Psychological Psychological Association Association Paper presented in Montreal. Montreal. in Stern, Measuring person-environment congruence in education Stern, G. G. G. G. (1970). (1970). People Peopleinincontext: context: Measuring person-environment congruence in education industry.New NewYork: York:Wiley. Wiley. and industry. Vernon, P. P. E. E. (1964). (1964). Personality Personalityassessment: assessment: critical survey. New York: Wiley. Vernon, AA critical survey. New York: Wiley. P. L. L. (1973a). (1973a). Psychodynamics, Psychodynamics, behavior behavior therapy, therapy, and the implacable implacable experimenter: experimenter: Wachtel, P. Wachtel, and the An inquiry inquiry into into the the consistency consistency of of personality. personality. Journal of Abnormal Psychology, Psychology, Journal of An 82, 324-334. 324-334. 82, Wachtel, P. P. L. L. (1973b). (1973b). On On fact, fact, hunch, hunch, and and stereotype: stereotype: A A reply reply to to Mischel. Mischel. Journal Journal of of Wachtel, Abnormal Psychology, Psychology,82, 82,537-540. 537-540.

Wallach, M. M.A., A.,& &Leggett, Leggett,M. M.I.I.(1972). (1972).Testing Testingthe thehypothesis hypothesis that a person will consistent: Wallach, that a person will bebe consistent: in size size of of children's children's drawings. Journal Stylistic consistency consistency versus versus situational situational specificity specificity in Stylistic drawings. Journal of Personality, Personality,40, 40,309-330. 309-330. of Wiggins, J. J. S. (1964). A review of Vernon, P. E., Personality assessment: A critical survey. Educational 24, 24, 983-985. Educationaland andPsychological PsychologicalMeasurement, Measurement, 983-985. and prediction: Principles ofpersonality assessment. Reading, Wiggins, Wiggins, J. J.S.S.(1973). (1973).Personality Personality and prediction: Principles of personality assessment. Reading, MA: MA: Addison-Wesley. Wiggins, J. J. (1971). The psychology ofpersonality. Wiggins, J. J. S., S.,Renner, Renner,K. K.E., E.,Clore, Clore,G. G.L., L.,&&Rose, Rose,R.R. (1971). The psychology ofpersonality. Reading, MA: MA: Addison-Wesley.

This Page Intentionally Left Blank

CHAPTER S5

INDIVIDUALS AND THE DIFFERENCES BETWEEN THEM JAMES T. T. LAMIELL GEORGETOWN UNIVERSITY UNIVERSITY GEORGETOWN

Sechrest once observed that to field of to too too great great an an extent extent the the field of personality personality has has been been dominated dominated .. . .. by psychologists pursuing pursuing the the ubiquitous ubiquitous but but elusive, elusive, and and maybe maybe even even chimerical, chimerical, psychologists differencesbetween betweenpersons. persons. Paradoxically,these thesepsychologists psychologistshave haveoperated operated differences Paradoxically, more often often than than not not from from aa theory theory that that posits posits not not differences differencesbut butuniversals. universals. more Freud's developmental developmental states, states, Adler's Adler's striving striving for for superiority, superiority, Jung's Jung's animusFreud's animus anima, Maslow's Maslow's need need hierarchy, hierarchy, and and many many other other concepts concepts were were meant meant to to apply apply anima, everyone everywhere, everywhere, but butpsychologists psychologistsbecame becamebogged boggeddown downininstudying studying to everyone differences in in aa way way that that has has never never been very productive. productive. differences been very (1976,p.p.4,4,emphasis emphasisadded) added) (1976,

own position on personality psychology's My own psychology's current currentstate statetakes takesthese theseremarks remarks My Contra the theview view that that has hasdominated dominated the thethinking thinking of of by Sechrest very seriously. Contra mainstream personality investigators for for much of the the greater part of the the present century, I have have been been arguing arguing that that the the assessment assessment and and study study of ofindividual individual differences differences to the is fundamentally and and irremediably ill-suited to the task of advancing personality As an an alternative to what many have, for years, years, been pleased to to believe theory. As theoretiqualifies as nomothetic inquiry, I have further argued that the overarching theoreti be better served by by inquiry that cal objectives of personality psychology would be proceeds "idiothetically" "idiothetically" (Lamiell, 1981). However infelicitous that neologism may have been, been, what what II have have meant meant to to suggest suggest by by it it is is an an approach approach to to the the investigation investigation of of have phenomena which whichrespects respectsthe theindividuality individualityofofthose thosephenomena-the phenomena—the psychological phenomena fact that that every every perception, perception, emotion, emotion, cognition, cognition, and and action action is is someone's-without someone's—without fact H ANDBOOK OF HANDBOOK OF PERSONALITY PERSONAUTY PSYCHOLOGY PSYCHOLOGY

117 117

COPYRIGHT COPYRIGHT C © 1991 1997 BY BY ACADEMIC ACADEMIC Puss PRESS.. ALL ALL lIGHJ'S RIGHTS OF OF REPIODUcnOIi REPRooucnoN III w ANY ANY FOIM FORM IESElVlD. RESERVED.

118 118

JAMES T. T.LAMIELL JAMES LAMIELL

compromising the search for truly general or nomothetic principles, that is, principles compromising in terms of ac of which one might understand the perceptions, emotions, cognition, ac1990a, 1990b). "Idiothetic" tions, and so forth of everyone everyone (Lamiell, 1987, 1987,1990a, *'Idiothetic" inquiry accommodates accommodates individual differences differences without making those differences differences themselves themselves focus of inquiry. the focus Consistent with the editors' vision of the present volume as a handbook, my argument as it has been been objective in this chapter is to trace the major lines of my argument developed to date. The reader reader interested in a more detailed detailed exposition is referred referred developed Psychology of of Personality: Personality: An An Epistemological Epistemological Inquiry Inquiry (1987; see also LaLa to my Psychology miell, 1990a).

I. WHY WHY INDmDuAL INDIVIDUAL DIFFERENCES RESEARCH CANNOT ADVANCE PERSONALITY THEORY Actually, the case against individual differences differences research as a framework framework for the advancement advancement of personality psychology's theoretical concerns is logically quite straightforward, and can be stated concisely as follows: 1: Any Thesis A n y theory of personality is a conceptual framework framework designed to Thesis 1: individual behavior/psycho behavior/psychoprovide explanations for and hence an understanding of individual functioning.^ logical functioning.1 Thesis 2: Except under hypothetical conditions conditions never realized in practice, the Thesis 2: reliability and validity coefficients coefficients and other statistical indices generated by studies studies of individual differences differences variables (alone or in combination with "situational" or of "treatment" variables; see Cronbach, 1957) bear bear no legitimate legitimate interpretation interpretation of of any any kind kind whatsoever whatsoever at the level level of of the individual. individual. Conclusion: Conclusion: Knowledge Knowledge of the sort yielded by individual differences differences research is fundamentally fundamentally and irremediably ill-suited to the task of advancing theories of individual behavior/psychological behavior/psychological functioning-however functioning—however useful useful that same knowl knowledge might be for other purposes—and purposes-and the discipline of personality psychology is therefore therefore in need of a viable alternative research paradigm. T o the best of of my knowledge, no one has ever seriously questioned Murray's To (1938) contention that, in the psychology of of personality, "the objects of study individual organisms organisms and not not aggregates aggregates of of organisms" organisms'' (p. 127, 127, emphasis are individual added). Thus, we may take Thesis 1 as one for which there is and has always been general agreement. But if this is so, and if Thesis 2 is also true, then the force of logic, and the debate over whether investigators investigators conclusion follows by force Thisisisnot nottotodeny denylegitimate legitimatetheoretical theoreticaldivergences divergencesconcerning concerningsuch suchmatters mattersasasthe thesources sources I^This of and the the dynamics dynamics of of its its development. development. It It is is merely merely to to point point out out that that individual individual persons persons of personality and the loci loci of of personality functioning, whatever whatever its its presumed presumed sources sources and and dynamics. dynamics. This This isis true, true, for for are the personality functioning, are example, even for for theorists theorists who who would would emphasize emphasize various various facets facets of of socialization socialization in in their their theoretical theoretical example, even conception 1984; Hurrelmann, of personality personality and and its its development development (see, (see, e.g., e.g.,Harre, Harr6,1984; Hurrelmann, 1988). 1988). conception of

CHAPTER CHAPTER 55

INDIVIDUALS AND DIFFERENCES

119 119

should abandon traditional "nomothetic" inquiry in favor of a more apposite framework framework is over. This is only how things ought to be, since, as a matter of of logical fact, Thesis 2 above is true. Alas, it is just this point with which apologists for conventional "nomothetic" inquiry cannot or will not reconcile themselves for (see, (see, e.g., Dar & Serlin, 1990; 1990; Ozer, 1990; 1990; cf. Lamiell, 1990b), and it is at least for this reason that the debate continues. partly for A. Knowledge Knowledge about about Individual Individual Differences Differences Variables Variables Is Is Not Not A. Knowledge about about Individuals Individuals Knowledge difficulty here is vividly illustrated by the following example, which, though The difficulty anecdotal, is not apocryphal: Several years ago I attended a lecture given by a personaUty investigator who, in the course of his senior and still very prominent personality for making generalizations about individuals on the basis comments, chided a peer for of experimental treatment group means. The lecture then proceeded with a very enthusiastic report of numerous validity coefficients coefficients he had obtained in a longitudi longitudinal study of selected individual difference difference variables. I later asked the speaker if he was in any way troubled by the fact that the Pearson product-moment product—moment correlation coefficients by which he was placing such store as grounds for his generalizations coefficients themselvesgroup groupmeans. means.He Hereplied, replied,"Well, "Well,there thereare aregroup group about individuals were themselves means and then there are group means." And so there are. Specifically, and with regard to the legitimacy of using group means as empirical grounds for generalizations about individuals, there are those around which the variances are zero, and then there are those around which the logically support support at at least least some somegeneralizations generalizations variances are not zero. The former will logically product-moment about individuals, and the latter will not. So too with the Pearson product-moment coefficients that, in the (all too) appropriate words of Bern and Allen correlation coefficients (1974), are "the sacred coin of the realm" in mainstream personality research: there are correlations which are perfect, and then there are correlations which are not perfect. The former former will bear at least some interpretation at the level of the individual precisely because they are, in effect, group means around which the variances are zero. The latter, that is, correlations which are not perfect, will not bear any interpretation at the level of the individual precisely because they are, in effect, effect, group means around which the variances are not zero. The problem here is not that a statement about an individual based on a group mean around which the variance is nonzero, or on a Pearson r of which the absolute value is less than 1.00, is knowably false for for all individuals. The problem any is that such a statement is not knowably true for for any individual, individual, and and this this isis because because the statement is certainly false for some individuals-though we could certainly false for some individuals—though we couldnot notsay saywhich which of ones without investigating the matter case by case-and case—and possibly false for for all of the individuals. Within a discipline in which the overriding objective is to explain and understand the behavior/psychological functioning functioning of individuals, it is difficult difficult after whom to imagine an epistemologically worse state of affairs. That the person after personality psychologists' most prized statistic has been named would have been

120 120

JAMES LAMIELL JAMEST. T.LAMIELL

untroubled by this state of affairs affairs is properly seen not as a failure of discernment on the part of Karl Pearson (1857-1936) but instead as a reflection reflection of of the fact that not interested interested in individuals: individuals: he was not It It is is almost almost impossible impossible to to study study any any type type of of life life without without being being impressed impressed by by the the small importance importance of of the the individual individual.. .. .. Evolution Evolution must must depend depend upon upon substantial substantial small changes in in considerable considerable numbers numbers and and its its theory theory therefore therefore belongs belongs to to that that class class changes of phenomena which statisticians statisticians have havegrown grownaccustomed accustomedto torefer referto toas asmass massphepheof phenomena which nomena. nomena.

(Pearson, p. p. 3; 3; quoted in in Porter, 1986, p. 306) (Pearson, 190111902, 1901/1902, quoted Porter, 1986, p. 306)

It is arguably one of of the great ironies of late twentieth century psychology that Pearson's most visible legacy has become the linchpin of a paradigm widely fancied individuality! as appropriate for handling questions about—of about-of all things—human things-human individuality! But irony or not, since the problem identified above is a logical and not an empirical one, it is not going to vanish or wilt in the face of any findings of any empirical study. One either recognizes that knowledge about individual differences variables is neither equivalent to nor substitutable for knowledge about individuals or one violates logic. There are no other choices, even where there seem to be. For example, many (perhaps most) contemporaries are pleased to believe that the traditional paradigm can be rescued from this critique by using group means or other aggregate statistics as the basis for probabilistic probabilistic statements about & Jackson, 1986a, pp. 471-473). Let us consider individuals (see, e.g., Paunonen & epistemological standpoint. the matter from an epistemological Take a statement of the form, "The probability is p p that Smith will do X." of p could such a statement be empirically The question is, under what values of verified or falsified falsified as a claim to knowledge about Smith? Clearly, the values 1.0 and 0 will "work" here, because substituting either one of of those values for p would amount to an assertion of certainty that Smith will (if (if p = = 1.0) or will not (if p = = 0) do X. As a claim to knowledge about Smith, either statement statement logically admits the possibility of disconfirmation in the face of evidence evidence that Smith does or does not do X. Brief Brief reflection will reveal, however, that the values 1.0 and 0 are the only values that will "work" in this context. Let .9, for for example. example. As As aa claim claim to to knowledge knowledge about about Smith, Smith, how how is is the the statement statement p p == .9, "The probability is .9 that Smith will do X" possibly to be empirically evaluated? If Smith in fact does X, is the statement to be considered verified or falsified, and in either case why? If Smith does X, it is not possible that the probability that he would have done so was 1.0 all along, and not .9? If so, then what is the truth value of the assertion that the probability was .9? Alternatively, if Smith does not do X, of would this disconfirm the assertion that the probability was .9 that he would? If so, why? After all, the assertion made no claim to certainty about what Smith would do. And yet if not, why not? If Smith has in fact not done X, then perhaps the probability that he would do X was in fact 0 all along and not .9 at all. Again, when probabilistic statements about individuals are based on the results of of individual individual differences research, how are such statements to be evaluated as claims to scientific differences scientific knowledge about those individuals?

CHAPTERS5 CHAPTER

INDIVIDUALS AND DIFFERENCES INDMDUALS

121 121

The answer, of course, is that they are are not to be so evaluated. The reason is that they cannot be so evaluated, and that is the problem. When it is based on the findings of individual differences differences research, a statement such as "the probability is .9 that Smith will do X " simply means that given Smith and 99 other individuals X'* identical to Smith with respect to the predictor variable(s), 90 will do X and 10 will left untouched by the not. Just which 90 will and which 10 will not is a question left statement not probability statement. This means that not only is the probability statement claim—least of all a scientific scientific one-to one—to knowledge about Smith, it is not really a claim-least really a claim to knowledge about any one of the other 99 individuals either. It is quite literally a claim to knowledge about no one, one, and that is why it fails to get at anything consequential in the domain of personality theorizing. Now some would point out, as did Paunonen and Jackson (1986b, p. 472, footnote 1), that a statement statement of the form form "The probability is .9 that Smith will do X X"" might be based on evidence pertaining to Smith's behavior across many situa situations, where it has been observed that in 90% 90% of those situations Smith has engaged in behavior X and in 10% 10% of those situations he has not. Under these conditions, verified or there is a sense in which the probability statement could be empirically verified falsified as a claim to knowledge about Smith. But that is because all of the data falsified from which the statement issues refer refer to observations about Smith, which means that we are no longer discussing individual differences differences research, either as the source verification/falsification. In other of the statement or as the locus of its subsequent verification/falsification. words, to achieve this sort of knowledge about individuals one must step outside of the individual differences differences framework. framework. B. Be B. Individual Individual Behavior Behavior Is Is Not Not Caused Caused by by and and Cannot Cannot Be Explained in in Terms Terms of of the the Difference(s) Difference(s) between between That Explained That Individual and and Others Individual Others

For all of the foregoing, many will cling to the intuitively appealing notion that the psychological differences differences between individuals are relevant relevant to and must therefore therefore somehow be incorporated into explanations for their respective actions. As compel compelling as this notion seems to be, it is found wanting on close inspection. Let us say, for for example, that among his other personality characteristics Smith is, among his his other personality character characteris an extraverted individual, and that Jones is, istics, an introverted individual. Thus, along the dimension of of individual differences differences known as introversion-extraversion, Smith's status is E while Jones' status is I. Let us further further suppose that as a direct result of his "E-ness," Smith's behavior is "1consistently chatty (C) at social gatherings, and that as a direct direct result of his Alness," Jones' behavior is consistently quiet (Q) in such settings. Now from a theoretical standpoint, the concern in this little scenario would be to explain Smith's "e-ness" "C-ness" in terms of his "E-ness," and to explain Jones' "Q-ness" in terms of his "I-ness." In individual differences differences research, however, ("predictor what an investigator is actually looking at on the psychological ("predictor variable") side is neither E nor I per se, but instead and quite literally at the

122 U2

JAMES LAMIELL JAMEST. T.LAMIELL

I). Similarly, what the investiga.tor difference between the two, [E - I]. investigator is actually difference looking at on the behavioral ("criterion variable") side is neither C nor Q per se, but instead and quite literally at the difference Q). difference between the two, [C - Q]. As a result, the focus in individual differences differences research is actually on the relationship between the two dif ferences, and it is finally for this reason that, differences, in such research, one's vision of of the individual persons between whom differences differences are being studied becomes blurred. ference between the respective psychologi The problem is this: How can the dif difference psychological constitutions of of Smith and Jones possibly be said to influence, determine, or cause Smith's behavior or or Jones' behavior? If one adheres to the logic of the traditional individual differences differences approach, one is eventually forced to concede that difference between Smith's extravertedness (E) and Jones' introvertedness (I) the difference nor for Jones. Presumably, what exists exists psychologically neither for Smith nor psychologically for Smith is simply his extravertedness (E)-and not the difference difference - I]. between his extravertedness and Jones' introvertedness, [E I). Similarly, what presumably exists psychologically for Jones is simply his introvertedness (I), difference between his introvertedness and Smith's extravertedness, and not the difference [E - 1I). ] . Hence, Hence, to to try try to to explain explain Smith's Smith's and and Jones' Jones'respective respective behaviors behaviors by by reference reference to [E - I) I] is to ground each of of the respective explanations in an entity which, though it might well be said to exist in some sense for for an onlooker onlooker of of Smith and Jones, cannot be said to exist for either Smith or Jones. Of course, one might object at this point that the entity [E ~ I) I] could serve some sort of explanatory function if one sets as one's task explaining neither Smith's difference between chattiness (C) nor Jones' quietude (Q) per se, but instead the difference the two [C — Q]. Q].The Theresponse responsetotothis thisobjection objectionisisnot notmerely merelyto toconcede concedeititbut buttoto underscore it as the point: the coherence that such individual differences research explanations can in principle principle ever offer requires that the discussion be limited to the differences differences between between individuals, and that it never be permitted to lapse over into a discussion about individuals. individuals. What the "laws" embodying embodying such such "explanations" "explanations" would "explain" is merely between-person variance in the criterion variable(s), and the "explanation" would be between-person variance in the predictor vari variable(s). Such "explanations" might well serve to advance purely demographic or actuarial agendas. They might even serve to advance theories of data. But they will persons, and this is troublesome because that happens never advance theories of persons, to be what personality theories are. No person's psychological constitution is between-person variance on the of a regression equation, and no person's behavior is predictor-variable side of between-person variance on the criterion side of a regression equation. Indeed, as they have been conceptualized by the ersatz "nomotheticists" of our discipline, it individual differences variables variables do do not not existfor for individuals. individuals. is necessarily the case that individual of individual differences variables can be a discussion of Hence, no discussion of individuals. Moreover, would-be general laws, the terms of which are individual

CHAPTERS5 CHAPTER

DIFFERENCES INDIVIDUALS AND DIFl'ERENCES

123 123

differences variables, cannot possibly be laws that explain individual behavior. differences differences, and because individual differ differBecause such laws pertain to individual differences, ences do not exist for individuals, one is forced to conclude that such laws pertain, quite literally, to no one. Laws that pertain to no one cannot possibly be laws that pertain to everyone, and it is to the latter that any nomothetic personality psychology worthy of the name would aspire.

II. II. SOME ISSUES IN NEED OF CLARIFICATION CLARMCATION A. A. Relationship Relationship of of the the Present Present Argument Argument to to Allport's AUport's Views Views

Many readers will detect in what has been said thus far traces of the so-called "nomothetic versus idiographic" controversy which has dogged the field now for for over 50 years. Following the German philosopher Wilhelm Windelband, Gordon Allport dis AUport (1937) introduced the terms "idiographic" and "nomothetic" into the discourse of personality psychologists as a way of highlighting, among other things, of person the difference difference between the study of oi persons on the one hand and the study oi variables, variables,that thatis,is,variables variableswith withrespect respecttotowhich whichpersons personshave havebeen beendifferentiated, differentiated,on on AUport noted, quite properly, that mainstream personality research was the other. Allport thoroughly thoroughly dominated by studies of the latter sort-which sort—which in his view qualified as nomothetic-and he steadfastly nomothetic—and steadfastly insisted that a proper understanding of personalities would require that the knowledge yielded by such studies be supplemented by knowledge that could be obtained only through studies of the former sort, that is, studies of of the sort he labeled idiographic. Given (a) the points of convergence between Allport's views and my own, and (b) the fact that shortly before his death in 1967 Allport AUport quite explicitly "cried uncle and retired to his corner" comer" (cf. Allport, AUport, 1966, p. 107), 107), the reader might wonder what reason(s) I might have for 1966, for presuming that some gain could be realized through yet another seance with Holt's (1962) "Teutonic ghost." Put briefly, my presumption in this regard stems from the convic conviction that, as forceful forceful and incisive as Allport AUport was in certain respects, he failed in several important ways to prosecute effectively effectively and/or correctly the case against the established individual differences differences framework. framework. To begin with, I doubt that Allport AUport aided his own cause by labeling the sort of inquiry he advocated "idiographic." For even if his usage of this term was consistent with Windelband's (1894/1904), the latter had coined the expression to refer refer to knowledge about unique, historically configured configured events or phenomena, had identified (die identified such knowledge as the goal of inquiry within the humanities {die Geisteswissenschaften), Geisteswissenschaften),and andhad hadexplicitly explicitlydistinguished distinguishedsuch suchknowledge knowledgefrom fromthe the Naturwissenschaften),At At sort of knowledge sought within the natural sciences (die Naturwissenschaften). stiU quite sensitive about about a time when mainstream academic psychologists were still scientists—on the natural science model, of course-Allport course—AUport their credentials as scientists-on

124 124

JAMEST. T.LAMIELL JAMES LAMIELL

of "antiscientist," and it is one which his wittingly or otherwise invited the charge of critics pressed swiftly swiftly (e.g., Skaggs, 1945) and sometimes harshly (e.g., Nunnally, 1967).2 AUport's use of the term "idiographic" was ill-advised, his use of of the But if Allport's term "nomothetic" was simply wrong wrong in the sense that that it was not not consistent consistent with the meaning that Windelband Windelband had intended. For Windelband, the term term nomothetic nomothetic referred to knowledge that could be expressed in the form of allgemeine Gesetze— referred form of allgemeine Gesetze general laws. Any such law specifies was immer ist—what always is—in immer ist-what is-in some specified empirical domain. It specifies, in other words, what what obtains in each specified each and every recurrent instance of the event or phenomenon it putatively governs, what is every recurrent instance of phenomenon putatively thus common common to all of those recurrent recurrent instances, and it is precisely a law's alleged generality that is thrown into doubt doubt by its failure failure to perform perform in this way. The German German expression for for general general is allgemein, allgemein, a word which itself itself derives from the expression allen alien gemein-common gemein—common to all. all Without Without doubt, this is the mean meanfrom Windelband attached attached to the term "nomothetic." *'nomothetic." It is most unfortunate unfortunate that this ing Windelband AUport branded branded as "nomothetic" the kind of of meaning was not preserved when Allport knowledge about about personality produced by inquiry conducted conducted within the traditional traditional individual differences differences framework. framework. We have already already noted noted that that such inquiry inquiry produces produces aggregate statistics that can be interpreted interpreted in a scientifically scientifically meaningful meaningful way/(9r no way for no individual (cf. (cf. Danziger, 1990). Such "laws" as can be formulated formulated on the basis of of such individual statistics, therefore, cannot possibly be regarded as laws found for found to hold recurrently for each of of many persons, and for this reason cannot cannot possibly be nomothetic of nomothetic laws of personality in the sense intended by Windelband Windelband (1894/1904).3 (1894/1904).^ personality Nevertheless, Allport AUport called the traditional traditional individual differences differences approach of personality "nomothetic," and in so doing only threw his intellectual to the study of adversaries into the proverbial briar patch. For given Windelband's Windelband's conception conception of of nomothetic knowledge as knowledge of the sort sought by and produced produced within the natural sciences psychology was trying to desperately to emulate, Allport's AUport's contemporaries contemporaries could scarcely scarcely have wished for for better better than to be accused accused (sic) by a critic of pursuing their their subject matter in a way that conformed conformed to the methods and and knowledge objectives of the natural sciences. Had Allport appreciated the AUport more fully fully appreciated foregoing foregoing considerations (or more vigorously pursued pursued their logical implications), he would never have labeled the traditional paradigm paradigm "nomothetic" to begin with, and the entire history of of the nomothetic versus idiographic controversy controversy might have AUporthimself himselfseems seemstotohave havebeen beensensitive sensitiveto tothis thischarge, charge,judging judgingby byhis hisattempt attemptinin1962 1962to to 2^Allport substitute the term "morphogenic" (borrowed from the scientifically respectable discipline of biology biology)) for the term "idiographic." **idiographic." It seems that by then, however, the damage had been done. The proposed terminological graft never took, and and despite despitehis hisefforts effortsAllport Allportnever neversucceeded succeededininaltering alteringthe thewidespread widespread perception of idiographic inquiry as antiscientific. 3 ^That That Windelband Windelband himself himself would would have have seen seen no no contradiction contradiction whatsoever whatsoever inin speaking speaking about about nomothetic knowledge knowledge within within the the domain domain of of personality personality is is clear clear from from other other portions portions of of the the original original nomothetic text. Alas, but without doubt, the number of authors who have written on the nomothetic versus idiographic controversy exceeds greatly the number of those who have read Windelband's text or familiarized themselves with its actual contents (Windelband, 11894/1904; 1992a). 89411904; cf. Lamiell, 1992a).

CHAPTER 5 CHAPTERS

I^a)IVIDUALS AND AND DIfFERENCES DIFFERENCES INDIVIDUALS

125 125

taken a different different (and rather more productive) course. But But all all speculation on on this do not not see see how the the fact fact can can any any longer be ignored that count aside, I personally do called the the "nomothetic" approach to the the study of personality is is not not what Allport called and has has never been anything of the the sort under the the meaning of the the term now and by Windelband. nomothetic intended by B. The The Need Need to to Distinguish Distinguish between between Uniqueness Uniqueness B. and Individuality Individuality and my view, Allport and and others who who have followed followed his his thought have made yet In my yet sufficient conceptual consequence to to warrant separate a third strategic error of sufficient discussion. The difficulty difficulty to which I allude here is nicely illustrated by the following passages, which appeared in a monograph published in 1955 within a section subtisubti tled "The Dilemma of Uniqueness": If If there there is is to to be be aa science science of of personality personality at at all all itit should should do do better better than than itit has has in in the of personality personality that that is is most most outstanding-its outstanding—its manifest manifest the past past with with the the feature feature of uniqueness uniquenessofoforganization. organization. (p. added) (p, 21, 21yemphasis emphasis added) Nor the example Nor is is itit helpful helpful to to take take refuge refuge in in the example of of other other sicences sicences .. .. .. (On (On the the contrary contrary .. .. ..)) we we should should refuse refuse to to carry carry over over the the indifference indifference of of other other sciences to of individuality. individuality. to the the problem problem of (p. 22, added) (p, 22,emphasis emphasis added)

from What I would highlight in these passages is the ease with which Allport moves from the term "uniqueness" to the term "individuality," with no apparent inclination to draw any distinction between the two. Space permitting, countless other examples be cited, and and not not only from from the writings of Allport. of this phenomenon could be an issue of uniqueness as as he argued his his case against the the By making such an adequacy of traditional "nomothetic" inquiry, I think that Allport succeeded succeeded— unfortunately—in creating the the impression that, in his his view, the the inability of such unfortunately-in inquiry to accommodate the possibility that in certain respects each person would be found to be like no other person was its critical flaw. But criticalflaw. But as as confirmed confirmed "nomothet "nomothetthe individual differences differences paradigm is is in fact not icists" have long well known, the not the phenomenon of uniqueness, at least in logically incapable of accommodating the term. a certain sense of that term. by the the notion Historically, "nomotheticists" have been guided in their work by is structured by by a finite number of underlying attri attrithat "the" human personality is butes in some amount of which every individual is endowed by nature and/or nurture. In accordance with this notion, it has been assumed that once the elements or components of the presumed generic structure have been isolated, the particular features of of any any one one individual's individual's personality personality will will be specifiable as features be comprehensively comprehensively specifiable as that individual's individual's measured measured coordinates coordinates or or "location" "location" within the structure structure (consider, (consider, that within the for example, example, the the recent recent work work of of McCrae McCrae & & Costa, Costa, 1986, 1986,1987; see also Angleitner, 1987; see also Angleitner, for 1991; Asendorpt, Asendorpt, 1995; McCrae & & Costa, Costa, 1995). 1995). 1991; 1995; McCrae

126 126

JAMES LAMIELL JAMEST. T.LAMIELL

Since it is logically possible that each measured individual would be found to take up a position occupied by no other measured individual within the multidimen multidimensional space, it is clear that, in this view, there is at least one sense in which uniqueness-and uniqueness—and individuality if that just means uniqueness-is uniqueness^—is a phenomenon that can be accommodated by traditional "nomotheticism." Thus have many "nom "nomotheticists" found license to proceed with business as usual without having to either concede the validity of Allport's assault on conventional practices or explicitly reject his thesis concerning individuality qua uniqueness. Not altogether satisfied with this ploy, Allport's strategy was to point out that "nomotheticism's" '*nomotheticism's" accommodation of uniqueness would "work" only insofar as the putative elements or components of the presumed generic structure were known to be universally applicable, and only insofar as those elements or components could be specified comprehensively. If either or both of these conditions failed to obtain, an investigator would inevitably fail to capture Smith's uniqueness because (a) the inves investigator would characterize Smith in terms that did not apply to Smith, (b) the investiga investigator would fail to characterize Smith in terms that did apply to Smith, or (c) both. The mainstream response to this position has typically been that the issues raised are empirical matters best handled in accordance with the established principles of & construct validation as set forth by Campbell and Fiske (1959; see also Cronbach & Meehl, 1955). The notion here has been that competently competently executed research focused on the reliability, convergent validity, and discriminant validity of of various individual differences constructs would eventually reveal those attribute dimensions necessary differences and sufficient for identifying· any given individual's personality characteristics (again, sufficient identifying personaUty Costa for current examples). see the previously cited work by McCrae & & Now, for reasons already mentioned, investigations into the reliability, reUability, conver convergent validity, and discriminant validity of various measures of individual differences of differences cannot Allport, no matter how cannot logically resolve the aforementioned concerns of of AUport, competently those investigations are executed, and no matter what findings they unearth. The questions Allport AUport raised concerned the grounds on which it would be X could or could determined whether or not some putative personality attribute X not be meaningfully applied to any given individual. The correlation coefficients coefficients in which evidence concerning reliability, convergent validity, and discriminant validity subsists simply beg such questions, because when they are less than +/-1.00-and +/-1.00—and they always are-they are—they are uninterpretable for individuals. Thus, the traditional "nomothetic" response to Allport's concerns regarding the celebrated assumption of universal applicability (Bem (Bern & & Allen, 1974) is inadequate at its very epistemological core-persistent core—persistent and widespread beliefs to the contrary notwithstanding (see, e.g., Paunonen & 1986)-and Allport & Jackson, 1986a, 1986b; cf. Lamiell & & Trierweiler, 1986)—and AUport need not have retreated one inch on this score. But if I would criticize traditional "nomotheticists" for pretending to meet chaUenges Allport AUport mounted when in fact they never have done and never can the challenges AUport himself himself for making such an issue of uniqueness do so, I would also criticize Allport

CHAPTER55 INDIVIDUALS INDIVIDUALSAND ANDDIFFERENCES DIFFERENCES CHAPTER

U7 127

place."* To see why, let us grant for just a few moments, and solely for in the first place.4 of McCrae and Costa's (1987) claim to have isolated discursive purposes, the validity of five basic dimensions necessary and (for now at least) sufficient sufficient as terms in which to describe the major features of any given individual's personality. Let us now suppose that the requisite five measurements have been made of the person Smith. Now if these five measurements serve their intended purpose at all, then they convey convey valid information about what is, rather than what is not, the case as regards Smith's personality. They do this in virtue of the fact that the intersection of the five measurements within the multidimensional space locates Smith at a particular point and not at any of the other possible points within that space. Ensembled, the measurements state that Smith is "here" "here" and not "there" or anywhere else in the space. It is for statements of of just this sort, that is, statements about what is, rather than what is not, the case as regards the personality of of a specific individual, that I would propose we reserve the term individuality. individuality,^5 But now with this putative knowledge about Smith's individuality at hand, in what research direction would we be led by concerning ourselves with the question of Smith's uniqueness? In the spirit of traditional "nomotheticism," for example, of suppose that we wished to know whether the location occupied by Smith in the five-dimensional space proposed by McCrae and Costa is or is not also occupied five-dimensional by one or more other individuals. It is important to see here that in posing this question a distinction between Smith's individuality on the one hand, and his uniqueness on the other, has already and necessarily been made, if only implicitly. For in order to determine if what is personality is the case for Smith and no one else-Le., else—i.e., the case regarding Smith's personality uniquely—one must first have at hand some knowledge of is the case for Smith uniquely-one individuality—what is what is the case for Smith. Thus, knowledge about Smith's individuality-what Smith—is both distinct from the logically rather than what is not the case concerning Smith-is thereof. prior to any knowledge about Smith's uniqueness or lack thereof. A A second crucial observation to be made is that if knowledge about Smith's individuality is at hand, then some basis exists for pursuing questions concerning how Smith came to be as he is, what keeps him as he is, and what might change the way he is, and how answers to these questions might be used to explain why Smith currently acts as he does and, perhaps, perhaps, to to predict how how he he will will act act in incertain certainfuture future circumstances. circumstances. Hence, given some initial knowledge about Smith's individuality, an investigator might well opt not not to pursue the question of uniqueness at all, and to pursue instead questions of the sort that Leon Levy (1970, p. 29) identified as central to personality of

^IIshould should note note here here that that on on this thisscore score Allport's AUport'sviews views were werevery verymuch muchininline linewith withthose those of of 4 Windelband, for for whom whom the the thought thought that that he he might might not be be unique unique was was utterly utterly abhorrent. abhorrent. Windelband, S ^This Thisisisnot notto tosay saythat thatIIregard regardthe thetraditional traditionalindividual individualdifferences differencesparadigm paradigmasaswell-suited well-suitedtoto the formulation of such statements. Indeed, II do not. However, since this issue is tangential to our immediate concerns, I let it pass here.

128 128

JAMES LAMIELL JAMEST. T.LAMIELL

psychology's overriding theoretical concerns. By the same token, if the question of uniqueness (which, it may be noted, is not to be found among those cited by Levy) is to be pursued, then the questions concerning personality dynamics (development, maintenance, change, etc.) must inevitably be held in abeyance. Bearing the above in mind, let us now suppose with McCrae and Costa (1986), and in traditional "nomothetic" fashion, that the five attribute dimensions in terms of of which Smith's individuality has been articulated can be meaningfully applied to all of the other individuals with reference to whom the question of of Smith's uniqueness is of of to be settled. Of course, in granting this assumption we already finesse one of AUport's major concerns with respect to the dilemma of of uniqueness (see above). Allport's But ignoring this fact for now, let us focus instead on the fact that even if the assumption is granted, another very serious problem immediately immediately arises, namely, that of determining how many-and many—and which-other which—other individuals should be compared with Smith before before concluding concluding that he is or is not unique in the sense under discussion. To say that someone is unique is to say that there is no one else just like him. But is the phrase "no one else" to mean, literally, not one other human being who has ever lived, is now living, or who will ever live? If so, then in taking up the question of of Smith's uniqueness we have, to say the least, set a rather formidable task for ourselves. On the other hand, if this is not what the phrase "no one else" of what the phrase is to mean remains. Furthermore, is to mean, then the question of any answer that we might offer to this question will demand a rationale, that is, an explanation for why the phrase "no one else" should be given any meaning other than its literal meaning as expressed above. But just to be sporting, let us suppose that somehow all of of these matters have been resolved: (1) that the population of individuals with reference to whom the ques question of Smith's uniqueness is to be settled has been specified in a conceptually defensi defensible and practically workable way; (2) that the assumption that McCrae and Costa's five attribute dimensions can be applied meaningfully meaningfully to every individual individual within that properly, and (3) that measurements of ofthose population has been justified justifiedproper/y, of each one of those individuals with respect to the five attribute dimensions are at long last available.6 available.^ Now let us us suppose that, as it has turned out, none of of the other n individuals within the investigated population has been found to occ upy the same position in occupy m the multidimensional space where we had previously located Smith. At long last, empirical grounds would exist for claiming that, in in at least one sense of of the term, term. Smith is unique. about Smith's personality But now a new question arises: What do we know about personality 6 ^Incidentally Incidentally(and (andas asififwe wewere werenot notininthis thisup upto toour ournecks necksalready), already),how howlong longshould shouldwe weexpect expect if and when it is completed, should we expect to find Smith the data-gathering process to take? And if Or might he have moved by at just that location in the multidimensional space where we left him? Or if by then he has moved, what, what, if if any, any, problems problemswould would this thiscreate? create?As Asimportant importantand anddifficult difficult then? And if as these questions are, it would be a shame to let them block our passage at this late stage of our discussion, so let us pretend that these questions, too, can be satisfactorily answered: answered: that all of the needed measurements measurements have been gathered within a reasonable length of time, and that Smith, God bless him, has stayed put at the precise coordinates within the multidimensional space where we had located him originally.

CHAPTER CHAPTERS5 INDIVIDUALS INDIVIDUALSAND ANDDIFFERENCES DIFFERENCES

129 129

that we did not know before the question of his uniqueness was settled? Is he any more the person we originally found him to be by virtue of the subsequently more established fact that there is no one else just like him? Surely, the answer to this question must be "no," for if it is "yes," then it follows that Smith is no longer the person he was, in which case the validity of the conclusion that he is unique is thrown back into doubt. Alternatively, suppose that our investigation has turned up indisputable evidence that there exists at least one other individual, Jones, whose personality is is organized organized exactly exactly like like Smith's. Smith's. The The question question is, is, is is Smith Smith any any less less the the personality person we we originally originally found found him him to to be be by by virtue virtue of of the the subsequently subsequently established established fact fact person that he he is is not not unique unique in in this this regard? regard? Once Once again, again, the the answer answer to to this this question question must must that be "no," "no," for for if if it it is is "yes," "yes," then then our our original original measurements measurements of of Smith Smith no no longer longer be represent his his individuality, individuality, in in which which case case he he really really might might be be unique unique after after all, all, and and represent in any case can no longer be regarded as identical to Jones. in any case can no longer be regarded as identical to Jones. The lesson here is not difficult to see: whatever the fidelity of our original assertions concerning Smith's individuality, that fidelity can be neither compromised nor enhanced by the results of inquiry into Smith's uniqueness. Smith cannot be any more the the person person he he is is simply simply by by virtue virtue of of our our establishing establishing (as if, if, somehow, somehow, we we ever ever more really could) that he is unique, and he cannot be any less the person he is even if it it really could) that he is unique, and he cannot be any less the person he is even if is discovered that he is not unique. Smith's individual personality—his individuality—is is discovered that he is not unique. Smith's individual personality-his individuality-is what it it is is whether whether he he is is unique unique or or not. not. Moreover, Moreover, the the questions questions of of genuine genuine theoretical theoretical— what and practical—consequence cited earlier concerning personality development, mainteand practical-consequence cited earlier concerning personality development, mainte nance, and change remain whether in the important features of his personality Smith nance, and change remain whether in the important features of his personality Smith is, to paraphrase Murray and Kluckholm's (1953) much too celebrated observation, is, to paraphrase Murray and Kluckholm's (1953) much too celebrated observation, "like all all other other persons, persons, like like some some other other persons, persons, or or like like no no other other persons." persons." "like AUport called "the dilemma of uniqueness" should be laid permanently What Allport to rest. It is not only unproductive but actually counterproductive counterproductive to continue to level against "nomotheticism" the charge that it fails to accommodate the possibility of individual uniqueness. In the first place, there is at least one logical sense in which the charge simply is not true, and committed "nomotheticists" will always & Paunonen, 1988). In the second be quick to reassert that fact (see, e.g., Jackson & place, while there is also a sense in which the charge is true, it is also true that, place, when all all is is said said and and done, done, the the point point is is moot. moot. It It is is the the fact fact of of individuality, individuality, not not the the when altogether separate separate and and finally finally inconsequential inconsequential matter matter of of uniqueness, uniqueness, that that under underaltogether mines conventional conventional "nomotheticism" "nomotheticism" as as aa framework framework for for the the advancement advancement of of per permines agenda.^ sonality psychology's pantheoretical agenda.' 7 This is as good a place of ^This place as any to point out that in calling for an an approach to the study of respectsthe theindividuality individualityofof psychologicalphenomena, phenomena,one one need not—and I do psychological need not-and I do notnot—seek -seek personality that respects to promote a kind of individualism in the traditional sense of that term (see in this regard the excellent mdividual psychology comes immediately article by Sampson, 1988). 1988). In this connection Alfred Adler's individual respectful of individuality but is anything to mind as a very clear example of a view which is at once respectful but a celebration of what Sanford calls "self-contained mdividualism." Adler*s concept of das das individualism." Indeed, Adler's Gemeinschaftsgefllhl very antithesis such individualism, and fact much more spirit Gemeinschaftsgefahl is is thethe very antithesis ofof such individualism, and is is in in fact much more in in thethe spirit individualism/' Similarly, William Stern's Stem's critical criticalpersonalism personalism(Stern, (Stem, of what Sanford calls "ensembled individualism." 1917,1918,1924; 1996,1992b) 1906, 1917, 1918, 1924; cf. Lamiell, 1996, 1992b) offers a comprehensive framework for conceiving of of the human person in a way that is mindful of individuality yet disdainful of individualism.

130 130

JAMES LAMIEU JAMEST. T.LAMIELL

m. "ImOTHEllC" INQUIRY AS AN ALTERNATIVE TO i n . "IDIOTHETIC" TO TRADmONAL "NOMOTHEllCISM" TRADITIONAL "NOMOTHETICISM^' Bern (1982) observed that, historically, the fatal problem with recommending recommending some something like what Allport AUport thought thought of as idiographic (or morphogenic) inquiry has always been that "one is never quite sure what to do next" (p. 23). In view of of this problem, and as my own views are clearly akin to Allport's-even AUport's—even though, as I from Allport's AUport's in several significant significant re rehave tried to make clear, they deviate from spects-I spects—I intend to focus in the remainder remainder of of this chapter on some basic principles of of "idiothetic" *'idiothetic" inquiry which I believe can and should guide us as we move from from the traditional individual differences differences paradigm toward a more adequate alternative. Rorer and and Widiger's (1983) worthwhile recommendation, I will proceed Following Rorer insightful epistemological commentary commentary by relating my views to a very sobering but insightful offered sometime ago by Paul Meehl (1978). offered of Meehl's MeehPs Views on the Slow Progress of A. Synopsis of "Soft" Psychology following passage nicely conveys Meehl's views concerning the notion that The following theoretical assertions in the so-called **soft'* "soft" areas of of psychology-among psychology—among which he explicitly included personality psychology—can corroborated adequately adequately by psychology-can be corroborated of conventional tests of of statistical significance carried out against the null means of hypothesis, in accordance with the inferential inferential principles set forth forth by the renowned Sir Ronald Fisher: I suggest to you that Sir Ronald has befuddled befuddled us, us, mesmerized us, us, and led us down the primrose path. I believe that the almost universal reliance on merely refuting the null hypothesis as the standard method for corroborating substantive substantive refuting theories in the soft soft areas is a terrible mistake, is basically basically unsound, unsound, poor scientific scientific strategy, and one of the worst things that ever happened in the history of psy psychology. (1978,p.p.81 817) (1978, 7)

The major major problem problem here, Meehl notes, is the fact that the null hypothesis is almost always false, and knowably almost knowably so with virtual certainty a priori (see also on this point, Bakan, 1966). Population Population means are virtually virtually never precisely equal, and population population correlation coefficients coefficients are virtually never precisely precisely zero. Con Conreach that magical p p < .05, sequently, a statistical relationship large enough to reach reject the null hypothesis, is virtually guaranteed guaranteed regardless of of the and thus to reject of the theoretical proposition under putative test, provided substantive merits of only that one is able and willing to make a sufficient sufficient number of of observations. properly concludes from from this that much of of what has historically Meehl quite properly passed for for theory testing in "soft" "soft*' psychology reduces to "meaningless substantive constructions on the properties of of the statistical power function" function" (p. 823).

CHAPTERS CHAPTER S

INDIVIDUALSAND ANDDIFFERENCES DIFFERENCES INDMDUALS

131 131

philosophy of of science, Meehl Adopting what might be termed a "neo-Popperian" philosophy for all practical purposes, it is simply simply oxymoronic to speak of of thus argues that, for reject the "risking" theoretical propositions against the possibility of of failing failing to reject null hypothesis. His conception conception of of a more apposite approach through approach is conveyed conveyed through the following following example:

If I tell you that Meehl's theory of climate predicts that it will rain sometime next April, and this turns out to be the case, you will not be much impressed with my "predictive success." Nor will you be impressed if I predict more rain < .(01) .001) in my t-test r-test (for p < in April than in May, even showing three asterisks (for table! If I predict from from my theory that it will rain on 7 of the 30 days of April, 7, you might perk up your ears a bit, but still you would and it rains on exactly 7, be inclined to think of this as a "lucky coincidence." But suppose that I specify specify which 7 days in April it will rain and ring the bell; then you will start getting seriously interested in Meehl's meteorological conjectures. Finally, if I tell you that on April 4th it will rain 1.7 inches (66 cm), and on April 9th 2.3 inches (90 cm) and so forth, and get seven of these correct within reasonable tolerance, for it. You you will begin to think that Meehl's theory must have a lot going for may believe that Meehl's theory of the weather, like all theories, is, when taken literally, false, since probably all theories are false in the eyes of God, but you if will at least say, to use Popper's language, that it is beginning to look as if verisimilitude,that thatis,is,truth-likeness truth-likeness. An Meehl's theory has considerable verisimilitude, . . .. . .. An unphilosophical chemist or astronomer or molecular biologist would say that this was just good sensible scientific practice, that a theory that makes precise predictions and correctly picks out narrow intervals or point values out of the range of experimental possibilities is a pretty strong theory. (Meehl, 1978, 1978,pp. pp.817-818, 817-818,emphasis emphasisand and parentheses original) (Meehl, parentheses in in original) With Meehl's thoughts in mind-and mind—and most especially those expressed in the last statement of the above quotation-let quotation—let us now turn to a consideration consideration of of studies conducting over the past several years in that I and various colleagues have been conducting the area of of SUbjective subjective personality judgments.

B. Studies in the Epistemology of Subjective Personality Judgments of background. If If one digs beneath beneath the reliability and validity First, a bit of coefficients (and, occasionally, other aggregate statistical indices) issuing from from coefficients differences to the conceptual conceptual core of of the paradigm, one studies of individual differences bedrock an unwavering conviction that meaningful meaningful statements about about the finds at bedrock endowed with one or extent to which any given individual's personality is endowed more underlying attributes (traits, predispositions, etc.) can be derived only by that individual with others. In this view, our knowledge of of who Smith comparing that is is necessarily and inextricably tied to our knowledge knowledge of of who others are. Though manifest manifest in a variety of of ways, this conviction conviction finds finds its most visible and Though

132 132

lAMES LAMIELL JAMEST. T.LAMIELL

of the normative measurement operations which formal expression in the logic of have long been regarded as the sine qua non of scientifically negotiable statements about personality.8 personality.® Several years ago, however, I arrived (via studies focused on the so-called & Cavenee, "illusory correlation" phenomenon; see Lamiell, 1980; Lamiell, Foss, & 1980; see also Shweder, 1975, 1980) at the hypothesis that, in formulating and expressing subjective judgments about their own and one another's personality characteristics, lay persons do not not routinely rely on a normative reasoning process. Instead, I hypothesized, hypothesized. Smith relies on what I later came to appreciate as an essentially dialectical dialectical reasoning process by which his judgments of, say, Jones, are framed not by contrasting Jones with others, but by contrasting Jones with a conception of who Jones is not but might otherwise be, i.e., with a conception of of Jones. the negation of A A question which often arises here is, would not Smith's conception of who itself require prior knowledge (existing "now" as memory traces of Jones is not itself "yes" answer here would, of course, previous experiences) of of who others are? A A "yes" imply that the judgments Smith makes of Jones (or, for that matter, of of himself) himself) are grounded in normative considerations after all. My answer, however, is "no." The thesis is that Smith's ability to conceive of who Jones is not but might otherwise be does not require prior knowledge of who others are. Indeed, the contention here is that were it not for the capacity to frame judgments dialectically, and hence of considerations about individual differences, it would be impossible independent of differences to be framed. The claim here, in other for knowledge about individual differences words, is that it is dialectical reasoning that makes normative (and for that matter, ipsative) reasoning possible, and not the other way around. Nor, if this is true, could it be so only for the lay person. To the contrary, it would have to be true as well for personality investigators themselves, and elsewhere I have tried to explain both that and why this is the case (Lamiell, 1987, chap. 5). Suffice it for now to say that it is here where one finds the most important methodological implications of of inquiry for those investigators who are interested in trait measurement this line of and in search of a viable technique (see Equation 2 in Lamiell, 1981, p. 282) with heretofore favored normative operations.9 operations.^ which to replace the heretofore

^To Tobe besure, sure,investigators investigatorshave havelong longrecognized recognizedthe thepossibilities possibilitiesoffered offeredbybyipsative ipsativemeasurement. measurement. 8 differences research, and hence inquiry However, prevailing wisdom has always been that individual differences grounded in the logic of normative measurement, is a logical precursor to any viable attempt to measure 1954; Falk, 1956). individuals ipsatively (see, e.g., Beck, 1953; 1953; Eysenck, 1954; Warning: those those investigators investigators who who are are merely merely looking lookmg for for aa better better way way to to measure measure individual ' Warning: 9 1986b; see also differences (as appears to be the case, for example, with Paunonen & Jackson, 1986a, 1986b; Asendorpf, 1988) will simply have to look elsewhere. The objective, in my view, is not to improve Asendorpf, individual differences differences research research but but to to abandon abandon it. it. In In this this same same vein, vein, the the notion notion should should be be resisted resisted that my proposed methodological alternatives can properly properly be be evaluated evaluated according according to to the the psychometric psychometric my proposed methodological alternatives can criteria of of the the individual individual differences differences paradigm. The realization realization that that those those criteria criteria are are inadequate inadequate to to the the criteria paradigm. The task(s) at hand hand is is what what gave gave rise rise to to the the proposed proposed methodological methodological alternatives alternatives in in the the first first instance. instance. task(s) at

CHAPTER 5 INDIVIDUA� ERENCES CHAPTERS INDIVIDUALSAND ANDDIFF DIFFERENCES

133 133

Given the hypothesis that lay persons reason dialectically rather than norma normatively in formulating subjective personality judgments, the challenge, ak la Popper, was to devise a means of exposing that hypothesis to the risk of of disconfirmation disconfirmation in empirical observations. To this end, several studies have to date been carried out (Lamiell & Durbeck, 1987; Lamiell, Foss, Larsen, & Hempel, 1983; Lamiell, Foss, Trierweiler, & Leffei, 1983). In all of those studies, the subject's task was very simple: he or she was presented with a series of 30-40 activity protocols and told that each such protocol information about the extent to which one of conveyed valid information of his or her peers of a number number of typically invests his or her time or effort in each of of activities. The subject was requested to consider the information displayed in each protocol, to form a judgment about the degree to which the indicated activity pattern reflected each of a number of underlying personality attributes (e.g., warm versus cold, judgsociable versus unsociable, industrious versus lazy), and then to express each judg ment by marking a rating scale. of this sort the subject's judgments are of normative nature, Now if in a task of of a normative then the actual ratings made of the targets by a particular subject should be well derived nonactuarially predicted by formal measurements of those same targets derived arithme1983; cf. Conger, 1983; Woody, 1983) via the arithme (Lamiell, Trierweiler, & Foss, 1983; tic proper to normative measurement operations. Alternatively, if the subject's reasoning process is not of a normative nature but is instead dialectical, and hence interactive measurement, patterned after the logic of what Cattell (1944) once called interactive then measurements of the targets derived nonactuarially via the arithmetic proper to such measurement should better predict the subject's actual ratings of of those targets. carefully the approach that was taken here: on the basis of of two competing Note carefully of the psychological psychological process en enand precisely articulated theoretical conceptions of specific point predictions were derived nonactuari nonactuarigaged by the experimental task, specific ally for where a given subject's ratings of the targets would Uterally literally fall on a specified scale. The sensible thing to do at this point was not to fashion some sort of of a null hypothesis test, and certainly not to set about analyzing individual differences differences in the subjects' perceptions of the targets (e.g., by engaging one or another of of the various data analysis procedures discussed by Schneider, 1973). The sensible thing to do was to check, for each subject, the degree of of correspondence between each of the two sets of of predicted ratings and that subject's actual ratings. Using for this purpose the well-known index of profile dissimilarity devised by Cronbach and GIeser Gleser (1953), the findings obtained with one subject who participated in the study by Lamiell and Durbeck (1987) are displayed in Table L I. Focusing for the moment on the dissimilarity indices shown in the first row of of Panel IV in the table, one can see that, for Target 1, the index resulting from a comparison of the subject's actual ratings to the interactively derived point predic predictions (.11) was lower (indicating less dissimilarity) than that resulting from a compar comparison of of the subject's actual ratings to normatively derived point predictions (.55). The dialectical theoretical conception of the subjective judgment process thus showed verisimiUtude in this instance, and it is important to see that this conclusion greater verisimilitude

TABLE II TABLE Profile Dissimilarities (One Subject, 40 Targets) Predicted Ratings, Actual Ratings, and Proportional Profile

Panel I: Predictions from normative model (N) Attribute

Panel II: Predictions from interactive model (I) Attribute

Panel III: Actual Ratings (A) Attribute

Target

11

2

3

11

2

3

11

2

3

11 2 3 4 5 6 7 8 9 10 111 1 12 13 13 14 15 16 17 18 19 20 21 21 22 23 24 25 26 27 28 29 30 31 31 32 33 34 35 36 37 38 39 40

17.32 8.65 13.57 9.09 0.00 20.00 14.27 7.90 10.95 3.67 14.09 7.57 8.08 8.63 10.60 3.91 2.83 2.79 9.51 16.71 16.71 12.84 11.26 7.24 15.21 15.81 15.81 8.01 9.02 13.34 10.19 15.70 19.67 13.46 13.80 9.42 9.90 10.16 5.53 17.19 6.43 6.10

0.00 19.81 19.81 16.39 5.40 4.23 4.07 6.76 20.00 12.72 7.16 9.88 2.34 4.94 6.11 10.25 8.67 14.48 8.46 18.64 12.50 5.37 8.91 7.96 2.13 6.08 11.57 7.53 11.02 12.07 7.68 8.21 2.13 11.05 5.57 10.25 12.84 7.44 7.44 9.38 0.74 4.56

10.39 .22 8.41 19.34 13.99 9.56 6.43 5.54 11.36 12.38 12.30 8.46 12.49 7.23 6.78 1.28 4.43 19.58 11.14 5.94 14.54 11.21 11.21 7.78 14.13 10.65 5.73 5.99 4.90 0.00 7.38 7.35 12.77 8.10 8.58 11.64 10.29 14.94 12.08 20.00 9.90

12.01 12.01 9.19 10.79 9.33 6.37 12.88 11.01 11.01 8.95 9.94 7.57 10.96 8.34 9.00 9.18 9.82 7.64 7.30 7.28 9.47 11.81 11.81 10.55 10.04 8.73 11.32 11.52 8.98 9.31 9.31 10.72 9.69 11.48 12.78 10.75 10.87 9.44 9.60 9.68 8.17 11.97 8.47 8.36

7.14 12.67 12.67 8.65 8.32 8.28 9.03 12.72 10.69 9.14 9.90 7.80 8.52 8.84 10.00 9.56 11.18 9.50 12.34 10.63 10.63 8.64 9.63 9.36 7.73 8.84 10.37 9.24 10.22 10.51 10.51 9.28 9.43 7.73 10.22 8.70 10.00 10.73 10.73 9.22 9.76 7.35 8.42

12.54 7.95 7.95 16.58 14.16 12.16 10.75 10.35 12.98 13.66 13.40 11.67 13.49 11.11 11.11 10.91 10.91 8.42 9.85 16.69 12.88 10.53 10.53 14.42 12.91 12.91 11.36 14.23 12.66 10.44 10.55 10.06 7.84 11.18 11.17 13.62 111.50 1.50 11.72 13.10 12.49 14.60 13.30 16.88 12.32

10 55 4 8 8 9 14 7 7 6 7 8 8 7 7 6 6 8 9 7 12 12 12 8 10 10 10 10 8 8 7 6 13 111 1 8 8 7 10 8 7 9 7 7

8 15 9 8 12 12 14 8 13 13 8 5 8 8 13 10 10 13 14 8 14 9 7 8 12 7 12 12 12 13 13 13 13 13 7 7 13 13 8 8 8 14 111 1 12 12

13 13 4 8 12 11 11 8 8 7 7 8 8 7 8 8 12 12 4 8 14 9 11 11 9 11 11 12 12 14 10 7 4 6 7 5 8 12 12 11 11 11 11 11 11 15 18 9

Means of the profi le dissimilarity values profile Standard deviations of the profile dissimilarity values t value for differences between correlated correlated means (N vs I)

Panel IV: Panel 1Profile Profile dissimilarities N vs A

I vs A

.55 .27 .53 .38 .57 .69 .09 .32 .34 .24 .40 .27 .43 .21 .21 .31 .31 .22 .21 .21 .35 .24 .57 .28 .06 .28 .37 .43 .22 .22 .28 .31 .31 .26 .40 .41 .41 .34 .23 .12 .26 .35 .41 .45 .36

.11 .11 .23 .36 .23 .25 .38 .19 .17 .33 .30 .32 .22 .33 .20 .15

.33 .33 .13 5.71

.22 .07 (p (P < < .01 .01))

.23 .23

.16 .14 .20 .25 .28 .16 .14 .13 .18 .10 .22 .31 .31 .20 .26 .30 .29 .19 .12 .15 .17 .28 .17 .20 .24

Source: Source:"Whence "WhenceCognitive CognitivePrototypes PrototypesininImpression ImpressionFormation? Formation?Some SomeEmpirical EmpiricalEvidence Evidencefor forDialectical DialecticalReasonReasoning as a Generative Process," and Behavior, 8, pp. 223-244. Process," by byJ.J.T. T.Lamiell Lamielland andP. P.Durbeck, Durbeck,1987, 1987,Journal JournalofofMind Mind and Behavior, 8, pp. 223-244. Reprinted with permission of publisher.

CHAnER 5 CHAPTERS

INDIVIDUALSAND ANDDIFFERENCES DIFFERENCES INDIVIDUALS

135 135

p value, or to any inferential inferential statistic of does not appeal to any p of any sort. There is no null hypothesis here to reject reject or fail to reject. reject. Of discussed represent Of course, a critic might object object that the results just discussed represent those obtained in one instance only. And And so they do. But by scanning the remainder remainder of of based on the interactive Panel IV in the table, one can see that point predictions based than did point predictions (dialectical) model approximated approximated actual ratings better than 31 instances, and failed failed to do so in only 9 instances. based on the normative model in 31 Here, the possibility for for putting a test of of the null hypothesis to good good use does arise. For For example, a t test comparing the two arrays arrays of of dissimilarity dissimilarity values values might might be carried out (see bottom of Table I). Alternatively, a simple chi-square analysis might be conducted, in which we would enter 31 31 tallies in one cell of of the chi-square table to represent the 31 31 "hits" for for the dialectical theory, and 9 tallies in another cell to represent the 9 "misses" for for that theory. The distribution distribution of of "hits" versus turn be tested for "misses" thus obtained/or obtained/or this individual individual subject subject could in turn for statistical significance significance against chance expectations, which in this example would be a distribu distribution of of 20 "hits" and 20 "misses." For the record, the obtained 12.1, obtained value of of chi-square in this instance equals 12.1, of freedom, would occur by chance alone much less a value which, at one degree of often than than 1 time in 100. It is vitally vitally important important to recognize, however, that that in this often context the statistical analysis did not serve as a test of the substantive theoretical not of context proposition. It served instead as a means of of determining determining whether whether or not tests of of that theoretical proposition already accomplished accomplished by other and entirely independent other and entirely independent means have confirmed confirmed or disconfirmed disconfirmed the proposition proposition with a degree degree of of regularity sufficient to regard it as empirically corroborated. sufficient corroborated. But But now what of of the objection objection that, for for all of of this, we have still considered theoretical but one subject? The argument, of course, is that scientific scientific theories or theoretical fall on results obtained obtained with just one subject. And And so propositions cannot stand or fall & they cannot. But suppose that I had at hand evidence (and I do; see Lamiell & Durbeck, 1987) that in the study from from which the data displayed displayed in Table I were obtained, which involved involved a total of 67 subjects (investigated (investigated individually, of of course), there were 57 for for whom the hypothesis that the subjective subjective reasoning process is dialectical was confirmed, confirmed, 10 for whom that hypothesis was not not clearly confirmed, confirmed, and none-not none—not one-for one—for whom the competing hypothesis (that the subjective subjective rea reasoning process is normative) was confirmed confirmed instead. This distribution of of "hits" and "misses" could likewise be submitted to a chi-square analysis, but to what end? The obtained ratio of of "hits" to "misses" appears to be overwhelming for for the simple and very good reason that it is, and any conventional statistical test at this point would be merely gratuitous.

c. C. Implications One thing I certainly do not wish to suggest by the foregoing foregoing is that studies of of center stage in the activity of of subjective personality impressions should now take center worthy function function personality investigators. Those studies do serve, in my view, the worthy

136 136

JAMEST. T.LAMIELL JAMES LAMIELL

of of demonstrating that individuals can be characterized, in terms of of their salient personality personality attributes, through the exercise of of a reasoning process that is at once coherent and exceedingly systematic, but which at no point requires the comparison of one individual to another. With respect to the enterprise of trait measurement, of is. If lay persons can proceed in the manner the question raised by these findings is, just described as they formulate and express their "subjective" impressions, then why could not personality personality investigators proceed in like fashion as they formulate formulate and express their "objective" impressions, in the form form of of personality profiles? The answer is that they could, and in my view they should. should. It is also to be hoped that the studies mentioned above will awaken at least some contemporaries to the theoretical possibilities that are opened up by taking seriously the concept of dialectical reasoning. The possibility suggests itself, for example, that the subjects of our inquiries reason in this fashion not only when they are engaging a fairly sterile experimental task involving personality ratings, but routinely, in the appraisal of the circumstances of their day-to-day lives (consider, for of the laws example, the dialectical themes that emerge in Frijda's [1988] discussion of of emotion; see also Rychlak, 1981, 1981,1988). of 1988). In short, there is, potentially at least, a theoretical consequence here than might immediately great deal more of genuine theoretical meet the eye, and it is unlikely that those possibilities are going to be vigorously explored so long as the majority of investigators within the field are busy searching for—or celebrating the "discovery" of-personality of—personality psychology's answer to the for-or periodic table of of elements. I believe, with Meehl, that if basic research in the psychology of personality is ever to prove successful in advancing personality theory, there is going to have to be a sharpening of hypotheses to the point that they enable one to make point predictions, or at least narrow interval predictions, from the range of experimental possibilities which are presented to the subjects. When one is positioned to do this, as we were in our studies of subjective personality judgments, tests of statistical significance are, as we have just seen, either obviated altogether or relegated to an epistemological role entirely different different from-and from—and decidedly more limited than than— that which such tests have played heretofore. The major advantages of this alternative approach vis-a-vis vis-^-vis the objective of fashioning a genuinely (rather than merely pseudo-) nomothetic paradigm for basic personality research are not difficult difficult to see. First of of all, and as the studies mentioned above illustrate, it is possible to carry out theoretically relevant studies of of individuals without either resorting to radical behaviorism or compromising compromising methodological rigor. The "trick" here, if that is what it is, is to bring two or more divergent and well-articulated of predicting what well-articulated theoretical theoretical propositions to bear on the task of Smith will literally do in specified specified circumstances. Hypothesis testing is then a matter of of investigating the correspondence between those divergent predictions and what Smith actually does. A test of this particular sort need engage no inferential inferential statistics of a null hypothesis. because it is not a test of Second, since empirical research designed in this way neither involves nor invites any comparison between what Smith does and what others do, the investiga-

CHAPTER 5 CHAPTERS

INDIVIDUALSAND ANDDIFFERENCES DIFFERENCES INDIVIDUALS

137 137

tion of each individual becomes a coherent study unto itself. This is just as it should be in a discipline that would advance our theoretical understanding of individuals. Third, the foregoing logically compromises not at all the search for general or nomothetic principles. To the contrary, in a discipline where individuals are the entities over which generality generaUty is sought, the rigorous examination of theoretical propositions at the level of the individual is not only not antithetical to the quest for generality but is in fact logically central thereto. If reliance on conventional conventional null hypothesis testing procedures is one of the worst things that has ever happened in the history of psychology, then another is surely the ascendance of the notion that genuinely nomothetic principles of of personality could somehow be established estabUshed to individuals.^^ without studying individuals. Fourth, nothing in the approach I am advocating requires one to reject the fact of individual differences. As mentioned in the introductory comments difference between "idiothetic" inquiry and the to this chapter, the crucial difference accommodates individual traditional paradigm is that while the former simply accommodates differences, the latter makes those differences the focus of investigation. For example, in the impression formation studies discussed previously the subjects of the most certainly did differ in their ratings of the targets, and no feature of employed methodology rendered inadmissable the possibility (nay, the certainty) differences that such would be the case. Nor, however, was the study of those differences the point of the research. The point of the research was to empirically evaluate alternative theoretical conceptions of the reasoning process through which each individual subject arrived at his or her ratings, and there is simply no analysis of the differences differences between the ratings that could have shed light on that question (see Lamiell, 1987, chapter 6, for a further discussion of this point). Here and elsewhere I have discussed the problems which arise when individual differences differences are made the focus of investigation. I stand by the conviction that those problems are fatal to personality psychology's overriding scientific objectives, and that those same problems are irremediable within the traditional paradigm (cf. Dar & Serlin, 1990; Lamiell, 1990a, 1990b; Ozer, 1990). Finally, I should not fail to acknowledge that as attempts to proceed along the lines Hues I have sketched are put into practice, great challenges are bound to arise. It is, after all, one thing to generate theoretically based point predictions in studies of of subjective personality judgments—even judgments-even that was not so simple as it might now appear to have been—and been-and quite another thing to generate comparable predictions in many of the other substantive areas in which personality investiga investigators have legitimate and important concerns. At the very least, however, I hope to have provided the outlines of an ideal toward which I believe we should be 10 ^^ItItisisinteresting interesting totonote note that, that,atatits itsbeginnings, beginnings,experimental experimental psychology psychology was wasvery verymuch muchan an general "N = 11" " affair in the domain of research methodology even at its objective was to discover general UN either seeks seekssuch such knowledge knowledge or or studies studies laws of human psychological functioning. The notion that one either individuals but but not not both both is is aa myth myth that that developed developed later later on on (see (see Danziger, Danziger, 1990). individuals =

138 138

JAMEST. T.LAMIELL JAMES LAMIELL

striving. Knowing what one is trying to achieve can be very helpful, even if if one cannot at a given point in time claim to have achieved it fully. fully. When all is said and done, perhaps the most pressing current current problem problem with personality psychology is that, over the years, the majority majority of of those who animate the field has gradually lost the ability to distinguish between between the genuinely scientific business of of formulating formulating and testing theoretically derived propositions concerning the behavior/psychological behavior/psychological functioning functioning of of individuals on the one hand and the for variance in countless measures of of individ individmerely actuarial business of accounting for differences in behavior/psychological behavior/psychological functioning functioning on the other. Or ual differences Or perhaps the distinction remains, but the role of of the actuary actuary is simply preferred preferred ability to make this distinction for one am greatly greatly disturbed disturbed by the witting in much larger numbers. In any case, I for or unwitting ascendance of the notion, evidence of of which abounds in the literature, subjects are not so much beings to be understood, to the end of of enlightening enlightening that our subjects wagered on us not only about them but about ourselves, as they are objects to be wagered Paunonen & & Jackson, 1986a, pp. 471-472), like so many many horses. "Idiothetic" (cf. Paunonen about placing bets, with the objective objective of of maximizing payoffs payoffs in the inquiry is not about persons as things things or as mere matter long run. It is not about regarding persons matter (see the business of above cited works by Stern). It is about the serious business of advancing theoretical conceptions of of individual behavior-psychological functioning, toward toward the end of of behavior-psychological functioning, for improving our understanding of ourselves and one another. I do not know for framework as I have sketched it up to now will get us where we certain that the framework want to go. I do know that the long-dominant long-dominant individual differences differences paradigm will for doing so is at hand, even not. It is time to move on. The epistemological basis for of all of of the territory territory is not. if an exquisitely detailed map of

REFERENCES Allport, (1937). Personality: York: Holt, AUport, O. G. W. (1937). Personality:AApsychological psychologicalinterpretation. interpretation.New New York: Holt,Rine Rinehart & Winston. psychology New Allport, Becoming:Basic Basicconsiderations considerationsfor fora a psychologyofofpersonality. personality. New AUport, O. G. W. (1955). (1955). Becoming: Haven, Cf: CT: Yale University Press. of Allport, AUport, O. G. W. (1962). The general and the unique in psychological science. Journal of

Personality, Personality,30, 30,405-422. 405-422.

Allport, Psychologist,21, 21,1-10. 1-10. AUport, O. G. W. (1966). (1966). Traits revisited. American Psychologist, Angleitner, A. (1991). European Journal Journal (1991). Personality psychology: Trends and developments. European

of Personality,5,5,185-197. 185-197. of Personality, Asendorpf, J. B. B. (1988). (1988). Individual Individualresponse responseprofiles profiles in in the thebehavioral behavioral assessment assessment of ofpersonal personality. European Personality,2,2,155-167. 155-167. European Journal of of Personality, Asendorpf, J. B. B. (1995). PersBnlichkeitspsychologie: Pers6nlichkeitspsychologie: Das empirische Studium der individuel individuellen Besonderheit aus spezieller spezieller und differentieUer differentieller Perspektive Perspektive [Personality [Personality psychology: psychology: Psychologifrom special and different different perspective]. Psychologi The empirical study of individuality from 46, 235-247. sche Rundschau, Rundschau, 46, PsychologicalBulletin, Bulletin, significance in psychological psychological research. Psychological Bakan, D. (1966). The test of significance

66, 423-437. 66,

CHAPTER 5 CHAPTERS

INDIVIDUALS INDIVIDUALS AND ANDDIFFERENCES DIFFERENCES

139 139

Beck, S. S. J. (1953). The science of personality: Nomothetic or idiographic? Psychological Psychological Review, Review,60, 60,353-359. 353-359. Bern, D. J., & Allen, A. (1974). On predicting some of the people some of the time: The Psychological Review, Review, 81, search for cross-situational consistencies in behavior. Psychological 81, 506-520. Bern, D. J. (1982). Toward Bem, Toward aa response response style style theory theory of of persons persons in in situations. situations. In In R R. A. A. Dienstbier Dienstbier and M. onon Motivation 1982: Personality-Current M.M. M.Page Page(Eds.), (Eds.),Nebraska NebraskaSymposium Symposium Motivation 1982: Personality—Current theory theory and research, research, Volume Volume30, 30,pp. pp.201-231. 201-231.Lincoln: Lincoln:University UniversityofofNebraska NebraskaPress. Press. Campbell, D. T., & Fiske, D. W. (1959). (1959). Convergent and discriminant validation by the multitrait-multimethod multitrait-multimethod matrix. Psychological PsychologicalBulletin, Bulletin,56, 56,81-105. 81-105. R. B. B.(1944). (1944).Psychological Psychologicalmeasurement: measurement:Normative, Normative,ipsative, ipsative,interactive. interactive.Psycholog PsychologCattell, R ical Review, Review,51, 51,292-303. 292-303. ical Conger, A. A. J. J. (1983). (1983). Toward Toward aa further further understanding understanding of of the the intuitive intuitive personologist: personologist: Some Some Conger, subjective psychometrics. Journal of of Percritical evidence on the diabolical quality of SUbjective Per sonality, 51, 51,248-258. 248-258. sonality, Psychologist, Cronbach, L. J. (1957). (1957). The two disciplines of scientific psychology. American Psychologist, 12, 671-684. 12, & GIeser, Gleser, G. (1953). Assessing similarity between profiles. Psychological Psychological Cronbach, L. J., & Bulletin,550, 456-473. Bulletin, 0, 456-473. J., & & Meehl, Meehl,P. P.E. E.(1955). (1955).Construct Constructvalidity validityininpsychological psychologicaltests. tests.Psychological Psychological Cronbach, L. J., Bulletin,52, 52,281-302. 281-302. Bulletin, Danziger, K. K. (1990). (1990). Constructing Constructingthe thesubject: subject:Historical Historical origins of psychological research. origins of psychological research. Danziger, New York: Cambridge University Press. Dar, R, R., & & Serlin, Serlin, R R. C. C. (1990). (1990). For For whom whom the the bell bell curve curve toils: toils: Universality Universality in in individual individual Dar, differences research. In D. N. Robinson & L. P. Mos (Eds.), Annals of of theoretical theoretical differences psychology (Vol. 6, pp. 193-199). New York: Plenum Press. PsychologicalReview, Review,661, Eysenck, H. J. (1954). The science of personality: Nomothetic! Psychological 1, 339-342. from nomothetic approaches to personality Falk, J. (1956). Issues distinguishing idiographic from PsychologicalReview, Review,63, 63,53-62. 53-62. theory. Psychological theory. Frijda, N. H. (1988). (1988), The laws of emotion. American Psychologist, Psychologist,43, 43,349-358. 349-358. Harre, R being.Cambridge, Cambridge,MA: MA:Harvard HarvardUniversity UniversityPress. Press. R. (1984). Personal Personal being. R. (1962). Individuality and generalization in the psychology of personality. Journal Holt, R of of Personality, Personality,30, 30,377-404. 377-404. Hurrelmann, K. development. New York: Cambridge K. (1988). Social Socialstructure structureand andpersonality personality development. New York: Cambridge University Press. Jackson, D. N., & Paunonen, S. V. (1988, (1988, August). The measurement measurement of of uniqueness uniqueness in personality. personality.Paper Paperpresented presentedatatthe the96th 96thannual annualconvention conventionof ofthe theAmerican AmericanPsychologi Psychological Association, Atlanta, GA. J. T. T.(1980). (1980).On Onthe theutility utilityoflooking of lookingininthe the"wrong" "wrong"direction. direction.Journal Journalof ofPersonality, Personality, Lamiell, J. 48, 82-88. 82-88. 48, Toward an an idiothetic idiothetic psychology psychology of ofpersonality. personality. American American Psychologist, Psychologist, Lamiell, J. T. (1981). Toward 36, 276-289. 36, Lamiell, J. J. T. T. (1987). (1987). The Thepsychology psychologyof ofpersonality: personality: Anepistemological epistemological inquiry. York: Lamiell, An inquiry. NewNew York: Columbia University Press. (1990a). Explanation in the psychology of personality. In D. N. Robinson & Lamiell, J. T. (199Oa). & of theoretical theoretical psychology psychology (Vol. (Vol. 8,8, pp. pp. 153-192). 153-192). New New L. P. Mos (Eds.), Annals of York: Plenum.

140 140

JAMES LAMIELL JAMEST. T.LAMIELL

Lamiell, J. J. T. T.(199Ob). (1990b).Let's Let'sbe becareful careful out outthere: there:Reply Replytotocommentaries. commentaries.In InD. D.N. N.Robinson Robinson&& L. P. Mos (Eds.), Annals of theoretical psychology psychology (Vol. (Vol. 6,6, pp. pp. 219-231). 219-231). New New of theoretical York: Plenum. Lamiell, J. T. (1992a, August). What did Windelband mean by nomothetic? nomothetic? Paper Paper presented presented at the 100th annual convention of the American Psychological Association, Washing Washington, DC. William Stern, Stern, differential differentialpsychology psychology and and the theproblem problem of of Lamiell, J. T. (1992b, August). William individuality.Paper Paperpresented presentedatatthe the100th 100thannual annualconvention conventionof ofthe theAmerican AmericanPsycho Psychoindividuality. logical Association, Washington, DC. Lamiell, J. T. (1996). William Stern: More than the LQ. I.Q. Guy. In G. A. Kimble, C. A. Boneau, & M. Wertheimer (Eds.), Portraits Portraitsof ofpioneers pioneersininpsychology: psychology:Volume VolumeIIII(pp. (pp. 72-85). Hillsdale, NJ: Erlbaum. Lamiell, J. T., & Durbeck, P. (1987). Whence cognitive prototypes in impression formation? formation? of for dialectical reasoning as a generative process. Journal of Some empirical evidence for Behavior, 8,8,223-244. 223-244. Mind and Behavior, Lamiell, J. T., Foss, M. A., & Cavenee, P. (1980). On the relationship between conceptual , 54-73. schemes and behavior reports: A closer look. Journal of of Personality, Personality,48 48, 54-73. J. T., T.,Foss, Foss,M. M.A., A.,Larsen, Larsen,R. R.J., J.,&&Hempel, Hempel,A. A.(1983). (1983).Studies Studiesininintuitive intuitivepersonology personology Lamiell, J. from an idiothetic point of view: Implications for personaiity personality theory. Journal Journalof of Person Personfrom ality, 51, 51,438-467. 438-467. ality, Lamiell, J. T., Foss, M. A., Trierweiler, S. J., & Leffel, G. M. (1983). Toward a further further understanding of the intuitive personologist: Some preliminary evidence for the dialeti dialetiunderstanding cal quality quality of of subjective subjective personality personality impressions. impressions. Journal of of Personality, Personality,53, 53,213-235. 213-235. cal Lamiell, J. T., & & Trierweiler, Trierweiler, S. S.J.J.(1986). (1986).Idiothetic Idiotheticinquiry inquiryand andthe thechallenge challengetotoconventional conventional "nomotheticism." Journal of Personality,54, 54,460-469. 460-469. of Personality, Lamiell, J. T., Trierweiler, S. J., & Foss, M. (1983). Theoretical vs. vs. actuarial analyses of M. A. (1983). of of Personality, Personality, 51, 51, personality ratings, and other rudimentary distinctions. Journal of 259-274. Conceptions of of personality: personality: Theories Theories and research. research. New NewYork: York:Random Random Levy, L. (1970). Conceptions House. McCrae, R. R., & & Costa, Costa,P. P.T., T.,Jr. Jr.(1986). (1986).Clinical Clinicalassessment assessmentcan canbenefit benefitfrom fromrecent recentadvances advances in personality psychology [Comment]. [Comment]. American Psychologist, Psychologist,41, 41,1001-1003. 1001-1003. McCrae, R. R., & Costa, P. T., Jr. (1987). (1987). Validation of the five-factor model of personality across instruments and observers. Journal of Personality and andSocial SocialPsychology, Psychology,52, 52, of Personality 81-90. McCrae, R. R., & Costa, P. T., Jr. (1995). Trait explanations in personality psychology. European Journal of of Personality, Personality,9,9,231-252. 231-252. Theoretical risks risks and and tabular tabular asterisks: asterisks: Sir Sir Karl, Karl, Sir Sir Ronald, Ronald, and and the the slow slow Meehl, P. E. (1978). Theoretical 46,46,806-834. 806-834. Psychology, progress of soft psychology. Journal of of Consulting Consultingand andClinical ClinicalPsychology, personality. London: London:Oxford Oxford University University Press. Press. Murray, H. A. (1938). Explorations in personality. Murray, H. A., & Kluckholm, Kluckhohn, C. (1953). (1953). Outline of a conception of personality. In C. Kluckholm, H. A. Murray, & D. Schneider (Eds.), Personality Personalityininnature, nature,society, society,and and culture culture (2nd ed., pp. 3-52). New York: Knopf. Nunnally, J. C. (1967). Psychometric Psychometric theory. theory.New NewYork: York:McGraw-Hill. McGraw-Hill. differences and the explanation of behavior. In D. N. Robin RobinOzer, D. J. (1990). Individual differences & L. P. Mos (Eds.), Annals of of theoretical theoreticalpsychology psychology(Vol. (Vol.6,6,pp. pp.201-209). 201-209).New New son & York: Plenum Press.

CHAPTERS CHAPTER 5

INDIVIDUALSAND ANDDIFFERENCES DIFFERENCES INDIVIDUALS

141 141

Paunonen, S. V., & & Jackson, D. N. (1986a). Idiothetic inquiry and the toil of Sisyphus. Journalof ofPersonality, Personality,5454, 470-477. Journal , 470-477. Paunonen, S. V., & Jackson, D. N. (1986b). Nomothetic and idiothetic measurement in personality. 447-459. personaHty. Journal Journalof ofPersonality, Personality,54,54, 447-459. Pearson, K. (1901-1902). Editorial. Biometrika, Biometrika,1,1,3.3. Princeton, Princeton Porter, T. M. (1986). The statisticalthinking, thinking,1820-1900. 1820-1900. Princeton,NJ:NJ: Princeton The rise rise of of statistical University Press. Rorer, L. G., & & Widiger, Widiger, T. T. A. A. (1983). (1983). Personality Personality structure structure and and assessment. assessment. Annual AnnualReview Review of of Psychology, Psychology,34, 34,431-463. 431-463. Rychlak, J. J. F. (1981). A philosophy of of science sciencefor forpersonality personalitytheory theory(2nd (2nded.) ed.)Malabar, Malabar, FL: Krieger. The psychology psychology of of rigorous rigoroushumanism humanism(2nd (2nd ed.). New York: New Rychlak, J. J. F. F. (1988). (1988). The Rychlak, ed.). New York: New York University Press. Sampson, E. E. E. E.(1988). (1988).The Thedebate debateon onindividualism: individualism:Indigenous Indigenouspsychologies psychologiesofofthe theindividual individual and their role in personal and societal functioning. American 15-22. American Psychologist, Psychologist,43,43, 15-22. Schneider, D. J. (1973). Implicit personality theory: A review. Psychological PsychologicalReview, Review,79,79, 294-309. Sechrest, L. (1976). Personality. Annual Annual Review Reviewof ofPsychology, Psychology,27,27,1-27. 1-27. Shweder, R. A. (1975). How relevant is an individual differences theory of personality? Journal 455-484. Journalof ofPersonality, Personality,43,43, 455-484. Shweder, R. A. (1980). Factors and fictions fictions in in person person perception: perception: A A reply reply to to Lamiell, Lamiell, Foss, and Cavenee. Journal 74-81. Journalof ofPersonality, Personality,48,48, 74-81. 234-238. Skaggs, E. B. (1945). Personalistic psychology as science. Psychological PsychologicalReview, Review,52,52, 234-238. Stern, W. (1906). Person Weltanschauung, BandBand I: I: Person und Sache: Sache:System Systemder derphilosophischen philosophischen Weltanschauung, Ableitung philosophical world view. Grundlehre [Person [Personand andthing: thing:Systematic Systematic philosophical world view. Ableitung und Grundlehre Leipzig: Barth. Volume foundations]. VolumeI:I:Philosophical Philosophical foundations]. Leipzig: Barth. Stern, W. (1917). Die Psychologie Psychologic und der Personalismus Personalismus[Psychology [Psychologyand andpersonalism]. personalism], Leipzig: Barth. Stern, W. (1918). Person der philosophischen Welftanschauung, BandBand II: II: Person und undSache: Sache:System System der philosophischen Welftanschauung, Die menschliche [Person andand thing: Systematic philosophical world world view, view. menschlichePersonlichkeit Persdnlichkeit [Person thing: Systematic philosophical Leipzig: Barth. Volume personality]. VolumeII: II:The Thehuman human personality]. Leipzig: Barth. System desdes kritischen Personalism us, Band III: Wert philo Stern, W. W. (1924). (1924).Person Personund undSache: Sache: System kritischen Personalismus, Band III: WertphiloSystem ofof critical personalism, Volume III: Philosophy of of sophie Sophie [Person [Personand andthing: thing: System critical personalism, Volume III: Philosophy values]. values].Leipzig: Leipzig:Barth. Barth. Windelband, W. (1904). Geschichte (3rd unaltered ed.). Strassburg: Geschichteund undNaturwissenschaft Naturwissenschaft (3rd unaltered ed.). -Strassburg: Heitz. (Original work published 1894). Woody, E. Z. (1983). The intuitive personologist revisited: revisited: A critique of dialectical person perception. Journal Journalof ofPersonality, Personality,51,51,236-258. 236-258.

1 1 1 This Page Intentionally Left Blank

1 1 1 1 1 1 1 1 1 1 1 1 1 1

1

CHAPTER 6 CHAPTER 6

PERSONALITY MEASUREMENT AND VALIDITY ISSUES RELIABILITY AND

STEPHEN G . WEST STEPHEN G. ARIZONA STATE STATE UNIVERSITY UNIVERSITY ARIZONA

F . FINCH JOHN F. TEXAS A&M A & M UNIVERSITY UNIVERSITY TEXAS

The present chapter introduces researchers to some of the classic work work in in personality personality The measurement. We We consider three fundamental issues in personality measurement: the nature of constructs, reliability, reliability,and andvalidity. validity.Throughout Throughoutthe thechapter, chapter,our ourpresen presenthe on techniques that hold promise of making future contributions tation will focus on the enhancement of our basic, theoretical understanding of personaUty. to the personality. ConseConse least aa rudimen rudimenquently, we will also generally favor approaches that begin with at least tary theory of the construct. No No attempt will be made to address issues that arise solely in applied personality personaHty research. Our Our focus on on fundamental issues also prepre cludes consideration of more advanced statistical models for testing and structure of personality personaUty measures (see Ozer & Reise, 1994; West & Finch, 1996, for reviews).

I. I. THE THE NATURE NATURE OF OF PERSONALITY PERSONALITY CONSTRUcrs: CONSTRUCTS: BASIC BASIC ISSUES ISSUES To appropriately interpret psychometric evidence, it is important to understand the "theory" "theory" of the the construct being investigated (Ozer & & Reise, 1994). Here, we the theory of will simply introduce several questions that should be addressed in the the construct that are are relevant to the the psychometric issues addressed later in in this HAIVDBOOK OF PERSONALITY PSYCHOLOGY

HANDBOOK OF PERSONALITY PSYCHOLOGY

143

143

CoPYRIGHT 991 BY COPYRIGHT Cl© t1997 BY ACADEMIC ACADEMIC Puss. PRESS. ALL RIGHTS OF REPRODUCTION IN ANY FORM RESERVED.

ALL RIGHTS OF REPRooucnoN IN ANY FORM RESERVED.

144 144

WEST AND AND FINCH

chapter. More complete discussions of construct theory issues can be found in Cronbach and Meehl (1955), and Wiggins Wiggins (1973). (1973). (1955), Loevinger Loevinger (1957), (1957),Messick Messick (1989), (1989),and 1. What is the expected degree of relationship among items that constitute the measure of the construct? Researchers often assume that items should have an adequate degree of intercorrelation (high internal consistency). For the trait and ability measures typically utilized by personality researchers, this assumption is influnearly always reasonable. According to this conception, each item should be influ enced to a degree by the underlying trait construct, giving rise to a pattern of positive intercorrelations so long as all items are oriented (worded) in the same direction. For example, the Beck Depression Inventory (Beck, Ward, Mendelson, Mock, & & Erbaugh, 1961) includes items assessing negative mood, sleep disturbance, and lack of energy, which are expected to be positively related. Alternative conceptions exist that do not lead to expectations of positive examples: (a) The original relations among items, among which are the following examples: & Rahe, 1967) treated life events such as death conception of life events (Holmes & of a spouse and getting a mortgage as being virtually random occurrences; little if BoUen & & Lennox, 1991). The any relation among them would be expected (see Bollen the life events critical measure is the total amount of life stress associated with all of the period, (b) Frederiksen and Ward that occur to each person in a specified time period. (1978) proposed that the mean level of creativity on a series of intellectual tasks does not not sharply sharply differentiate differentiate highly highly creative creative from from less less creative creative people. people. Rather, Rather, it it does is the the maximum maximum level level of of creative creative work work on on the the tasks tasks (capability) (capability) that that should should be be is hydraulic utilized, (c) (c) Psychodynamic Psychodynamic and and motivational motivational perspectives perspectives often often assume assume aa hydraulic utilized. in which which one one mode mode of of expression expression of of aa conflict conflict or or motive motive may may be be in in competition competition model in model with another another mode. mode. A A measure measure of of compulsive compulsive behaviors, behaviors, for for example, example, may may find with find that items items measuring measuring avoidance avoidance of of stepping stepping on on cracks cracks in in sidewalks sidewalks and and repetitive repetitive that handwashing could could even even be be negatively negatively related. related. handwashing 2. What is the structure of the construct? Personality researchers typically (and often implicitly) assume that a single dimension underlies each construct. The dimension may be assumed to be bipolar as in a mood scale that is anchored by "high degree of positive mood" and "high degree of negative mood" as its two as in a mood scale that is anchored ends. Alternatively, dimensions may be unipolar as not present" present" and and "high "highdegree degreeof ofnegative negativemood." mood."More Morecomplex complex by "negative mood not dimensional structures of single constructs may be also proposed, most commonly hierarchical structures. For example, Costa and McCrae (1992) have proposed that each of of the the Big Big Five Five dimensions dimensions of of personality personality has has an an underlying underlying hierarchical hierarchical each For example, extraversion is composed of lower order dimensions (facets) structure. For example, extraversion is composed of lower order dimensions (facets) of warmth, gregariousness, assertiveness, activity, excitement seeking, and positive of warmth, gregariousness, assertiveness, activity, excitement seeking, and positive emotions. Finally, conceptions conceptions have have been been proposed proposed in in which which types types or or latent latent classes classes emotions. Finally, are believed to underlie the measures. Gender is a straightforward example: There are believed to underlie the measures. Gender is a straightforward example: There are two discrete types (male, female). Latent class conceptualizations of other are two discrete types (male, female). Latent class conceptualizations of other personality variables (e.g., self-monitoring) have also recently been proposed (Ganpersonality variables (e.g., self-monitoring) have also recently been proposed (Gan gestad & & Synder, Synder, 1985, 1985,1991). gestad 1991).

CHAPTER 66

PERSONALITY MEASUREMENT PERSONAUTY

145 145

3. What is the stability of the construct? Researchers studying traits and abilities have typically assumed that their constructs were stable over time. However, However, other researchers have studied state variables (e.g., mood) that would be expected to vary from from day to day. Still other researchers have taken developmental perspec perspectives in which some constructs are expected to be stable only within a specified specified period of development. 4. What is the pattern pattern of relationships of measures of of the construct of interest with other measures of the same construct and with measures of other constructs? Cronbach Cronbach and Meehl (1955) long ago proposed the model of a nomological net in which laws relate ''(a) "(a) observable properties of quantities to each other; or (b) theoretical constructs to observables; or (c) different different theoretical constructs to each other." (p. 290). 290). Although only the rudiments of a nomological net may be specifiable for a new construct, the nomological net should become increasingly well specified specified as research proceeds. Specification Specification of such a a net provides a blueprint for identify for testing the construct theory. At the same time, researchers should ideally identify other constructs that can be proposed as competing interpretations. For example, investigators studying assertiveness may need to clearly distinguish their construct from aggression. and its measures from These questions are introduced introduced so that readers will recognize that evidence about internal consistency, structure, stability, and relations with other measures and other constructs must be evaluated in terms of the theory of the construct under investigation. High internal consistency and stability are typically desirable for traits; they are not desirable properties for measures of life stress in properties for structerms of the original Holmes and Rahe conception. Some conceptions of the struc ture of self-esteem self-esteem propose a single general underlying dimension, whereas other conceptions emphasize several related dimensions. The more clearly the construct specify answers to each of the above questions, the easier it is for for theory can specify straightforward tests of the theory. researchers to collect evidence that provides straightforward

II. RELIABILITY ll.

A. Classical Classical Test Test Theory Theory Perspective Perspective A. An important property of good measures is reliability. rehability. At its heart, reliability is a simple concept-it concept—it is an index of the reproducibility or dependability of measure measurePersonality researchers have traditionally treated issues of reliability within ments. Personality framework of classical test theory (Crocker & & Algina, 1985; GuUiksen, 1950; the framework 1985; Gulliksen, 1968; Nunnally & & Bernstein, 1993), which will be emphasized here. & Novick, 1968; Lord & rehability and certain data structures that severely However, certain aspects of reliability violate the assumptions of this framework, which we shortly describe, may best be for a comprehen comprehen& Brennan, 1989, 1989, for treated using alternative approaches (see Feldt & sive overview).

146 146

WESTAND ANDFINCH FINCH WEST

(X) Classical test theory begins by partitioning each observed measurement (X) into two components: "true score" (T) and measurement error (E). Each true score individual. represents the mean of a very large number of measurements on a specific individual. In contrast, measurement error lumps together all of the transient influences that can affect affect test scores. Such influences are presumed to fluctuate randomly from measurement occasion to measurement occasion. Feldt and Brennan (1989; see also Stanley, 1971; 1971; Thorndike, 1951) have identified some of of the potential general sources of these influences: 1. Subject-related characteristics (e.g., health, concentration, recent life events) 2. Characteristics of of testing situation (e.g., noisiness of room) 3. Examiner characteristics (e.g., examiner race, idiosyncrasies or subjectivity in ratings or observations) 4. Characteristics of of instruments (e.g., equipment problems, sampling of items)

Classical test theory begins with the notion of parallel forms of a test (or measure). Parallel forms implies that the different different versions of the test have the same mean, variance, and distributional characteristics, and correlate equally with each other as well as with external criteria in large samples (see Lord & & Novick, 1968). Under these assumptions, true score and measurement error can be treated as independent. This implies that the variance of the observed test scores will equal the sum of the variance of of the true scores and the variance of the measurement a^-ai-^ai. Reliability (Pxx (pxx)) is then defined as the ratio of the variance error, u1 = of + �. Reliability of the true scores to the variance of the observed scores, Pxx pxx = = orlo^* orf/�. different ways, each of which In practice, reliability is assessed in several different makes somewhat different different assumptions; is prone to different different biases; and has a different different meaning. Two procedures, internal consistency and test-retest, are utilized most commonly in personality research. 1. Internal consistency. Questions about the degree of relationship among items that constitute a measure are typically addressed using an an index of internal consistency. The internal consistency of a measure can be evaluated by dividing each subject's test into two halves according to a specified procedure (e.g., odd versus even numbered items) and then correlating the score on the two halves. Cronbach's (1951) coefficient coefficient a a is the most commonly used of of these measures and is equal to the mean of of the correlations between all possible split halves of of the test. An equivalent statistic for dichotomous items is provided by Kuder-Richardson's Kuder-Richardson's Coefficient a a provides a good estimate of of reliability in terms (1937) Formula 20. Coefficient of the sampling of items from the content area, often the major source of measure measureof (Nunnally & & Bernstein, 1993), and normally provides a good estimate ment error (Nunnally of the reliability of an alternate form of a test. However, coefficient coefficient a a does not of intake into consideration fluctuations in the subjects, situation, examiner, or in between testing occasions. Feldt, Woodruff, Woodruff, and Salih struments that may occur between

CHAPTER CHAPTER 66

147 147

PERSONALIlY PERSONALITY MEASUREMENT MEASUREMENT

(1987) present statistical tests of coefficient coefficient a a in one-sample and multiple-sample cases. Coefficient Coefficient a a also increases with increasing test length. The value of coefficient coefficient a a for a longer test can be estimated using the Spearman-Brown prophecy formula formula (see Crocker & Algina, 1986). The formula is fJp p ^PP

=

kPxx 1 + + ((A: - 1)Pxx l)pxx' k -

where fJp pppp is the estimated internal consistency of the projected (long) test, k is the pxx is the reliability of the ratio of the length of the long to the short test, and Pxx 10-item test has a coefficient a = = .70 and a short test. For example, if an existing lO-item coefficient a 20-item test is proposed, the estimated internal consistency of the new test is .82. for the projected long test would be sampled from from This estimate assumes that items for the same content domain and have the same distributional properties as the existing coefficient a a is often often regarded as an estimate of reliability 10 items. Consequently, coefficient specified length.^ for a test of a specified length. 1 Test-retest Measures of test-retest test-retest reliability are typically used to address 2. Test-retest. questions about the stability of personality constructs. In test-retest reliability, the Pearson correlation between subjects' scores on a measure at Time 1 and the scores of the same group of subjects on the same measure (identical items) at Time 2 is computed. This approach makes two strong assumptions: (a) subjects' levels on the administrations,^ and ability or trait in question should not change between test administrations,2 useful memory for the items that could affect affect their (b) subjects should have no useful responses on the second administration. Since stability of traits and abilities tends to decrease over time (Conley, 1984), relatively short intervals between test and retest are normally recommended. However, very short test-retest intervals can day-toeasily lead to overestimates of the reliability of the test with respect to day-to day influences because of implicit pressures on the respondent to give consistent responses (see McClelland, 1980). It should be noted that these two forms of reliability focus on different ques different questions. Internal consistency addresses the sampling of the items and within-test fluctuations in subject characteristics (e.g., changes in concentration) (Nunnally & Test-retest measures address day-to-day variations in subject Bernstein, 1993). Test-retest characteristics, the testing situation, and the examiner. Given that a constant set of items is utilized, the adequacy of sampling of items is not addressed. Personality Technically,coefficient coefficientaaassumes assumesthat thattests testsare areatatleast leasttau tauequivalent equivalent(Lord (Lord&&Novick, Novick,1968), 1968), 1^Technically, is identical. Consequently, coefficient coefficient a a will underestimate meaning that the factor loading for each item is the internal consistency of tests composed of items with unequal factor loadings (see BoUen, Bollen, 1989). 1989). At coefficient a a does not address fluctuations between testing occasions which can reduce the same time, coefficient the reliability of a test. Pearsoncorrelations correlations are arenot not influenced influenced by byshifts shifts ininmean meanlevel, level,only onlyby byshifts shifts ininthe therelative relative 2^Pearson ranking of the subjects in z-score terms from Time 11 to Time 2. When mean shifts are theoretically intraclass correlation coefficient coefficient is more appropriate. also considered to be sources of error, the intrac1ass

AND fiNCH FINCH WEST AND

148 148

researchers normally focus on trait and ability measures which should ideally be high on both types of reliability. However, as noted earlier in The Nature of of Personality Constructs (Section I), other patterns may be reasonable for certain constructs. A measure of angry mood would be expected to have high internal consistency, but low test-retest reliability. A measure of daily life events that have occurred during the past week might be expected to have low internal consistency (e.g., school events and relationship events will tend to be minimally related) and low test-retest reliability of the events that occur during non-overlapping time periods. periods. Nonetheless, short-term retrospective recall of the events of a previous week should exhibit high reliability when compared with the original reports (Sand(Sand ler & & Guenther, 1985). 1985). Thus, Thus,the thenature natureof ofreliability reliabilityevidence evidencesought soughtfor foraa measure measure should depend on the theoretical conception of the measure (Ozer & & Reise, 1994). Reise, 1994). B. B. Generalizability Generalizability Theory Theory Perspective Perspective Classical test theory lumps together all influences that may produce measurement error. Generalizability theory (Cronbach, GIeser, Gleser, Nanda, & & Rajaratnam, 1972; & Webb, Webb, 1991; 1991; Wiggins, 1973) 1973) is a general alternative to classical test Shavelson & theory that yields more refined measures of reliability and accommodates a wider variety of data structures. The theory also serves as an important heuristic tool for thinking about issues of reliability and generalizability. The application of generalizability theory begins by designing measurement studies in which whichfacets facets that thatpotentially potentiallyinfluence influence the the observed observedscores scoresare are deliberately deliberately varied. An observational study might collect measurements of aggressiveness on each child in the sample under each possible combination of the following facets: (a) observers, (b) days of the week, and (c) classroom topics (e.g.. (e.g., Math, English). Drawing on extensions of analysis of variance models (Cronbach et al., Lind al., 1972; 1972; Lindquist, 1953), 1953), the variance attributable to each of the facets and their two-way and three-way interactions can be estimated. It may tum turn out that some of the facets facets account for a trivial percentage of the variance and can be neglected, whereas others are are very very important important and and must mustbe be considered consideredin in any any study. study.Intraclass Intraclasscorrelation correlation coefficients coefficients (analogous to reliability reliability coefficients) coefficients) can can be be calculated calculated that, that,for for example, example, describe the relationship between the observation of one or more of the observers and a universe of similar observers. These These coefficients coefficients are are useful useful in in planning planning subse subsequent studies. For example, if a follow-up study can collect data only three days per week or can only utilize two instead of four observers, the theory provides clear methods of estimating the level of expected reliability of the measurements under each measurement plan. Generalizability Generalizability theory theoryaccommodates accommodatesaawide widevariety variety of data structures that cannot be be addressed in in classical test test theory theorywithout withoutsignificant significant modification modification of the formulas. For example, the Attributional Style Questionnaire (Peterson et al., 1982) has subjects give three separate responses to each item, meaning that these responses are not independent. Generalizability theory permits straightforward col straightforward calculation of appropriate reliability reUabihty coefficients coefficients once the data collection design is known.

CHAPTER CHAPTER 66

PERSONAUTY PERSONALITY MEASUREMENT MEASUREMENT

149 149

Generalizability theory has often often been utilized by educational researchers, but only rarely by personality researchers in assessing the reliability of measures (see, e.g:, e.g., Farrell, MarioUo, Mariotto, Conger, Curran, & & Wallander, 1979). 1979). In part, this may (see, reflect reflect the discomfort discomfort that many psychologists appear to experience in mixing correlational and analysis of variance approaches (Cronbach, 1957). 1957). There are, in addition, appreciable difficulties difficulties in utilizing the approach. Among these are the complexities in understanding and specifying analysis of variance models with sev several random factors, the instability of the variance estimates with small sample sizes, and the possibility of obtaining negative estimates for for some of the variance 1989; Shavelson & components ((Jones Jones & & Appelbaum, 1989; & Webb, 1991). 1991). Despite these redifficulties, generalizability theory continues to hold considerable promise for re carefully probing the dependability of their measures. In searchers interested in carefully usefulness of generalizability theory has received special emphasis particular, the usefulness self-other agreement (Kenny, 1993). from researchers studying consensus and self-other c. C. Importance Importance of of Reliability Reliability

Researchers originally trained in experimental psychology or social psychology frequently underestimate the importance of reliability, often often using one-item mea meafrequently reUability in their research. In a simple correlational sures or measures of unknown reliability .5 correlation and the reliability study, if the true scores of two measures have a .5 0.3. This of each measure is .6, .6, the obtained correlation can be expected to be 0.3. underestimated and that a much larger sample means that degree of relationship is underestimated size will be needed to detect the relationship between the two variables than would perfect reliability. Unreliability of the be the case if they were measured with perfect randomdelivery of the treatment or in measurement of the dependent variable in a random ized experiment also leads to underestimation of the magnitude of the treatment effect. In multiple regression analyses, the inclusion of one highly reliable predictor among a set of correlated predictor variables that are measured with less reliability can easily lead to overestimation of the importance of the more reliable predictor. multitrait-multimethod matrices to be discussed in a later section, In the analysis of multitrait-multimethod differential reliability of the measures can lead to mistaken conclusions about differential vahdity. In short, unreliable measures have consider considerconvergent and discriminant validity. able potential to bias all results to which they contribute. A perhaps less obvious example of this problem occurs when a single measure for a personality of behavior is collected and used as a "gold standard" criterion for self-report measure of aggressiveness might be correlated measure. For example, a self-report with the "gold standard" of the intensity of electric shocks ostensively delivered confederate as punishment for for mistakes. In such studies, no information information is to a confederate typically provided provided about about the the reliability reliability of of the the behavioral behavioral measure measure of of aggressiveness aggressiveness typically across days, days, across across confederates, confederates, or or across across types types of of punishment punishment (e.g., (e.g., electric electric shock shock across versus aversive aversive noise). noise). If If the the two two measures measures do do not not show show the the expected expected degree degree of versus of correlation, this this lack lack of of information information makes makes it it impossible impossible to to define define whether whether the the correlation, source of of the the problem problem is is aa failure failure of of the the construct construct theory theory or or lack lack of of dependability dependability source

150 150

WEST WEST AND AND FINCH FINCH

of Gormly (1984) of the behavioral behavioral measure. Moskowitz Moskowitz and Schwarz Schwarz (1982) (1982) and Gormly (1984) cogent examples of of the use of of different different forms forms of aggregation to produce produce provide cogent of aggregation more measures of the same more reliable measures which then correlate with with other other measures of the construct. Epstein Epstein (1983) and Rushton, Brainerd, and Pressley Pressley (1983) present present more complete discussions of of this issue. complete

Ill. in. VALIDITY Once issue of Once it is established established that a measure measure has adequate adequate reliability, the the issue of validity arises. Closely define validity as the degree Closely following following Messick Messick (1989, p. 13), we we define degree to which empirical empirical evidence evidence and theoretical theoretical rationales rationales support the adequacy support adequacy of interpretations based on test scores or other measures. This definition underscores interpretations definition underscores theory and validity the close relationship between construct theory validity and highlights the field's increasing emphasis on construct construct validity (American (American Psychological Associafield's increasing Psychological Associa tion, 1985; Angoff, Angoff, 1988; Cronbach, 1989). Based on on this definition, we we next next identify identify of validity. validity. a few of the characteristics and complexities of Validity is a property property of of the interpretation interpretation of of the measure measure 1. Validity of a measure, not not of T o illustrate, there is considerable considerable body body of of evidence evidence supporting supporting the the interpreta interpretaitself. To of scores on the Stanford-Binet Stanford-Binet test as a measure measure of intelligence in young young tion of of intelligence of evidence evidence supporting supporting the interpretation of of these these scores scores children. There is a paucity of the interpretation as a measure of weaker evidence of creativity creativity in children and relatively relatively weaker evidence supporting supporting the interpretation adult college college students. interpretation of of these scores as intelligence in adult 2. Validity involves the interpretation of hypothesis about) the of (a hypothesis the meaning meaning of of scores scores based on a measure. The better developed developed the theory theory of of the hypothesized hypothesized underlying underlying construct and the specification specification of of alternative alternative underlying underlying constructs, the easier it will be to collect clear empirical against a particular empirical evidence evidence for for or against interpretation of of a test score. on the current current preponder 3. The validity validity of of an interpretation interpretation is always always based based on preponderance of of evidence and is subject to change. New N e w evidence evidence may may arise to challenge challenge an existing of the the existing evidence may existing interpretation interpretation or a new new alternative alternative account account of existing evidence be proposed. The validity of an interpretation of a measure is never be validity of interpretation of measure never established; it is only currently supported to the degree warranted by by the the empirical empirical evidence. supported degree warranted 4. The validity present in the form validity of of an interpretation interpretation has been been difficult difficult to to present form of difficulty, researchers researchers all of a convenient convenient quantitative quantitative index. Perhaps because because of of this difficulty, too often often report report quantitative quantitative evidence evidence of of reliability reliability (indices (indices of consistency, too of internal internal consistency, test-retest correlations, or both), but fail to report report evidence evidence supporting the validity supporting the of their preferred preferred interpretation interpretation of of their measures. When When validity evidence has been been of validity evidence have traditionally traditionally relied on on either either a qualitative summary of of reported, researchers have qualitative summary evidence provided provided by the body body of of available research research or a listing of a selected selected the evidence listing of of correlations with other measures. At A t least least for for some some validity validity questions, string of newer techniques techniques including including generalizability generalizability theory theory (Kane, meta-analysis newer (Kane, 1982) and meta-analysis

PERSONALITYMEASUREMENT MEASUREMENT CHAPTER 66 PERSONALITY

151 151

(Hedges, 1988; Mabe & & West, 1982; Schmidt, Hunter, Pearlman, & & Hirsh, 1985) offer the promise promise of of providing more more comprehensive, quantitative estimates. offer of validity, its historicl:l1 historical evolution Readers wishing more in-depth discussions of as a concept, and and its philosophy philosophy of of science underpinnings should refer refer to Angoff Angoff (1988), Cronbach (1989), and Messick (1989). Forms of of Validity Evidence A. Fonns

As with any hypothesis, a variety of of different different forms forms of of evidence may be sought of the interpretation interpretation of of the measure. Several with which to probe the validity of commonly utilized forms forms of of validity evidence are briefly briefly presented imporcommonly presented next. The impor of each of of the forms forms of of evidence will vary depending on the nature of of the tance of validity question being addressed addressed and the construct theory, issues we raised earlier chapter. in the chapter. 1. i. Content

Evidence about about the validity validity of of the content content of of a measure involves comparisons of of coverage of of the items with the domain domain implied by the construct theory. the actual coverage of the content will be This means that the clearest evidence about the adequacy of available when a well-specified well-specified theory theory specifies specifies the domain and facets of of the con construct. Three potential problems with measures may arise. First, important important facets facets of the construct may be underrepresented underrepresented by having too few or no items. For of proposed that six facets facets example, Jackson, Ahmed, and Heapy (1976) originally proposed (competitiveness, concern for for excellence, status with experts, status with peers, achievement via independence) independence) constitute need for for achieve achieveacquisitiveness, and achievement ment. In terms of of Jackson et al.'s conception, a measure that did not include items on status with peers would underrepresent underrepresent the construct construct of of need for for achievement. achievement. Second, a facet facet or even a small portion of of a facet facet may be represented represented by a large of items relative to other facets. Severe oversampling of of items can increase number of of an aspect of of the construct, turning what is in reality the perceived importance of major one (sometimes termed a "bloated "bloated a minor dimension into an apparent major specific factor"). Third, additional additional dimensions dimensions not specified specified by the construct theory specific For example, if most of of the items of of a measure of need may be reliably measured. For for achievement achievement were worded worded in a socially desirable manner or if several of of the for fear of of failure, then the measure could be contaminated by these items measured fear construct-irrelevant dimensions. construct-irrelevant The adequacy of of the measure's coverage of of the content of the construct construct is normally assessed in two ways. First, judges can review the items for for completeness completeness of coverage of of the domain of of the construct. Depending on the type and evenness of of measure being developed, experts in the content area or representatives of of the of subject population or both may serve as judges. As a first step in providing providing evidence subject concerning the validity of of the measure, the judges should reach consensus that concerning adequate coverage of of the domain domain of of the construct construct has been achieved. Indeed, with

152 152

WEST WEST AND AND FINCH FINCH

a construct theory theory that clearly specifies specifies the domain and the facets of the construct, item writers working independently should ideally be able to produce highly related (and in the limit, interchangeable) measures of the construct. Second, psychometric investigations should indicate that the data are consisconsis tent with the structure hypothesized by the construct theory. For example, Snyder's which was wasoriginally originallyproposed proposedto tohave haveaasingle singleunderly underly(1974) self-monitoring scale, which ing dimension (factor), was later found by Briggs, Cheek, and Buss (1980) to have three underlying dimensions. Jackson et al. (1976) found that data were not consistent with their hypothesis that need for achievement is a hierarchically orgaorga nized construct with six distinct facets (second-order factor model). Failures to confirm the structure of the hypothesized construct typically indicate that the meamea revision. Data consistent with with the hypothe hypothesure, the construct, or both need further revision. sized structure support the continued use of both the measure and the construct theory. Given the importance that has historically been placed on measures having only one underlying factor (dimension) and the frequency with which one factor isuseful useful to toexamine examinethe theissue issueof ofsingle singlefactoredness factoredness structures have been proposed, it is in greater detail. Several cases exist in the literature in which a one-dimensional structure was hypothesized, but was not initially subjected to a strong empirical test. Rotter's (1966) internal-external (I-E) locus of control scale and Snyder's (1974) self-monitoring (S-M) scale, to cite two examples, have been criticized in grounds that thatseveral severallargely largelyindependent independentfactors factorsactually actuallyunder underthe literature on the grounds lie the items (see, for example, Collins, 1974; 1974; Mirels, Mirels, 1970, 1970,for forI-E; I-E;Briggs Briggs&&Cheek, Cheek, 1988; 1988; Briggs et aI., al, 1980, for S-M). In these cases, one underlying dimension of the measure can be responsible for correlations with one set of criterion variables, whereas another underlying dimension is responsible for correlations with a second set of criterion variables. Worse still, it is theoretically possible for the two dimendimen sions to correlate in in opposite directions with a set of criterion variables, producing an overall 0 relationship. Zuckerman and Gerbasi (1977) provide a nice illustration of some of these problems, showing that many of the ambiguities in a portion of of the I-E literature could be clarified by consideration of the multifactor structure of the measure. measure. Neuberg, Neuberg,Judice, Judice,and andWest West(in (inpress) press)show showhow howgreater greaterempirical empiricaland and & conceptual clarity can be brought to the literature on need for closure (Webster & Kruglanski, 1994) through careful consideration of the two largely independent dimensions that underlie this measure. Unidimensional measures are clearly prepre ferred unless a well-developed theory precisely details the relationships among the firstmultiple dimensions (or classes) that compose the construct as well as their first order and and interactive interactive relationships relationships with with external external criteria. criteria. Masculinity-femininity Masculinity-femininity order scales provide provide an an important important illustration illustration of of how how aa construct construct originally originally conceived conceived scales of as as unidimensional unidimensional (Constantinople, (Constantinople, 1973) has been been reconceptualized reconceptualized as as being being of 1973) has composed of two separate dimensions (Bem, 1974; Spence & Helmreich, 1978). composed of two separate dimensions (Bern, 1974; Spence & Helmreich, 1978). Carver (1989), (1989), Hull, Hull, Lehn, Lehn, and and Tedlie Tedlie (1991), (1991), Messick Messick (1989), (1989), and and Jackson Jackson and Carver and Paunonen (1985) present general discussions of some of these issues; Sternberg Paunonen (1985) present general discussions of some of these issues; Sternberg and Weil Weil (1980) (1980) offer offer an an empirical empirical illustration illustration of of how how well-developed well-developed theory theory can and can

CHAPTER 66 PERSONALITY PERSONALITYMEASUREMENT MEASUREMENT CHAPTER

153 153

guide tests of the construct's hypothesized structure and its relationship with exterexter nal criteria. The desirability and ease of interpretation of unidimensional tests have long of unidimensional been recognized (e.g., Cronbach, 1951; 1947, 1948; McMc 1951; Guilford, 1954; Loevinger, 1947,1948; Nemar, 1946; 1946; see Walker, 1931, for an early mention). Numerous indices that purport to assess unidimensionality unidimensionality have been developed developed over the years (Hattie, 1985). Nonetheless, the index most commonly utilized utilized by personality researchers for this purpose is an inappropriate a, which, as inappropriate one, Cronbach's (1951) coefficient coefficient a, discussed previously if previously is a measure of internal consistency. Typically in practice, if a scale has a coefficient a a of about .70 or better for a reasonable length test, it is taken as an adequate measure of the underlying dimension. A A unidimensional unidimensional scale will necessarily produce a high coefficient coefficient a. a. Unfortunately, Unfortunately, scales having multiple underlying factors can also easily produce high levels of coefficient a. of coefficient a. For example, example. Green, Lissitz, and Mulaik (1977) have shown in a simulation that a lO-item 10-item test in which scores were produced by five underlying underlying factors can produce values of of coefficient a greater than .80! Consequently, we recommend that investigators .80! coefficient a utilize more sensitive techniques to detect departures from unidimensionality. These techniques include confirmatory confirmatory factor analysis (BoUen, (Bollen, 1989) for ratings, and item response theory (Hambleton, Swaminathan, & & Rogers, 1991) for dichotomous items (see Ozer & & Reise, 1994; & Finch, 1995, for overviews). Coefficient Coefficient a a should 1994; West & be restricted to its intended use as a measure of of internal consistency in line with Cronbach's (1951) original recommendations. 2. External External Criteria Criteria 2.

A A second source of validity evidence comes from the degree to which the measure can predict external criteria that are theoretically expected to be related to the imporconstruct being measured. Loevinger (1957) has in particular emphasized the impor tance of external criteria in validational efforts: "It seems reasonable to require of any test include a demonstration of that complete validation of of some non-zero relationship with a non-test variable" (p. 675). The criteria may occur simultaneously with the measurement of of the construct or be expected to occur in the future. Potential criteria may come from a wide variety of different sources, such as behav of different behavioral samples in laboratory or naturalistic settings, ratings by knowledgeable infor informants or clinicians, biographical data, physiological data, and other self-report measures of of the same construct. The usefulness of of each potential criterion source will depend on the nature of the construct and the type of of of measure being validated vaUdated (Moskowitz, 1986). As a simple illustration of the validation of a construct against external criteria, a measure of extraversion might be expected to discriminate current members of social clubs from nonmembers. Or, it might be expected to predict the likelihood that an individual would initiate future future conversations conversations with strangers. Or, it might be expected to predict that individuals scoring as extroverts on the measure would be rated as being extraverted by their spouses and their employers. Such types of the measure as extraversion. of evidence support the preferred interpretation of

154 154

WEST WEST AND AND FINCH FINCH

However, it is possible that other theoretically distinguishable characteristics (e.g., anxiety) may also be assessed by the measure that also predict the external criteria. In such cases, the measure will better predict the criterion (e.g., membership in social clubs) to the extent it reflects all of the factors (e.g., high extraversion and efforts in low anxiety) that influence the external criterion. Extended validation efforts measuresuccessfully successfully predicts predictsdifferent different external externalcriteria criteriaininseveral severalseparate separate which the measure studies make a much more convincing case for the interpretation, particularly if a strong nomological network is developed linking the construct to a diverse set of of predicted outcomes (Cronbach & & Meehl, Meehl, 1955). 1955).Yet, Yet,even evensuch suchprograms programsof ofvalida validation still must be carefully scrutinized to determine whether other variables may form aa common common basis basis for for group group membership membership or or the the occurrence occurrence of of the the predicted predicted form behavior (Houts, Cook, & Shadish, 1986). behavior (Houts, Cook, & Shadish, 1986).

3. 3.

ExperimentalManipulations Manipulations Experimental

in which whichthe thesubject's subject'slevel levelon onthe theconstruct constructisisdirectly directlymanipu manipuExperimental designs in lated can also provide important validity information about measures of certain Measuresof ofstate statevariables variablessuch suchasas negative mood should show significant negative mood should show significant constructs. Measures films change when subjects are exposed to an appropriate manipulation such as sad films stories. Predictions about trait traitvariables variablessuch suchasasdepression depressionare areless lessclear clear or news stories. and depend on the strength of the manipulation and the degree of overlap between the state and the trait constructs. Exposing subjects to a sad film is sadfilm isunlikely unlikelyto toresult result in significant changes on those facets of a measure of depression related to sleep disturbance and lack of energy. In contrast, psychotherapeutic interventions that tobe beeffective effective should shouldproduce producesignificant significantchanges changesrelative relativetotoananuntreated untreated are known to control group on a new measure of depression. B. Validation B. Convergent Convergent and and Discriminant Discriminant Validation Validation of a measure requires a twofold approach. First, as outlined in the previous sections, evidence bearing on predictions made by the construct theory needs to be collected. A measure of assertiveness might be validated through a variety of methods such as asking knowledgeable informants to rate each person on assertiveness, behavioral observations in a standard assertiveness test situation, or correlation with scores scores on onanother another existing existingmeasure measureof ofassertiveness. assertiveness.This Thisseeking seeking of confirmatory confirmatory evidence for the proposed interpretation interpretation of the measure has been convergentvalidation. validation. Second, evidence is also needed showing Second, evidence is also needed showing thatthat thethe termed convergent measure under consideration differs from measures of other constructs. For examexam ple, it should be possible to empirically distinguish a new measure of assertiveness discrimifrom existing measures of aggressiveness. This second approach, termed discrimi nant validation, validation,helps helps justifythethe proposal a new construct. nant to to justify proposal ofof a new construct. Studies of convergent and discriminant validity need to be planned carefully as a number of biases can distort the results. results. Measures purporting to assess the different definitions or theories of the same construct can be based on radically different different measures of self-esteem construct. For example, correlations between different

PERSONALITYMEASUREMENT MEASUREMENT CHAPTER 66 PERSONALI1Y

ISS 155

from high (.6 to .8) to low negative ((-.2) Different range from .2) (Briggs & Cheek, 1986). Different facets of of measures may reflect reflect the same or different different facets of the construct. Co Cotypes of worker, spouse, and peer informants of agreement informants may reach a relatively relatively high level of agreement concerning concerning a target person's degree of of extraversion, extraversion, whereas they may reach no agreement agreement concerning the individual's recent recent stressful stressful life Ufe events since each individ individThe degree of ual observes only one domain of of the person's life. The of discriminant validity will reflect construct measures that are reflect the diversity diversity of of the sample of of construct included. For example, Backteman found that teacher Backteman and Magnusson Magnusson (1981) found teacher ratings strongly discriminated between aggressiveness and but provided provided a much and timidity, but weaker differentiation differentiation between between aggressiveness, motor motor disturbance (fidgeting), lack of discriminant validity studies of concentration, and lack of of school motivation. Ideally, discriminant should provide a strong test of of those specific of the new measure: Measures of specific con constructs that represent the most plausible alternative interpretations of of the new measure should always be included. Convergent Convergent and discriminant discriminant validity studies are most most typically typically conducted conducted at of tests (measures), but can also be conducted conducted at the item level. Indeed, the level of Jackson procedure in which measures of Jackson (1971) has advocated a test construction construction procedure of developed simultaneously, with item selection two or more constructs are developed selection based on both convergent convergent and discriminant discriminant evidence. Powerful Powerful statistical techniques techniques based on confirmatory confirmatory factor factor analysis have been been developed developed that permit strong tests of of that permit strong convergent and discriminant discriminant validity at either either the item item or the test level (see Bollen, or the test level 1989; Finch & & West, 1996). -

C. Multitrait-Multimethod Multitrait-Multimethod Matrix c.

Campbell and Fiske (1959; see also Marsh, 1989b) developed developed a stringent technique for probing the convergent convergent and discriminant validity vaUdity of of measures. They argue that for systematic variance in test responses may be divided into trait-related trait-related and method methodmethodrelated components. To the extent that two measures share the same method intercorrelation will be inflated. For For example, if self selfrelated components, their intercorrelation of humor humor and driving ability both reflect individuals' report measures of sense of both reflect tendencies to bias their reports in a positive direction, then the correlation correlation obtained inflated. between these two measures would be seriously inflated. To address this problem, Campbell Campbell and Fiske (1959) proposed proposed the strategy of trait -multimethod (MTMM) matrix. of the multi multitrait-multimethod matrbc. In this approach, several constructs (typically traits or abilities) are measured using multiple measurement measurement techniques. For example, as part of of a larger study by Gersten, Beals, West, and Sandler (1987), large sample of children were assessed using three different different methods (structured (structured a large interviews with each child, child reports, and parent reports) to measure three different dimensions of of symptoms (anxiety, depression, and and conduct different conduct problems). In different methods of of measuring a single construct construct produce general, to the extent that different high correlations relative to those obtained using a single method method of of measuring different constructs, convergent and discriminant validity are demonstrated. different discriminant demonstrated.

156 156

WEST WEST AND AND FINCH FINCH

To illustrate the use of this technique more concretely, consider the three idealized examples shown in Table I, panels A, B, and C, C, which are based on hypothetical data. Three traits, neuroticism (N), extraversion (E), and openness to (O) are measured using two two methods, self-reports (1) and spouse reports experience (0) (2). (2). We assume that all measures have equal reliabilities in examples lA IA and IB and introduce the complication of unequal reliabilities in example IC.

TABLE TABLE II Illustration of Hypothetical Multitrait-Multimethod Multitrait-Multimethod Matrices with Three Traits and Two Methods 2b Method 2^

Method 1^ 14 Traits

Nl N1

El E1

Ol 01

N2

E2

02 02

A. Convergent and discriminant validity with minimal methods effects effects A. Method 1 Self-report

Nl Neuroticism N1 El Extraversion E1 Ol Openness 01

Method 2 Spouse-report

N2 Neuroticism E2 Extraversion 02 Openness

(.9) .4 4

2 .2 .6 .6 .2 .1 .1

(.9)

.2

(.9)

.2

.1 .1

.6 .6

.1 .1

.4

(.9) .4 4

2.2

(.9)

2 .2

(.9)

(.8) .5 .5

(.8) (.8)

effects B B.. Strong methods effects Method 1 Self-report

Nl Neuroticism N1 El Extraversion Ol Openness 01

Method 2 Spouse-report

N2 Neuroticism E2 Extraversion 02 Openness

(.8) (.8) .6 .6 .5 .5

(.8) (.8) .5 .5

(.8) (.8)

.1 .1

.3 3

.0 .0

.2 .2

.0 .0

.1 .1

.0 .0

.0 .0

.2 2

(.8) (.8) .6 .6 .5 .5

C. Effects of unreliability and lack lack of discriminant validity (02) Method 1 Self-report

Nl El Extraversion E1 01 Ol Openness

Method 2 Spouse-report

N2 Neuroticism E2 Extraversion 0 2 Openness 02

(.4) 3 3

.1 .1 A .4 .1 .1 .1 .1

(.9)

.2 2 .2 2 .6 .6 .1 .1

(.9) ..11 .1

..4 4

(.9)

.4 4 .2 2

(.9) .5 .5

(.9)

Note: Reliabilities are are printed on diagonal in parentheses. Validity coefficients coefficients are printed in boldface boldface Note: type. Correlations between different different traits measured with the same method are printed in italics. different traits measured with the different different methods are printed in standard Correlations between different typeface. Self-report. 4« Self-report. b* Spouse report. report.

CHAPTER 66

PERSONALITYMEASUREMENT MEASUREMENT PERSONALIlY

157 157

Following four criteria are normally Following Campbell Campbell and Fiske (1959, pp. 82-83), four normally multitrait-multimethod matrices. utilized in examining multitrait-multimethod correlations of of different different methods of of measuring measuring the same same trait (conver (conver1. The correlations validity coefficients) coefficients) should be statistically statistically significant significant and large enough enough in gent validity magnitude to justify justify further further consideration. For example, in Table Table lIA A the values of rEl.E2 (.6), (.4; values printed in bold) are all statistically rNl.N2 rm;s2 (.6), (-6), rEi,E2 (-6), and r01,02 roi,o2 (-4; statistically significant and of of reasonable magnitude. magnitude. significant 2. Each of of the convergent validity coefficients coefficients should should be be higher than the of different different methods of of measuring measuring different different traits located in the corre correcorrelations of sponding column and row. For example, in Table llA, A , rEl.E2 rEi,E2 (-6) sponding (.6) is larger than other = .2; -2; rEl.02 rEi,o2 = = .1) and row row (rE2,Ni = .2, -2, rE2.01 rE2,oi = = rEl.N2 = (rE2.Nl = values in the same column ((rEi,N2 for all analogous comparisons comparisons within within the different-trait, different-trait, .1). This should hold true for different-method block of of correlations that appears in print in Table Table IA. lA. in standard print different-method of the convergent convergent validity coefficients coefficients should be higher 3. Each of higher than the of the same method of of measuring different different traits located in the corre correcorrelations of sponding column and row. For example, in Table llA, A , TNI N2 = = .6 6 is larger than the rNl.N2 sponding of rNi,Ei 4 ) and TNI.QI of rN2 rN2,E2 (-4) value of rNl.El ((.4) rNl.01 (.2) in the same column and the the values of .E2 (.4) rN2,o2 (.2) (-2) in the same row. (Some thought may may be be required to locate these and rN2.02 triangular form form in which MTMM latter two values in the triangular MTMM matrices are typically for all analogous presented.) This relationship should hold true for analogous comparisons comparisons within different-trait, same-method same-method triangles that are printed in italics in Table IA. lA. the different-trait, Campbell and O'Connell O'Connell (1982) have more more recently recently noted noted that that this this third Although sharing the criterion may be overly stringent. Although the same method method may inflate correlations between different different traits, the use of of different methods different methods to assess the coefficients may produce nonshared nonshared variance that will same trait in the validity coefficients minor failures failures to meet meet this criterion tend to attenuate these correlations. Thus, minor carefully rather than leading to an automatic conclusion of should be interpreted carefully automatic conclusion a lack of discriminant validity. validity. 4. Finally, the same general pattern of of relationships relationships should should hold for for each of the triangles in the different-trait, different-trait, different-method different-method block block (standard print) and the different-trait, different-trait, same-method same-method block block (italics). Such a finding finding suggests suggests that the correla correlafor each of of the traits are independent of the method method tions between the true scores for independent of of measurement. measurement. of Applying these criteria to Table lA, l A , we see that they Applying they are all easily met. Hence, evidence for the convergent convergent and discriminant there is strong evidence discriminant validity of of the three traits. In Table IB, we see an idealized example in which a strong method example which strong method effect effect confounding leads to inflated correlations has been added to the MTMM matrix; this confounding inflated correlations whenever two traits are measured by the same method method and hence produces problems whenever E 1 - E 2 correlation (.3) (.3) in Table IB is in meeting the third criterion. Thus, the E1-E2 lower than the E1-N1 E1~N1 (.6), the E1-01 E l - O l (.5), the E 2 - N 2 (.6), or the E2-02 E 2 - 0 2 (.5) (.5) E2-N2 the relatively relatively small of the the validity correlations. This result combined with the small values of

158 158

ANDFINCH FINCH WEST AND

coefficients coefficients suggests serious problems with the convergent convergent and discriminant discriminant validity of the measures. of Finally, in Table IC two alterations have been introduced into the MTMM been introduced relative to Table IA. First, the reliability of self-report measure of matrix lA. of the self-report of neuroticism (.4) is very low and substantially attenuates all of of the correlations in the N1 N l column. Second, a higher higher correlation between between spouse reports of of extraversion extraversion = .5) is reported. Applying Applying the four 'E2.02 = four Campbell and Fiske and openness ((rE2,o2 but criteria 3 and 4 are criteria to Table IC, we see that criteria 1 and 2 are passed but not. The validity vaUdity correlation correlation for for neuroticism, 'NI,N2 rNi,N2 (.4), (-4), does not exceed 'N2,E2 rN2,E2 (.4). (-4). artifact of of the low reliability of of N1 N l and and would not not be a problem problem if if the the This is an artifact corrected for for attenuation attenuation prior (Althauser & & correlation matrix were corrected prior to analysis (Althauser Heberlein, 1970; Jackson, 1969). validity correlation correlation for 1969). On the other hand, the validity roi,o2 (-4), is exceeded by ro2,E2 (.5), which is is not not an an artifact artifact of of differential differential '02,E2 (.5), openness, '0 1 ,02 (.4), Method 2-Method 2-Method 2 reliability. Note also that the pattern of correlations in the Method different-trait triangle differs differs from from the pattern pattern of of correlations in all of of the other different-trait different-trait triangles. This pattern of of results with respect respect to criteria criteria 3 and 4 different-trait suggests that that openness openness fails fails to to exhibit exhibit discriminant discriminant validity validity with respect to to suggests with respect spouse reports. reports. spouse illustrate the utility utility of These examples clearly illustrate of the Campbell-Fiske Campbell-Fiske approach of convergent and discriminant discriminant validity. However, the Campbell Campbellin probing issues of approach has several limitations of of which investigators should Fiske approach should be aware in interpreting their results. Campbell-Fiske approach approach makes a strong strong assumption that the mea mea11.. The Campbell-Fiske sures have equal reliabilities and no restriction of of range. As we saw in our our analysis of of Table IC, these problems can lead to artifactual artifactual failures failures to satisfy satisfy the four four criteria of attenuation. unless the correlation matrix is corrected for for these sources of 2. The cookbook cookbook nature of of the Campbell-Fiske Campbell-Fiske approach approach has unfortunately unfortunately facilitated facilitated choices of of methods and traits for for inclusion in the MTMM on the basis of construct theory of convenience rather than theory. Investigators need to use the construct of of the traits as a guide in the choice of of traits and methods to be included. Theoreti Theoretically, which traits is it important important to discriminate between? independent comparisons that 3. The large number number of non nonindependent that are required required for for the statistical analysis of of MTMM matrices leads to potential potential problems problems of of quantifying quantifying and interpreting the results. Steiger (1980) and Meng, Rosenthal, and Rubin Rubin (1992) have proposed methods that may be adapted to the basic analysis of of MTMM MTMM matrices. Confirmatory Confirmatory factor factor analysis approaches approaches (Marsh & & Grayson, 1995) provide provide direct tests of of the fit of of the data to MTMM models. 4. The Campbell-Fiske criteria assume that trait and method method factors factors are uncorrelated, uncorrected, that trait and method method factors factors do not not interact, that method method factors factors are uncorrelated, uncorrected, and that all traits are influenced influenced equally by the method method factors. These assumptions may be violated in some applications, making the results difficult difficult to confirmatory factor factor analysis approaches approaches have begun begun to provide provide interpret. Again, confirmatory of dealing with these violated assumptions. some methods of

CHAPTER 66 CHAPTER

PERSONALITY MEASUREMENT

159 159

Despite these limitations, the Campbell and Fiske (1959) criteria provide an excellent, though stringent, strategy for the study of MTMM matrices. As Messick (often humbling) heuristic (1989) notes, the Campbell-Fiske approach "is a tough (often device that forces the investigator to confront confront simultaneously simultaneously both convergent convergent and discriminant evidence, or the lack thereof" (p. 47). 47). Indeed, Fiske and Campbell (1992) have recently lamented that psychologists have made but little progress in improving the convergent and discriminant validity of their measures during the past 35 35 years. Careful Careful consideration of method effects effects can potentially lead to substantive interpretations that further further inform inform our understanding of the trait constructs (Ozer, 1989). More sophisticated analytical techniques for MTMM matrices are currently being developed; however, the Campbell-Fiske criteria provide an important and from which the results of such analyses can be more easily informative baseline from interpreted (Marsh, 1989a).

IV. rv. CONCLUSION Two underlying themes have emerged in this chapter that are appearing with increasing frequency frequency in modem modern writings on measurement.

1. This chapter moves away from a cookbook approach approach and advocates strong theoretical guidance of all investigations. Even consideration of reliability should be guided by the construct theory. The definition, domain, and theoretical structure affect the sampling of items and even the types of of measures that of the construct affect should be sought. The construct theory determines the types of criteria that should of the measure. be sought in attempting to establish the validity of identifying and testing 2. This chapter also emphasizes the importance of identifying of constructs. Houts et al. (1986) alternative hypotheses about the interpretation of emphasize the importance of examining one's own work and even entire research literatures for the possibility of common biases that may contribute to the results. Probing research literatures for common assumptions, asking what aspects of the carefully to one's critics are construct theory have not been tested, and listening carefully identifying such biases. For example, nearly all of the research on the methods of identifying Big Five Personality Traits has involved self-reports or the reports of knowledgeable knowledgeable others rather than observational measures (Digman, 1990). Similarly, the stringent tests provided by the inclusion of the most serious contending interpretations of a multitraitconstruct in convergent and discriminant validity studies (ideally utilizing multitrait multimethod matrices) offer offer promise in refining our understanding of these concon offer the strongest evievi structs. Tests of competing models of personality structure offer dence for or against a hypothesized structure. We believe that the increased emphasis on the theoretical bases of personality constructs advocated here will have a salutatory effect effect on personality measurement. from attempts to routinize approaches to measurement may A movement away from foster more careful careful planning and implementation implementation of measurement measurement designs. help foster

160 160

AND FINCH WEST AND

Modern M o d e r n measurement m e a s u r e m e n t techniques techniques are a r e reemphasizing reemphasizing the interdependence i n t e r d e p e n d e n c e of of theo theomethodological, measurement, m e a s u r e m e n t , and statistical statistical analysis choices. choices. retical, methodological,

ACKNOWLEDGMENTS ACKNOWLEDGMENTS We thank Peter Peter Bentler, Sanford Sanford Braver, William Chaplin, Clifford Clifford Clogg, William Graziano, Joseph Joseph Hepworth, Jay Hull, Douglas Jackson, John Johnson, Herbert Herbert Marsh, Robert Robert McCrae, Hee-choon Shin, Jeffrey Jeffrey Tanaka, Howard Howard Tennen, and Jenn-Yun Jenn-Yun Tein for their comments comments Hee-choon on an earlier earher version version of this chapter. A major major section section that was originally part of earlier earlier chapter is now included included in another another article (Finch & &,West, West, 1996). 1996).One Oneof ofthe the versions of this chapter authors (S. (S. G. W.) was partially partially supported supported by National National Institute of Mental Health Grant Grant authors during the writing writing of this chapter. chapter. P50MH39246 during

REFERENCES REFERENCES P., & Heberlein, T. A. (1970). Validity and the multitrait-multimethod Althauser, Ahhauser, R. P., multitrait-multimethod matrix. matrix. Borgotta & G. W. Bohrnstedt Bohmstedt (Eds.), Sociological Sociological methodology methodology 1970 (pp. In E. F. Borgotta 151-169). San Francisco: Jossey-Bass. American Na American Psychological Psychological Association, American American Educational Educational Research Research Association, & National Council on Measurements Measurements Used in Education. (1985). Standards Standards for for educational educational and psychological psychological tests. Washington, DC: Association. DC: American Psychological Association. and Angoff, Angoff, W. H. (1988). Validity: An evolving concept. In H. Wainer Wainer & H. I. Braun (Eds.), Test validity validity (pp. 19-32). Hillsdale, NJ: Erlbaum. Erlbaum. Backteman, G., & Magnusson, D. (1981). Longitudinal stability of personality personality characteristics. characteristics. Journal of Personality, Personality, 49, 49, 148-160. Journal of Beck, A. T., Ward, C. H., Mendelson, M., M., Mock, J., & Erbaugh, J. (1961). (1961). An inventory for measuring depression. Archives Archives of of General General Psychiatry, Psychiatry, 4, 4, 561-571. Bern, Bem, S. L. (1974). The measurement measurement of psychological androgyny. Journal Journal of of Consulting Consulting and and Clinical Clinical Psychology, Psychology, 42, 42, 155-162. Bollen, BoUen, K. A. (1989). Structural Structural equations equations with latent variables. New York: Wiley (Inter (Interscience). Bollen, BoUen, K. A., & Lennox, R. (1991). Conventional wisdom on measurement: A structural structural equation equation perspective. Psychological Psychological Bulletin, Bulletin, 110, 110, 305-314. factor analysis in the development development and Briggs, S. R., & Cheek, J. M. (1986). The role of factor evaluation of personality personality scales. Journal Journal of of Personality, Personality, 54, 54, 106-149. evaluation with Briggs, S. R., & Cheek, J. M. (1988). On the nature of self-monitoring: Problems with Journal of of Personality Personality and Social Social Psychology, Psychology, assessment, problems with validity. Journal 54, 663-678. self-monitoring scale. Briggs, S. R., Cheek, J. M., M., & Buss, A. H. (1980). An analysis of the self-monitoring Journal of of Personality Personality and Social Social Psychology, Psychology, 38, 38, 679-686. Journal Convergent and discriminant discriminant validation by the Campbell, D. T., & Fiske, D. W. (1959). Convergent multitrait-multimethod matrix. Psychological Psychological Bulletin, Bulletin, 56, 56, 81-105. multitrait-multimethod as diluting trait relationships rather rather Campbell, D. T., & O-Connell, E. J. (1982). Methods as than adding irrelevant irrelevant systematic variance. In D. Brinberg & L. Kidder Kidder (Eds.), Forms Forms than of validity validity in research (pp. (pp. 93-111). San Francisco: Jossey-Bass. of

CHAPTER 66 CHAPrER

PERSONALITYMEASUREMENT MEASUREMENT PERSONALITY

161 161

Carver, C. S. (1989). How should multifaceted multifaceted personality constructs be tested? Issues illustrated by self-monitoring, attributional style, and hardiness. Journal Journal of of Personality Personality Social Psychology, Psychology, 56, 577-585. and Social Collins, B. E. (1974). Four components of the Rotter Internal-External Internal-External Scale: Belief Belief in a difficult difficult world, a just world, a predictable world, and a politically responsible world. Journal Journal of of Personality Personality and Social Psychology, Psychology, 29, 381-391. Conley, J. J. (1984). The hierarchy of of consistency: A review and model of of longitudinal differences in intelligence, personality, and self-opinion. self-opinion. findings on adult individual differences Personality and and Individual Individual Differences, 5, 11-25. Personality Constantinople, A. (1973). Masculinity-femininity: An exception to a famous dictum. Psycho Psychological logical Bulletin, 80, 389-407. Costa, P. T., Jr., & McCrae, R. R. (1992). Revised Revised NEO NEO Personality Personality Inventory Inventory (NEO (NEOPI-R) professional manual. PI'R) and and NEO NEO Five-Factor Inventory Inventory (NEO-FFl) (NEO-FFI) professional manual Odessa, FL: Psychological Assessment Resources. modern test theory. Fort Worth, Introduction to classical and and modern Crocker, L., & Algina, J. (1986). Introduction TX: TX: Harcourt Brace Jovanovich. Coefficient alpha and the internal structure of Psychometrika, Cronbach, L. J. (1951). Coefficient of tests. Psychometrika, 16, 297-334. 16, Cronbach, L. J. (1957). Beyond the two disciplines of scientific scientific psychology. American American Psychol Psychologist, 30, 116-127. Cronbach, L. J. (1989). Construct validation after thirty years. In R. L. Linn (Ed.), Intelligence: Measurement, theory, and public public policy. policy. Proceedings Proceedings of of a symposium symposium in honor honor of of Lloyd Lloyd G. Humphries Humphries (pp. 147-171). Urbana: University of of Illinois Press. c.,, Nanda, H., & Rajaratnam, N. (1972). The dependability Gleser, G. C dependability Cronbach, L. J., GIeser, of behavioral behavioral measurements: measurements: Theory of of generalizability generalizability for for scores scores and profiles. profiles. New of York: Wiley. Cronbach, L. J., & Meehl, P. E. (1955). Construct validity in psychological psychological tests. Psychological Psychological Bulletin, 52, 281-302. Digman, J. M. (1990). Personality structure: Emergence Emergence of of the five-factor model. Annual Annual Review of of Psychology, Psychology, 41, 417-440. Review Epstein, S. (1983). Aggregation and beyond: Some basic issues of of the prediction of of behavior. Journal Journal of of Personality, Personality, 51, 360-392. Farrell, A. D., Mariotto, M. J., Conger, A. J., Curran, J. P., & «& Wallander, Wallander,J.J.L. L.(1979). (1979).Self Selfof heterosexual social anxiety and skill: A A generalizability ratings and judges' ratings of Journal of of Consulting Consulting and and Clinical Psychology, Psychology, 47, 164-175. study. Journal Educational measure measureFeldt, L. S., & Brennan, R. L. (1989). Reliability. In R. L. Linn (Ed.), Educational ment (3rd ed., pp. 105-146). New York: Macmillan. ment coefficient alpha. Feldt, L. S., Woodruff, D. J., & Salih, F. A. (1987). Statistical inference for coefficient Applied Psychological Psychological Measurement, 11, 93-103. Applied Finch, J. F., West, S. G. (1996). The investigation investigation of of personality personality structure: Statistical Statistical models. models. Department of of Psychology, Tempe. Unpublished manuscript, Arizona State University, Department Fiske, D. W., & Campbell, D. T. (1992). Citations do not solve problems. Psychological Psychological Bulletin, 112, 393-395. Frederiksen, N., & Ward, W. C. (1978). Measures for the study of of creativity in scientific scientific Applied Psychological Psychological Measurement, Measurement, 2, 1-24. problem solving. Applied Gangestad, S. W., & Snyder, M. (1985). "To carve nature at its joints": On the existence of of Psychological Review, Review, 92, 317-349. discrete classes in personality. Psychological

162 162

AND FINCH WEST AND FINCH

Gangestad, S. W., & Snyder, M. (1991). Taxonomic analysis redux: Some statistical considera consideraand Social Social Psychology, tions for testing a latent class model. Journal Journal of of Personality Personality and Psychology, 61, 141-146. C , Beals, Deals, J., West, S. G., & Sandler, I. (1987, March). A A measurement measurement model model of of Gersten, J. c., major constructs constructs of of child psychopathology. psychopathology. Paper presented at the meeting of the major Society for for Research Research in Child Development, Development, Baltimore, MD. Society Gormly, J. (1984). Correspondence Correspondence between between personality personality trait ratings and and behavioral events. Journal Journal of of Personality, Personality, 52, 220-232. Green, S. B., Lissitz, R R. W., & Mulaik, S. A. (1977). Limitations Limitations on coefficient coefficient alpha as an index of Measurement, 37, 827-838. of test unidimensionality. Educational Educational and Psychological Psychological Measurement, Guilford, J. P. (1954). Psychometric methods methods (2nd ed.). New New York: McGraw-Hill. Guilford, (1954). Psychometric GuUiksen, H. (1950). Theory Theory of of mental mental tests. New York: Wiley. Gulliksen, R. K., Swaminathan, H., & Rogers, H. J. (1991). Fundamentals Fundamentals of of item response response Hambleton, R Newbury Park, CA: Sage. theory. Newbury Hattie, J. (1985). Methodology Methodology review: Assessing Assessing unidimensionality unidimensionality of of tests and items. Ap Applied Psychological Psychological Measurement, Measurement, 9, 139-164. plied Hedges, L. V. (1988). The The meta-analysis of of test test validity validity studies: Some new new approaches. In Wainer & H. I. Braun (Eds.), Test validity validity (pp. 191-212). Hillsdale, NJ: Erlbaum. Erlbaum. H. Wainer R. H. (1967). The social social readjustment readjustment rating scale. Journal Journal of of Holmes, T. H., & Rahe, R Psychosomatic Research, 14, 213-218. Psychosomatic C , Cook, T. D., & Shadish, W. R, R., Jr. (1986). The person-situation A Houts, A. c., person-situation debate: A Journal of of Personality, Personality, 54, 52-105. critical multiplist perspective. Journal J. G., Lehn, D. A., & Tedlie, J. C. (1991). A A general approach approach to testing multifaceted multifaceted Hull, 1. Journal of of Personality Personality and Social Social Psychology, Psychology, 61, 932-945. personality characteristics. Journal Jackson, D. N. (1969). Multimethod Multimethod factor factor analysis in the evaluation evaluation of of convergent convergent and discriminant validity. Psychological Psychological Bulletin, Bulletin, 72, 30-49. Jackson, D. N. (1971). The dynamics of of structured structured personality personality tests. Psychological Psychological Review, Review, 78, 78, 229-248. Jackson, D. N., Ahmed, S. A., & Heapy, N. A. (1976). Is achievement achievement a unitary construct? construct? Journal Journal of of Research in Personality, Personality, 10, 1-21. Jackson, D. N., & Paunonen, S. V. (1985). Construct Construct validity and and the the predictability predictability of of behavior. Journal Journal of of Personality Personality and Social Social Psychology, Psychology, 49, 49, 554-570. Jones, L. V., & Appelbaum, Appelbaum, M. I. (1989). Psychometric methods. Annual Annual Review Review of of Psychol Psychology, 40, 23-43. 23-43. Kane, M. T. (1982). A sampling sampling model for validity. Applied Applied Psychological Psychological Measurement, Measurement, 6, 125-160. Kenny, D. A. (1993). A A coming of of age for for research research on interpersonal interpersonal perception. Journal Journal of of Personality, 61, 789-807. Kuder, G. F., & Richardson, M. W. (1937). The theory theory of of the estimation estimation of of test reliability. Psychometrika, 2, 151-160. Psychometrika, Lindquist, E. F. (1953). Design and and analysis of of experiments experiments in psychology psychology and and education. education, Houghton-Mifflin. Boston: Houghton-Mifflin. Loevinger, J. (1947). A systematic systematic approach approach to the construction construction and and evaluation evaluation of of tests of of Psychological Monograph, Monograph, 61 6i(No. (No. 4, Whole No. 285). ability. Psychological Loevinger, J. (1948). The technique of of homogeneous homogeneous tests. Psychological Psychological Bulletin, Bulletin, 45, 507-529. Loevinger, J. (1957). Objective tests as instruments of of psychological theory. Psychological Psychological Reports, Reports, 3, 635-694.

CHAPTER 66 PERSONALITY PERSONALITYMEASUREMENf MEASUREMENT

163 163

Lord, F. M., M., & Novick, M. M. R. (1968). Reading, MA: (1968). Statistical Statisticaltheories theoriesofofmental mentaltest testscores. scores. Reading, MA: Addison-Wesley. Mabe, P. A., Ill, III, & West, S. G. (1982). (1982). Validity of self-evaluations self-evaluations of of ability: A review and meta-analysis. Journal Journal of ofApplied Applied Psychology, Psychology,67, 67,280-296. 280-296. Confirmatory factor factor analyses of of multitrait-multimethod multitrait-multimethod data: Many Marsh, H. W. (1989a). Confirmatory problems and a few solutions. Applied Psychological 13,13, 335-361. PsychologicalMeasurement, Measurement, 335-361. Marsh, H. W. (1989b). Multitrait-multimethod Multitrait-multimethod analyses. In J. P. Keeves (Ed.), Educational Educational researchmethodology, methodology,measurement measurement evaluation: international handbook. Ox Oxresearch andand evaluation: An An international handbook. ford, England: Pergamon Press. ford, (1995). Latent variable models of of multitrait-multimethod multitrait-multimethod data. Marsh, H. W., & Grayson, D. (1995). Structuralequation equationmodeling: modeling: Concepts, issues, and applications In R. R. H. H. Hoyle Hoyle (Ed.), Concepts, issues, and applications In (Ed.), Structural (pp. 177-198). Thousand Oaks, Oaks, CA: Sage. McClelland, D. C C. (1980). (1980). Motive dispositions: The merits of operant operant and respondent mea meaof Personality Personalityand andSocial SocialPsychology, Psychology,I, 1, 10-41. sures. Review of 10-41. (1946). Opinion-attitude methodology. Psychological Psychological Bulletin, Bulletin,43, 43,289-374. 289-374. McNemar, Q. (1946). X.-L., Rosenthal, R., & Rubin, D. B. B. (1992). (1992). Comparing correlated correlation coeffi coeffiMeng, x.-L., PsychologicalBulletin, Bulletin,Ill, 111, 172-175. cients. Psychological 172-175. Messick, S. (1989). Validity. In R. L. Linn (Ed.), Educational measurement measurement (3rd ed., pp. 13-104). New York: Macmillan. of Consulting Consulting Mirels, H. L. (1970). Dimensions of internal versus external control. Journal of ClinicalPsychology, Psychology,34,34,226-228. 226-228. and Clinical (1986). Comparison of self-reports, reports by knowledgeable informants, Moskowitz, D. S. (1986). of Personality, Personality,54, 54,294-317. 294-317. and behavioral observation data. Journal of C. (1982). (1982). Validity comparison of of behavior counts and Moskowitz, D. S., & Schwarz, J. C ratings by by knowledgeable knowledgeable informants. informants. Journal Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, ratings 43, 518-528. 43, for closure scale Neuberg, S. L. Judice, T. N., & West, S. G. (in press). What the need for Toward differentiating differentiating among among related related epistemic epistemic motives. motives. measures and what it does not: Toward

Journalof ofPersonality Personalityand andSocial Social Psychology. Journal Psychology.

C, & Bernstein, I. I. H. (1993). (1993). Psychometric Psychometric theory (3rd ed.). New York: Nunnally, J. C, McGraw-Hill. (1989). Construct validity in personality assessment. In D. M. Buss & N. Cantor Ozer, D. J. (1989). trends andand emerging directions (pp. (pp. 224-234). (Eds.), Personality Personalitypsychology: psychology:Recent Recent trends emerging directions 224-234). New York: Springer-Verlag. Annual Review of Psychology, Ozer, D. D. J., & Reise, S. P. (1994). Personality assessment. Annual of Psychology, 45, 45, 357-388. Peterson, C, C, Semmel, A., von Baeyer, C, C., Abramson, L. Y., Metalsky, G. I., & Seligman, M. Cognitive Therapy Re M. E. P. (1982). The Attributional Style Questionnaire. Cognitive Therapy and Research, search, 6,6,287-299. 287-299. Rotter, J. B. B. (1966). (1966).Generalized Generalizedexpectancies expectanciesfor forinternal internalversus versusexternal externalcontrol controlof ofreinforce reinforceWhole No. 609). PsychologicalMonographs, Monographs,80(1, 80(1, Whole No. 609). ment. Psychological C. J., & Pressley, M. M. (1983). (1983). Behavioral development development and construct Rushton, J. P., Brainerd, C validity: PsychologicalBulletin, Bulletin,94, 94,18-38. 18-38. validity: The principle of aggregation. PsychologicaL Sandler, I. N., & Guenther, R. T. (1985). Assessment of life stress events. In P. Karoly (Ed.), Measurementstrategies strategies health psychology 555-600). New York: Wiley (Interin in heaLth psychology (pp.(pp. 555-600). New York: Wiley (Inter Measurement science). Schmidt, F. L., Hunter, J. E., Pearlman, K., & Hirsh, H. R. (1985). Forty questions about validity generalization and meta-analysis. Personal Personal Psychology, Psychology,3838, 697-798. , 697-798.

WEST AND WEST ANDFINCH FINCH

164 164

(1991). Generalizability Generalizahilitytheory: theory:AAprimer. primer,Newbury NewburyPark, Park, Shavelson, R. J., & Webb, N. M. (1991). CA: Sage. Snyder, M. (1974). Personalityand andSocial Social (1974). Self-monitoring Self-monitoring of expressive behavior. Journal of of Personality Psychology, Psychology, 30, 30,526-537. 526-537. Spence, J. T., & Helmreich, R. L. (1978). (1978). Masculinity Masculinity and andfemininity: femininity: Their Theirpsychological psychological University ofof Texas Press. dimensions, dimensions, correlates, correlates,and andantecedents. antecedents.Austin: Austin: University Texas Press. Stanley, J. C. (1971). Reliability. In R. L. Thorndike (Ed.), Educational Educational measurement measurement (2nd (2nd ed., pp. 356-442). Washington, DC: National Council on Measurement in Education. Steiger, J. H. (1980). Tests for for comparing elements of a correlation matrix. Psychological Psychological Bulletin, Bulletin, 87, 87,245-251. 245-251. Sternberg, R. J., & Weil, E. M. (1980). An aptitude-strategy interaction in linear syllogistic reasoning. Journal of of Educational Educational Psychology, Psychology,72 72,226-234. 226-234. Thorndike, R. L. (1951). Reliability. In E. F. Lindquist (Ed.), Educational Educationalmeasurement measurement(pp. (pp. 560-620). Washington, DC: DC: American Council on Education. (1931). Answer-pattern Answer-pattern and score scatter in tests and examinations. British Walker, D. A. (1931). Journal of of Psychology, Psychology, 22, 22,73-86. 73-86. (1994). Individual differences differences in need for for cognitive Webster, D. M., & Kruglanski, A. W. (1994). Personality and andSocial SocialPsychology, Psychology,67, 67,1049-1062. 1049-1062. of Personality closure. Journal of (1973). Personality Personality and prediction; prediction; Principles Principles of of personality personality measurement. measurement Wiggins, J. S. (1973). Reading, MA: Addison-Wesley. (1977). Belief Behef in internal control or belief belief in a just world: Zuckerman, M., Gerbasi, K. C. (1977). The use and misuse of the I-E scale in prediction of attitudes and behavior. Journal Personality,45, 45,356-379. 356-379. of Personality, of ,

CHAPTER 77 CHAPTER

PERSONALITY INFLUENCES ON THE CHOICE OF SITUATIONS THE WILLIAM ICKES ICKES UNIVERSITY OF OF TEXAS TEXAS AT AT ARLINGTON ARLINGTON UNIVERSI1Y

MARK SNYDER SNYDER MARK UNIVERSI1Y UNIVERSITY OF OF MINNESOTA MINNESOTA

STELLA GARCIA GARCIA UNIVERSI1Y UNIVERSITY OF OF TEXAS TEXAS AT AT SAN SAN ANTONIO ANTONIO

Chip and Priscilla, a Yuppie couple from Chicago, have just moved to Dallas and are sampling sampHng some of the trendier nightspots on Lower Greenville Avenue. As As they they push push through through the the swinging swinging doors doors of of what what appears appears to to be be aa quaint quaint little little Western saloon right out of the TV Gunsmoke, they theyare areconfronted confrontedby by TV series Gunsmoke, six huge bikers bikersfrom fromthe the motorcycle motorcyclegang gangLos LosDiablos, Diablos,who whoturn turnon ontheir theirbarstools barstools to glare at them. The bikers have an average height of more than six feet, an average weight of more than 250 pounds, pounds, an an average average beard beard growth growth of of more more than six days, and an average of more than two tattoos and three missing teeth. The fumes they emit smellflammable. flammable. Two Two of them stare with contempt at Chip, and one leers evilly at Priscilla. "This doesn't look like our kind of place," place," Chip Chip says to Priscilla, as they prepare to beat a hasty retreat.

Just as in this fictional example, real people in their everyday lives deliberately choose to enter some situations and to avoid others. The goal of of this chapter is to review a rapidly growing body of of data which suggests that these choices are deter determined, at least in part, by the degree to which people perceive certain situations IIANDBOOK PERSONA.LITY PSYCHOLOGY HANDBOOKOF OF PERSONAUTY PSYCHOLOGY

165165

COPI'JI/GIIT COPYRIGHT C © 1997 1997 BY BY ACADEMIC ACADEMIC PRESS PRESS.. ALLAIIGIITS OF IEPIOCUcnON IN ANY FORMFORM IESElVED. L L RIGHTS OF REPRODUCTION IN ANY RESERVED.

166 166

ICKES, ICKES, SNYDER, SNYDER, AND AND GARCIA GARQA

as either "fitting" or failing to "fit" such aspects of their own personalities as their traits, their attitudes, and their self-conceptions. Because this chapter is intended primarily as a review of the available literature, the major theoretical perspectives that bear on this work are described only in their broadest outlines. For a more detailed discussion of these theoretical perspectives, the reader is referred to the original sources as well as to the theoretical integration by Snyder and Ickes (1985).

I. WHAT DOES "CHOOSING ^THOOSING SITUATIONS" MEAN? actively choose to be in situations that that best best "fit" "fit" their theirpersonali personaliThe idea that people actively ties is hardly a novel concept. In fact, its relevance to the matter of behavioral theorists. For example, as as early early consistency has long been recognized by personality theorists. G. W. Allport AUport noted that individuals play an active role in seeking out as 1937, O. with their theirdispositions. dispositions.Some Somedecades decadeslater, later,Mischel Mischel environments that are congruent with (1969,1977) situa1977) and Block (1968) argued that personality may be a function of situa (1969, (1981, 1988, 1990; 1990; see also Endler, 1988) tional contingencies, and Magnusson (1981, asserted that an understanding of human behavior requires an understanding of the situations in which humans behave. Taking this argument full circle back to AUport environ Allport (1937), Bowers (1973) stated that "people foster consistent social environments which then reciprocate by fostering behavioral consistency" (p. 329), and Bandura (1982) noted that by constructing their environments, individuals achieve Bandura some regularity in their behavior. Recently, Snyder and Ickes (1985) have proposed the settings in which to live one's life life.. .. . may reflect features that "one's choices of the of one's conceptions of self, one's characteristic dispositions, one's attitudes and (p. 915). Similarly, Caspi, Bern, Bem, and values, and other attributes of personality" (p. Elder (1989) have argued that "a person's selection and creation of environments the most individuating and and pervasive expressions of his or her personality" is one of the (p. 377). Common to most, if not all, of these statements is the assumption that, in the natural course of their lives, individuals can freely choose to be in certain sitsit uations and to avoid others. These preferred environments provide opportunities for personal dispositions to be manifested and reinforced. Once individuals are in their chosen situations, their words and actions are genuine reflections of their personalities, and the fact that they display these behaviors in in settings they substantial degree degree of ofconsistency consistency in in their theirbehavior. behavior. have specifically specifically chosen ensures aa substantial The interest of personality psychologists in the relation between personality and situational choices is due, in no small measure, to their long-standing interest in understanding the processes that govern temporal stability and cross-situational in behavior. As Snyder Snyderand and Ickes Ickes(1985) (1985)have havenoted, noted,personality personalitypsychol psycholconsistency in dispositional strategy strategy that viewed traits and other ogists were guided first first by a dispositional relatively stable and enduring dispositions as responsible for behavioral consis-

CHAPTER 77 CHAPTER

PERSONALITY AND CHOICE OF SITUATIONS PERSONALITY

167 167

} Later, in part as a reaction to critiques of the trait approach, especially that tency tency.^ of Mischel (1968), the statistical or mechanistic mechanistic version version of of an an interactional interactionalstrategy strategy emerged. This approach viewed behavior as the product of the interaction between the person and the situation. Such interactions were assumed to be unidirectional influenced such that personal and situational variables could influence, but not be influenced & Endler, 1977; by, individuals' behavior (Cantor & Kihlstrom, 1980; Magnusson & Snyder & Ickes, 1985). The most recent phase of of this endeavor has seen the emergence of a dynamic interactionalstrategy strategythat thatviews viewsbehavioral behavioralconsistency consistencyasasthe theproduct productofofthe therecipro reciprointeractional cal cal causal relation between personality and environment. It is assumed that people have a tendency to choose to enter and participate in those situations that they perceive to be most conducive to the behavioral expression of their own traits and reinforcement dispositions. Because these situations are typically the ones in which reinforcement of the expressed behaviors is most likely to occur, the choice of situations is an important cause of the temporal stability in individuals' behavior.

n. n. CONCEPTUAL CONCEPTUAL AND AND MEmODOLOGICAL METHODOLOGICAL ISSUES ISSUES If the notion that peoples' personalities influence their choice of situations is not a new idea, why have personality psychologists only recently begun to study this phenomenon? Argyle, Furnham, and Graham (1981) have proposed one answer to this question. They have noted that, whereas personality theorists have been successful in making conceptual distinctions among person concepts (traits), fairly successful potentially more difficult) difficult) task they have not yet undertaken the analogous (but potentially for appropriate of developing an appropriate taxonomy of situations. Similar calls for conceptual work have been made by Runyan (1978), Duke and Nowicki (1982), and Endler Endler (1983).2 (1983).2 Throughoutthis thischapter, chapter,we weintend intendaatraditional traditionalusage usageofofthe theterms termstraits traitsand anddispositions, dispositions, I^Throughout defining them as relatively stable internal structures that guide, and and therefore partially determine, behavioral acts. acts. This usage is consistent with Allport's Allport*s (1931, 1966) view of traits as "neurodynamic structures" structures" whose reality can be presumed even if their precise forms and functions cannot yet be is reserved forfor those specified (see also Funder, 1991). 1991). Our use of the term cross-situational cross-situationalconsistency consistency is reserved those topographically similar or dissimilar behaviors that appear to have the same general meaning across aa wide range of situations to the actor who performs them, and can therefore be viewed viewed as a patterned patterned manifestation of an an underlying trait traitor or disposition disposition(e.g., (e.g.,punctuality). punctuality).Obviously, Obviously,the theterm termcross-situational cross-situational consistency consistencycannot cannotbe beapplied appliedtotocases casesininwhich whichthe thebehaviors behaviorsdisplayed displayedinindifferent differentsituations situationshave have substantially different different meanings for the actor, despite their topographical similarity (e.g., taking money money from someone someone can can variously variously be be interpreted interpreted as as "receiving "receiving one's wages," "getting "getting aa rebate," rebate," "taking "taking aa from one's wages," bribe," or or "stealing"). bribe," "stealing"). Giventhe theconceptual conceptualdifficulties difficultiesinvolved, involved,we wedo donot notpropose proposeto tosolve solvethe theproblem problemof ofdefining defining 2^Given situation situationhere. here.We Wewill willsimply simplynote notethat thatsituations situationscan canbe beconceptualized conceptualizedasas(a) (a)multidimensional multidimensionalfields fields of action action in in which which (b) (b) behavior behavior is is to to some some degree degree constrained constrained by by environmental environmental press, press, and and which of which (c) can be be viewed viewed from from any any of of aa number number of of different different perspectives perspectivesthat thatare areavailable availableto toaa given givenperceiver. perceiver. These These perspectives, perspectives, discussed discussedas as "the "the forms formsof ofsocial socialawareness" awareness"by byWegner Wegnerand andGiuliano Giuliano(1982), (1982),include include tacit and and focal focal self-awareness, self-awareness, tacit tacit and and focal focal other other awareness, awareness, and and tacit tacit and and focal focal group group awareness. awareness. tacit

168 168

ICKES, IcKES, SNYDER, SNYDER, AND GARCIA

The lack of relevant conceptual work is not the only problem, however. Another, and possibly more serious, problem is that most researchers are still committed to the assumptions of mechanistic interactionism and the conven conventions of traditional empirical methods such as the laboratory laboratory experiment. As Snyder and Ickes (1985, pp. 914-915) have pointed out, the assumptions un underlying the experimental method are, in at least two respects, inconsistent with those underlying a dynamic interactional approach wherein the individuals' choice stabiUty of situations is both a cause and a consequence of the disposition-based stability in their behavior. The limitations imposed by this inconsistency are discussed below. A. Limitations of the Experimental Approach

First, a key component for conducting a sound experiment is the direct experimental manipulation manipulation of independent variables which, in most psychological experiments, are aspects of the situation to which the subject is exposed. Great care is taken to ensure that the different different levels of the independent (Le., (i.e., situational) variable are made sufficiently sufficiently distinct from from each other to guarantee effects effects on the behavior of of the participants assigned to these different different conditions. This, in fact, is the major major strength of the experimental methodology: it provides maximal opportunities for researchers to witness the impact of situations on behavior. At the same time, however, controlling some aspects of the situation through manipulation of the independent variables tends to work against the goals of the personality researcher. experimenSpecifically, to the extent that the psychological "strength" of the experimen tal situation constrains the subjects' behavior in a way that minimizes individual differences, the very phenomena which personality theorists seek to observe differences, 1982; Snyder & & Ickes, and understand become increasingly less evident (Ickes, 1982; 1985). Second, another essential component for conducting a sound experiment is the random assignment of participants to the different different treatment conditions. This for individual differences differences across conditions so that a step is necessary to control for more valid causal conclusion can be made about the impact of the independent howvariable on the dependent variable. Despite its methodological importance, how major ever, randomly assigning participants to treatment conditions eliminates one major particivehicle by which individuals' attributes are manifested. That is, by assigning partici opportunity pants to specific treatment conditions, the experimenter denies them the opportunity to choose whether or not to be exposed to the assigned condition instead of to one frequently of the remaining conditions, of whose nature and even existence they are frequently kept unaware. Further, once they are in a treatment condition, they can exert minimal influence on the situation and can only react to the manipulated indepen independent variable(s). Thus, the experimental procedure effectively and intentionally dent variable(s). Thus, the experimental procedure effectively and intentionally minimizes the the extent extent to to which which behavior behavior in in the the experimental experimental situation situation is is aa reflection reflection minimizes of the individual's attributes. of the individual's attributes.

CHAPTER 77 CHAPTER

PERSONALITY AND CHOICE OF SITUATIONS PERSONAUTY

169

The manipulation of variables and the random assignment of subjects to treatment conditions are, of course, course, the the essential essential and and defining defining features features of of the the experi experimental method. They are precisely the features that make this method so well suited to testing casual hypotheses about the effects effects of manipulated independent independent variables on measured dependent dependent variables. However, in real world settings, people are not typically denied the freedom freedom to be where they want to be, when to be there, and with whom to be there. In real world settings, people are most often often found in situations of their own choosing. Viewed from from this perspective, conventional experimental paradigms can be seen as highly unusual and constraining situations that eliminate the usual opportunities by which people choose to be in certain set settings. Practitioners of the situational strategy for for studying personality differences differences & Ickes, 1985). are thus faced with an important dilemma (Snyder & 1985). On the one hand, by experimentally controlling features of the situation situation and randomly assigning participants to the treatment conditions, investigators of of the dynamic interactional approach are assured of attenuating, and possibly even eliminating, the personality processes that are presumably of greatest interest to them. On the other hand, because the same experimental procedures typically guarantee the internal validity of of any findings obtained, researchers may be understandably reluctant to give them up.

B. Alternative Alternative Research Research Methods Methods B. The obvious challenge, then, is to identify identify and understand the consistencies in behavior that are expressed through the selection of of situations without breaking the rules of conventional empirical investigation. Fortunately, through the efforts efforts of a growing number of researchers, several novel methodologies have been develdevel oped to achieve this goal. These methodologies all reflect reflect the fundamental fundamental assumption of the dynamic interactional approach. As we have noted, dynamic interactionism is a reciprocal or transactional model which assumes that the situation is both a cause cause and a consequenceof ofthe theperson's person'sbehavior behavior(Endler (Endler& & Edwards, Edwards,1986). 1986).Given Giventhis thisassump assumpconsequence tion of bidirectional causality, it is possible either (a) to view some aspect of the situation as the independent (or predictor) variable and some aspect of the person as the dependent (or criterion) variable, or (b) to view some aspect of the person as the predictor variable and some aspect of the situation as the criterion variable. The first view is the traditional one embodied in most most experimental research; the of the empirical research on how second view is one that is currently guiding much of affect their choice of situations. individuals' personalities affect The methodologies most consistent with the second view can be grouped grouped into three broad categories: (1) studies of actual situational choices in the real world, in the laboratory, and (3) of hypothet(2) studies of actual situational choices in (3) studies ofhypothet After ical situational choices (e.g., in response to survey or questionnaire items). After of these three categories, we will reviewing some representative studies in each of

170 170

IcKES, SNYDER, SNYDER, AND AND GARCIA GARQA IcKES,

briefly note a number of other empirical methods that can be used by practitioners of the dynamic interactional strategy. /. Studying Actual Actual Situational Situational Choices Choices in the Real Real World 1. In research consistent with the view that personality traits constrain people's every everypreferday choices of situations, Furnham (1981) studied how individuals' activity prefer ences vary as a function function of their personality. He found that extraverts were more likely than introverts to report being in situations that invited competition, intimacy, (1966, 1985) has shown (in research to be described and assertion. Similarly, Holland (1966,1985) later in this chapter) that people's occupational preferences preferences are also determined to a significant significant degree by their personality. In conceptually related research, the experience sampling method (Csikszent (Csikszent1982; Csikszentmihalyi & mihalyi & & Figurski, 1982; & Kubey, 1981; 1981; Hormuth, 1986) or similar techniques (Wheeler, Reis, & & Nezlek, 1983) have been used to study the naturally occurring selection of situations. For example, Emmons, Diener, and Larsen (1986) had people keep records of specific behaviors and moods in a number encountered in their own environments. One of their of situations they typically encountered findingsrevealed revealedthat thatextraverts extravertsspent spentmore moretime timeand andfelt feltmore morepositive positiveininsituations situations findings that provided social, as opposed to solitary, recreation activities. In another study using the experience sampling method, Diener, Larsen, and Emmons (1984) found found for order chose to be in common or "typical" that people who had a high need for often than in novel situations. situations more often identify situations that It should be noted that this approach can also be used to identify individuals choose to avoid. avoid. For For example, example,Furnham Furnham (1981) (1981)reported reported that that neurotics neurotics tended to avoid situations that were high in social stimulation and provided opportu opportunities for extended social interaction. 2. 2. Studying Actual Actual Situational Situational Choices Choices in the Laboratory Laboratory

It is not necessary, however, to abandon laboratory procedures when conducting research of For example, example,in inaalaboratory laboratorystudy studyof ofself-monitoring self-monitoring processes, processes, of this type. For Snyder and Gangestad (1982) found found that high self-monitors preferred preferred situations that provided them with precise and unambiguous specifications of the type of of person called for in the setting to ones that provided them with a minimally defined defined preferred situations that permitted permitted them character. In contrast, low self-monitors preferred to act in accordance with their own dispositions. Obviously, laboratory studies such as this are correlational rather than experiexperi mental, since the participants are free to choose the situation or "condition" they wish to enter rather than being randomly assigned to it. On the other hand, because the experimental convention of random assignment forces some participants to be in a situation that they would otherwise not choose to enter, all truly experimental situations can justifiably justifiably be described as "artificial" in the sense that they are determined by the experimenter instead of being chosen by the participants them themselves (Diener et aI., al, 1984; 1984; Wachtel, 1973).

CHAPTER 77

PERSONALITY AND CHOICE CHOICE OF SITUATIONS PERSONALIlY

3. 3.

171

Hypothetical Situational Situational Choices Studying Hypothetical

As an alternative to presenting people with a choice between actual actual situations, some researchers have developed paper-and-pencil measures designed to assess preferences for, hypo hypovariations in individuals' habitual choices of, or characteristic preferences thetical situations situations that that are are described described in in less less immediate, immediate, more more abstract abstract terms. terms. thetical strategist is interested in people's choices of situations Because the situational strategist as a function function of their attributes, a valid assessment device would be one that measures the strength of situational preferences. Although most existing scales preferences (cf. Furnham, 1981; 1981; L. A. Pervin, were not designed to measure such preferences 1981), at least a few promising scales of this type have been developed. For example, preferFurnham (1982) has used the Social Situation Scale to study how situational prefer ences vary as a function function of psychoticism and social desirability. Crozier (1979) and Wolpe and Lang (1964) have used the Fear Survey Schedules to assess the degree fearful in various situations (e.g., situations to which people would be anxious or fearful involving criticism or negative evaluation). In other research, W. H. Jones, Russell, identify the type and Cutrona (1985) have used the Shyness Situations Measure to identify dispositional^ shy are most likely to feel shy (e.g., eating of situations in which the dispositionally in a restaurant or giving a speech).

for Studying Studying Situational Situational Choices 4. Other Methods for Still another method for investigating the type of situations certain individuals Bem and Funder choose is the template-matching technique proposed by D. J. Bern technique. Lord (1982) proposed (1978; D. J. Bem, Bern, 1981). In one study applying this technique, that a person's behavior should be consistent across situations to the degree that there is similarity in the person's templates for for the situations. Lord's (1982) study revealed that cross-situational consistency emerged when the perceived situational idiographic but but not not when when ititwas wasnomothetic nomothetic(i.e., (i.e.,consensual). consensual).These These similarity was idiographic results not only demonstrated that the template-matching technique can be used differences in situation selection but provided further further support to study individual differences is manifested manifested for D. D. J. Bem Bern and Allen's (1974) assertion that consistency in behavior is at the idiographic level. Finally, there are a number of other empirical methods that can be used by the situational strategist. The researcher can obtain the subjects' verbal or written comfort in and preference for for a particular situation expressions about their level of comfort (Mehrabian, 1978). of 1978). The researcher can also obtain cognitive representations of chosen situations to determine if there are characteristic schemas that people hold for for the settings they are likely to spend time in (Price, 1981). And, to complement these self-report of self-report techniques, the researcher can obtain behavioral measures of preference (e.g., degree of exploration, the physical movement of avoid avoidsituational preference ing or approaching tasks/persons in the available settings, and the length of stay).

c. C. Some Some Unresolved Unresolved Problems Problems

Although all of the methods described above can profitably be employed by situational strategists, some final words of caution are in order. First, as D. J.

172 172

IcKES, SNYDER, SNYDER, AND AND GARCIA GARCIA IcKES,

Bern and Allen (1974) have noted, the English language presents us with a rich vocabulary for describing traits traits but an impoverished vocabulary for describing situations. Asking people to identify their choices of situations or to indicate situations. on aa questionnaire their their likelihood likelihood of of entering entering aa particular particular situation situation presumes presumes on semantically distinguish distinguish one one situation situation from from another. another. Researchers Researchers that they can semantically must be aware of this linguistic obstacle and find find ways to overcome it. Second, and in a related vein, individuals (a) may not have much experience in discriminat discriminating one situation from another, (b) may interpret descriptions of situations quite differently from the way the experimenter intends, (c) may view rating scales as not particularly relevant to their cognitive representations of situations, (d) may use aa diff different of reference reference than than that that of of the the experimenter, experimenter, (e) (e) may may view view use erent frame frame of past situations as as dissimilar to future situations, (f) may presume that they are responding to to functionally functionally equivalent equivalent situations situations when out aa questionnaire, questionnaire, responding when filling filling out or (g) (g) may may view view situations situations in in fairly fairly nonspecific nonspecific and and global global terms terms (Argyle (Argyle et et al., or aI., 1981; Fumham, Furnham, 1982; 1982; Fumham Furnham & & Jaspars, Jaspars, 1983; 1983; Lord, Lord, 1982; 1982; Runyan, Runyan, 1978). 1981; 1978). These considerations considerations can can pose pose formidable, formidable, though though not not insurmountable, insurmountable, challenges challenges These to the the study study of of situational situational preferences. preferences. to m. ASSESSING mE THE DIMENSIONS OF SOCIAL SITUATIONS m

Few dynamic dynamic interactionists interactionists would would question question the the assertion assertion that that people people choose choose to to Few enter and spend time in situations that allow and/or enable them to behave in a manner that that reflects reflects features features of of their their personalities. personalities. Inherent Inherent in in such such an an assertion, manner assertion, however, is is the the assumption assumption that that individuals individuals can can assess assess the the behavioral behavioral opportunities opportunities however, provided by by aa given given setting. setting. In In other other words, words, people people are are aware aware of of and and can can identify identify provided features of the situation that are particularly particularly conducive conducive to the behavioral behavioral expression expression of their personal attributes. Considerable evidence is available to suggest that individuals can indeed assess the appropriateness of specific behaviors for specific situations, and that they can also assess the constraints that certain situations place on certain behaviors (e.g., Argyle et aI., Cantor, Mischel, Mischel, & &Schwartz, Schwartz,1982a,b; 1982a,b; al, 1981; Cantor, & Bouffard, 1974; Smith-Lovin, 1979; & Glick, 1985; Price, 1974; Price 1985; 1974; & 1974; 1979; Snyder & Gangestad, 1982). Armed with this type of knowledge about situations, people ought to be able to determine the specific setting(s) that will permit and promote the expression of their own personalities. A number of theories have been proposed to specify the mechanisms that individuals use to aid them in choosing disposition ally relevant situations. For our dispositionally present purposes, it is convenient to group these theories into three categories: (1) goal-based motivational theories (e.g., Argyle et aI., repre al, 1981), (2) cognitive representational theories (e.g., Cantor et al., aI., 1982a,b), and (3) the affect congruence affect model (Emmons, Diener, & & Larsen, 1985). The reader should note, however, that there is a substantial degree of conceptual overlap among the three approaches, and that the assumptions and processes that are explicit in one approach may be implicit in the other two.

CHAPTER CHAPTER?7

PERSONALITY AND CHOICE OF SITUATIONS A. A.

173

Goal-Based Goal-Based Motivational Motivational Theories Theories

Some of the theories proposed to specify specify the mechanisms underlying situational choice are variants of the goal-based motivational theory, which is premised on the notion that situations consist of relatively well-defined well-defined goal structures (Argyle, 1980; 1980; Argyle et aI., al., 1981; 1981; Cantor, Cantor, Norem, Norem, & & Langston, Langston, 1991; 1991; Emmons Emmons & & Diener, Diener, 1986a; 1986a; Emmons et aI., 1981). Goal structures refer to the opportunities al, 1986; Furnham, 1981). provided by the situation that may aid people in accomplishing their goals. People with specific needs or objectives are presumably aware that such goal structures therefore motivated to actively seek out the specific situations that exist, and are therefore they believe will best facilitate the attainment of their goals. These goal-structure theories of situational choice have already received at for example, presented subjects with least some empirical support. Gorta (1985), (1985), for verbal descriptions of situations and asked them the purposes that would lead them described a set of particular goals to to be in each of those settings. She then described subjects, in each case asking them to respond by describing the situations they would most likely choose in order to achieve these specific objectives. The general finding of ofthe thestudy studywas wasthat thatthe thespecific specificgoal goalthat thatsubjects subjectshad hadininmind mindhighlighted highlighted finding affected their percep percepparticular aspects of the situation as important, which in turn affected preference for for entering it. Although Gorta did tion of the situation as well as their preference affect situational choice, not specifically investigate personality factors that might affect her methodology provides one possible way of testing how situational choices are guided and channeled by personal attributes. time-samphng In a study using the more compelling methodology of the time-sampling 1982; Csikszentmihalyi & technique (e.g., Csikszentmihalyi & & Figurski, 1982; & Kubey. Kubey, found the relation between an individual's 1981), Emmons and Diener (1986b) found objectives and the goal structure of the situation to be stronger in those cases in which the individual had the opportunity to choose the situation rather than having findings, the results of of it imposed on him or her. Consistent with Gorta's (1985) findings, of Emmons and Diener's (1986a; see also 1986b) study indicate that the choice of situation is linked to the goal(s) a person wishes to achieve and to the goal structure of the social setting(s) which he or she chooses to enter. B. Cognitive Cognitive Representational Representational Theories Theories B. Other theories of situational choice focus on the cognitive representations that formed regarding the features of different different situations. Theorists and people have formed schematiresearchers in the cognitive social learning tradition assume that people schemati cally encode, store, and retrieve information information about the specific aspects of a setting 1982a;Lord, Lord, 1982; 1982; Mischel, Mischel, Cantor et et aI., al, 1982a; that are of particular interest to them (e.g., Cantor identify 1977). Given this assumption, a number of investigators have tried to identify dimensions of soc;:ial social settings that contribute to the type of inferences people make regarding the behavior that can be expressed in particular social situations (Wish, 1975; & Kaplan, 1976; & Kaplan, 1977). 1975; Wish, Deutsch, & 1976; Wish &

174 174

ICKES,SNYDER, SNYDER,AND ANDGARCIA GAROA IcKES,

al. (1976) identified situational dimensions that people For example, Wish et a1. perceive as relevant to the conduct of certain kinds of social relationships. Specifi Specifically, they found that the dimension of cooperative/friendly cooperative/friendly versus competitive/ hostile was a more salient aspect of the social situation for older than for younger subjects, for married than for single persons, and for politically right than for individuals. With With regard regard to to another another situational situational dimension, dimension, they theyfound found politically left individuals. that inequality was important for subjects with unconventional religious beliefs whereas equality was identified as pertinent for Christians. In many cases, the most salient feature of a given setting is the the type type of of person person most likely to be found in that setting. The findings findings of some studies (e.g., (e.g.. Cantor, 1980; 1980; Cantor & & Mischel, 1979) 1979) suggest that individuals share prototypic images of : the person most likely to be found in a given setting. When the participants in : these studies were asked to describe the characteristics of the "ideal" person in a particular setting, the results revealed that regardless of the level of situational superordinate categories categories such such as as "social" "social" or or "stressful" "stressful" situations situations abstraction (e.g., superordinate versus less abstract categories such as "party" or "interview"), there was a consenconsen person-for-a-situation prototype for each category. sual person-for-a-situation Before leaving this topic, we should note that other advocates of the social learning approach place somewhat less emphasis on cognition and somewhat more emphasis on the notion that behavioral regularities are the conditioned (and condi conditional) products of situational reinforcements. An article by L. A. Pervin (1981) provides a useful discussion of this perspective. C. Model C. The The Affed Affect Congruence Congruence Model Afinal final theory that attempts attempts to toaccount accountfor for the theprocesses processesthat thatindividuals individualsuse usetotoassess assess whether situations are conducive to the behavioral expression of their attributes is the affect a1. (1985). affect congruence model proposed by Emmons et al. (1985). As they have noted, "affect "affect experienced in the situation will partially determine future future decisions to enter or avoid that situation .. .. .. thus, to the extent that individuals experience affect affect which is compatible with their psychological predispositions, the probability of choosing that situation again in the future will be increased" (p. (p. 695). 695). Studies reported by Emmons and his colleagues provide some preliminary support for this theory. For example, Emmons and Diener (1986b) found that affect (satisfaction) was associated with goal attainment in both chosen positive affect situations,whereas whereasnegative negativeaffect affectwas wasmarginally marginallyrelated relatedtotothe thenonat nonatand imposed situations, only. In related studies, (a) extraverts tainment of goals in imposed situations only. affect when studying in the library than when reported experiencing more positive affect studying at home, (b) highly sociable persons reported feeling more positive when they were were in inchosen chosensocial socialrecreational recreationalsituations, situations,and and(c) (c)neurotics neuroticstended tendedtotoreport report they negative affect in any situation they entered (Diener et al., 1984; Emmons & Diener, negative affect in any situation they entered (Diener et aI., 1984; Emmons & Diener, 1986a; Emmons Emmons et et aI., al, 1985,1986). 1986a; 1985, 1986).

CHAPTER CHAPTER 77

PERSONALnY PERSONALITY AND AND CHOICE CHOICE OF OF SITUATIONS SITUATIONS

175 175

D. D . Commonalities and Contrasts Contrasts In general, the theories we have just reviewed suggest that individuals attempt to identify identify and selectively enter those situations that encourage and reinforce reinforce the expression of their own particular attributes. A common assumption of these theories is that the more the person's disposition "fits" the situation, the better the outcome the person can expect to obtain. A second common assumption is that specific motives underlie people's choices of of dispositionally congruent situa situations. The theories differ, goal differ, however, in the specific motives they posit. In the goalfor example, the person is presumably motivated motivated to accomplish or based theories, for satisfying the need for for affiliation, affiliation, esteem, or status. attain a particular goal, such as satisfying In the cognitive representational theories, the person is presumably motivated to future choices of understand dimensions of situations so that future of dispositionally relevant affect congruence model, the person situations will be easier to make. Finally, in the affect form of affective satisfaction as a conseis presumably motivated to obtain some form affective satisfaction conse quence of having made an appropriate situational choice. differences in the assumed motives for for situational situational choice imply corre correThese differences differences in the individual's perception of of the situations that he or she sponding differences of the encounters. The goal-based theories emphasize the individual's perception of for goal attainment available in different opportunity structures for different situations (e.g., al., 1981). 1981). In contrast, the cognitive representational theories emphasize Argyle et aI., features and dimensions that signal the individual's perception of those abstract features the potential congruence/incongruence between the situation and the individual's affect congruence model emphasizes the individual's own disposition(s). Finally, the affect perception of the anticipated positive or negative affect derived from entering affect to be derived and taking part in the situation.

IV. UNDERSTANDING UNDERSTANDING INDmnuALS INDIVIDUALS IN IN TERMS TERMS OF OF IV. CHOICES SITUATIONAL CHOICES

If people can assess situations according to the opportunities available for the fact systematically systematically choose to behavioral expression of their dispositions, do they in fact enter and spend time in such opportune settings? Our review of the literature influence reveals that there are several categories of studies which document the influence of individual dispositions on situational choices:

1. The first and largest category includes studies that have investigated investigated dif diffunction of specific personality traits such as ferences in situational choices as a function 1973; Zuckerman, 1978), self-monitoring self-monitoring (e.g., Snysensation-seeking (e.g., Segal, 1973; Sny & Gangestad, 1982; 1982; Snyder & & Kendzierski, 1982), 1982), introversion/extraversion der &

176 176

IcxES, IcKES,SNYDER, SNYDER,AND AND GARCIA GARCIA

(e.g., Eysenck, 1974; 1974; Furnham, 1981), and shyness (e.g., A. K. Watson & & Cheek, 1986). 2. A second category includes studies documenting instances in which people choose to enter those specific settings that help them sustain existing self-concepts, & Jones, 1978; 1978; Tesser, for example, studies of self-esteem maintenance (Berglas & 1988; Tesser & & Moore, 1987) and self-verification self-verification (Swann, 1987; 1987; Swann & & Read, 1988; & Giesler, 1992; 1992; Swann, Wentzlaff, & & Krull, 1992). 1981; Swann, Stein-Seroussi, & 3. A third category includes studies that document how situational choices Berman, 1979; 1979; Snyder & & can vary as a function of social attitudes (e.g., Kahle & & Herman, Kendzierski, 1982; 1982; Wilson & & Nias, 1975). 4. A fourth category includes studies designed to investigate the settings within which whichpeople people choose choose to to conduct conduct particular particulartypes typesof ofrelationships relationships(e.g., (e.g.,Argyle, Argyle, 1980; Argyle & &Furnham, Furnham,1982; 1982;Glick, Glick,1985; 1985;Jellison JeUison&ifcIckes, Ickes,1974; 1974;Snyder, Snyder,Gange Gange1980; Argyle stad, & & Simpson, 1983; 1983; D. G. Winter, 1973). studies that thatexamine examinethe thepersonal personalorigins originsof ofchoices choices 5. Afifth fifth category includes studies &Witt, Witt, 1970; 1970;Furnham, Furnham,1981), 1981),educational educationalsettings settings (e.g.. Bishop & of leisure situations (e.g., & Sinnett, 1973; & Rubin, 1967; & Bloom, (e.g., Eddy & 1973; B. A. Pervin & 1967; Stern, Stein, & 1956), and occupational situations (e.g., Atkinson, 1958; 1958; Holland, Holland, 1985; 1985;Rosenberg, Rosenberg, 1957; 1957; Vroom, 1964). 6. In addition to the various categories of studies that have demonstrated the choice of dispositionally congruent situations, there is a smaller category of studies incongruentsituations. situations. that have explored people's responses to dispositionally incongruent dispositionThese studies reveal, as expected, that people typically choose to avoid disposition ally incongruent situations (e.g., Furnham, 1981). They also reveal, however, that in those cases in which people enter disposition ally incongruent situations, they dispositionally often attempt to alter the situations to make them more congruent with their own personalities (e.g., Srull 1975; D. Watson & SruU & & Karabenick, 1975; & Baumal, 1967). In a theoretical treatment of this issue, Snyder and Ickes (1985) have proposed some conditions in which people deliberately choose to enter dispositionally incongruent situations in order to effect desired changes in themselves, in the situation, or in other people. . 7. Finally, to complement the relatively large category of studies that have much smaller smaller examined the influence of dispositions on situational choice, there is a much have examined examined the thereciprocal reciprocalinfluence influence of of situational situational choices choices category of studies that have and personality development. These studies suggest that the press of the different situations encountered throughout an individual's life can dramatically shape the behaviors that that eventually eventually become become part part of of that that individual's individual's personality personality (e.g., (e.g., Caspi, Caspi, behaviors 1987; Caspi & Bem, 1990; Caspi et al, 1989; Caspi & Herbener, 1990; Caspi, Caspi & Bern, 1990; Caspi et aI., 1989; Caspi & Herbener, 1990; Caspi, 1987; Herbener, & Ozer, Ozer, 1992; 1992; Ickes Ickes & & Turner, Turner, 1983; 1983; Runyan, Runyan, 1978). 1978). Herbener, &

Unes of evidence for the proposition In summary, there are several different lines spendtime time in insituations situationsthat thatpromote promotethe theexpression expression that people choose to enter and spend of their own trait-relevant behavior. A more detailed look at this evidence is provided in the following sections.

CHAPTER? CHAPTER 7

PERSONALITY AND ANDCHOICE CHOICEOF OFSmJATIONS SITUATIONS PERSONALITY

177 177

A. Characteristic Dispositions and Situational Choices Many studies have investigated the relation between personality dispositions and situational preferences. In general, the results of these studies offer offer compelling evidence that people choose to enter and spend time in situations that will foster, foster, promote, and encourage the behavioral manifestations manifestations of their own traits and dispo dispositions. For example, people with low arousal-seeking tendencies appear to prefer prefer settings that are relatively low in complexity, and they are likely to avoid highly informative informative situations. People with high arousal-seeking tendencies, on the other hand, tend to seek out environments that are highly informative and relatively complex (Mehrabian, 1978; Mehrabian & & Russell, 1973). 1973). According to the findings findings of several studies, individuals with an internal locus of control are more likely to choose test settings that require skill or ability than are individuals with an external 1975; D. SruU & & Karabenick, 1975; D. Watson Watson & & locus of control (Kahle, 1980; Rotter, 1966; Srull In other other research, research, extraverts extraverts have have been been found found to to seek seek out out stimulating stimulating Baumal, 1967). In physical pursuits) whereas whereas introverts prefer prefer leisure situations (e.g., ones that involve physical more passive recreational situations (e.g., reading; Furnham, 1981). Fumham (1982) has reported that psychotics choose to interact in spontaneous Furnham formal and stable ones. With regard to and volatile situations as opposed to more formal see also also Russell, Russell, Cutrona, Cutrona, & & the dispositionally shy, A. K. Watson and Cheek (1986; see found that novelty is perhaps the most salient saUent factor factor eliciting Jones, 1986) have found shyness reactions in social settings (e.g., engaging in conversation with strangers). Snyder and his colleagues have shown that low self-monitors self-monitors In another domain, Snyder incongruent with their are unwilling to enter into situations that are relatively incongruent beHefs and attitudes about certain issues, whereas high self-monitors self-monitors may personal beliefs different roles (Snyder, 1979; actively choose situations that cast them in different 1979; Snyder & & 1982; Snyder & & Kendzierski, 1982). Gangestad, 1982; further evidence of the congruence between a person's person's Other studies provide further situational choices and his or her characteristic dispositions. Machiavellians MachiaveUians prefer prefer face-to-face face-to-face interactions, because encounters of this type enable them to optimally 1970; Geis, 1978). 1978). In apply their tactics for exploiting others (Christie & & Geis, 1970; the context of therapy situations, field-dependent field-dependent patients seem to be happiest in supportive therapies with well-defined well-defined structures, whereas field-independent field-independent pa patients prefer prefer to play a more active role in the content and progress of their therapy (Karp, Kissin, & 1970). Related & Hustmyer, 1970). Related research research has has shown shown that that field-dependent field-dependent individuals are more likely than field-independent field-independent individuals to rely on other people for Goodfor useful useful information information in otherwise ambiguous settings (Gates, 1971; Good enough, 1978; Greene, 1973; Mausner & 1978; 1973; & Graham, 1970; 1970; Nevill, 1974). Research on high sensation-seekers has provided extensive evidence of of their preference for situations that are novel and intensely arousing. For example, relative For example, relative to low sensation-seekers, high sensation-seekers are more likely to volunteer for unusual experiments (e.g., drug studies) and to report having experienced a greater 1974,1978). In studies studies conducted conducted in in business business variety of sexual activities (Zuckerman, 1974, 1978). In

178 178

ICKES, IcKES, SNYDER, SNYDER, AND AND GARCIA GARQA

environments, extraverts with an internal locus of control have been shown to & actively participate in the the interior design of their offices (McElroy, Morrow, & Ackerman, 1983; 1988). Finally, 1983; Osborn, 1988). Finally,people peoplewith withaahigh highneed needfor forapproval approvalseem seem to value highly those situations that provide them with cues that help to guide and and & Jacobson, 1978). organize their behavior (Millham & B. B. Self-Conceptions Self-Conceptions and and Situational Situational Choices Choices Research investigations have focused on other categories of dispositions in addition to traits. These studies provide evidence that people choose situations for the sake of congruency with these other classes of dispositions as as well. well. A number of these studies have shown, for example, that people also actively choose to enter and spend time in settings that preserve, sustain, or maintain their conceptions of self. The "self-handicapping" studies by by E. E. Jones and Berglas (Berglas & & Jones, 1978) provide evidence that people who are are concerned concerned about about threats threats to to their their self selfand to sustain their self-image by perceived competence will attempt to protect and by can explain away their failures and putting themselves in situations in which they can and take credit for their successes. On the the other hand, if situations fail to protect and, and, instead, threaten their self-competent image, there there is is evidence evidence that that individuals individuals will instead, threaten their self-competent image, will to leave leave these these settings settings (Conolley, (ConoUey, Gerard, Gerard, & & Kline, Kline, 1978). choose to 1978). choose In some situations, people strive to maintain or increase their self-evaluation in a comparison or a reflection process. process. According to Tesser and and his his by engaging in 1982, 1983; Tesser colleagues (Tesser, 1984, 1985, 1988; Tesser 1984,1985,1988; Tesser & & Campbell, Campbell, 1980, 1980,1982,1983; Tesser&& 1987; Tesser & & Paulhus, 1983; 1983; Tesser & & Smith, 1980), 1980), certain factors such Moore, 1987; as the relevance relevance of of another another person's person's performance performance on on aa task, task, the the perceived perceived closeness closeness of the other, and and the the other's actual performance determine the particular strategy one uses to preserve one's self-esteem. For example, in order to protect or regain their self-esteem, people may either increase their efforts efforts on on a task or distort the perception of their own own performance relative to the performance performance of others (Tesser, Campbell, & & Smith, 1984). In other cases, people will seek to validate their self-concepts by by using cer cer1961, 1965) "self tain "congruency" (Backman, 1988; 1988; Secord & & Backman, 1961, 1965) or **selfverification" (Swann, 1987; 1987; Swann, Stein-Seroussi, Stein-Seroussi, & & Giesler, Giesler, 1992) 1992) strategies strategies when when interacting with those whom they encounter in the the situations they have selected. Both Secord and ap and Backman's congruency theory and and Swann's self-verification self-verification approach assert that people not only choose, but at times actively construct, certain the purpose of maintaining a stable self-image. For For features of their situations for the and Read Read (1981, (1981, Study Study 1) 1)found found that that people people who who saw saw themselves themselves example, Swann and as likable likable spent spent more more time time reading reading an an interaction interaction partner's partner's appraisal appraisal of of them, as particularly if the interaction partner viewed them as likable. In another study, particularly if the interaction partner viewed them as likable. In another study, Swann and Hill (1982) reported that self-conceived dominants who had been labeled Swann and Hill (1982) reported that self-conceived dominants who had been labeled as submissive submissive by by their their interaction interaction partners partners subsequently subsequently reacted reacted with with some some veheas vehe mence toward their partners in order to reassert their dominant personalities. mence toward their partners in order to reassert their dominant personalities.

CHAPTER CHAPTER 77

PERSONALITY AND CHOICE CHOICE OF SITUATIONS PERSONAUIY

179 179

"selfOf course, not all people are highly motivated to use "congruency" or "self verification" strategies to help ensure that their self-conception can be sustained self-conception in the face of situational pressures that might threaten it. As Snyder and his colcol (for a review, see Snyder, 1987), some people are leagues have repeatedly shown (for so sensitive and responsive to situational pressures that they strive for congruency by presenting themselves as being whatever kind of person the situation apparently self-monitors seem to adopt different different personalities to match demands. These high self-monitors the different different situations they enter. In contrast, low self-monitors self-monitors are more likely to selectively enter those situations that allow them to behave in a manner consistent with their typical self-conception 1982). For self-conception (Snyder and Gangestad, 1982). For more more detailed detailed discussions of congruency theory and its relevance to the question of how situational function of personality, see Backman (1988) and Snyder and Ickes choices vary as a function (1985, pp. 921-932). C. Social Social Attitudes Attitudes and and Situational Situational Choices Choices C. focused on Another set of studies that examine people's situational choices have focused for the specific purpose of behaviorally expressing one's atti attithe settings chosen for underlying assumption assumption is that people prefer prefer to enter settings settings that provide provide tudes. The underlying them with opportunities to express and act upon their opinions and beliefs. For example, Kahle and Berman Herman (1979) reported that people with favorable attitudes toward particular candidates for a political office actively seek out situations in which they will be exposed to messages favorable to their candidate. Similarly, low found to gravitate toward settings that dispose them to self-monitors have been found behave in ways that will enhance the congruence between their attitudes and their & Kendzierski, 1982). behavior (Snyder & As another another example, several studies have shown that authoritarians tend to information that may change their attitude toward a choose not to accept any information particular object (Dillehay, 1978; 1978; Katz, 1960; 1960; Katz, McClintock, & & Sarnoff, 1957). Other Other research suggests that that extraverts, being more permissive than introverts in their attitudes about social behavior, choose to engage in behaviors that involve more risks, such as having sexual intercourse more frequently, in more different different ways, and with more different 1975; see also Gangestad & & different partners (Wilson & & Nias, 1975; Simpson, 1990). 1990). Extraverts are also more likely to break institutional rules and wind up in prison more often often than introverts (Eysenck, 1971). Other research indicates that smokers with an external locus of control tend to be more "chance oriented" than smokers with an internal locus of of control. That is, believers in external control are more likely to continue such behavior despite warnings about the consequences of heavy smoking. In addition, fraternities and less likely to they are more likely to be members of social fraternities attend church than are nonsmokers or smokers with an internal locus of control & Werner, 1965; 1965; Straits & & Sechrest, 1963). ((James, James, Woodruff, & 1963). In a similar vein, it has been reported that women with an an external locus of control are less likely

Icm, IcKES»SNYDER, SNYDER, AND AND GARCIA GARQA

180 180

to practice practice effective effective birth birth control control than than are are those those with with an an internal internal locus locus of of control control to (Lundy, 1972). (Lundy, 1972). D. Choices D. Social Social Relationships Relationships and and Situational Situational Choices A A number number of of empirical empirical studies studies have have begun begun to to explore explore how how situational situational choices choices can can vary as as aa function function of of one's one's relationship relationship with with others. others. In In fact, fact, the the types types of of relationship relationship vary that exist exist between between people people can can have have aa profound profound impact impact on on their their choice choice of of situations situations that 1981). For For example, example, Argyle Argyle and and Furnham Furnham (1982) (1982) found found that that (cf. Duck Duck & & Gilmore, Gilmore, 1981). (cf. type of of relationship relationship could could be be used used to to predict predict the the type type of of situation situation chosen. chosen. Marital Marital type the most most powerful powerful influence influence on on choice choice of ofsituations, situations,with withdomestic, domestic, relationships had had the relationships informal, intimate, and recreational settings being chosen most frequently. Friendinformal, intimate, and recreational settings being chosen most frequently. Friend ships exerted the next most powerful influence on situational choices, with less ships exerted the next most powerful influence on situational choices, with less domestic and more task-oriented situations selected for these relationships. domestic and more task-oriented situations selected for these relationships. On On the other other hand, hand, the the situations situations chosen chosen for for interacting interacting with with disliked disliked colleagues colleagues the were relatively relatively short short in in duration, duration, infrequent, infrequent, and and formal formal (e.g., (e.g., morning morning coffee, coffee, were meetings). meetings). Other studies studies have havebegun begunto toaddress addressthe thequestion questionof ofhow howpersonality personalityinfluences influences Other bothsituations situationsand andtypes typesofofrelationships. relationships.For Forexample, example,Glick Glick(1985) (1985) preferences for for both preferences studied the the conditions conditions chosen chosen by by high high and and low low self-monitors self-monitors for for initiating initiating relation relationstudied ships. He He found found that that low low self-monitors self-monitors chose chose potential potential dating dating partners partners on on the the basis basis ships. of the the partners' partners* personality personality characteristics, characteristics, whereas whereas high high self-monitors self-monitors selected selected of potential dating dating partners partners on on the the basis basis of of their their physical physical attractiveness. attractiveness. In In addition, addition, potential revealed some some systematic systematic differences differences in in situational situational preferences preferences for for Glick'sfindings Glick's findings revealed the particular type type of of partner partner chosen. chosen. Low Low self-monitors self-monitors preferred to initiate initiate the preferred to the the particular relationship with with aa partner partner who who had had desirable desirable personality personality characteristics characteristics in in aa roman romanrelationship tic situation; situation; they they avoided avoided romantic romantic situations situations if if they they were were forced forced to to interact interact with with tic partners partners who who had had less-than-desirable less-than-desirable characteristics. characteristics. High High self-monitors, self-monitors, by by the the same token, token, chose chose to to interact interact with withphysically physicallyattractive attractivepartners partnersin inromantic romanticsettings settings same settings ifif the the partner partner was was physicaUy physicallyunattractive. unattractive.Further Furtheranalysis analysis and avoided avoided such such settings and revealed that that romantic romantic situations situations provide provide specific specific opportunities opportunities that that can can be be strategi strategirevealed cally employed employed to to facilitate facilitate romantic romantic relationships relationships with with one's preferred type type of cally one's preferred of partner. For For example, example, these these situations situations involved involved less less structured structured and and longer longer interacpartner. interac tions, took took place place at at aa different different time time of of the the day, day, and and provided provided future future opportunities opportunities tions, to pursue pursue the the relationship. relationship. to In aa similar similar study, study, Snyder Snyder et et a1. al. (1983) (1983) found found systematic systematic differences differences between between In high and and low low self-monitors self-monitors in in the the type type of of partner partner chosen chosen for for leisure leisure activities. activities. High High high self-monitors chose chose leisure leisure partners partners according according to to their their particular particular skills skills (i.e., (i.e., being being self-monitors "specialist") in in the the activity activity domain, domain, whereas whereas low low self-monitors self-monitors preferred preferred activity activity aa "specialist") partners who were were nonspecialists but whom they particularly particularly liked. liked. In In terms terms of partners who nonspecialists but whom they of the preferred preferred situations situations in in which which to to conduct conduct such such relationships, relationships, high high self-monitors self-monitors the preferred to to interact interact with with aa specific specific partner partner for for particular particular activities activities only. only. Low Low self selfpreferred monitors, on the the other other hand, were less less likely likely to to differentiate differentiate their their social social worlds worlds monitors, on hand, were

CHAnER CHAPTER 77

PERSONAUTY nONS PERSONALITY AND AND CHOICE CHOICE OF OF SITUA SITUATIONS

181 181

well-Uked partners regardless of the this way, and instead chose to spend time with well-liked type of activity involved. E. E. Choice Choice of of Leisure, Leisure, Educational, Educational, and and Occupational Occupational Settings Settings Finally, there is considerable evidence that people's dispositions can influence their choices of leisure activities, educational settings, and occupational situations. jluences on Influences onChoice ChoiceofofLeisure LeisureSituations Situations i. Dispositional Dispositional ln 1.

With regard to leisure activities, Furnham (1981) found that extraverts preferred preferred a distinctively different different pattern of recreational activities than did introverts. For example, extraverts preferred to engage engage in activities that were affiliative, stimulating, physical, and informal. Introverts, on the other hand, chose leisure activities that required a sense of order and planning. In a related study, Emmons et al. (1986) found that extraverts not only preferred preferred social recreational settings but also reported found affect when they were in chosen social settings than in imposed more positive affect found that people with a high need nonsocial situations (e.g., reading). They also found affect in chosen recreational situations and were less for play experienced positive affect happy in imposed work settings.

2. Dispositional Dispositional ln Influences onChoice Choiceofof 2. jluences on Educational Situations Educational With regard to educational settings, people with a strong power motivation are Ukely to choose courses deaUng dealing with the application of direct and legitimate more likely poHtics; D. G. Winter, 1973; 1973; D. D. G. G. Winter Winter & & Stewart, Stewart, power (e.g., science, law, or politics; often officers officers in 1978). Moreover, students with a high need for power are more often student newspaper newspaper and and radio radio university student organizations, dormitory counselors, student faculty-student committees (D. & Stewart, workers, and members of faculty-student (D. G. Winter & 1978). vein. Stern Stem et a1. al. (1956) have reported that people with an 1978). In a related vein, authoritarian disposition disproportionately choose to enter military academies. Other research suggests that extraverted, action-oriented students are particu particularly likely to spend time in those areas of the campus that encourage affiliation affiliation and social interaction (e.g., lobbies and parks; Eddy & & Sinnett, 1973). And, as we prefer have noted previously, field independents, who tend to value cognitive skills, prefer to enter and achieve success in academic and vocational areas, whereas field depen depenprefer areas that deal with other dents, who do not particularly value such skills, prefer 1969; Winter, Moore, Goodenough, & & Cox, Cox, 1977). 1977). 1978; Levy, 1969; people (Goodenough, 1978; 3. 3.

Dispositional jluences on Dispositional ln Influences onChoice Choiceof of Occupational Situations

With regard to choices about occupational situations, people who believe it is important to be autonomous and independent on the job tend to select occupations that are characterized by relatively high levels of uncertainty and worker autonomy (Morse, 1975). In In other other examples, examples, individuals individuals high high in inpeople-orientation people-orientation are are particuparticu-

IcKES,SNYDER, SNYDER, AND AND GARCIA GARQA ICKES,

182 181

larly likely to choose such jobs as social work, medicine, and teaching; individuals highneed needfor forself-expression self-expressiontend tendto toprefer prefercareers careersininarchitecture, architecture,journalism, journalism, with aa high prefer to enter occupations that and art; and individuals high in reward-orientation prefer involve sales, finance, and management (Rosenberg, 1957). 1957). Finally, achievement achievementoriented people are especially likely to choose companies that emphasize individual advancement over those emphasizing power and status (Andrews, 1967). In a recent book, Holland (1985) has proposed six personality types that are associated with distinctly different different vocational interests. For instance, because the investigative type is task-oriented and and prefers prefers thinking thinking rather rather than than acting acting out out prob problems, this individual is more likely to choose a scientific occupation (e.g., design engineer). In contrast, the social type can be found in human services jobs jobs because because of this individual's trait of sociability. Whereas the enterprising type is dominating and adventurous and is therefore well-suited to business or sales positions, the conventional type is conforming and conservative and tends to gravitate toward occupations such as asbookkeeping. bookkeeping.The Therealistic realistictype typefavors favorstackling tacklingconcrete concreteinstead instead of abstract problems and is therefore found in mechanical occupations. Finally, the artistic type is sensitive to emotions and expressions of individuality and is more likely to to be be aa playwright playwright or or aa commercial commercial artist. artist. The The results results of of aa number number of of studies studies likely are reviewed as being consistent with Holland's theory (e.g., Benninger & Walsh, are reviewed as being consistent with Holland's theory (e.g., Benninger & Walsh, 1980; Costa, McCrae, & Holland, 1984; Walsh, Horton, & Gaffey, 1977). 1980; Costa, McCrae, & Holland, 1984; Walsh, Horton, & Gaffey, 1977).

Situations F. Reactions Reactions to to Dispositionally Dispositionally Incongruent Incongnient Situations F. We have marshalled many empirical examples to illustrate the proposition that reflecting people are particularly likely to enter and to spend time in situations reflecting their self-conceptions, their attitudes, their personal tastes in leisure, educational, and occupational activities, and the types of relationships they have. have. We now consider the complementary issue of how people respond to disposition ally incon incondispositionally gruent gruentsituations. situations. In general, general, reactions reactionsto todispositionally dispositionallyincongruent incongruentsituations situationstake takeone oneofofthree three forms: (a) choosing deliberately to avoid dispositionally incongruent settings, avoid dispositionally incongruent settings, (b) (b) inadvertently entering such settings and then attempting to to cope copewith with the thelack lack of of congruity between the setting and one's own dispositions, and (c) choosing deliber deliberentersuch suchsettings, settings,totochange changeeither either one's one'sown owndispositions dispositionsor orfeatures features of of ately to enter the setting itself.

/• Avoiding AvoidingIncongruent Incongruent Situations 1. Situations First, and most obviously, people may choose not to enter and spend time in dispositionally incongruent settings. settings. The The act act of ofdeliberately deliberately avoiding avoidingcertain certainsettings settings and situations can at times be just as as revealing and reflective of a person's disposi disposienteringother othersettings settingsand andsituations situations(cf. (cf. tions and traits as the act of deliberately entering E. E. E. Jones Jones & & Davis, Davis, 1965; 1965; E. E. E. E. Jones Jones & & McGillis, McGillis, 1976). 1976). For For example, Furnham Furnham (1982) reported reported that that people people with with high high social social desirability desirability scores scores preferred preferred to to enter enter (1982) situations that are socially simple and informal but to avoid ones in which they

CHAPTER?7 CHAPTER

PERSONALITY AND CHOICE OF SITUATIONS

183 183

could become the focus of attention. Research by Efran Efran and Boylin (1967) suggests that they also tend to avoid situations that are highly evaluative or that call for some degree of assertiveness (e.g., group discussions). For another example, people with an avoidant as avoidant personality disorder are asrejection or humiliation. sumed to desire social contact but at the same time to fear rejection As a result, they choose not to expose themselves to social settings in which there is the potential for rejection and humiliation. If they are caught in such a setting, they & Everly, 1985). 1985). Similarly, individuals act socially detached and withdrawn (Millon & personaHty disorder may be *'in" physidiagnosed as having a schizoid personality "in" a situation physi cally but avoid it psychologically by being unresponsive to the behavior of other interpersonally passive and bland; Millon & & Everly, 1985). people (e.g., by being interpersonally In educational settings, people high in trait anxiety tend to report a higher level of of state anxiety about situations characterized by social and academic failure and, & Felling, 1970). 1970). It has also been shown thus, tend to avoid such situations (Hodges & from academic settings because of that neurotic extraverts are likely to withdraw from of medical/ fear of failure whereas neurotic introverts are likely to withdraw for medical! 1973). Lastly, Cox, Endler, Swinpsychiatric reasons (Eysenck, 1974; 1974; Wankowski, 1973). Swin son, and Norton (1992) have noted that individuals with panic disorders typically son, avoid situations that are agoraphobic in nature (e.g., walking alone in busy streets). 2. Attempting Attempting to to Cope Cope with with Incongruent IncongruentSituations Situations 2.

foresight that people frequently frequently display in their choice of situations, Despite the foresight they may at times still find themselves in situations that are incongruent with their own dispositions. Some research evidence suggests that, when inadvertently confronted by such situations, people will react negatively to the situation and!or and/or confronted try to change some aspect(s) of the situation in order to make it more congruent with their own dispositions. Both of these processes were evident in a study of of disposiinterracial interactions reported by Ickes (1984). In this study, whites whose disposi tion was either to seek out or to avoid interaction with blacks inadvertently found themselves in a situation that required them to wait for an indefinite period in the preferred to seek out interaction company of a black person. Relative to whites who preferred preferred to avoid interaction with blacks (a) reported with blacks, those who preferred heightened feelings of anxiety and concern about the interactions, and (b) tried to minimize their psychological involvement with their black partners by looking and smiling at them less. 3. 3.

Choosing Choosing Incongruent IncongruentSituations Situations

The last, and seemingly most paradoxical, reaction to dispositionally incongruent situations is to deliberately choose to enter and spend time in them. According to Snyder and Ickes (1985), this course of action may not actually be paradoxical at all. reflect people's desires to use the personality-shaping properties all. Instead, it may reflect of such dispositionally incongruent settings either (a) to change themselves (e.g., effect either a temporary or more permanent change in their own dispositions) to effect itself (e.g., convert a or (b) to exert their influence so as to change the situation itself

184 184

icKES, ICKES, SNYDER, SNYDER, AND AND GARCIA GAROA

vacant lot into a playground) or to change the behavior of of other people in the situation (e.g., convert gang members into a basketball team). There is, unfortunately, a conspicuous dearth dearth of of empirical studies concerning dispositionally incongruent incongruent situations. As a stimulus these last ways of reacting to dispositionally to the collection of relevant data, we have previously offered offered some theoretical & Ickes, 1985). speculations about the dynamics of these activities (Snyder & 1985). Our intent was to explain, among other things, why we might find shy people in swinging sursingles' bars, snake phobics in reptile houses, paunchy sybarites in wilderness sur front rows of church. We noted that, in all of these vival courses, and sinners in the front incongruent with their current cases, the situations these people choose to enter are incongruent dispositions and attributes at the time that they enter them. We proposed, however, congruent with the dispositions and attri attrithat these same situations may be quite congruent Uke to possess at some time in the future. future. butes these people would like Our theoretical speculations also concerned the motives of the preacher in pacifist in the the house of ill repute, the temperance lady in the local tavern, the pacifist for persons with AIDS (e.g., Omoto & & Pentagon, and the healthy individual caring for for these individuals' behavior Snyder, 1990). 1990). We proposed that a common motive for situation in order to change may be their desire to exert influence in and on the situation itself or to change other people. People whose choice of the setting itself of incongruent situations is determined by this motive tend to cast themselves in the role of social reformer on a scale that may vary greatly in ambition and in degree of of influence. reformer Missionaries, proselytizers, social workers, reformers, revolutionaries, and radical activists provide relatively dramatic examples, but more mundane examples (e.g., identified. the slum-reared gatecrasher of a high society social affair) affair) can also be identified.

G. The The Role Role of of Choosing Choosing Situations Situations in in G. PersonaDty Personality Development Development The reciprocal causal relationship between personality and situation obviously raises the important questions of (a) how personality traits assessed during childhood influence situational choices made in adolescence and adulthood, and (b) how situational influences at different develop different points in the life cycle contribute to the development of personality. These questions have recently been addressed addressed in research reported by Caspi, Elder, and Bem (e.g., Caspi, 1987; 1987; Caspi, et al., aI., 1989, 1989, 1992; 1988; Caspi & Caspi & & Elder, 1988; 1988; Caspi, Elder, & & Bem, 1987, 1987,1988; & Herbener, 1990; Elder & & Caspi, 1988; 1988; Elder, Caspi, & & Burton, 1988). of the processes that help These investigators have begun to document some of trait-relevant behavior. Using life records of promote life-course continuities in trait-relevant of effects of situational influences on particular individuals (L-data) to explore the effects personality across the life span, Caspi and his colleagues have compiled some for the stability and consistency in individuals' situational impressive evidence for function of dispositions and traits. For example, they found found that males choices as a function tended—as adolescents and later as who were dispositionally shy as children tended-as adults—to be aloof, withdrawn, and lacking in social poise. These men were also adults-to

CHAPTER? PERSONALITY AND ANDCHOICE CHOICEOFOF SITUAIIONS CHAFfER 7 PERSONAUTY SITUATIONS

185 185

found to experience greater marital instability and less occupational stability, and appeared to have more difficulty difficulty making transitions to adult roles than the other males in their cohort. Dispositionally shy females, on the other hand, were found found to follow or choose, in later adult life, a traditional, conventional, or domestic lifestyle of of marriage and family. Other research in this tradition has further documented the reciprocal develop developmental influences of of situational factors and personality personaUty traits. For example, Ickes and cf. Ickes, 1983) found that college-age men who had the experience of Turner (1983; cf. being raised with one or more older sisters received more eye contact and greater liking from their female partners in an initial interaction than did men who had only younger sisters. Block, Block, and Keyes (1988) reported that young girls exposed early in life to a family environment that emphasized dependency, disorder, constant disappointment from parents, and less emphasis on propriety, convention, personaHties (e.g., dysphoric, distancing, distrustful, and religion, tended to develop personalities and defensive) that were consistent with the type of situations they subsequently chose to be in (e.g., drug use). And Elder and Caspi (1988) reported that males who were assessed to have an above average tendency for losing control before before the Great Depression of of the 1930s developed explosive personalities after being exposed to this particular situational (i.e., societal) stress. Behavioral manifestations manifestations of their explosive personalities were subsequently evident in their marital and of family relationships. family Caspi and Elder (1988) have argued that the reciprocal relation between personality and situation is so strong and far reaching in its effects effects that unstable personaUties can be reproduced across generations through the maintenance of of personalities particular situational conditions such as unstable family relationships, marital tenten sion, and ineffective ineffective parenting. They reported that the styles of of behavior expressed of the fourth generation by grandparents can still be seen in the behavior patterns of of children. of CONCLUSION V. CONCLUSION V. of contemporary psychology that behavior is a function of It is practically a truism of of the situation. Beginning with Lewin's early assertion that "every the person and of psychological event depends upon the state of the person and at the same time on the environment, although their relative importance is different different in different different cases" (1936, p. 12), the task of of defining the precise nature of the function function that joins person and situation has been central to the theoretical and empirical missions of of researchers in personality and social psychology. Although many approaches have been taken to defining the "person by situa situation" function (for a review, see Snyder & & Ickes, 1985), in this chapter we have function—the dynamic interactionist chosen to focus on one approach to specify this function-the personaUty and social behavior. This or "situational" strategy for the study of personality situations. approach concentrates on the reciprocal influences of individuals and situations,

186 186

SNYDER, AND AND GARCIA GARCIA IcKES, SNYDER,

examining especially the active role that individuals play in choosing to enter and spend time in particular social situations, as well as the consequent opportunity that chosen situations gain for guiding and directing the behavior of those who have placed themselves therein. That people actively gravitate toward some types of of situations situations and and deliberately deliberately void others, and that their choices of situations are reflections of features of their avoid a personal identities (including such dispositions as attitudes, traits, and conceptions of self), may may constitute major sources of the regularities, stabilities, stabilities, and and consistencies consistencies in behavior that are typically regarded as the defining characteristics of personality. As evidence in support of these propositions, we have reviewed research findings findings field studies of both actual from a variety of sources, including both laboratory and field and hypothetical behavior. This accumulated (and still rapidly growing) body of evidence is substantial, and it convincingly implicates a wide range of traits and dispositions, conceptions of self and identity, attitudes and beliefs, and other perper sonal attributes in the processes of choosing situations. In addition, researchers have begun to explore some of the mechanisms by which people choose situations, in particular the dimensions along which situations are perceived, categorized, and ultimately responded to. Although our chapter has concentrated on the choosing of situations, we wish to emphasize that such activities do not constitute the sum total of individuals* individuals' active structuring of the circumstances in which they operate. Space permitting, we could have surveyed the research literature indicating that, just as people choose their situations, they may also systematically choose their roles and their personal and social relationships. Moreover, they may make these choices in ways that allow them to take on roles and to participate in relationships that provide opportunities to act upon their attitudes, values, traits, dispositions, self. Space dispositions, and and conceptions conceptions of of self. limitations have also prevented us from dealing with the related phenomenon of hmitations of people's attempts to influence and modify situations (either ones that they have previously chosen to enter or ones in which theyfind find themselves through no choice of their own) in ways that make these situations more supportive of their personali personaUties and identities (see Snyder, 1981; 1981; Snyder & & Ickes, 1985). Together, through their choices of and influence upon situations, roles, and relationships, people may construct social worlds that are conducive to their own personalities. personaUties. As much has been accomplished in demonstrating that people do actively choose their situations, in in probing the determinants of these choices, and in specify specifying their consequences, even more remains to be done. As we have indicated, comparatively little is known about people's choices of discrepant situations, that is, those instances in which people actively place themselves in situations that tend to dispose behaviors that run counter to their own current personal attributes. Because hypotheses about the nature, antecedents, and consequences of these choices are are readily generated, we we anticipate that such processes processes will will become become produc productive areas of study for future research. We We also expect that the the psychology psychology of of development will willprove provehighly highlyamenable amenable to inquiries guided by dynamic interactionist considerations of choosing situations.

CHAPTER CHAPTER 7 7

187 187

PERSONAUTY PERSONALITY AND AND CHOICE CHOICE OF OF SITUATIONS SITUATIONS

Theoretical analyses have pointed to the importance of of considering considering people's choices of situations and the constraining influences these chosen situations have on the course of of development, either through facilitating facilitating certain developmental developmental trajectories trajectories or by closing the doors on alternative developmental developmental courses (cf. Block, 1971; from the Gangestad Gangestad & & Snyder, 1985). Moreover, using archival data data from the Berkeley Guidance Study, Caspi et al. (1989) have examined examined the ways in which which behavior behavior patterns can be sustained other mechanisms, sustained across the life course by, among other individuals' entry into environments that reinforce reinforce and and sustain interactional interactional styles and behavioral dispositions. As this review has clearly demonstrated, the notion notion that that individuals choose of their personality personality traits and dispositions is a well wellsituations on the basis of and other other dispositions established with this fact, estabUshed fact. As the field of of personality reckons with fact, two important outcomes are guaranteed. The first outcome is that that the field will inevitably inevitably gravitate toward a dynamic interactional interactional approach to the study of toward of personality—one personality-one in which situations both influence influence and are influenced influenced by the personalities personalities of of the individuals found in them. The second outcome is that who are found that the developing interest in influences will increasingly personstudying these dynamic, causally reciprocal influences increasingly take person of their laboratories and into the real world world situations in which ality researchers out of individuals live their lives.

ACKNOWLEDGMENT for the preparation of this chapter was supported by funds funds provided by Library research for the Department of Psychology, University of Texas at Arlington.

REFERENCES AUport, G. W. (1931). (1931). What is a trait of personality? Journal of of Abnormal Abnormal and Social Social Allport, Psychology, Psychology, 25, 25,368-372. 368-372. AUport, G. W. (1937). Personality: Personality:AA psychological psychologicalintepretation. intepretation.New NewYork: York:Holt, Holt,Rine RineAllport, hart & Wilson. Allport, AUport, G. W. (1966). Traits revisited. American Psychologist, Psychologist,21, 21,1-10. 1-10. Andrews, J. J. D. W. W. (1967). (1967). The Theachievement achievement motive motiveand and advancement advancement in in two twotypes typesof oforganiza organizations. Journal of of Personality Personalityand andSocial SocialPsychology, Psychology,6,6,163-168. 163-168. Argyle, M. (1980). (1980). The analysis of social situations. In M. Brenner (Ed.), The structure structureof of action. action. New York: St. Martin's Press. Argyle, M., & SL Furnham, Fumham, A. (1982). (1982). The ecology of relationships: Choice of situations as a function of relationship. British British Journal Journalof of Social SocialPsychology, Psychology,21, 21,259-262. 259-262. function Fumham, A., & Graham, J. A. (1981). (1981). Social situations. situations.New NewYork: York:Cambridge Cambridge Argyle, M., Furnham, University Press. Atkinson, J. J. W. W. (Ed.). (Ed.). (1958). (1958). Motives Motivesininfantasy, fantasy,action, action,and and society. Princeton, Van Nossociety. Princeton, NJ:NJ: Van Nos Atkinson, trand. self: A A dialectical dialectical approach. approach. In In L. Berkowitz (Ed.), (Ed.), Advances in Backman, C. C. (1988). (1988). The The self: Backman, L. Berkowitz experimentalsocial socialpsychology psychology(Vol. (Vol.2121San SanDiego, Diego,CA: CA:Academic AcademicPress. Press. experimental 0

188 188

ICKES, IcKES,SNYDER, SNYDER,AND ANDGARCIA GARQA

Bandura, A A. (1982). The psychology of chance encounters and life paths. American American Psycholo Psychologist, gist, 37, 747-755. Bern, D. J. (1981). Assessing situations by assessing persons. In D. Magnusson (Ed.), Toward Toward psychology of of situations: An An interactional perspective perspective (pp. 295-257). Hillsdale, NJ: a psychology Erlbaum. Bern, D. J., & & Allen, A A. (1974). On predicting some of of the people some of of the time: The Psychological Review, Review, 81, 81,506-520. search for cross-situational consistency in behavior. Psychological 506-520. Bern, D. J., & Bem, & Funder, D. C. (1978). Predicting Predicting more of the people more of of the time: Assessing Psychological Review, Review, 85, 485-501. the personality of situation. Psychological Benninger, W. B., & non-college-degree working & Walsh, W. B. (1980). Holland's theory and non-college-degree men and women. Journal Journal of of Vocational Vocational Behavior, Behavior, 17, 18-88. Berglas, S., & & Jones, E. E. (1978). Drug choice as an externalization strategy in response to noncontingent success. Journal Journal of of Personality Personality and and Social Social Psychology, Psychology, 36, 405-417. Bishop, D. W., & & Witt, P. A. (1970). Sources of behavioral variance during leisure time. Journal of of Personality Personality and and Social Social Psychology, Psychology, 16, 352-360. Journal (1968). Some reasons for the apparent inconsistencies of of personality. Psychological Psychological Block, J. (1%8). Bulletin, 70, 210-212. Lives through time. time, Berkeley, CA: Bancroft. Bancroft. Block, J. (1971). Lives & Keyes, S. (1988). Longitudinally foretelling foretelling drug usage in adoles adolesBlock, J., Block, J. H., & environmental precursors. Child Child Development, Development, cence: Early childhood personality and environmental 59, 336-355. Bowers, K. (1973). Situationism in psychology: An analysis and a critique. Psychological Psychological Review, 80, 307-336. Review, of situations: Situation prototypes and person situation proto protoCantor, N. (1981). Perceptions of Toward a psychology psychology of of situations: An An interactional interactional types. In D. Magnusson (Ed.), Toward perspective perspective (pp. 229-244). Hillsdale, NJ: Erlbaum. Cantor, N., & & Kihlstrom, J. F. (1980). Cognitive and social processes processes in personality: Implica Implications for behavior therapy. In C. M. Franks & & G. T. Wilson (Eds.), Handbook Handbook of of behavior behavior therapy. New York: Guilford Press. Cantor, N., & & Mischel, W. (1979). Prototypes in person perception. In L. Berkowitz (Ed.), Advances Advances in experimental experimental social social psychology psychology (Vol. 12 pp. 3-52). New New York: Aca Academic Press. Cantor, N., Mischel, W., & & Schwartz, J. (1982a). A A prototype analysis of of psychological situations. Cognitive Cognitive Psychology, Psychology, 14, 45-77. Cantor, N., Mischel, W., & Schwartz, J. (1982b). Social Social knowledge: Structure, content, use, and abuse. In A A. M. Isen & &A A. H. Hastorf Hastorf (Eds.), Cognitive Cognitive social social psychological, psychological, (pp. 33-72). New York: ElsevierlNorth-Holland. Elsevier/North-HoUand. Cantor, N., Norem, J., & & Langston, C. (1991). Life tasks and daily life life experience. Journal Journal of of Personality, Personality, 59, 425-451. A. (1987). Personality in the life course. Journal Journal of of Personality Personality and and Social Social Psychology, Psychology, Caspi, A 53, 1203-1213. & Bern, Bem, D. J. (1990). Personality continuity and change across the life course. In Caspi, A., & Handbook of of personality: personality: Theory Theory and and research (pp. 549-575). New L. Pervin (Ed.), Handbook York: Guilford Press. Caspi, A Bern, D. J., & interac A.,, Bem, & Elder, G. H., Jr. (1989). Continuities and consequences of of interacJournal of of Personality, Personality, 57, 375-406. tional styles across the life course. Journal Caspi, A, A., & Elder, G. H., Jr. (1988). Emergent Emergent family patterns: The intergenerational intergenerational of problem behavior and relationships. In R. A A. Hinde & & J. Stevensonconstruction of

CHAPTER? CHAmR 7

PERSONALITY AND ANDCHOICE CHOICEOf OFSITUATIONS SITUATIONS PERSONAUlY

189 189

Hinde families: Mutual influences. Oxford, England: Oxford Hinde (Eds.), (Eds.), Relationships Relationshipswithin within families: Mutual influences. Oxford, England: Oxford University Press. Caspi, A., Elder, G. H., Jr., & Bern, D. J. (1987). Moving against the world: Life-course Life-course patterns of explosive children. Developmental DevelopmentalPsychology, Psychology,23, 23,308-313. 308-313. Caspi, A., Elder, G. H., Jr., & Bern, D. J. (1988). (1988). Moving away from from the world: Life-course Life-course patterns of shy children. Developmental , 824-831. DevelopmentalPsychology, Psychology,2424, 824-831. Caspi, A., & Herbener, E. S. (1990). (1990). Continuity and change: Assortative marriage and the Psychology, consistency of personality in adulthood. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 58, 58, 250-258. Caspi, A., Herbener, E. S., & Ozer, D. J. (1992). Shared experiences and and the similarity of of personalities: A longitudinal study of of married couples. Journal of of Personality Personality and and Social Social Psychology, Psychology,62, 62,281-291. 281-291. Christie, R, R., & Geis, F. L. (Eds.). (1970). (1970). Studies Studies in Machiavellianism. Machiavellianism. New NewYork: York:Aca Academic Press. Press. demic Conolley, E. E. S., S., Gerard, Gerard, H. H. B., B., & & Kline, Kline, T. T. (1978). (1978). Competitive Competitive behavior: behavior: A A manifestation manifestation Conolley, of motivation motivation for for ability ability comparison. comparison. Journal Journalof ofExperimental ExperimentalSocial Social Psychology, of Psychology, 14, 14, 123-131. Costa, P. J., McCrae, R R. R, R., & Holland, J. L. (1984). Personality and vocational interests in an adult sample. sample. Journal Journalof ofApplied AppliedPsychology, Psychology,42,42,390-400. 390-400. an R. P., & Norton, G. R R. (1992). (1992). Situations and specific coping Cox, B., Endler, N. S., Swinson, R strategies associated with clinical and nonclinical panic attacks. Behavior Research Researchand and Therapy,330, 67-69. Therapy, 0, 67-69. Crozier, W. R Socialand and R. (1979). Shyness as a dimension of personality. British British Journal Journal of of Social ClinicalPsychology, Psychology,18,18,121-128. 121-128. Clinical Csikszentmihalyi, M., & Figurski, T. J. (1982). (1982). Self-awareness and aversive experience in Journalof ofPersonality, Personality,50,50,15-28. 15-28. everyday life. Journal R. (1981). (1981). Television and the rest of life: A systematic Csikszentmihalyi, M., & Kubey, R comparison of of subjective experience. Public Opinion Quarterly, Quarterly, 45, 45,317-328. 317-328. Diener, E., Larsen, R R. J., & Emmons, R R. A. (1984). (1984). Person x situation interactions: Choice of situations situations and and congruence congruence response response models. models. Journal Journalof ofPersonality Personalityand andSocial Social PsyPsy chology, chology, 47, 47,580-592. 580-592. of Dillehay, R (Eds.), Dimensions of R. (1978). Authoritarianism. In H. London & J. E. Exner (Eds.), personality personality (pp. 85-127). New York: Wiley. Duck, S., & Gilmour, R R. (1981). Personal Personal relationships relationships (4 vols.). London: Academic Press. Duke, M. P., & Nowicki, S., Jr. (1982). (1982). A social learning theory analysis of interactional theory concepts and a multidimensional model of human interaction constellations. psychotherapy. New In In J. J. C. C. Anchin Anchin & & D. D.J.J.Kiesler Kiesler(Eds.), (Eds.),Handbook Handbookof ofinterpersonal interpersonal psychotherapy. New York: Pergamon Press. (1973). Behavior setting utilization by emotionally disturbed college Eddy, G., & Sinnett, E. (1973). students. Journal Journalof of Consulting Consultingand andClinical Clinical Psychology, 210-216. Psychology, 40, 40, 210-216. R. (1967). Social desirability and willingness to participate in a Efran, J. S., & Boylin, E. R group discussion. discussion. Psychological PsychologicalReports, Reports,20,20, 402. group 402. (1992). Studying lives in a changing society: Sociological and Elder, G. H., Jr., & Caspi, A. (1992). Aronoff & personological explorations. In Robert A. Zucker, Albert I. Rabin, J. Aronoff S. J. J. Frank Frank (Eds.), Personality Personalitystructure structureinin course: Essays on personality thethe lifelife course: Essays on personality in thein the S. Murray tradition tradition (pp. (pp. 276-322). 276-322). New New York: York: Springer. Springer.

190 190

ICKES, IcKES,SNYDER, SNYDER,AND AND GARCIA GAROA Elder, G. H., Jr., Caspi, A., & Burton, L. M. (1988). Adolescent Adolescent traditions in development development perspective: Sociological and historical insights In M. Gummas & W. A. Collins (Eds.), Minnesota Minnesota Symposia Symposia on on Child Child Psychology Psychology (Vol. 120, pp. 151-179). Hillsdale, NJ: Erlbaum. Emmons, R. A., & Diener, E. (1986a). A A goal-affect goal-affect analysis of of everyday everyday situational choices. Journal Journal of of Research in Personality, Personality, 20, 309-326. Emmons, R. A., & Diener, E. (1986b). Situation selection selection as a moderator moderator of of response response 1013-1019. consistency and stability. Journal Journal of of Personality Personality and Social Social Psychology, Psychology, 51, 51,1013-1019. Emmons, R. A., Diener, E., & Larsen, R. J. (1985). Choice of of situations and and congruence congruence models of of interactionism. Personality Personality and Individual Individual Differences, 6, 693-702. Emmons, R. A., Diener, E., & Larsen, R. J. (1986). Choice and avoidance of of everyday everyday situations and affect affect congruence: Two models of reciprocal interactionism. Journal Journal of of Personality Personality and Social Social Psychology, Psychology, 51, 815-826. A personality personality model, but not yet a theory. In M. M. Endler, N. S. (1983). Interactionism: A Nebraska Symposium Symposium on Motivation Motivation (pp. 155-200). Lincoln: University University Page (Ed.), Nebraska of Nebraska Press. of Interactionism revisited: "On the role of of situations in personality personality Endler, N. S. (1988). Interactionism {E6&,)yApplications Applications of ofinteractional interactionalpsychol psycholresearch." In S. G. Cole & R. G. Demaree (Eds.), Essays in honor honor of of Saul Saul B. B, Sells (pp. 179-188). Hillsdale, NJ: Erlbaum. Erlbaum. ogy: Essays Interactionism in personality personality in the twentieth century. century. Endler, N. S., & Edwards, J. M. (1986). Interactionism Personality and and Individual Individual Differences, 7, 7, 379-384. Personality Readings in extraversion-introversion: extraversion-introversion: II. II. Fields of of application. application. Lon LonEysenck, H. J. (1971). Readings don: Staples. Personality and learning: The experimental approach. Proceedings, Proceedings, Eysenck, H. J. (1974). Personality Association Association of of Educational Educational Psychologists Psychologists Conference. Conference. Funder, D. C. (1991). Global traits: A A neo-Allportian neo-AUportian approach to personality. Psychological Psychological Science, 2, 31-39. Furnham, Fumham, A. (1981). Personality and activity preference. British British Journal Journal of of Social Social Psychology, Psychology, 20, 57-68. Furnham, Fumham, A. (1982). Psychoticism, social desirability, and situation selection. selection. Personality Personality and and Individual 43-51.. Individual Differences, 3, 43-51 Furnham, A., & Jaspars, JJ.. (1983). The evidence for for interactionism iinn psychology: A critical critical analysis of ferences, of the situation-response inventories. Personality Personality and and Individual Individual Dif Differences, 4, 627-644. ^,627-644. Gangestad, S., & Simpson, J. (1990). Toward an evolutionary history of female sociosexual sociosexual variation. Journal Journal of of Personality, Personality, 58, 69-96. Gangestad, S., & Snyder, M. (1985). "To "To carve carve nature nature at its joints": On the existence of of Gangestad, of personality. Psychological Psychological Review, Review, 92, 317-349. discrete classes of Verbal conditioning, transfer transfer and operant level "speech style" as Gates, D. W. (1971). Verbal functions of of cognitive style (Doctoral dissertation, City University University of New York). functions Dissertation Dissertation Abstracts Abstracts International, 32, 3634B. London & J. E. Exner, Jr. (Eds.), Dimensions Dimensions Geis, F. L. (1978). Machiavellianism. In H. London of personality, personality, (pp. 305-363). New York: Wiley. /I of toward relationships: Choosing a situation in which to begin begin Glick, P. (1985). Orientations toward a relationship. Journal of of Experimental Experimental Social Psychology, Psychology, 21, 544-562. 544-562. Goodenough, D. R. (1978). Field dependence. In H. London & J. E. Exner, Jr. (Eds.), Dimensions Dimensions of of personality, personality, (pp. 165-216). New New York: Wiley.

CHAPTER 77

CHOICEOF OFSITUATIONS SITUATIONS PERSONALITY AND CHOICE

191 191

European Journal Journal of of Social Social Psychology, Psychology, Gorta, A. (1985). Choosing situations for a purpose. European 15, 15, 17-35. Greene, L. R. (1973). Effects of field independence, physical proximity and evaluative feed feedback on affective affective reactions and compliance in a dyadic interaction (Doctoral disserta dissertation, Yale University). Dissertation Dissertation Abstracts Abstracts International, International, 34, 2284B-2285B. Hodges, W. F., & Felling, J. P. (1970). Types of stressful situations and their relation to trait anxiety and sex. Journal Journal of of Consulting Consulting and Clinical Psychology, Psychology, 34, 333-337. Holland, J. L. (1966). The psychology psychology of of vocational vocational choice. Waltham, MA: Blaisdell. Holland, J. L. (1985). Making Making vocational vocational choices: A theory theory of of vocational vocational personality personality and work work environments environments (2nd ed.). Englewood Cliffs, NJ: Prentice Hall. 262-293. Hormuth, S. E. (1986). The sampling of experience in situ. Journal Journal of of Personality, Personality, 54, 54,262-293. Ickes, W. (1982). A basic paradigm for the study of personality, roles, and social behavior. In W. Ickes & E. S. Knowles (Eds.), Personality, Personality, roles, and and social social behavior behavior (pp. 305341). New York: Springer-Verlag. Ickes, W. (1983). Influences of past relationships on subsequent ones. In P. Paulus (Ed.), Basic Basic group group processes processes (pp. 315-337). New York: Springer-Verlag. of interaction in interracial Ickes, W. (1984). Compositions in black and white: Determinants of dyads. Journal Journal of of Personality Personality and Social Social Psychology, Psychology, 47, 330-341. of having an older, opposite-sex Ickes, W., & Turner, M. (1983). On the social advantages of Journal of of Personality Personality and Social sibling: Birth-order influences in mixed-sex dyads. Journal Psychology, 45, 210-222. Psychology, James, W., Woodruff, A., & Werner, W. (1965). Effects of of internal and external control of Consulting Consulting Psychology, upon changes in smoking behavior. Journal of Psychology, 29, 184-186. Jellison, J. M., & Ickes, W. (1974). The power of the glance: Desire to see and be seen in cooperative and competitive situations. Journal Journal of of Experimental Experimental Social Social Psychology, Psychology, 10, 444-450. Jones, E. E., & Davis, K. E. (1965). From acts to dispositions: The attribution process in Advances in experimental experimental social psychology person perception. In L. Berkowitz (Ed.), Advances social psychology (Vol. 2, pp. 219-266). New York: Academic Press. A Jones, E. E., & McGillis, D. (1976). Correspondent inferences and the attribution cube: A New directions directions comparative reappraisal. In J. H. Harvey, W. Ickes, & R. Kidd (Eds.), New attribution research (Vol. 1, pp. 389-420). Hillsdale, NJ: Erlbaum. in attribution Jones, W. H., Russell, D., & Cutrona, C. (1985). {\9%5). A personality personality congruent analysis of of situations. situations. Unpublished manuscript, manuscript. University of Tulsa, Department of of Psychology, Tulsa, OK. Kahle, L. (1980). Stimulus condition self-selection by males in the interaction of of locus of control and skill-chance situations. Journal of of Personality Personality and and Social Social Psychology, Psychology, 38, 50-56. Kahle, L., & Berman, J. J. (1979). Attitudes cause behaviors: A A cross-lagged panel analysis. Journal Journal of of Personality Personality and and Social Social Psychology, Psychology, 37, 315-321. Karp, S. A., Kissin, B., & Hustmyer, F. E. (1970). Field dependence as a predictor of of alcoholic 77-83. therapy dropouts. Journal Journal of of Nervous Nervous and Mental Mental Disease, Disease, 150, 150, 11-%3, Katz, D. (1960). The functional approach to the study of of attitude. Public Public Opinion Opinion Quarterly, Quarterly, 24, 163-204. C , & Sarnoff, I. (1957). The measurement of ego defense defense as related Katz, D., McClintock, c., to attitude change. Journal Journal of of Personality, Personality, 25, 465-474. independence-field dependence and and occupational occupational interests. Unpublished Levy, S. (1969). Field independence-fi eld dependence master's thesis, Cornell University, Ithaca, NY NY.. dynamic theory theory of of personality. personality. New York: McGraw-Hill. Lewin, K. (1936). A dynamic

ICKES, IcKES, SNYDER, SNYDER, AND AND GARCIA GARCIA

192

Lord, C. C. G. G. (1982). Predicting Predicting behavioral behavioral consistency consistency from from an an individual's individual's perception Lord, perception of of 42,42, 1076-1088. situational similarities. Journal Journalof of Personality Personalityand andSocial SocialPsychology, Psychology, 1076-1088. Lundy, J. J. R. R. (1972). Some Some personality personality correlates correlates of of contraceptive contraceptive use use among among unmarried unmarried Lundy, Journalof ofPsychology, Psychology,80, 80,9-14. 9-14. female college college students. students. Journal female Magnusson, D. D. (1981). Problems Problems in in environmental environmental analyses-an analyses—an introduction. introduction. In In D. D. Magnus MagnusMagnusson, son (Ed.), Toward aapsychology psychology of of situations: situations:An Aninteractional interactional perspective, 3-7). son (Ed.), Toward perspective, (pp.(pp. 3-7). Hillsdale, NJ: NJ: Erlbaum. Hillsdale, Erlbaum. Magnusson, D. D. (1988). On On the the role role of of situations situations in in personality personality research: research: An An interactional interactional Magnusson, perspective. In In S. S. G. G. Cole Cole & & R. R. G. G. Demaree Demaree (Eds.), (Eds.), Applications Applications of of interactional perspective. interactional psychology: Erlbaum. psychology: Essays Essaysininhonor honorof ofSaul SaulB.B.Sells Sells(pp. (pp.155-178). 155-178).Hillsdale, Hillsdale,NJ: NJ: Erlbaum. Magnusson, D. (1990). Personality development from an interactional perspective. In L. Pervin (pp. 193-222). New York: personality:Theory Theoryand andresearch research (pp. 193-222). New York: Pervin (Ed.), (Ed.), Handbook of ofpersonality: Guilford Press. Magnusson, Magnusson, D., D., & & Endler, Endler, N. N. S. S. (1977). Personality Personality at the crossroads: crossroads: Current Currentissues issuesinin interactional interactionalpsychology. psychology.Hillsdale, Hillsdale,NJ: NJ:Erlbaum. Erlbaum. Mausner, B., B., & & Graham, Graham,J.J.(1970). (1970).Field Fielddependence dependenceand andprior priorreinforcement reinforcementasasdeterminants determinants Mausner, of social interaction in judgment. Journal Journal of of Personality Personalityand andSocial SocialPsychology, Psychology,16,16,

486-493.

McElroy, c., Morrow, McElroy, J. J. C, Morrow, P. P. W., & & Ackerman, Ackerman, R. R. J. J. (1983). Personality Personality and and interior interior office office Journalof ofApplied Applied Psychology, Psychology, design: the accuracy accuracy of of visitor visitor attributions. attributions. Journal design: Exploring the 68, 68, 541-544. Mehrabian, haracteristic individual Mehrabian, A A. (1978). C Characteristic individual reactions reactions to to preferred preferred and and unpreferred unpreferred envi environments. Journal Journalof ofPersonality, Personality,46,46,717-731. 717-731. Mehrabian, A., & & Russell, J. (1973). A questionnaire measure of arousal seeking tendency.

Environment Environmentand andBehavior, Behavior,5,5,315-333. 315-333.

Milham, J., J., & & Jacobson, Jacobson, L. L. (1978). The The need need for for approval. approval. In In H. H. London London & & J. J. E. E. Exner Exner Milham, (Eds.), Dimensions Dimensionsof ofpersonality, personality,(pp. (pp.365-390). 365-390).New NewYork: York:Wiley. Wiley. Millon, T., T., & & Everly, Everly, G. G. S. S. (Eds.). (Eds.). (1985). Personality Personalityand andits itsdisorders: disorders:AAbiosocial biosocial learning Millon, learning approach. approach. New NewYork: York:Wiley. Wiley. New York: Wiley. Mischel, Mischel, W. W. (1968). Personality Personalityand andassessment. assessment. New York: Wiley. Mischel, W. (1969). Continuity Continuity and and change change in in personality. personality. American AmericanPsychologist, Psychologist,24,24,1012Mischel, W. 1012-

1018. 1018.

Mischel, W. (1977). The The interaction interaction of of person personand andsituation. situation. In In D. D. Magnusson Magnusson & & N. N. S. S.Endler Endler issues in interactional psychology, (pp. (pp. (Eds.), Personality Personalityatatthe thecrossroads: crossroads:Current Current issues in interactional psychology, 333-352). Hillsdale, NJ: NJ: Erlbaum. Morse, J. J. (1975). (1975).Person-job Person-jobcongruence congruenceand andindividual individualadjustment adjustmentand anddevelopment. development.Human Human

Relations, Relations,28, 28,841-861. 841-861.

Nevill, Nevill, D. D. (1974). Experimental Experimental manipulation manipulation of of dependency dependency motivation motivation and and its its effect effect on on Psy eye eye contact contact and and measures measures of of field dependency. dependency. Journal Journalof ofPersonality Personalityand andSocial Social Psy-

chology, 29, 29,72-79. 72-79. chology,

Omoto, M., & Snyder, Snyder, M. M. (1990). Basic Basic research research in in action: action: Volunteerism Volunteerism and and society's society's Omoto, M., & 16,16, 152-165. response to AIDS. Personality Personalityand andSocial SocialPsychology PsychologyBulletin, Bulletin, 152-165. Osborn, Osbom, D. D. R. R. (1988). Personality Personality traits traits expressed: expressed: Interior Interior design design as as behavior-setting behavior-setting plan. Personalityand andSocial SocialPsychology, Psychology,14,14, 368-373. Personality 368-373.

Pervin, L. L. A, A., & & Rubin, Rubin, D. D. B. B. (1967). Student Student dissatisfaction dissatisfaction with with college college and and the the college college Pervin, dropout: A A transactional transactional approach. approach. Journal Journalof ofSocial SocialPsychology, Psychology,72,72, 285-295. dropout: 285-295. Pervin, L. L. A A. (1981). The The relation relation of of situations situations to to behavior. behavior. In In D. Magnusson (Ed.), Toward Pervin, D. Magnusson (Ed.), Toward of situations: situations:An Aninteractional interactional perspective, 343-360). Hillsdale, (pp.(pp. 343-360). Hillsdale, NJ:NJ: a psychology of perspective, Erlbaum. Erlbaum.

CHAPTER 7 CHAPTER?

ANDCHOICE CHOICEOF OFSITUATIONS SITUATIONS PERSONALITY AND

193 193

Price, R. H. (1974). The The taxonomic taxonomic classification classification of of behaviors behaviors and and situations situations and and the the problem problem of behavior-environment Relations, 27, 27,567-585. 567-585. behavior-environment congruence. Human Relations, Price, R. H. (1981). Risky situations. In D. Magnusson (Ed.), (Ed.), Toward of Toward a psychology of situations: perspective, (pp.(pp. 103-112). Hillsdale, NJ:NJ: Erlbaum. situations:An Aninteractional interactional perspective, 103-112). Hillsdale, Erlbaum. Bouffard, D. L. (1974). Behavioral Behavioral appropriateness appropriateness and and situational situational constraint constraint Price, R. H., & Bouffard, , as dimensions of social behavior. Journal Journal of of Personality Personalityand andSocial SocialPsychology, Psychology,3030, 579-586. Rosenberg, M. (1957). Occupations Occupationsand andvalues. values.Glencoe, Glencoe,IL: IL:Free FreePress. Press. B. (1966). (1966). Generalized Generalized expectancies expectanciesfor forinternal internalversus versusexternal externalcontrol controlof ofreinforce reinforceRotter, J. B. ment. Psychological PsychologicalMonographs, Monographs^ 8(?(Whole 609). 80(Whole No.No. 609). personRunyan, M. W. (1978). The life course as a theoretical orientation: Sequences of person of Personality, Personality,46, 46,569-593. 569-593. situation interaction. Journal of (1986). A trait-situational analysis of shyness. Russell, D., Cutrona, C. E., & Jones, W. H. (1986). Shyness: Perspectives Perspectivesononresearch research In W. H. Jones, J. M. Cheek, & S. R. Briggs (Eds.), Shyness: treatment.New NewYork: York:Plenum PlenumPress. Press. and treatment. Secord, P., & & Backman, Backman, C. C. (1961). (1961).Personality Personalitytheory theoryand andthe theproblem problemof ofstability stabilityand andchange change PsychologicalReview, Review,68, 68,21-32. 21-32. in individual behavior: An interpersonal approach. Psychological Secord, P., & Backman, C. (1965). An interpersonal approach to personality. In B. Maher (Ed.), Progress Progressininexperimental experimental personality research (Vol. 2, pp. 91-125). York: personality research (Vol. 2, pp. 91-125). NewNew York: Academic Press. Segal, R B. (1973). Sensation-seeking and anxiety: Assessment of response to specific stimulus Journal of of Consulting Consultingand andClinical Clinical Psychology, 135-138. situations. Journal Psychology, 41, 41, 135-138. Smith-Lovin, L. (1979). Behavior settings and impressions formed formed from social scenarios. Social 42,42, 31-43. Social Psychological PsychologicalQuarterly, Quarterly, 31-43. M. (1979). (1979).Self-monitoring Self-monitoringprocesses. processes.In InL.L.Berkowitz Berkowitz(Ed.), (Ed.),Advances Advancesininexperimental experimental Snyder, M. social social psychology psychology (Vol. (Vol. 12, 12, pp. pp. 85-128). 85-128). New New York: York: Academic Academic Press. Press. influence of individuals on situations. In N. Cantor & J. F. Snyder, M. (1981). On the intluence Kihlstrom (Eds.), and social interaction, (pp.(pp. 309-329). Hillsdale, (Eds,), Personality, Personality,cognition, cognition, and social interaction, 309-329). Hillsdale, NJ: Erlbaum. appearances/private realities: psychology of self-monitoring. private realities: TheThe psychology of self-monitoring. Snyder, M. (1987). Public appearances/ New York: Freeman. selfSnyder, M., & Gangestad, S. (1982). Choosing social situations: Two investigations of self monitoring processes. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 123-135. Psychology, 43,43, 123-135. & Simpson, Simpson,J.J.A. A. (1983). (1983).Choosing Choosingfriends friendsas asactivity activitypartners: partners:The The Gangestad, S., S., & Snyder, M., Gangestad, Journalof ofPersonality Personalityand andSocial Social Psychology, 1061-1072. Psychology, 45,45, 1061-1072. role of self-monitoring. Journal Personality and and social social behavior. behavior. In In G. G. Lindzey Lindzey & & E. E.Aronson Aronson Snyder, M., & Ickes, W. (1985). Personality of social psychology psychology (Vol. (Vol. 2, 2, pp. pp. 883-947). 883-947). New New York: York: Ran Ran(Eds.), The handbook of dom House. Snyder, M., & Kendzierski, D. (1982). Choosing social situations: Investigating the origins of Personality, Personality,50, 50,280-295. 280-295. of correspondence between attitudes and behavior. Journal of Srull, T. K., & Karabenick, S. A. (1975). Effects SruU, Effects of personality-situation locus of control Journalof ofPersonality Personalityand andSocial Social Psychology, 617-628. congruence. Journal Psychology, 32,32, 617-628. Stem, G. G., Stein, M. I., & Bloom, B. S. (1956). Methods Methods in inpersonality personalityassessment assessmentofofhuman human behavior in incomplex complexsocial socialsettings. settings. New York: Free Press. behavior New York: Free Press. Straits, B. C, c., & Sechrest, L. (1963). Further support of some findings about characteristics Journalof of Consulting ConsultingPsychology, Psychology, of smokers and non-smokers. Journal 27,27, 282.282. Swann, W. (1987). Identity negotiation: Where two roads meet. Journal of of Personality Personalityand and Social Psychology, Psychology,53, 53,1038-1051. 1038-1051. Social

194 194

IcKES,SNYDER, SNYDER,AND ANDGARCIA GARQA IcKES, Swann, W., & Hill, c. C. (1982). When our identities are mistaken: Reaffirming Reaffirming self-conceptions self-conceptions 43,43, 59-66. through social interaction. Journal Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, 59-66. W., & Read, S. S. J. (1981). Self-verification Self-verification processes: How How we we sustain our self selfSwann, W., conceptions. Journal Psychology, 17, 17, 351-372. Journalof ofExperimental ExperimentalSocial Social Psychology, 351-372. B. (1992). (1992). Why Why people self-verify. Journal of Swann, W., W., Stein-Seroussi, A., A., & Giesler, R. B. Personalityand andSocial SocialPsychology, Psychology,62,62,392-401. 392-401. Personality Swann, W., W., Wentzlaff, R. M., & & Krull, D. D. S. S. (1992). Allure of negative feedback: Self Selfverfication strivings among depressed persons. of Abnormal Abnormal Psychology, Psychology, 10, 10, verfication persons. Journal of 293-306. Self-evaluation maintenance processes: Implications for relationships and Tesser, A. (1984). Self-evaluation development. In J. Masters & K. Yarkin (Eds.), Boundary areas areasof ofpsychology: psychology:Social Social New York: Academic Press. and development. development New Tesser, A. (1985). Some effects effects of self-evaluation self-evaluation maintenance on cognition and action. In R. M. M. Sorrentino & & E. E. T. T. Higgins (Eds.), The The handbook handbookof ofmotivation motivationand andcognition: cognition: York: Foundations Foundationsof ofsocial socialbehavior. behavior.New New York:Guilford GuilfordPress. Press. A. (1988). Toward a self-evaluation maintenance model of social behavior. In L. Tesser, A. Berkowitz (Ed.), Advances in in experimental experimentalsocial socialpsychology psychology(Vol. (Vol.21). 21).New NewYork: York: Academic Press. Tesser, A., & & Campbell, J. (1980). Self-definition: Self-definition: The The impact of the relative performance performance and similarity of others. Social Psychology Psychology Quarterly, Quarterly,43, 43,341-347. 341-347. Self-evaluation maintenance and the the perception perception of of friends friends Tesser, A., & Campbell, J. (1982). Self-evaluation and strangers. Journal Journal of ofPersonality, Personality,59,59,261-279. 261-279. Tesser, A., A., & & Campbell, J. (1983). Self-definition Self-definition and self-evaluation self-evaluation maintenance. In J. & A. Greenwald (Eds.), Research Research on onmotivation motivationinineducation: education:The The classroom Suls & classroom New York: Academic Press. milieu. New milieu. Tesser, A., Campbell, J., & & Smith, M. M. (1984). (1984). Friendship choice and and performance: Self Selfevaluation maintenance in in children. Journal Journal of Personality Personality and andSocial SocialPsychology, Psychology, 46, 46, 561-574. Tesser, A., A., & & Moore, J. (1987). On the convergence of public and private aspect of self. In R. Baumeister (Ed.), Public self self and andprivate privateself self(pp. (pp,99-116). 99-116).Berlin: Berlin:Springer-Verlag. Springer-Verlag. Tesser, A., A., & Paulhus, D. D. (1983). The The definition of self: Private and public self-evaluation self-evaluation Psychology, 44,44, 672-682. management strategies. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 672-682. Tesser, A., A., & Smith, J. (1980). Some effects effects of friendship and task relevance on helping: one you you like. Journal Journalof ofExperimental ExperimentalSocial Social Psychology, You don't always help the one Psychology, 16, 16, 582-590. V. H. (1964). Work and motivation. motivation. New New York: Wiley. Vroom, V. Wachtel, P. L. (1973). Psychodynamics, behavior therapy, and the implacable experimenter: of Abnormal Abnormal Psychology, An inquiry into the consistency of personality. Journal of 82, 324-334. 82, Walsh, W. A., & W. B., Horton, J. A., & Gaffey, R. L. L. (1977). (1977). Holland's theory and college degreed workingg men and women. Journal Journalof ofVocational VocationalBehavior, Behavior, 180-186. 10,10, 180-186. workin Wankowski, J. A. A. (1973). (1973). Temperament, Temperament,motivation, motivation, and academic achievement Birmingham, Wankowski, and academic achievement. Birmingham, AL: University of Birmingham Educational Survey and Counselling Unit. & Cheek, J. M. M. (1986). Shyness situations: Perspectives of a diverse sample K., & Watson, A. K., of shy females. Psychological PsychologicalReports, Reports,59, 59,1040-1042. 1040-1042. & Baumal, Baumal,E. E. (1967). (1967).Effects Effects of oflocus locusof ofcontrol controland andexpectation expectationof offuture futurecontrol control Watson, D., & upon present present performance. performance. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 6, 212-215. upon Psychology, 6, 212-215.

CHAPTER CHAPTER 7 7

PERSONALITY PERSONALITY AND AND CHOICE CHOICE OF OF SITUATIONS SITUATIONS

195 195

D. M., & Giuliano, T. (1982). (1982). The forms of social awareness. In W. Ickes & E. S. Wegner, D. Knowles (Eds.), Personality, Personality, roles, roles,and andsocial socialbehavior behavior(pp. (pp.165-198). 165-198).New New York: York: Springer-Verlag. Wheeler, L., Reis, H. T., & Nezlek, J. (1983). (1983). Loneliness, social interaction, and sex roles. Journal of Personalityand andSocial SocialPsychology, Psychology,45, 45,943-953. 943-953. of Personality Wilson, G. D., & Nias, D. K. B. (1975). Behavior, 2, 330-332. (1975). Sexual types. New Behavior, Winter, D. G. (1973). The The power motive. motive. New York: Free Press. Winter, D. G., & Stewart, A. J. (1978). The power motive. In H. London Lx)ndon & J. E. Exner (Eds.), Dimensions of of personality, personality, (pp. (pp.391-447). 391-447).New NewYork: York:Wiley. Wiley. Winter, H. A., Moore, C. A., Goodenough, Goodenough, D. R., & Cox, Cox, P. (1977). (1977). Field-dependent Field-dependent and andfield fieldindependent cognitive styles and their educational implications. Review Review of of Educational Research, Research,47, 47,1-64. 1-64. Wish, M. (1975). Subjects' expectations about their own interpersonal communication: A multidimensional approach. Personality 1, 1, 11-20. Personalityand andSocial SocialPsychological PsychologicalBulletin, Bulletin, 11-20. (1976). Perceived dimensions of of interpersonal rela relaWish, M., M., Deutsch, M., M., & Kaplan, S. J. (1976). of Personality Personalityand andSocial SocialPsychology, Psychology,33,33,409-420. 409-420. tions. Journal of (1977). Toward an implicit theory of interpersonal communication. Wish, M., M., & Kaplan, S. (1977). Sociometry, 40, 40, 234-246. 234-246. Sociometry, (1964). A fear survey schedule for for use in behavior therapy. Behavior Wolpe, J., & Lang, P. J. (1964). Research and and Therapy, Therapy,2,2,27-30. 27-30. Research B. Maher (Ed.), (Ed.), Progress Progressininexperi experi(1974). The The sensation sensation seeking seeking motive. motive. In In B. Zuckerman, M. (1974). personalityresearch research(Vol. (Vol.7,7,pp. pp.79-148). 79-148).New NewYork: York:Academic AcademicPress. Press. mental personality Sensation-seeking. In In H. H. London London & J. J. E. Exner, Jr. Jr. (Eds.), (Eds.), Dimensions Dimensions Zuckerman, M. (1978). Sensation-seeking. of personality, personality, (pp. (pp. 487-559). New York: Wiley. of

This Page Intentionally Left Blank

PART III PART in

DEVELOPMENTAL ISSUES

This Page Intentionally Left Blank

CHAPTER S8 CHAPTER

STAGES OF OF PERSONALITY DEVELOPMENT LOEVINGER JANE LOEVINGER WASHINGTON WASHINGTON UNIVERSITY UNIVERSITY

One of the first and field of of the first and still greatest achievements of psychologists in the the field individual differences was Binet's breakthrough toward the measurement of intelli intelliFor Binet, in order to measure aa child's intelligence, one held him him or her her gence. For the average adult as as aa standard, and and determined how how far he or she up against the she the adult in in achievement. There have been many refinements, but approached the but by using the the achieved adult status as the measure intelligence is still measured by as the and standard. Piaget broke with that tradition. For him, the the child's mistakes, which did not not seem to interest Binet, were just what was was interesting. When the child achieved adult status, Piaget lost interest. By studying the the child's mode of thought via the the evolution of errors, Piaget discerned stages in in the development of intelligence: sensorimotor, preoperational, concrete operational, and and formal operational. Piaget's and point of view, his his search for the the sense in the child's nonsense, have methods and influenced much of contemporary psychology. L MEASUREMF;NT MEASUREMENT OF OF PERSONALITY PERSONALITY I.

Success in measuring intelligence, particularly in Binet-type tests, has led psycholo psychologists to believe believe that that they they can can also also measure measure personality by by similar techniques. techniques. In In this this is not not like intelligence in its formal properties, they have gone astray. Personality is the differences all all favor the measurability of intelligence and the resistance resistance of and the any easy measurement. personality to any IIAIVDBOOK HANDBOOK OF OF PERSONALITY PERSONAUTY PSYCHOLOGY PSYCHOLOGY

199 199

CopYRiGirr C © 1997 1997 IY BY ACADEMIC ACADEMIC Puss PRESS.. CoPYRIGHT ALL RIGHTS OF REPRODUcnON IN ANY FORM RESERVED.

ALL RIGHTS OF REPRODUOION IN ANY FORM RESERVED.

200 200

JANE JANE LoEVlNGER LOEVINGER

disanalogies broaden the discussion The dis analogies between personality and abilities (to brQaden beyond the controversial topic of of "intelligence") "intelligence*') are just those qualities quaUties that render intelligence and other abilities relatively measurable. In the first place, abilities tend to increase with age, monotonically monotonically during the ages when their measurement is most studied, that is, during youth and early adult life. In the second place, all abilities tend to be positively correlated. That is, abilities may be correlated posi positively or not at all, but one ability will rarely interfere with another one. Personality traits, on the other hand, may have any relation with one other: positive, negative, or curvilinear. Personality traits often, perhaps typically, increase with time up to some consepoint, and then tend to decrease. This curvilinear relation with age has as its conse personality traits. Two personal personalquence possible curvilinear relations with other personality manifestations of of the same underlying developmental continuum ity traits that are manifestations can have any correlation whatsoever, provided that they maximize at different different stages (Loevinger, 1984). The ensuing complexities are great. Because the tools of the psychometrician are almost entirely based on linear hypotheses, psychometric or statistical solutions to the riddle of of personality structure are quite ineffective, ineffective, no matter how potent they are in the study of abilities. To be sure, there have been a number of of effective effective studies of of personality using strictly psychometric approaches, and even some modest but real predictions of particuparticular outcomes. But for the above reasons, there are also limitations, particu larly when one is seeking the fundamental fundamental structure of of personality rather than specific predictions. n. PERSONALITY STAGES AND TYPES D.

A A totally different different approach to the structure of personality is based on the idea of personality stages and types. The idea that there are types of of people and stages in the development development of of personality is as old as recorded history and older than any attempts at measurement. For example, stages can be found in the Bhagavad Bhagavad Vita and types of of people in the Iliad. However, many of the older typologies have been discredited by modern psychology. For one thing, they were interpreted as arguing for a discontinuity between types. All the evidence lies in the other direction, favoring continuous variation. different from a trait, there must be some evidence For a type to be anything different that diverse, separable traits or aspects vary more or less together. Similarly, for stages, there must be various aspects that change more or less simultaneously. inteUigence Synchronous development has come under criticism, even in relation to intelligence (Fischer, 1980); even Piaget (1972) does not hold to an extreme claim. Indeed, Piaget (1932) never asserted that there are stages in the moral domain, which is about as close as he came to personality. For a type or stage theory to be plausible, there must be some logical or at least intelligible structure to the diverse elements constituting a type or stage. That

CHAPTER CHAPTER 88

PERSONALITY PERSONALITY DEVELOPMENT DEVELOPMENT

201 201

is the feature that theorists have delighted in demonstrating. Their arguments, however, are mainly theoretical. Without empirical evidence, other psychologists are not easily convinced of the force of their arguments.

i n . KoHLBERG: STAGES OF OF MORAL MORAL JUDGMENT JUDGMENT m. KOHLBERG: STAGES In relation to personality, stage theories in recent years have been dominated by (1969). His theory of of a neo-Piagetian school whose leader was Lawrence Kohlberg (1969). stages of development of moral judgment is well known. The distinctive feature of of his theory is that knowledge of the highest or end state is used to define previous states, a feature borrowed from from Piaget's theory of intelligence. That is not not the same, same, however, as Binet's definition definition of intelligence in terms of percentage of adult status achieved. Kohlberg, like Uke Piaget, allows for a more dialectical course of evolution. The logical endpoint is necessary as the guiding principle of stage order. Many followed this aspect of the Kohlbergian investigators studying related variables have followed neo-Piagetian methodology. Kohlberg and many other developmental theorists have used story completion methods to study personality development. Typically they have used stories present presenting a moral dilemma. The stage assigned to a subject depends not primarily on the for the dilemma but rather on the structure of the subject's reason reasonsolution chosen for ing. This method has the advantage that the interviewer can press the subject to clarify his meaning when there is an ambiguous response. The disadvantage is that which attends all interview methods: it is time consuming and likely to be somewhat forms of the unreliable in both its administration and its scoring. When written forms interview are used, both the advantages and the disadvantages of the interview are foregone. IV. IV. A A STAGE.TYPE STAGE-TYPE THEORY: THEORY; EGO EGO DEVELOPMENT DEVELOPMENT personalIn studying ego development as a broad aspect (or collection of aspects) of personal followed a different different methodology. In the first place, ity, my colleagues and I have followed 1966). To that I have argued that types represent the trace of stages (Loevinger, 1966). extent, I am following Binet in in taking advantage of age changes as a partial criterion for personality development as a dimension of individual differences. differences. Taking stages as a major major clue or criterion allows the use of age as a guiding some principle; this has the merit that age can be measured almost without error, something that cannot be said for any general facet of personality. Thus the empirical approach begins by looking for for average age changes. Then, by observing which changes occur together in some diverse sample, one has a clue as to what can reasonably be called personality types. after all, Types can, in turn, be used to refine our conception of stages. For, after many kinds of development occur simultaneously, and some art is required to

202 202

JANE LOEVINGER JANE LOEVINGER

discern what elements go together to define a stage. Height is not an element of ego development, even though both increase together in time. The two variables are separated logically, of of course, and empirically by their lack of of correlation within any age cohort taken by itself. Kohlberg and the neo-Piagetians also use age changes differences as their guide to stages, but they are less clear about studying individual differences with age held constant. Although Kohlberg and his fellow fellow neo-Piagetians have used age changes, they have taken logical coherence as their ultimate criterion for stage structures. The study of of ego development, by contrast, has been built around an empirical, partly compsychometric approach, seeking the common thread in stages and types. This com mon element defines ego development as something more than an age progression of individual differences. The and something more than a one-dimensional trait of content of of those stage types is best denoted by describing the stages. To study ego development, our major instrument has been the Sentence & Wessler, 1970; Loevinger, Wessler, & & Red RedCompletion Test (SCT) (Loevinger & more, 1970). The SCT has proven to be well adapted to the task of refining the definition of the stages in accord with data from thousands of cases. The theory of ego development has been criticized because it does not present each stage as a logically coherent whole, it does not prescribe higher stages as & Zahaykevich, 1977; better, and it "confuses" structure with content (Broughton & Kohlberg, 1981). These criticisms are approximately correct, factually; the question is whether these are strengths or weaknesses of the method. The root difference difference between the SCT method and the method of of the neo neoof the SCT method on masses of of data. Because the Piagetians is the reliance of definition of stages is derived from data, it cannot also be guaranteed to be strictly logical, any more than people are strictly logical. Stages are characterized in terms of of both structure and content of of the subjects' responses, because both vary with ego level, sometimes structure and sometimes content of thought supplying the best clue as to level. Methods have been evolved for parlaying responses of subjects at high levels to reveal the next higher level. But this too is empirical empirical and does not give us license to declare that the highest level is the "best." The psychological coherence of of the resulting stage pictures is impressive and intuitively evident to most of of those who have studied the material. This constitutes a kind of validation of the conception of ego development and of the SCT method. The stages of of moral development delineated by Kohlberg (Colby & & Kohlberg, 1986) and the stages of of ego development (Table I) are very similar at the lower stages; at the higher stages it is harder to draw exact parallels. The same comments hold for many other developmental stage theories propounded in recent years (Loevinger, 1976). V. STAGES OF EGO DEVELOPMENT V.

The lowest stage or stages are inaccessible to those whose research is based on of word usage and sentences. The foundations of ego development are lost analysis of

CHAPTER 8

203 203

PERSONALITY DEVELOPMENT DEVELOPMENT TABLE TABLE II

Some Characteristics of Levels Levels of Ego Development Characteristics Level

Impulse control

Interpersonal mode

Conscious preoccupation

Impulsive

Impulsive

Egocentric, dependent

Bodily feelings

Self-protective Self-protective

Opportunistic Opportunistic

Manipulative, wary

"Trouble," control

Conformist Conformist

Respect for rules

Cooperative, loyal

Appearances, behavior

Self-aware

Exceptions allowable

self-aware Helpful, self-aware

adjustment Feelings, problems, adjustment

Conscientious

Self-evaluated standards, Self-evaluated self-critical

Intense, responsible

Motives, traits, traits, achievements achievements Motives,

Individualistic

Tolerant

Mutual

Individuality, development, roles

Autonomous

conflict Coping with conftict

Interdependent Interdependent

Self-fulfillment, psychological Self-fulfillment, causation

Note: Adapted from from Loevinger (1976,1993c, & elsewhere). Note: (1976, 1993c, &

in the mists of infancy. The earliest eariiest stage that can be measured is the Impulsive Stage. Stage. At this stage the small child has achieved awareness of self as a separate person. Impulses are a kind of verification of that separateness, but the child has impulses. The emotional range is narrow. In small children at first first no control of the impulses. charming; when it persists into adolescence and adulthood, it is at best this stage is charming; maladaptive and in some cases psychopathic. At the next stage the child is aware of impulses as such, enough to exert some control in order to protect himself or herself and to secure at least immediate advantage. This Thisisisthe theSelf-Protective Self-Protective Stage. Stage.One Onemust mustalmost almostuse useaadifferent differentvocabu vocabulary in in describing describingits itsmanifestations manifestationsin inchildhood childhoodand andininadult adultlife, Ufe,though thoughininprinciple principle the core is the same. In small children there is a natural dependence, egocentricity, and calculation of advantage for self. The small child's love of ritual is probably part of the early effort effort at atself-control. self-control. In Inadolescents adolescentsand andadults adultsone onesees seesexploitation exploitation of other people, taking advantage, preoccupation with being taken advantage of, of, hostile humor, and related traits. W. C. C. Fields, or the characters he portrayed, is the perfect exemplar. The emotional range is limited and the conceptual range simple. Normally this stage is outgrown in childhood or adolescence. However, unlike unUke the Impulsive Stage, a person who remains at the Self-Protective Stage can sometimes become a big success in adolescence and adult Hfe. life. Stage; probably no stage or type has been The next stage is the Conformist Stage; so often described as the Conformist. The Conformist has progressed to the point

204 204

JANE LoEVINGER LOEVINGER lANE

of of identifying identifying self self with the group, however wide that group may be: family in childhood, and later peer group, school, and so on. Thinking tends to be in terms of of stereotypes. The emotional range is limited to standard cliches and banalities banalities— on—but it is wider than in previous stages. happy, sad, mad, glad, and so on-but The next level that has been thoroughly studied with the scr SCT is called the Self-Aware Self-Aware Stage. (Previously (Previously this was called a transitional level, Conscientious ConscientiousConformist, but it appears to be a stable position). This is perhaps the level that has been least often described previously, though it is the modal stage in late adolescence and adult life, at least in urban United States (Holt, 1980). The person at this level has gone beyond the simplified rules and admonitions of the Conformist convento see that there are allowable contingencies and exceptions, described in conven tional and broad terms. Although still basically a conformist, the person at this professed level is aware that he or she does not always live up to the group's professed standards. There is a wider emotional and cognitive range. There is greater aware awareof self self as separate from the group, sometimes leading to a characteristic ness of loneliness or self-consciousness. That does not mean, however, that persons at this level are any less adjusted than those of other levels. levels. Every stage has its its own strengths and weaknesses (which is why it is erroneous to refer to this continuum strengths as "ego "ego strength," as is sometimes done). At the next or Conscientious Stage, the person lives by his or her own ideals and standards, rather than merely seeking group approval. By this stage the person has acquired a richly differentiated differentiated inner life, with a wide vocabulary to express cognitively shaded emotions. Older adolescents and adults at earlier stages do not lack the vocabulary to understand such cognitively cognitively shaded emotions, but they are longnot as likely to use such terms spontaneously. By this stage, the person has long term goals and ideals. The elements of a mature conscience are present; in fact, the person may be oppressively conscientious. The level beyond the Conscientious Stage is termed the Individualistic Stage. At this level one begins to see an awareness of of paradoxes and contradictions in life. Persons become aware of of development development as a process and of of its place in their own life; they think in terms of of psychological causation; they take a broad view of of life as a whole. At the Autonomous Autonomous Stage there is a further development of of the characteristics that appear at the Individualistic level. Where the Self-Protective Self-Protective person uses hostile humor, the Autonomous person will often often display a kind of of existential humor, seeing the irony in life situations. Rather than seeing situations in in terms of diametri diametrically opposite choices of of good and bad, there is is awareness of of the multifaceted multifaceted complexity of of situations and life choices. Above all, there is respect for other persons and their need for autonomy, even people such as one's own children for whom one has some responsibility. To that extent there is a lessening of of the overburdened of persons at the Conscientious Stage. At the same time, there is a conscience of of wider social concerns. growing tendency to see one's own life in context of The theoretical highest stage, Integrated, is rarely seen in random samples from the general population. At this stage the characteristics of the Individualistic

CHAPTER88 PERSONALITY PERSONALITYDEVELOPMENT DEVELOPMENT CHAPTER

205 20S

and Autonomous Stages are more fully developed. In addition, there is some capacity to integrate the vital concerns of of one's own life with those of the wider society. A A good characterization of of this stage is Maslow's (1954) description of the Self-Actualizing Self-Actualizing person. Assuming that the names of stages are merely markers on what is a continuous line of of development-and evidence-one may ask how the development—and there is no contrary evidence—one lines between the stages are drawn. For ego development, the rule has been that vocabulary is needed to describe the characteristics of the stage, a new when a new vocabulary stage has appeared. Because our stage descriptions are arrived at and perfected perfected empirically, the descriptions are richest in the range where the most people are found, that is, at the level of of the Conformist, the Self-Aware, and the Conscientious Stages. Paradoxically, the greatest interest is often in the earliest and the highest stages. The earliest stages are important because when they persist into adult life, fascination for they are maladaptive. The highest stages, despite their rarity, hold a fascination many students in the field, perhaps because we all aspire to be such a person, or at least to see ourselves as such a person. Nothing would be more deceptive than to define the lowest stages and the highest from very low to very highest stages and encourage the inference that the path from high is a straight line progression. We do have some access to the lowest lowest stages by observing youngsters and by observing some persons who have trouble making it in society. The highest stages are more problematic, in part because, in principle, people (presumably including psychologists) can understand stages already passed through much more clearly than stages still beyond their own attainment (Rest, Turiel, & Kohlberg, 1969). It is only a slight exaggeration to say that one cannot understand fully fully the stages beyond one's own. Reviewing many expositions of of stages of of personality development (including Reviewing development and related variables), I have the impression that every theorist moral development of the highest stage a kind of of apologia apologia per per vita sua. projects onto his picture of developTo address this problem rather than to perpetuate it, the study of ego develop ment by means of of the SCT has used data to bootstrap the theory into its highest region. Although Although Kohlberg and other theorists undoubtedly used data informally to construct their theories, they did not have a formal technique for utilizing data in the refinement of of the stage types. That has been the distinctive contribution of of the ego development research method (Loevinger, 1993b, 1993c). VI. SOURCES OF ERROR VI.

Although the stages are empirically grounded, that does not make them infallible of personality development, for many reasons. The obvious reason is descriptions of that personality is much too complex to yield all its secrets to any single method. Indeed, many psychologists and psychoanalysts believe that only clinical and depth depthpsychological methods can yield important insights; our work, among others, proves otherwise.

JANE JANE LoEVINGER LOEVINGER

206 206

The weakness of of the method lies elsewhere, in confounding confounding factors. The universe, including the universe of of interpersonal interpersonal relations, does not present itself itself to us as a set of orthogonal variables. In life ego development development is intertwined intertwined with socioeconomic status, intelligence, and verbal ability. A A relatively relatively intelligent intelUgent person of of relatively high socioeconomic status with good good verbal ability will probably probably have One a higher level of ego development than a person less blessed in those respects. One can find enough exceptional individuals, either higher or lower lower than than one would higher or of such demographic variables, to guarantee concept guarantee that the concept expect on the basis of of of ego development development is not dispensible because of of those correlations in nature. But appears empirically empirically to be a sign the correlations do pose a problem, because what appears of ego level may in fact fact be a sign of of another another associated associated variable, such such as intelligence, of guarantee otherwise. Because social class, or verbal fluency. There is no way to guarantee of the long history history of of psychological such correlations are to be expected on the basis of interpreted as validation of of the developmental developmental continuum. testing, they can be interpreted hard-boiled psychometrician, however, they they are not of the success of of not signs of To a hard-boiled convergent validity so much as the failure failure of of discriminative discriminative validity validity (Campbell (Campbell & & convergent Fiske, 1959). Granted that that there are confounding confounding variables that make the mea meaGranted that will always make surement of of ego development development subject subject to error, just just as all other other personality measuresurement personality measure for psychological psychological theory theory from from this study, ments are, there still is a substantive gain for for no matter matter how long one stared at intelligence, social social status, and and verbosity verbosity data, for description of of these stages stages of development one would never come up with the rich description of development of which only a small glimpse has been given here. This is the case for for the study of of ego development development by the sentence completion completion method, particularly particularly when viewed of psychometrically based criticism. against psychometrically V n .. SSTAGES T A G E S VERSUS FACfORS FACTORS VII of trait theories have claimed In recent years some proponents of claimed the entire field of of measurement as their purview, and, in in particular, they have reduced personality measurement personality to five factors: extraversion, agreeableness, conscientiousness, neuroti neurotiintellect (Goldberg, cism, and a fifth factor factor called, among other things, openness or or intellect 1990; McCrae & & Costa, 1990). There are many differences differences between between the views of taken by the There of personality personality taken factor theorists and those taken taken by stage theorists in development in general, and and ego development factor particular (Westenberg & & Block, 1993). For For example, stage theories theorists in particular development is largely ignored ignored by are based on tracing development, whereas development factor-trait theorists. Developmental Developmental theories allow allow for somefactor-trait for dialectical growth, some of factor-traits. factor-traits. thing hardly admissable in the linear logic of of the most striking differences differences is that the factor-trait called conscientious conscientiousOne of factor-trait called ness confounds confounds characteristics of of the Conformist Conformist and the Conscientious stages. Most ness adults studied by psychologists, for for example, college students, fall fall within this range. A distinction distinction like that between the Conformist Conformist and the Conscientious person person is A

CHAPTERS8 CHAPTER

PERSONALITY DEVELOPMENT PERSONALITY

207 207

recognized recognized by most stage theories. This difference, difference, and the whole dimension of of ego development to which it gives access, apparently apparently is hidden from from factor factor theory, at development least in the current current five-factor five-factor and related related versions (Loevinger, 1993a). Traditional psychometric methods, based on objectively objectively scored paper-and paper-andpencil (or computerized) computerized) tests, easily adaptable to factorial factorial methods, will always play a large part part in psychological approaches to personality. I do not claim that a consensus is emerging in favor favor of of stage theories, nor nor do I predict predict that it ever will; nor do I wish it so. For, as Mill (1859) said, "He "He who knows only his own side of of nor little of of that." Stage and type theories will remain a rich and the case, knows little indispensable source source of of insights into personality personality structure. indispensable

REFERENCES (1977). Review Review of of J. J. Loevinger's Loevinger's Ego Ego development: development:Concep ConcepBroughton, J., & Zahaykevich, M. (1977). theories. Telos, Telos,32, 32,246-253. 246-253. tions and theories. Campbell, D. T., & Fiske, D. W. (1959). Convergent and discriminant validation by the PsychologicalBulletin, Bulletin,56, 56,81-105. 81-105. multitrait-multimethod matrix. Psychological multitrait-multimethod & Kohlberg, Kohlberg, L. L. (1986). (1986). The Themeasurement measurementof ofmoral moraljudgment. judgment New NewYork: York:Cambridge Cambridge Colby, A., & University Press. Fischer, K. W. (1980). A theory of cognitive development: The control and construction of of hierarchies of skills. PsychologicalReview, Review,87, 87,477-531. 477-531. skills. Psychological factor Goldberg, L. R. (1990). An alternative "description of personality": The Big-Five factor of Personality Personalityand andSocial SocialPsychology, Psychology,59, 59,1216-1229. 1216-1229. structure. Journal of Holt, R. R. (1980). Loevinger's measure of ego development: Reliability and national norms of Personality Personality and andSocial SocialPsychology, Psychology,39, 39, for male and female short forms. Journal of 909-920. Stage and and sequence: sequence:The Thecognitive-developmental cognitive-developmental approach approach to tosocializa socializaKohlberg, L. (1969). Stage tion. In D. A. Goslin (Ed.), Handbook of of socialization socialization theory theoryand andresearch research(pp. (pp.347347480). 480). Chicago: Rand McNally. Kohlberg, L. (1981). The meaning and andmeasurement measurementof ofmoral moraldevelopment. development.Worcester, Worcester,MA: MA: Clark University Press. Loevinger, J. (1966). The The meaning meaning and and measurement measurement of of ego ego development. development. American AmericanPsychol Psychologist, ogist, 21, 21,195-206. 195-206. Loevinger, J. (1976). Ego and SanSan Francisco: Jossey-Bass. Ego development: development:Conceptions Conceptions andtheories. theories. Francisco: Jossey-Bass. Loevinger, J. (1984). On the self self and predicting behavior. In R. A. Zucker, J. Aronoff, & A. I. Rabin (Eds.), Personality Personalityand andthe theprediction predictionof ofbehavior behavior(pp. (pp.347-480). 347-480).Orlando, Orlando, FL: FL: Academic Press. Loevinger, J. (1993a). Conformity Conformity and conscientiousness: One factor factor or two stages? In D. C. Funder, R. D. Parke, C. Tomlinson-Keasey, & K. Widaman (Eds.), Studying Studyinglives lives through time: time:Personality Personalityand anddevelopment development (pp. 189-205). Washington, DC: American through (pp. 189-205). Washington, DC: American Psychological Association. Psychological Loevinger, J. (1993b). Ego development: Questions of method and theory. Psychological Inquiry, 4, 4, 56-63. Inquiry, of personality: True or false. Psychological PsychologicalInquiry, Inquiry, Loevinger, J. (1993c). Measurement of 4, 1-16. 1-16. 4,

208

JANE LoEVINGER LOEVINGER lANE

& Wessler, R. (1970). Measuring Measuringego egodevelopment. development.1. LConstruction Construction Loevinger, J., & andand useuse of of San Francisco: Jossey-Bass. a sentence sentencecompletion completiontest. test, San Francisco: Jossey-Bass. Loevinger, J., Wessler, R, R., & & Redmore, Redmore, C. C. D. D. (1970). (1970).Measuring Measuringego egodevelopment. development2.2.Scoring Scoring manual manualfor for women womenand andgirls. girls.San SanFrancisco: Francisco:Jossey-Bass. Jossey-Bass. Maslow, A. H. (1954). Motivation Row. Motivation and andpersonality. personality.New NewYork: York:Harper Harper&& Row. McCrae, R R. R., & & Costa, Costa,P. P.T., T.,Jr. Jr.(1990). (1990).Personality Personality adulthood. New York: Guilford Press. inin adulthood. New York: Guilford Press. Mill, J. S. (1859). Of the liberty of thought and discussion. On liberty liberty (Chap. 2). The moral moraljudgment judgmentof ofthe thechild. child. New York: Free Press. Piaget, J. (1932). The New York: Free Press. Piaget, J. (1972). (1972). Intellectual Intellectual evolution evolutionfrom from adolescence adolescence to to adulthood. adulthood.Human HumanDevelopment, Development, Piaget, J. 15, 15, 1-12. Rest, J. R, R., Turiel, E., & & Kohlberg, L. (1969). Level of moral judgment as a determinant Journal of of of preference and comprehension of moral judgments made by others. Journal Personality, Personality,37, 37,225-252. 225-252. & Block, J. (1993). Ego development and individual differences in Westenberg, P. M., & 65.65. 792-800. personality. Journal Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, 792-800.

CHAPTER 9

THE EMOTIONAL OF EARLY BASIS OF PERSONALITY DEVELOPMENT IMPLICATIONS IMPLICATIONS FOR FOR THE THE EMERGENT EMERGENT SELF-CONCEPT SELF-CONCEPT

REBECCA REBECCA A. A. EDER EDER HOSPITAL Sr. ST. loUIS LOUIS CHILDREN'S CHILDREN'S HOSPITAL SARAH C. MANGELSDORF MANGELSDORF SARAH UNIVERSITY OF ILLINOIS, URBANA-CHAMPAIGN URBANA-CHAMPAIGN UNIVERSITY OF ILLINOIS,

I. THE THE EMOTIONAL EMOTIONAL BASIS BASIS OF OF EARLY EARLY PERSONALITY PERSONALITY DEVELOPMENT DEVELOPMENT 1

In recent years, years, personality personality psychologists psychologists have have made made aa great great deal deal of of progress progress toward toward advancing our our understanding of the the structure of personality. For example, investiga investigators have suggested that the structure of personality can be best described in terms :caUed *'Big of the "Big Five" traits of openness, conscientiousness, extraversion, the so so-called 1985, 1987). Despite these agreeableness, and neuroticism (see McCrae & & Costa, 1985,1987). advances in personality research, little is known about the emergence and develop developmental course of those traits described as characteristic of adult personality. We believe that the common thread between early personality development and later adult personality lies in individual differences in emotionality. We We propose that personality development in thefirst first 8 years of life occurs in five five basic steps in in which the child progresses (1) from organized patterns of behaviors (probably due HANDBOOK OF PERSONALII1' PSYCHOLOGY

HANDBOOK OF PERSONALITY PSYCHOLOGY

209

209

CoPYIIGHT C COPYRIGHT © 1997 1997 IY BY ACADEMIC ACADEMIC PRESS. PRESS. ALL RIGHTS OF REPRODUcnON IN ANY FOIM RESERVED.

ALL RIGHTS OF REPRooucnoN w ANY FORM RESERVED.

210 210

EDER EDER AND ANDMANGELSDORF MANGELSDORF

initially to innate temperament and early patterns of caregiving), (2) to a rudimen rudimentary nonverbal conception conception of emotional states, (3) to a rudimentary rudimentary verbal con conand, finally, emotional states, (4) to a verbal conception conception of dispositions, and, ception of emotional (5) to a full full metatheory metatheory of self. We conceive of the first step as a precursor precursor to personality and the latter latter four four steps as personality personality development development per se. se. Given the emergent self-concept's self-concept's importance to personality personality and its relative neglect by previ previemergent ous researchers, the main focus of our chapter is on dispositional conceptions of of Specifically, we will emphasize the child's early phenomenological phenomenological experience, self. Specifically, important aspects of the self-concept self-concept develop, and those mecha mechathe ages in which important that are involved in its development. nisms that THE II. INFANT PERSONALITY DEVELOPMENT: FIRST SIGNS OF mE SELF-CONCEPT

In hypothesizing how individual differences differences in emotionality emotionality emerge, it is useful useful to following influences: (1) the influence influence of the parent's personality personality on examine the following the developing developing child, (2) the parent's perceptions of the child, (3) (3) the child's own behavior and characteristics or temperament, and (4) the attachment relationship behavior between the child and the parent. It is clear that it is difficult difficult to disentangle the between independent effects effects of each of these influences. In our view they are mutually influential phenomenological experience. For influential and interact to produce the child's phenomenological example, early individual differences differences in children's behavior behavior may interact with paren parenfurther influence influence tal beliefs, expectations, and values regarding these behaviors to further herself as worthy or unworthy unworthy of love and affection. affection. the child's sense of him- or herself following section, we will examine how these varied influences influences interact In the following developing self-concept. We begin by discussing the relations to influence influence the child's developing between parental personality personality and temperament, parental personality personality and attach attachbetween ment, and infant infant temperament Thereafter, we elaborate on our our view temperament and attachment. Thereafter, that infant-caregiver infant-caregiver attachment attachment and infant infant temperament temperament are both fundamental fundamental to that the development self-concept. development of one's self-concept. Association between Parental Personality and A. Association Childhood Cliildhood Temperament A substantial number of researchers have noted a relation between parental person personA temperament.^l More specifically, parents who endorse negative ality and child temperament. Oneof ofthe themajor majorproblems problemsconfronting confrontingtemperament temperamentresearchers researchersisisthe thedefinition definitionofoftempera temperal^One Forthe thepurposes purposesof ofthis thischapter chapterwe weutilize utilizeMcCaU's McCall'sdefinition. definition.According According McCall, ^'temperament ment. For toto McCall, "temperament consists of of relatively relatively consistent, consistent, basic basic dispositions dispositions inherent inherent in in the the person person that that underlie underlie and and modulate modulate consists the expression expression of of activity, activity, reactivity, reactivity, emotionality, emotionality, and andsociability" sociability" (McCall, (McCall,in inGoldsmith GoldsmithetetaI., al.,1987, 1987, the p. 524). We We utilize utilizethis thisintegrative integrativedefinition definitionof oftemperament temperamentthroughout throughoutthis thischapter. chapter.For Formore more thorough thorough reviews of of these these defi definitional issues, readers readers should should refer refer to to Bates Bates (1987) and Goldsmith Goldsmith et etaI. al. (1987). (1987). reviews nitional issues, (1987) and

EARLYPERSONALITY PERSONALITYDEVELOPMENT DEVELOPMENT CHAPTER 99 EARLY

211 211

statements about themselves (particularly regarding anxiety, general negative affect affectivity, or depression) have been found their babies as having more found to also rate their negative temperamental 1987; Daniels, Plomin, & & Greenhalgh, 1984; temperamental traits (Bates, 1987; Mangelsdorf, Gunnar, Kestenbaum, Lang, & Andreas, 1990; Vaughn, Bradley, Mangelsdorf, Gunnar, & & Barglow, 1987). Similarly, mothers who rate themselves as high on Joffe, Seifer, Joffe, & mothers positive emotionality also tend to rate their babies as having high positive affect emotionality having high affect Campos, 1994). Recently, Goldsmith and his (Goldsmith, Losoya, Bradshaw, & & colleagues (1994) suggested four personality might relate to four reasons why parental parental personaUty child temperament: (1) parental personality may bias reports of of child temperament, temperament, leading to artifactual artifactual associations between these two variables; (2) similarities similarities may be due to genetic transmission transmission of of parental personality personality characteristics characteristics to their off offspring; (3) parental personality is part of the developmental developmental context for for children's personality development; and (4) child temperament influence parental person temperament may influence personaUty. The first of of these reasons is discussed discussed in detail by Bates (1987). Based on a ality. careful review of of the literature, he concludes concludes that that "tendencies to endorse negative careful oneself are correlated with tendencies to see more negative tem temstatements about oneself b a b y. .. . . .. However, while the effect peramental traits in one's baby effect has been replicated, stronger overall, than the evidence that correlates of of it has not been any stronger that there are correlates of infant infant behavior behavior (Bates & & Bayles, the perceptions in (objective) observations of 1984)" (Bates, 1987, p. 1136). cited by Goldsmith Goldsmith et al. (1994) regarding the The remaining reasons cited particular child's temperament temperament and and his or her parent's association between a particular straightforward. That That is, an anxious parent parent may personality are more or less straightforward. parent may have passed this trait on to his or her child. Likewise, an anxious parent her child and may communicate communicate be visibly anxious when interacting with his or her this anxiety, resulting in a more anxious child. Finally, an anxious fussy fussy child may make parents feel more anxious about themselves and their parenting parenting abilities. We believe that all of of these explanations have merit, and all of of them likely explain part of the observed observed correspondence correspondence between between parent parent and child personality. Hence, there must be a bidirectional bidirectional relation maternal relation between between maternal personality and child personality. A A mother's personality, for for example, can personality influence both her perception of of and her behavior influence behavior toward toward her her child, and can influenced by her child's temperament. temperament. also be influenced

B. Association between Parental Personality Personality and Infant-Caregiver Attachment Infant-Caregiver According to Bowlby (1969/1982), all infants become "attached" "attached" to the people who care for for them. He claims that humans have an innate need for for social interaction focused on a specific figure over the course of of the first year of of life. that becomes focused attachment to a specific figure (or figures) figures) has evolutionary significance in that This attachment evolutionary significance for the species. If If infants are attached it has survival value for attached to specific caregivers from predators. and seek to maintain proximity with them, they will be protected protected from

212 212

EDERANDMANGELSDORF EDER AND MANGELSDORF

of the However, these attachment relationships may vary in quality as a function of type of of caregiving that each infant receives. Evidence suggesting that maternal personality is related to mother-infant mother-infant attachment has been accrued by researchers using non-normative samples. For of mothers diagnosed with affective affective disorders are more likely to example, children of of normal parents (Gaensbauer, Harmon, be insecurely attached than children of & McKnew, 1984; Radke-Yarrow, Cummings, Kuczynski, & & Chapman, Cytryn, & offspring are at much 1985). Similarly, when parents are abusive or neglectful, their offspring & Sroufe, 1981; greater risk for insecure attachment (Crittenden, 1985; 1985; Egeland & Council, Grunembaum, Botein, & & Zoll, ZoU, 1984; Schneider-Rosen, Lyons-Ruth, Connell, & Cicchetti, 1985). Few relations, however, have been found Braunwald, Carlson, & of attachment. between normal range variations in maternal personality and security of found For example, using a large battery of measures, Egeland and Farber (1984) found variables—maternal maturity and complexity complexity of of thinking-predicted thinking—predicted that only two variables-maternal of attachment at 12 months. Similarly, Belsky and Isabella (1988) found security of of a large battery of measures were correlated with attachment: that only two of of secure (B) babies scored higher on interpersonal affection affection than (1) mothers of of insecure (A (A and C) infants, and (2) mothers of avoidant (A) infants mothers of of ego strength than those of were found to display significantly poorer levels of either secure (B) or resistant (C) infants. Weber, Levitt, and Clark (1986) used the of Temperament Scale (DOTS) to measure maternal temperament Dimensions of of group A A infants scored significantly significantly higher on intensity and reported that mothers of of reaction than mothers of of either B or C babies. Mothers of C infants, however, of differ from those of of B infants on any of the DOTS dimensions. did not differ The lack of of strong relations between attachment classifications and normal range maternal personality differences differences may be partly due to the categorical nature of of the infant attachment classifications. These classifications do not allow for substantial variability among children, particularly within the securely attached group.2 Another reason may be that the relation between these variables is not group.^ direct, but is instead mediated by other variables (e.g., child temperament). This would also account for the conflicting conflicting evidence regarding the relation between child temperament and attachment classifications. We discuss this topic in the next section. The categorical nature of of the attachment attachment coding system, the unbalanced distri distribution of subjects across the three categories of avoidant, resistant, and secure, and of the differences differences in the distribution between samples from different different Western countries greatly limit the nature of analyses performed using strange situation data. Hence, of performed attachment researchers have recently attempted to assess the construct in a variety of of other ways. For example, Waters and Deane (1985) developed a Q-Set for rating attachment behavior. 2 ^The TheQ-sort Q-sortmethod methodfor forassessing assessingattachment attachment(Waters, (Waters,1995; 1995;Waters Waters&&Deane, Deane,1985) 1985)allows allowsfor for more variability and should therefore remedy this problem.

CHAPTER CHAPTER 99 EARLY EARLYPERSONALITY PERSONALITYDEVEWPMENT DEVELOPMENT c. C.

213 213

Association between Temperament and Attachment

It has been argued that measures of attachment quality reflect characteristics of of the relationship relationshiprather ratherthan thancharacteristics characteristicsofofonly onlyone oneor orthe theother othermember memberof ofthe the dyad (Ainsworth, Blehar, Waters, & & Wall, 1978). Of Of course, one would imagine that extremely deviant characteristics of either the mother (as in the case of affectivl! affective disorders) or the infant infant might dictate the course of of the relationship. In this regard, several researchers have emphasized infant temperament, emphasized the crucial role played by infant and have proposed a number of ways in which temperament temperament might influence the infant-caregiver relationship (see Goldsmith, Bradshaw & infant-caregiver & Riesser-Danner, 1986; Lamb, Thompson, Gardner, & & Charnov, 1985). The most extreme position is that of Kagan (1982), who claims that that individual individual ces observed in the strange situation are probably due to endogenous differ differences differdifferen ences in infants rather than to variations in the quality of caregiving as proposed by attachment theorists (e.g., Ainsworth et al., aI., 1978). Another perspective on the differences among relation between temperament and attachment is that individual differences difficult to parent (Waters & & Deane, infants may make some babies more or less difficult difficulty in responding sensitively to 1982). For example, parents would have more difficulty a highly irritable or "difficult" baby and this would result in an insecure attachment relationship. There is some evidence that supports this perspective when behavioral & Isa Isaevaluations of newborns (rather than parental reports) are used (Belsky & bella, 1988). On the other hand, the very notion of "sensitive parenting" implies that the parent is attuned to the needs of the individual child, whatever these needs may be. Indeed, attachment attachment theorists suggest that the truly sensitive parent should be able to overcome obstacles in parenting presented of different different tempera presented by infants infants of temperaments (e.g., Ainsworth, 1983). This implies that the sensitive parent should be equally capable of being sensitive to all children. We believe that the effects infant-caregiver attach effects of infant infant temperament temperament on infant-caregiver attachment are indirect and are mediated by such variables as maternal personality personality and social support. Parents will be differentially differentially sensitive to particular temperamental characteristics in their infants. For example, some parents might be better able to tolerate irritable or "difficult" "difficult" children than others, and this will determine whether a secure or insecure attachment relationship results. This third perspective can be discussed as a "goodness of fit" model. Such a perspective perspective assumes that it is the fit fit between child temperamental characteristics and parental personality that deter determines relationship (e.g., attachment) outcomes. An implication of this perspective is that one would be unlikely, except perhaps in extreme cases, to find main effects effects of either infant infant characteristics or parental characteristics on attachment. However, it has been documented that in the case of extreme irritability (e.g., van den Boom, Boom, 1989) or prematurity (Mangelsdorf (Mangelsdorf et aI., al., 1996) in babies or maternal psychopathol psychopathol& Chapman, 1985), 1985), such main ogy (e.g., Radke-Yarrow, Cummings, Kucynski & effects effects are noted. Such a model is consistent with attachment theory in that it takes into account the evidence that consistent and sensitive maternal behavior is

214 214

EDER AND ANDMANGELSDORF MANGELSDORF EOER

associated associated with secure attachment attachment relationships (e.g., Ainsworth et al., ai., 1978; Belsky, Egeland & Farber, Farber, 1984; Isabella, 1993; 1993; Isabella & Belsky, Rovine, & Taylor, 1984; Egeland 1991). However, a goodness of of fit model would would suggest that that we need to examine how infant infant and parent parent characteristics interact to predict secure versus insecure at attachments. In support support of of a goodness of of fit model, Crockenberg Crockenberg (1981) reported reported that newborn irritable temperament temperament is predictive of of insecure attachment attachment only in conjunc conjuncnewborn maternal social support. That That is, her research research did not reveal significant significant tion with low maternal effects of of either temperament interaction be bemain effects temperament or social support. Instead, the interaction tween these two variables predicted predicted insecure attachment. Similarly, Mangelsdorf Mangelsdorf et al. (1990) found found that security security of of attachment, as assessed assessed at 13 months in Ainsworth's Ainsworth's strange situation, could be predicted predicted by an interaction between between infants' proneness pronenessto-distress temperament temperament measured measured at 9 months in a standardized standardized laboratory laboratory temper temperament assessment assessment (Matheny (Matheny & Wilson, 1981) and maternal personality. That is, ament mothers who scored high on the Constraint Constraint scale of of Tellegen's (1982) Multidi Multidimothers mensional Personality Personality Questionnaire Questionnaire (MPQ) and who had high-scoring high-scoring infants infants on proneness-to-distress at 9 months months were also likely to have insecurely attached attached infants infants proneness-to-distress at 13 months. As in the Crockenberg (1981) study, there was no significant significant main effect of either temperament or maternal personality; rather, it was the interaction interaction effect of either temperament or maternal personality; rather, it was the between the two that predicted insecure attachment. It is easy to imagine how the the between the two that predicted insecure attachment. It is easy to imagine how combination of a fearful and rigid mother (i.e., the profile of high scorers on MPQ combination of a fearful and rigid mother (i.e., the profile of high scorers on MPQ Constraint) with with an an easily easily distressed distressed infant infant could could result result in in aa less less than than optimal optimal rela relaConstraint) tionship. tionship. Thus, the findings findings from from these studies support support a goodness goodness of of fit model. What What remains to be examined, however, however, are the parameters parameters of of such a model. Are there certain infant infant characteristics such as irritability or proneness-to-distress proneness-to-distress that are more likely to lead to insecure attachment than others? Which parental characteris attachment characteristics are particularly important to attachment relationships? Future research must particularly important attachment examine more thoroughly thoroughly the particular particular variables that go into the "fit" between between infant infant and parental parental characteristics. Aspects of of infant infant temperament temperament have been found found to be predictive of of certain infant behaviors in the strange situation. For For example, Vaughn, Lefever, Lefever, Seifer, Seifer, infant found that mothers' ratings of of infants' negative emotionality emotionality and Barglow (1989) found "difficult temperament" temperament" were significantly significantly related to the negative negative emotionality emotionality or "difficult infants displayed displayed during the separation separation episodes of of the strange situation. that these infants temperament ratings ratings alone did not predict It is also noteworthy, however, that temperament emotionality during the reunion reunion episodes with the caregiver, nor did negative emotionality overall attachment classifications (i.e., A, B, or C). Gunnar, Gunnar, they relate to overall attachment classifications Mangelsdorf, Larson, and Hertsgaard Hertsgaard (1989) found found that proneness-to-distress proneness-to-distress Mangelsdorf, predicted proneness-to-distress proneness-to-distress in the strange situation at assessed at 9 months predicted 13 months, but not overall attachment attachment classifications. Hence, there is some stability in infants' behaviors across different different contexts, but this consistency, in of itself, cannot predict predict the quality of of infant-caregiver infant-caregiver attachment attachment relation relationand of ships. Rather, the attachment relationship emerges out of a complex interaction ships. Rather, the attachment relationship emerges out of a complex interaction

CHAPTER CHAPTER 99

EARLY PERSONALITY DEVELOPMENT EARLY

215 215

among a variety of factors, including (but not limited to) infant infant temperament, maternal personality, and maternal social support. Thus far far we have examined how maternal personality is related to infant temperament and infant-caregiver infant-caregiver attachment. We have also examined whether infant temperament and infant-caregiver infant-caregiver attachment constitute separate constructs and the extent to which these constructs are related. These previous topics are necessary for for apprehending the emergent self-concept self-concept because, in our view, one's self-concept arises from from the combination of one's attachment attachment relationships, one's self-concept influential figures (e.g., parents) temperamental characteristics, and the reactions of influential to one's temperamental characteristics. In the following sections, we will explore the infant's phenomenological experience. Specifically, we will address the role of of infant attachment and temperament in the emergent self-concept. infant

D. Theoretical Theoretical Relation Relation between between Attachment Attachment and and Infant Infant D. Self-Concept Self·Concept "affective bond" (Ainsworth, According to attachment theorists, attachment is an "affective 1973; 1979). According to Sroufe and Waters (1977) this term is meant as 1973; Sroufe, 1979). a metaphorical description of the construct suggesting that there is an enduring, stable quality to the attachment relationship. This metaphor is important because affective bond) endure despite the the patterns of behavioral organization (or the affective findingthat thatthe thediscrete discretebehaviors behaviorsreflective reflectiveofofattachment attachmentinininfancy infancy(e.g., (e.g.,proximity proximity finding seeking, contact maintaining, and separation protest) are likely to change or disapdisap separapear over time. The emotional distress observed in young children during separa from their attachment figures makes it clear that the bond is an affective affective one. tions from affective bond and one's In this section we examine the relation between this affective self-concept. developing self-concept. Attachment theorists propose that the self self is an organization of attitudes, expectations, and feelings which are derived from from the infant's interactions with his or her caregiver during the first year of life (Bowlby, 1973; 1973; Bretherton, 1991; 1991; Sroufe, 1990). Alternatively, 1990). Alternatively, one one could could conceive conceive of of the the self self as as developing developing out out of of the the affective affective bond called "attachment" (Bretherton, 1991; 1991; Sroufe, 1990). According to Bowlby (1973), infants come into the world prepared to engage in social relations and as a result of ongoing interactions, particularly with the primary caregiver, develop internal representations or "working models" of attachment attachment figures and of them themselves. These These models models help help the the young young child child to to perceive, perceive, interpret, interpret, and and predict predict events. events. For example, a central feature of the infant's working model of the world involves how the attachment figures are expected to respond to events. Thus, an internal representation of of working model of an attachment figure might be composed of a representation self and other, according that person as available to meet one's needs. Models of of self is, a cold, rejecting parent to Bowlby, tend to be reciprocal or complementary. That is, simultaneously viewing the parent as unloving and himlherself him/herself will lead to a child simultaneously as unlovable. A securely attached child is "likely to possess a representational him/herself as as at least a potentially lovable and valuable person" (Bowlby, model of himlherself

EDER AND MANGEUDORF EDERANDMANGELSDORF

116 216

1980, p. 242). 242). Note that these internal models are essentially evaluative: propositions propositions such as **my "my caregiver is good to me and can be trusted" coexist with notions such as "I am good." good." The concept of internal working models is very helpful helpful for imagining how the attachment relationship is experienced experienced by the child, and how it affects affects the child's attachment found moderate moderate emerging self-concept. In support of this point, Cassidy (1988) found between 6-year-olds' attachment attachment classification classification and their responses responses on sev sevrelations between **hypothesized self-concept measures. She concluded concluded that her results support the "hypothesized eral self-concept between quality of attachment attachment to mother and child's repre reprepresence of connections between self" (p. 130). sentation of self' (p. 130). One limitation with the concept of internal working models, however, is that that it can account account only for global evaluative self-concepts. That is, is, according to this conception of themselves as either good or bad, view, children will develop a conception their attachment attachment relationship. The notion of internal working models, depending on their helpful for understanding understanding the acquisition of global self-concepts self-concepts therefore, may be helpful therefore, versus (e.g., self-esteem) and most certainly would aid in apprehending typical versus self-concept development development (e.g., Bretherton, 1991). 1991). However, However, itit is isnot not helpful helpful atypical self-concept for predicting the rather rather substantial variability evidenced in normal children's self selffor concepts by 3 years of age (Eder, 1990). How then might children develop these differentiated conceptions of their world and their place in it? We suggest more differentiated phenomenological experience of their own that an understanding of children's phenomenological temperament is crucial to answering this question. question. temperament E. Theoretical Theoretical Relation between Infant Temperament Temperament and Self-Concept Infant Self-Concept

has been written about the influence of early temperament on the Very little has developing self-concept. However, we can speculate in much the same way as those who have written about the role of attachment on internal working models. One view of temperament is that it consists of individual differences differences in the expression of basic emotions such as fear, anger, and happiness (Campos, Barrett, Lamb, & Stenberg, 1983). For example, some children have lower thresholds Goldsmith, & for stressful events with fear than others. Such differences differences surely for responding to stressful must result in differences differences in infants' phenomenological experience. That is, is, with increasing cognitive development, continuity in these early emotional experiences eventually becomes incorporated into a system of understanding about the world. low thresholds for fear fear may may come to view the world as infants with low For example, infants a dangerous place. In this way, way, temperament and and self-concept are are thought thought to be separate experiences that interact and contribute to early personality development (see also Goldsmith et al., al., 1987). to, early individual differences differences may Parental perceptions of, and reactions to, influence the child's self-concept. Malatesta and Wilson (1988) suggest that the influence contingent response of parents to an infant's state (e.g., "You're feeling cranky factor in the development of self-awareness. today, aren't you?", p. 94) is a central factor

CHAPTER 99 EARLY EARLYPERSONALITY PERSONALITYDEVELOPMENT DEVELOPMENT CHAPTER

217 217

This explanation can account for children's understanding of of basic emotional states such as anger, fear, and happiness, but it does not explain the acquisition of of nonspe nonspecific affects (e.g., negative affectivity) affectivity) that are also a part of of one's self-concept self-concept (Watson & & Clark, 1984). More general evaluations of of the child's emotional states self-concept. For example, a child must also play a role in the child's developing self-concept. who is perceived as "difficult" may well come to internalize notions of of himself himself or herself herself as "difficult" or as "a failure." If this pattern continued into adulthood, such a child would score high on measures of of negative emotionality. Thomas and Chess (1977) suggest that "goodness of of fit" between the infant's temperament and the environment is important in predicting predicting child outcomes. Al Alinfluence of temperament on though they do not directly address the influence of an infant's temperament parental personality his or her developing self-concept, it is clear that variations in parental self-concept. We illus illusmight interact with infant temperament to affect the child's self-concept. thought to be a early emerging, trate this point using activity level because it is thought & Plomin, 1984). Imagine heritable trait that is somewhat stable over time (Buss & dependa very active infant. This child may develop very different self-conceptions depend ing on the environment in which he or she is raised. If this child is raised by parents who are also active, who appreciate the child's energetic approach to life, and who channel the child's energy into constructive activities, this child will develop a very different self-concept self-concept than if he or she is born bom into a family family in which high energy different frequently reprimanded for levels are considered a shortcoming and the child is frequently his or her behavior. In the first case, the child might come to feel "I like to run fast," "I am good at sports," "I am competent," and "I am good." In the second instance, the child might eventually believe "I am a bother," "I get in trouble a lot," "I am naughty," and "I am bad." This example also illustrates the difficulty difficulty of predicting adult personality traits from infant temperament. In the first instance, of achievement and well wellactivity level might result in an adult who scores high in achievement positive affect). In the second being (i.e., one who would score high on measures of of positive instance, activity level might result in an adult who is a low scorer on measures of constraint. of of infant temperamental number of Although there is little consensus on the exact number characteristics, there is agreement that temperament is reflected reflected in early individual differences in behavioral tendencies along several different differences different dimensions (e.g., activity level and emotionality). The findings of temperamental dimensions of several different different temperamental suggests that individual differences differences in infant temperament would result in children acquiring quite differentiated differentiated views of of the world (e.g., it is dangerous versus safe, fun versus tedious). Parental reactions to infant characteristics must also become self-concept (e.g., "1 "I am a sissy") and would result in incorporated into the child's self-cOlicept differentiation. still more differentiation. from the attachment attachment relationships In sum, whereas the working model derived from of themselves as good versus bad, early leads the child to form a global view of of under undertemperamental characteristics are speculated to contribute to one's system of differentistanding the world. Together, these processes should allow for the highly differenti ated self-concepts that are observed in older children and adults. In the next section,

218 218

EDER EDER AND ANDMANGElSDORF MANGELSDORF

we address the empirical evidence concerning the development development of the self-concept self-concept infancy. in infancy.

F. When Does the Self·Concept Self-Concept Develop? There is some debate concerning the age at which the infant infant can be said to possess a self-concept. The prevailing view, initially proposed by Freud (1950, 1955) 1955) and elaborated 1967; Mahler, Pine & 1975; elaborated upon by other other theorists (Mahler, 1967; & Bergman, 1975; Piaget, 1952; Sullivan, 1953), is that that infants infants go through a period when self self and other are undifferentiated undifferentiated and that the infant infant only gradually comes to see him/her him/her self self as distinct from from others and the rest of the world. This sense of oneself oneself as distinct is thought development of the self-concept. self-concept. thought to be a prerequisite for the development Mahler et al. al. (1975) outlined outlined a comprehensive comprehensive developmental developmental model of the phases the infant through in the development infant goes through development of the self. They proposed proposed that during the first first months of life children are in an undifferentiated undifferentiated phase during which they are in a state of symbiosis with their mothers. This phase is proposed proposed to last until infant emerges from from this symbiotic state and in the 7 or 8 months. Gradually, the infant next phase comes to achieve a sense of self self through the processes of separation and individuation. Separation involves the capacity to recognize that other other human beings are not part part of of oneself. Individuation Individuation entails developing a sense of of identity as an auton autonomous person. In this view, the child gradually emerges from from these processes (of (of which four identifiable sometime between 25 four subphases are proposed) with a self-concept self-concept identifiable and 36 months. This self-concept self-concept integrates integrates perception, memory, cognition, emotion, emotion, and other other ego functions functions into a unitary unitary personality. personality. Sroufe's 's (1975) in a number Sroufe's (1979) view is similar to Mahler Mahler et al. al.'s number of of ways. For 18 and 36 example, he proposes that the self-concept self-concept emerges sometime between 18 months, although he believes that self-assertion self-assertion is often often exhibited earlier in the second second year. In Sroufe's Sroufe's theory of socioemotional socioemotional development development the interdependence interdependence of af affective and cognitive development development is emphasized. It is not surprising, then, that he proposes that the emergence of of the self-concept self-concept coincides with cognitive changes such as those described by Piaget in the sixth and final stage of of sensorimotor thought, when the child is thought thought to achieve representational representational (Le., (i.e., symbolic) thought. Lewis (1987) proposes that the self self has two components that unfold unfold with development: (1) the existential self, which is the experience of oneself existential oneself as distinct from from other other persons and objects, and (2) the categorical categorical self, which consists of the ways that that we think about ourselves. These components are similar to those described described by William James (1892) as the "I" " I " and the "Me", "Me", respectively. In contrast contrast to **Me" (i.e., one's representation representation of oneself) oneself) could be James, who believed that the "Me" **I" (Le., (i.e., oneself oneself experienced) should be left left studied by psychologists whereas the "I" existential and the categorical selves as to philosophers, Lewis regards both the existential amenable to psychological inquiry. In line with the previous investigators, Lewis about 18 months, infants infants are able to think categorically categorically about suggests that at about support of this suggestion, he reports that infants infants demonstrate demonstrate selfselfthemselves. In support

CHAPTER 99 EARLY EARLYPERSONALITY PERSONALITYDEVELOPMENT DEVELOPMENT CHAFfER

219 219

self-representarecognition in a mirror by this age.-^ age.3 He claims that common early self-representa tional categories are gender, age, competence, and value (good or bad). A strikingly different different view is that of of Stern (1985), who places the age at which A infants have an organized sense of of self self much earlier than other investigators. Stern proposes that in order for for the infant to have a sense of core self, he or she must possess self-agency, self-coherence, self-affectivity, self-affectivity, and self-history. Self-agency Self-agency in involves controlling one's actions. Self-coherence Self-coherence entails having a sense of being a physical whole. Self-affectivity Self-affectivity involves experiencing emotions that are attached to physical concerning oneself. oneself. Finally, self-history self-history involves having a sense of of continuity events concerning self-experiences constitute what he with one's past. According to Stern, these four self-experiences calls a sense of of core self self that forms between 2 and 6 months. Some of of the differences differences among previous researchers can be attributed to variations in definitions. Most researchers view the self-concept self-concept as a system of & Sentis, 1982), understanding represented in memory (e.g., Eder, 1988; Markus & or what James called the "Me" and what Lewis calls the categorical self. Given that such a definition requires that infants have the cognitive ability necessary for representing information in memory, these researchers tend to place the emergence of the self-concept self-concept at approximately approximately 18 months. In contrast, a minority of investiga investigaof of self self (i.e., what James called the tors have emphasized one's experience or sense of of self self does not have to be represented and/or organized in memory, "I"). If a sense of then it can be present very early on (perhaps even at birth). What is the role of of a sense of of self versus a self-concept self-concept in early personality development? Stem Stern suggests that affective affective experiences in the first 6 months of of life are one of of the "invariants" in the developing self. This sense of of self is "prerepresen "prerepresentational," according to Emde (1983), in that it exists before before children have the ability for abstract mental representation. That is, early affective experiences (e.g., fear) become linked by a common thread that comes to characterize a particular individual individual of this view, very young infants have been found to be (e.g., fearful). In support of of expressing identifiable facial expressions of of emotion. For example, the capable of identified in expressions for joy, interest, disgust, and physical distress have been identified neonates (Izard, 1977). Others have also suggested that expressions of of sadness and & Walden, 1982). surprise were evident in neonates during a Brazelton exam (Field & At present, however, there is some disagreement among investigators on the degree to which these early expressions reflect reflect the infant's affective affective state. That is, some researchers claim that infants who express recognizable facial expressions of emotion also experience the corresponding emotional state. Proponents of this view, which is called "discrete emotions theory," suggest that very young infants, affective states (see Izard & & perhaps even neonates, experience several distinct affective Malatesta, 1987). Others, however, claim that although early expressions may be affective states, they do not necessarily signal the existence of of precursors to later affective

^Given Giventhat thatthis thisresearch researchisisnot not directly related psychological self-conceptions, it not discussed directly related toto psychological self-conceptions. it ititnot discussed 3 self-recognition in in the self-concept. in this chapter. See Harter (1983) for a discussion of the role of self-recognition

120 220

EDER AND MANGEL'lDORF EDERANDMANGELSDORF

affective affective states comparable to those experienced by adults with similar expressions expressions (e.g., Emde, Kligman, Reich, & & Wade, 1978; Kagan, 1984; Sroufe, l979). 1979). According to Sroufe, the differentiation differentiation of emotions from from these precursors precursors does not begin until after after 3 months of age. age. These viewpoints are not yet reconcilable, although physiological correlates of emotional expressions expressions may eventually eventually clarify clarify our under underphysiological affective experiences. standing of infants' affective Regardless of of whether whether infants infants experience distinct distinct emotional emotional states, we believe that that until affective affective states are organized organized and represented represented into a conceptual system, integrated into the child's self-concept. self-concept. We speculate on when they cannot become integrated this might of might occur occur in the next section. To anticipate, it is likely likely that because of constraints of cognitive development, this does not occur until after after the first first 6 further insight insight into these issues, we review the research research on months of life. To gain further development of infants in the next section. the conceptual development G. Conclusion: Infancy In our opinion, a sense of oneself-as oneself—as described by Stem Stern (1985)-is (1985)—is one one of the first phenomenological experiences of self, and as such is an important precursor to personality development. However, if these sense-of-self sense-of-self experiences are not understandintegrated into a conceptual system, they do not contribute to children's understand ing of their world and themselves. Stem also suggests that self-history or a sense of one's continuity over time is an important aspect of what he calls one's core self. We agree that memory is an important important aspect of the self-concept, for how can we we know who infant memory who we we are, are, unless we we know who who we we were? Research on infant indicates that infants as young as 55 months can recognize a photograph of a face face that they had seen a week earlier, and infants as young as 3 months of age reenact motor activities learned several days earlier (Rovee-Collier, 1987). However, little (Rovee-CoUier, 1987). is known about the organization of infant infant memory. Some investigators, however, have argued that most studies of infant infant categoric knowledge do not clearly demon demonstrate that infants are capable of forming non-perceptually non-perceptually based categories (Mark (Markman & & Callanan, Callanan, 1984). 1984). Hence, Hence,although althoughyoung younginfants infants are are able ableto torecognize recognizefamiliar familiar information, they may not be integrating information information into a system of understanding 6-month(i.e., a concept) in the first 6 months of life. We are intrigued by how 2- to 6-month infants might possess a sense of their continuity continuity over time without having old infants developed the cognitive skills necessary for organizing previous autobiographical developed events in in memory. memory. We We should should emphasize emphasize that that like like Stem, we stress stress the the importance importance events Stem, we of early early affective affective experiences. experiences. Our Our view view differs differs from from Stem's only regarding regarding when of Stem's only these experiences experiences become become integrated integrated into into the the child's self-concept. these child's self-concept. sum, during infancy, early affective affective experiences form form the basis for the In sum, child's developing self-concept. As soon as children can begin to represent and categorize their their experiences experiences (sometime (sometime after after 66 months months of of age), age), they they form form aa represen represencategorize tation of of their their world world and and their their place place in in it. it. This This representation representation grows grows out out of of the tation the interaction betweeen betweeen early early temperamental temperamental characteristics, characteristics, parental parental personality, personality, and and interaction the infant-caregiver infant-caregiver attachment attachment relationship. relationship. the

CHAPTER 9

EARLY PERSONALITY DEVELOPMENT EARLY

221

ni. THE TRANSmON TRANSITION BETWEEN INFANCY AND CmLDHOOD: CHILDHOOD: m. TODDLERS' SOCIAL UNDERSTANDING

An important development in the child's self-concept occurs during toddlerhood (i.e., age) when children begin to label their internal states (e.g., thoughts, (Le., 22 years of age) drives, and feelings). In several naturalistic studies children were observed to use "want")» and drives (e.g., "hungry") by 2 words for their thoughts, desires (e.g., "want"), years of age (Bretherton & 1982; Dunn, Bretherton, & & Munn, 1987). & Beeghly, 1982; Internal states are common to all persons and are therefore not idiosyncratic to particular individuals, whereas dispositions are frequent, enduring tendencies that can be used to characterize and differentiate specific persons. It is not surprising that the young child conceives of and talks about the more generally applicable internal states prior to articulating explicit dispositional concepts (see Wellman & & Gelman, 1987, for an elaboration of this point). Conceptions of the internal states of self and other seem to be acquired & Beeghly, 1982), providing simultaneously (Bretherton & providing support for for the notion that children jointly represent their social world and their place in it. Bretherton, Fritz, Zahn-Waxier, Zahn-Waxler, and Ridgeway (1986) believe that "the ability to think of one's own past or future is conceptually similar to the ability to attribute internal states to selfothers" (Bretherton et al., aI., 1986, p. 533). Hence, young children do have some self of awareness which they can articulate by 2 years of age, but do not yet conceive of themselves and their characteristics as enduring. Stem (1985) makes the intriguing suggestion that the onset of language may Stern make children's experience of themselves less direct than it was earlier. He specu speculates that the autobiography constructed by children during this period significantly alters their felt experiences. Theories of infantile amnesia suggest that in addition to being less direct, these experiences also become more memorable once they are encoded in language. For example, Schachtel (1947) proposed that the majority of of one's earliest memories are for experiences that occurred after the first first 3 years of of life as a function of the onset of language and the continuity provided to verbally encoded events. IV. PERSONALITY DEVELOPMENT IN IN EARLY CmLDHOOD CHILDHOOD IV.

As in the area of infant temperament, a great deal of research has been devoted differences in personality among preschool and kinder kinderto understanding individual differences garten children. Similar to the work on temperament, this research has primarily relied on the reports of other persons and/or behavioral observations. It may be roughly three three categories: categories:research researchon on control controland/or and/orinhibition, inhibition,research research divided into roughly on broader personality constructs (e.g., ego resiliency), and research on emotional expressiveness.

222 222

EDER AND EDER ANDMANGELSDORF MANGELSDORF

By 4 years of age, children demonstrate individual differences differences in the extent to which they are able to delay gratification. Moreover, individual differences differences are also apparent in the cognitive strategies that high versus low self-control children gratification (see Mischel, Shoda, & & Rodriguez, 1989, for a review). use to delay gratification Individual differences differences have also been uncovered in preschoolers' willingness to approach novel objects in unfamiliar contexts. Furthermore, these differences differences are related to a variety of different physiological measures of arousal (Kagan, Resnick, & & Snidman, 1988). Although this research is very informative about individual differ differences in preschool children's behavior, behavior, it is not clear whether these differences in behavior correspond to differences differences in children's self-concept self-concept or phenomenologi phenomenological experience. This information would be particularly informative for constructs such as inhibition that are thought to be related to adult personality traits. For example, Kagan et al. (1988) suggest that childhood inhibition is analogous to and possibly predictive of of adult introversion/extraversion (p. 167). Hence, it would be of of consider considerable interest to know whether children low on inhibition, for example, also view themselves as outgoing and talkative. Such information information would lead to a fuller picture of personality development in these preschool years because it would clarify the of emotional basis of of constructs such as inhibition. One component that is crucial to introversion/extraversion in adulthood is one's energy level (see Watson & & Clark, introversion/extraversion of the emotionality underlying extraversion). Kagan in this volume, for a discussion of and his colleagues (1988), however, suggest that children high on inhibition feel fearful in unfamiliar unfamiUar social settings. In the absence of of information about children's fearful actual experience of of unfamiliar social settings, we can only infer their experiences experiences from observing their responses to these events. Having children's self-report self-report would enable us to specify specify the emotion underlying these traits. For example, it would actually/ee/ allow us to determine whether children rated as high in inhibition actually feel fear in social settings or some alternative emotion (e.g., lethargy). of ego resiliency developed by the Blocks (e.g., J. H. Block & & The construct of personality construct used to describe the degree to which Block, 1980) is a broad personality individuals approach problems in a flexible manner. Children are identified identified as high or low on this dimension through the reports of others (e.g., teachers' Q-sorts). infer aspects of the children's self-concept self-concept through items Raters are also asked to infer of self self as 'bad.'" *bad."' Children rated as such as "Appears to feel unworthy; thinks of of ego resiliency at securely attached are found to score higher on Q-sort indices of & Sroufe, 1979). These results suggest that measures of of ages 4 to 5 (Arend, Gove, & of children'S children's ideas about themselves (i.e., their ego resiliency may tap some aspects of self-report measures were not "internal working model"). Unfortunately, because self-report obtained from the child, this connection remains speculative and must be established in future research. Individual differences differences have also been uncovered in children'S children's expressive be beIndividual particularly relevant to research which has defined haviors. These findings are particularly differences in emotional infant temperament as essentially reflecting individual differences (e.g.. Goldsmith & & Campos, 1986). For example, children rated as expressivity (e.g.,

EARLYPERSONALITY PERSONALITYDEVELOPMENT DEVELOPMENT CHAPTER 99 EARLY

223 223

having more expressive faces are also likely to have been rated rated as higher on extraver extraver& Walden, 1982). Hence, individual individual differ differsion by their teachers (Buck, 1977; 1977; Field & ences are apparent apparent in the degree to which children children spontaneously spontaneously express emotions. A recent study also revealed individual variability variability in the extent extent to which A children are able to exert control over their expressive behaviors. In this study, 3modified version of and 5-year-old 5-year-old children were rated by their teachers on a modified of Children scoring in the Snyder's (1974) self-monitoring self-monitoring scale (Eder & & Jones, 1989). Children top and bottom third of the measure were tested tested (in counterbalanced counterbalanced order) on their ability (1) to recognize recognize standardized standardized photographs photographs of of facial facial expressions of of emotional expres emotion (Ekman (Ekman & & Friesen, 1975), and (2) to produce six basic emotional expressions (e.g., happiness, sadness, anger). High scorers on the self-monitoring self-monitoring scale producing emotional expressions significantly better at both recognizing and producing were significantly differences found found in recogni recognithan low scorers. The usual developmental (i.e., age) differences production were not significant. However, self-monitoring interacted sig sigtion and production self-monitoring interacted nificantly with age such that high 3-year-olds, for for example, were better better at posing nificantly 5-year-olds were better emotional expressions than low 5-year-olds, and high 5-year-olds better than individual differences low adults. These results demonstrate that individual differences in the control of of apparent at an early age. expressive behavior are apparent Of future future interest is whether whether high self-monitoring self-monitoring children Of children are generally more expressive than low self-monitoring self-monitoring children, or are simply better expressive better at controlling (thus making them less expressive expressive in situations situations requiring their expressive behaviors (thUS of concern is the relation spontaneous display rules such as masking). Also of relation between between spontaneous felt emotions (i.e., one's feelings feelings about and controlled emotional expressiveness and felt would be expected expected to show oneself). If these two constructs are independent, they would different pattern of correlations with children's self-report. To speculate, children children a different feel more accepted accepted by themselves and others and who are high in expressivity may feel self-concepts that reflect reflect these feelings feelings (e.g., well-being). thus should demonstrate self-concepts other hand, children who are able to control their their expressive displays might On the other especially comfortable comfortable in social settings, and their self-reports self-reports that that demonstrate demonstrate feel especially corresponding differences differences in sociability. corresponding differences in In sum, although there is substantial research on individual differences of this research kindergarten children's personality, the majority majority of preschool and kindergarten of other has emphasized emphasized behavioral observations and/or and/or the reports of other persons. This research is very informative personality that that are noticed by informative for for those aspects of of personality other persons, but neglects children's own representation representation of of their personalities. Indeed, this self-representation developments during self-representation undergoes many many important important developments the preschool years; these changes are discussed next.

A. Development Development of of the Self-Concept Self-Concept in Preschool and Early School-Age Children Several features characterize the majority majority of of research on the self-concepts of pre preself-concepts of kindergarten children. Note that each of of these characteristics contrasts school and kindergarten personality development discussed prewith the traditional research on children's personality development discussed

224 224

EDER AND MANGEI3DORF EDERANDMANGELSDORF

viously. First, most researchers have focused on children's self-descriptions, rather than on behavioral behavioral or observer data. Second, these investigators have generally ignored emotions and have instead been concerned with the role of of cognitive development in children's self-understanding. self-understanding. Third, investigators have tended to emphasize age (i.e., group) rather than individual differences. differences. In a frequently cited study, Keller, Ford, and Meacham (1978) categorized tIle the responses of of 3- to 5-year old children to open-ended and fill-in-the-blank fiU-in-the-blank ques ques). More tions (e.g., "I am a boy/girl who ---""). More than than half half of of the the responses responses referred referred to actions (e.g., "I am a boy who plays"). Hart and Damon (1986) interviewed children from 6 to 16 years using a variety of free-response "What free-response questions such as **What are you like?" like?" Responses were categorized according to four types of content (i.e., developmental physical, active, social, and psychological) and for four possible developmental belevels, with higher levels indicating increasing abstraction. Strong correlations be tween age and developmental level were found. different approach, approach. Bannister and Agnew (1976) examined Using a slightly different the role of of memory in young children's self-concepts. In their study, 5- to 9-year 9-yearolds' responses to questions about themselves were tape recorded. All situational of the children were eliminated and the answers were cues about the identity of rerecorded in a single adult voice. When children were reinterviewed, 4 months identify their later, they heard several answers to each question and were asked to identify own previous answer. In addition, they were asked to provide reasons for their of the youngest youngest children choices. In providing reasons for their responses, most of of the simply said that they remembered their responses, whereas the majority of oldest subjects indicated that they based their answers on their general likes and Agnew concluded that young children rely more on their dislikes. Bannister and Agnew specific memories for information about themselves, but older children are capable of using a more general theory of of self. From findings such as those described of above, early reviewers concluded that the structure of young children's self-concepts self-concepts becomes more general with increasing age, and that the content of of these self selfconcepts becomes more psychological with age (see especially Harter, 1983). More recently, these conclusions have been reexamined in light of of work from cognitive and language development. Specifically, if if the self-concept self-concept is actually a concept, it should depend on a person's memory. Hence, age differences differences in young children's self-concepts self-concepts should be similar to those found in memory studies. Given that the ability to retrieve general memories is already present by 3 years of of age, whereas the ability to access specific specific memories emerges between 4 and 5 years of of age (Hudson, 1986), a similar pattern should also be expected for children'S children's self-concept. Perhaps early investigators, in asking only one type of question, confounded confounded structure and content in their studies. To examine this question, Eder and her colleagues asked 3- to 7-year-old children general and specific questions about their behaviors and internal states (Eder, 1989; Eder, Gerlach, & & Perlmutter, 1987). General questions involved frequent and/or typical activities and were not located temporally located. Behavior in one particular point in time. Specific requests were temporally

CHAPTER99 EARLY EARLYPERSONALITY PERSONALITYDEVELOPMENT DEVELOPMENT CHAPTER

225 225

questions concerned activities and involved action verbs; requests about internal states involved adjectives. The findings supported those from previous research on children's memory by revealing revealing that general memories about oneself oneself dominated the responses of the youngest children, whereas specific memories increased sub substantially from 3 to 6. The old view that young children do not evidence any psychological self selfunderstanding has also been disputed by recent investigators (Miller & & Aloise, 1989). For example, Wellman and Gelman (1987) suggest that two distinct types of of personological conceptions conceptions were confounded in earlier earUer work, namely, conceptions conceptions of internal states and of of dispositions. The distinction is important because the young child may possess an awareness of of internal states prior to acquiring dispositional concepts. Indeed, as we indicated earlier, even 2-year-olds describe internal states of themselves and others in their conversations, suggesting that they already have a rudimentary understanding of internal states by this age (e.g., Bretherton & & Beeghly, 1982). Sixty Sbcty percent of the 3-year-olds in Eder's studies (Eder, 1989; Eder et aI., al., 1987) responded responded to state questions with appropriate descriptions of their internal states and emotions. For example, in response to the question, "Tell me 3i-year-old girl said, "Usually how you've usually been when you're scared," one 31-year-old approfrightened." Even when the children did not generate state terms, they used appro priate behaviors in their responses, for example, responding to the question, "Tell me how how you you usually usually have have felt when you've you've been been happy," happy," with with "smiling." "smiling." Finally, Finally, me felt when even the the youngest youngest children children seemed seemed to to understand understand the the relation relation between between emotions emotions even and other other internal internal states states (e.g., (e.g., beliefs, beliefs, attitudes). attitudes). Thus, Thus, they they often often justified justified their their and responses by by referring referring to to aa belief belief or or attitude. attitude. For For example, example, one one 31-year-old 3i-year-old boy boy responses stated, "I don't feel that good with grownups, 'cause I don't like grownups." stated, "I don't feel that good with grownups, 'cause I don't like grownups." concepMost other research, however, indicates that full-blown dispositional concep (e.g.. Ruble & & Rholes, 1981; 1981; Shantz, 1983). tions do not emerge until 7 to 8 years (e.g., This raises the question of of what develops in children's self-understanding self-understanding between 2 years (when they understand emotional states) and 7 years (when they understand dispositions). Research on adults' conceptions of of states and traits reveal that two types of of features are evident in their personological conceptions: situational features (e.g., cross-situational consistency) and temporal features (e.g., duration; see Chaplin, John, & & Goldberg, 1988). Previous research has indicated that the situa situational aspects of of dispositions are not present in young children's personological conceptions. For example, Rholes and Ruble (1984) reported that children do not regard dispositions as the tendency to behave consistently across situations until they are 7 or 8 years old. In contrast, children probably possess an awareness of the temporal aspects of of dispositions by 3 years of age. That is, the general memories reported by young children in Eder's research (e.g., Eder, 1989) were representa representations of of behaviors and/or internal states that were stable and enduring, indicating young children conceive of of themselves and other persons as having characteristics that are enduring across time. These early general memories may form the basis for the more mature dispositional conceptions held by older children and adults.

226 226

EDER AND EDER ANDMANGELSDORf MANGELSDORF

Eder (1990) examined whether such general memories are organized into meaningful and consistent self-conceptions. She presented children, in a random order, with pairs of of statements reflecting 10 of Tellegen's lower-order personality dimensions (e.g., Well-Being; see Tellegen, 1985). It was (e.g.. Achievement, Alienation, Weil-Being; assumed that if children selected statements in a nonrandom fashion, then they must possess a corresponding corresponding organizing construct (i.e., a trait). To examine this, a new method was developed for assessing assessing young children's self-concepts. One hundred and eighty children between 3 and 8 years were pre prerepresenting the low and high endpoints of of Tellegen Tellegen's sented with pairs of statements representing 's dimensions by two puppets. For example, one puppet would state, "It's not fun to representing scare people," and the other would say, "It's fun to scare people," thus representing the high and low endpoints of aggression. Children were then asked to pick the statement that better described themselves. three-factor solution Responses were factor analyzed with varimax rotation. A A three-factor best described the responses for every age group, although although the constructs differed differed different age groups. The factors that emerged for for the 3-year-olds were for the different labeled Self-Control, General Self-Acceptance, and Rejection. The factors that emerged for the 5-year-olds were Self-Control, Self-Acceptance Self-Acceptance via Achievement, Self-Acceptance via Affiliation. Finally, those that were revealed for for the 7and Self-Acceptance Emotional Stability, Extraversion, and Determined Fearlessness. year-olds were Emotional Separate higher-order factor scales were computed for each age group and these meaningful and to demonstrate good internal were found to be psychologically meaningful consistency in every age group (e.g., average alphas = = .75, .78, and .78, for the consistency 3-, 5-, and 7-year-olds, respectively). of age, children possess common underly underlyThese results indicate that by 3 years of ing dispositional constructs for organizing information information about themselves. Further Furthermore, by 7 years two of the constructs resemble in content and degree of of complexity those that have been identified in adults. Specifically, the Emotional Stability factor corresponds to Eysenck's construct of Neuroticism (reversed) and to Tellegen's Negative Emotionality. The Extraversion factor is analogous to Eysenck's Extraver ExtraversionlIntraversion sion/Intraversion and, to a lesser extent, Tellegen's Positive Emotionality. New Approaches Approaches 1. New

Early research tended to neglect the content of young children's children'S self-conceptions, focusing more on the particular categories that they used to describe themselves. focusing self-conceptions or of these self-conceptions Very little information was gleaned regarding the nature of of of children's feelings about themselves. Unfortunately Unfortunately this provided a discontinuity of the between research with preschoolers and research on the evaluative aspects of self self (i.e., self-esteem) self-esteem) in infants mfants and older children and adolescents. self-conceptions largely focused focused on age Further, research on young children's self-conceptions differences in children's self-descriptions, whereas investigators of of the self-concepts self-concepts differences of older children and adults emphasize individual individual differences. of differences. For example, Harter differ from one another in per per(1982) demonstrated that by third grade, children differ self-competence (i.e., the extent to which they view view themselves as competent ceived self-competence

CHAPTER 99 EARLY EARLYPERSONALITY PERSONALITYDEVELOPMENT DEVELOPMENT CHAPTER

227 227

in a variety of different different domains) and in their global self-worth (i.e., the degree to which they view themselves as worthwhile persons). In spite of of their importance, previous investigators have largely ignored indi individual differences differences in preschool and kindergarten children's self-concepts. We suggest several reasons for this: First, developmentalists tend to place more importance on age differences, believing that individual differences differences may ultimately reflect error. Second, the measures used to study individual differences differences in older children and checkhsts) are not suitable for use with young adults (e.g., rating scales, adjective checklists) children. Third, as we indicated previously, early research suggested that young children did not demonstrate psychological self-conceptions. If young children do not have any trait-like personological conceptions, then there is little reason to differences in their self-understanding. self-understanding. seek out individual differences However, it is possible that conclusions derived from early studies underesti underestiof children's self-understanding. Previous investigators used mate the true level of open-ended questions to study children's self-concepts. One limitation of this format is that it depends on the language production skills of of the subject. Much research with preschool children shows that they comprehend more than they produce (e.g., & Maratsos, 1975). It is not surprising then that they perform better on Kuczaj & language comprehension, rather than language production, tasks (see Kuczaj, 1986, for a review). This may be especially true for studies on the psychological aspects of the self-concept, self-concept, because state and trait adjectives are relatively recent additions of & Kuczaj, 1985). to young children's vocabularies (Ridgeway, Waters, & For these reasons, Eder (1990) utilized a comprehension task to test children's self-conceptions. She found that when children listened to pairs of statements pre presented in a random order and were asked to pick the one that best described themselves, even 3-year-olds made their selections in a nonrandom fashion. It was concluded that these children made use of of underlying organizing constructs (e.g., traits) when evaluating information. The presence of of dispositional self-conceptions is a necessary, but not sufficient, sufficient, condition for the existence existence of individual differences in self-report. Even if children conceive of of themselves in terms of of a construct, for example, "well-being," they might not differ differ significantly significantly from one another in their self-view self-view on that construct (e.g., they might all conceive of of themselves as high in well-being). Hence, Eder also examined whether individual differences children 's self-conceptions. differences were evident in children's self-conceptions. Children's mean scores on both the dimensions and the higher-order factor scales were close to midpoint; moreover, the standard deviations revealed considerable variation, indicating their self-conceptions self-conceptions differed from child to child. Finally, these differences differences showed some stability over time. That is, 3i-, 3h-, 5i-, and 7i-year-old children's responses on every higher-order factor yielded statistically significant 1month test-retest correlations. These stability coefficients coefficients averaged .47, .60, and .65 for the 3i-, 5t5J- and 7i-year-olds, respectively. We are not suggesting suggesting that 3-year-old children conceive of of themselves as "High in Self-Acceptance" or even as "The type of person who feels accepted by others." Rather, it is likely that they are using an organizing principle which they

228 228

EDER EDER AND ANDMANGEUDORF MANGELSDORF

are unable to articulate in much the same manner as they use implicit language rules (see Bretherton & & Beeghly, 1982, for a similar point). These young children possess a working theory, but not a metatheory, about themselves. What is the organizing principle behind children's statements? Perhaps chil chilheadfirst and dren group together similar behaviors (e.g., that they go down slides headfirst climb things that are high) and then simply remember these clusters of of similar behaviors. Whereas this explanation may account for the consistency consistency on some (especially lower-order ones such as Harm-Avoidance), children also dimensions (especially *'When combined quite dissimilar items such as "I climb things that are really high," "When A more compelling alternative I get angry, I hit people," and "I get upset a lot." A underlying emotional emotionalexplanation is that children selected statements based on their underlying of the hedonic ity. An examination of Table I provides one with a strong sense of of children's tone underlying the scales that were formed based on the factor analyses of reflective of how young children/ee/ responses. Hence, these self-concepts are reflective of how children/eel about themselves, rather than necessarily indicating how they behave. In sum, 2-year-olds differentiated are able to label emotional states and 3-year-olds demonstrate very differentiated self-conceptions on which they show stable and meaningful meaningful differences. differences. dispositional self-conceptions These trait conceptions not only have characteristics similar to those observed in adults (e.g., they are internally consistent), but by 7 years, they begin to look strikingly similar to those dispositions that have been described in adult research. V. PUTIIN PUTTING TOGETHER: A A MODEL OF V. G IT ALL TOGETIlER: PERSONALDT DEVELOPMENT PERSONALITY

Traditional research on personality development has been subdivided into several distinct domains, with each domain developing and utilizing its own characteristic questions and methods. For example, temperament researchers have emphasized maternal and/or teacher reports to understand individual differences in biologically individual differences biologically based constructs. In contrast, attachment researchers have mostly depended on observations of mother-child interactions in the strange situation to arrive at a child's attachment classification. Finally, early self-concept as self-concept researchers have assessed children's own self-descriptions self-descriptions and emphasized age, rather than individ individual, differences. differences. We propose that these literatures have more in common than a superficial superficial analysis would indicate. Specifically, each of of these areas is essential for understand understanding the emotional basis of of early personality, and thus they jointly provide a satisfac satisfactory answer to the question, "What is important in personality development?" In this final section, we use this theme to integrate the bodies of of research that we have described previously. affective Recent research on adult personality has consistently stressed the affective basis of personality dimensions; that is, responses on self-report self-report are first and fore foreof how individuals feel about themselves, rather than how they most indicative of Lazarus, 1975; 1975;Tellegen, Tellegen, 1985; 1985;Watson Watson& &Clark, Clark, behave, for example (J. Block, 1989; Lazarus,

CHAPTER 9

EARLY PERSONALITY PERSONALITY DEVEWPMENT DEVELOPMENT EARLY

229 229

TABLE TABLE II Examples of Items on the Factor Scales Low items

High High items items 3l-year-olds 3i-year-olds 1: Self-control Factor 1: Self-control

When II get get angry, angry, II feel feel like like hitting hitting someone. someone. When I sometimes try to push in in front front of people on line. I get mad a lot. Sometimes I get in trouble for being bad.

When II get angry, I feel like being quiet. II don't ever try try to push in front front of people on line. II get get mad a little. I never get in trouble for being bad.

Factor 2: General self-acceptance self-acceptance

like work work that's not not very very hard. hard. II like It's more more fun fun to to do do things things by by myself myself than than with with other other It's people. people. I'm not not usuaIly usually very very happy. happy. I'm Sometimes myself. Sometimes II feel feel like like II just just don't don't like like myself.

like hard hard work. work. II like It's more more fun fun to to do do things things with with other other people. people. It's II am am usuaIly usually happy. I reaIly really like myself.

Factor 3: 3: Rejection Rejection Factor don't ever ever feel feel people people want want bad things things to to happen happen II don't to me. me. to When my my friends friends visit visit they they play play with with me me and and not not my my When toys.

sometimes feel feel people people want want bad bad things things to to happen happen II sometimes to me. to When my my friends friends visit visit they they play play with with my my toys toys and and not not When with me. with

People don't usuaIly usually say say mean mean things things to to me. People I am am the leader in "FoIlow "Follow the Leader." Leader."

People People always always say say mean mean things things to to me. Other Other people people are are the the leader leader in in "Follow "Follow the the Leader." Leader." 51-year-olds 5i-year-olds Factor 1: 1: Self-control

When II get get angry, angry, II feel feel like like hitting hitting someone. someone. When sometimes try try to to push push in in front front of of people people on on line. line. I sometimes get mad mad aa lot. lot. II get Sometimes I get in trouble for being bad.

When When II get get angry, angry, II feel feel like like being being quiet. don't ever ever try try to to push push in in front front of of people people on on line. line. II don't get mad mad aa little. little. II get never get get in in trouble trouble for for being being bad. II never

Factor 2: Self-acceptance Self-acceptance via via achievement achievement Factor I like work that's not very hard. hard. Other people pick the game to play. play. don't like like to to have have people people look look at at me. me. II don't Sometimes, II just just don't like like myself. myself. Sometimes,

like hard hard work. work. II like pick the the game game to to play. II pick like to to have have people people look look at at me. me. II like really like like myself. myself. II reaIly

Factor: Self-acceptance Self-acceptance via via affiliation affiliation Factor: People always say mean things to me. me. I don't have a best friend. friend. I'm happiest when I'm by myself.

People don't don't usuaIly usually say say mean mean things things to to me. People have aa best best friend. friend. II have I'm I'm happiest happiest when when I'm I'm around around people. people.

I am not a good boy (girl). (girl).

II am am aa good good boy boy (girl). (girl).

continues continues

EDER AND MANGELSDORF EDERANDMANGELSDORF

230

TABLE I continued continued Low items

High items 7i-year-olds Factor 1: 1: Emotional stability

When I get angry, I feel like hitting someone. I sometimes think no one cares what happens to me. A lot of things make me upset. Sometimes, I just don't like myself.

When II get angry, I feel like being quiet. II know that people care what happens happens to me. It's hard for me to get upset. I really like myself.

Factor 2: Extraversion It's not fun to scare people. I'm happiest when I'm by myself. I don't like to show things at "Show and Tell" at school. I don't like to boss people around.

Sometimes it's fun to scare people. I'm happiest when I'm around people. II like to show things things at "Show and Tell" Tell" at school. II like to boss people around.

3: Determined fearlessness Factor 3: When I'm scared, I run away. When I see something scary on TV, TV, I cover my face. It's not fun to ride in a fast car. I get scared a lot.

When I'm scared, I stand up to what scares me. II like to look at scary things TV. things on TV. It's fun riding in a fast car. II never get scared.

1984). 1984). Indeed, self-reported self-reported distress is one criterion for diagnosing pathology in the DSM framework framework (American Psychiatric Association, 1980). How do individuals come to feel the way they do? How is it that someone comes to be high on Negative Emotionality? Recall that we anticipated this discussion by suggesting earlier that personality development occurs in five steps. A. 1: Organized A. Step Step 1: Organized Pattems Patterns of of Behavior Behavior The first component in how humans feel about themselves is probably evidenced after birth. Neonates demonstrate substantial individual differences differences in their soon after behaviors. Such "temperamental" differences, emo differences, for for example, in activity level, emotional expressiveness, and attention, are thought to be genetic in origin and are thought to be evidence that aspects of one's personality structure are present at birth. In our view, many of these temperamental temperamental differences differences are necessary necessary antecedents of of certain later personalities, but none are themselves sufficient. sufficient. That is, the early structures will interact with future future contents in important ways. Hence, one cannot predict what a person will be like as an adult from from only knowing his or her infant infant temperament. This is because these differences differences act as constraints on potential individual differences fully accounting for them. differences in personality traits, rather than fully For example, infants infants who are very low on energy may never become extraverted

CHAPTER 9 CHAPfER

PERSONALITY DEVELOPMENT DEVELOPMENT EARLY PERSONALITY

231 231

adults. However, energetic infants may or may not become extraverts, depending attachment relationship). Relatedly, the content on other experiences (e.g., their attachment affect the original structure in fundamental fundamental ways. For example, children who may affect develop high positive affect affect from from their life experiences may show more energy than from their original temperament. would be expected from (e.g.. Goldsmith & & Campos, 1986) has emphasized that these One approach (e.g., first individual differences differences are experienced by the infant in terms of basic emotional states (e.g., happiness, happiness, anger anger,, fear). Whereas all all children children experience basic emotional emotional states (e.g., Ekman & & Friesen, 1975), they differ differ in the degree, frequency, and intensity with which these states are experienced (Izard, Hembree, & & Huebner, 1987). We We believe believe that, that,with withincreases increasesin incognitive cognitivedevelopment, development,these theseinitial initialtemper temperamental differences differences can lead to differences differences in children's views and feelings about the world they inhabit. For example, infants who have a low threshold for fear may develop into individuals who view their world as a dangerous place. differences also affect affect parents' reactions to Moreover, such temperamental differences further influenced influenced by parental personality. For their children. These reactions are further example, active infants may be perceived by parents as "energetic" *'energetic" or as "hyperac "hyperactive," depending on the parents' personalities (or their own activity level). level). Recent research provides indirect support support for for this suggestion. Mothers who score high on Negative Emotionality on Tellegen's (in press) measure are likely to also rate their 35-year-old children as being Under-Socialized, perhaps demonstrating an 3- and 5-year-old intolerance of these mothers for for the rather high activity level characteristic of most preschool and kindergarten children (Eder & & Mahmood, 1989). for example, would If complete random selection occurred, active children, for have an equal likelihood of having parents who perceive them as energetic as with those that view them as hyperactive. However, recent findings findings suggest that there is a genetic relation between children's temperament and their parents' personalities is (Goldsmith et al, at, 1994) that may constrain the actual combinations of parent and infant temperament that occur in the world. Temperamental factors in children and parents probably also set a range of of attachment relations. That is, is, secure attachments are less likely to occur for for certain combinations of parental personality and child temperament. However, a wide margin of possible combinations can allow for attachment. For example, mothers in activity are in in a better position to provide sensitive parenting who are intermediate in to children who are either slightly more lethargic or more active than average than to children at either extreme. The margin of this range also depends on another is, parentalflexibility, personality variable, that is, flexibility, withflexible flexible parents having a wider effective parents (Mangels (Mangelsrange of temperaments they can tolerate in order to be effective dorf et al., dorf at, 1990).

B. Step Step 2: 2: Nonverbal Nonverbal Conceptions Conceptions B. Whereas temperament may provide a child with a sense of his or her world (e.g., it is a dangerous place), attachment has been thought to provide children with a

AND MANGELSDORF EDn EDERANDMANGELSDORF

132 232

sense of themselves. The "working model" of that attachment of self self that attachment provides is exclusively self-evaluative self-evaluative (i.e., goodlbad). good/bad). Specifically, Specifically, children children who develop secure relationships with their caretaker caretaker are thought thought to develop working working models of of them themselves as "worthy "worthy of of love" (e.g., Bretherton, 1991). Conversely, those who develop Temperament, on the other insecure attachments have negative self-evaluations. Temperament, differentiated views of of the world. Together, hand, provides more differentiated Together, these provide the initial emotional basis of of the child's personality. By the second second step (when children have the capacity to form form nonverbal concepts), children children begin begin to have a sense of of the world and their place in it. Thus, a child who views the world world as a dangerous place may also feel that he or she is worthy of of the protection protection he or demands or she demands and receives from his or her parents. We might view such a child as both cautious and from her might such child both cautious and content. that we do not consider consider infant infant temperament sufficient We want to stress that temperament to be sufficient for for accounting for for adult personality traits. Rather, temperamental temperamental dispositions com combine with the child's emotional experience to produce produce these traits. For For example, temperamental disposition disposition of of activity activity must combine the temperamental combine with emotions such as positive affect affect and sociability to produce the adult trait of of extraversion (see also Watson and Clark, this volume, chap. 29). C. Step 3: 3: Verbal Conceptions The transition between these two steps occurs when when the child develops some labels (either articulated articulated or not) about his or her emotional emotional states. This is thought thought to around 2 years of of age (e.g., Bretherton Bretherton & & Beeghly, 1982). In addition addition to occur around felt emotions, 2-year-olds can pretend pretend to be experiencing experiencing a state other articulating felt illustrated in the following following example (from (from Dunn, 1988, p. 21): .• than their own as illustrated C: Bibby on. M: M: You don't want your bibby on. You're not eating. C: Chocolate cake. Chocolate cake. M: You're not having any chocolate cake either. Whines) T C: Why? ((Whines) Tired. ired. Ooh! M: You're tired? Oohl C: Chocolate cake. M: No chance.

4: Dispositions Dispositions D. Step 4: At this point (approximately (approximately 3 years of of age), children children use some underlying constructs At to organize and evaluate information of these constructs information about themselves. Some of might reflect reflect children's responses to the labels and/or and/or reactions of of others (e.g., For example, children who are fearful fearful and and receive negative reactions parents). For Children who were responded responded to more might view themselves as babies or sissies. Children self-righteous need need for for caution or care. positively, on the other hand, might feel as self-righteous caution or

CHAPTER 99 EARLY EARLYPERSONALITY PERSONALITYDEVELOPMENT DEVELOPMENT CHAPTER

233 233

We speculate that as they acquire the ability to control their emotions (proba (probably also at about 3 years; see Cole, 1985), children will begin to inhibit their expression of of poorly received emotions. For example, children who receive disap disapproval for their fearful fearful behavior behavior might inhibit future expressions of of fear. Children who generally receive negative reactions (e.g., abused and/or neglected children) might inhibit the expression of aI., 1988), whereas of most emotions (see Camras et al., others might only inhibit those that seemed especially offensive offensive to their parents. emotional basis of of personality is also manifested manifested These findings illustrate how the emotional in emotion expression (i.e., behavior). Although children have the ability to label their eemotions motions at this time, they metatheories do not have full meta theories of of themselves. That is, that they can state that they get scared a lot (versus a little), hide from thunder and lightning (versus go look at it), and so forth. However, they are not able to state that they are high (or low) fearfulness or Harm-Avoidance. Hence they employ employ an underlying construct on fearfulness to organize their feelings, but can articulate only specific feelings, not the construct itself. Furthermore, we believe that the construct they employ may actually be a of the world (e.g., it is dangerous), and not a view of of themselves themselves per se. view of of self self may be tied to their view of the world. Alternatively, this view of 5: Dispositional Conceptions E. Step 5:

At approximately 7 or 8 years of of age, children can articulate the constructs they use. They begin to describe themselves and others in terms of of traits (e.g., honest) and use trait information to make judgments about past and future behavior (see & Ruble, 1984). Rholes & Stern (1985) suggests that children's construction of of their experiences differs differs from the actual experiences. The implication is that the constructs evident during to—but not isomorphic with-the with—the earlier views held Step 5 are probably related to-but by the young child. For example, a child who scores high on constraint (and describes himlherself him/herself as "careful") may have felt as an infant (and still may actually feel) that the world is a dangerous place. Thus, the emotional basis of the self-concept self-concept is directly experienced but seldom directly stated. It is probable that the emerging ability to articulate one's self-conception self-conception affects the way that the child construes new experiences. Epstein (1973) has sug suggested that an implication of of having a self-theory self-theory is that one engages in theory confirmation. For example, once children conceive of of themselves as "generous" they begin to remember specific examples of their generosity and tend to overlook cases in which they behaved ungenerously. Some adult-like dimensions seem to be present in 7-year-olds. The dimensions of of Emotional Stability and Extraversion have been revealed in children's responses of age (Eder, 1990). The degree of of continuity between these early by 7 years of emerging traits and similar adult structures must still be established. Nonetheless, it appears that personality development is at least partially completed by 7 years of age-as age—as Freud originally concluded. of

EDER EDER AND AND MANGELSDORF MANGELSDORF

234 234

VI. VI. CONCLUSION Whereas previous accounts of of early personality personality development typically stress those of the child that are noticed by other persons, we were interested in under underaspects of standing children's own phenomenological phenomenological experience. Hence, we explored the emergence of of dispositional self-conceptions self-conceptions in infancy and early childhood. Based on previous theoretical and empirical work, we suggested that these conceptions of behavior, (2) to nonverbal conceptions of of unfold from (1) organized patterns of emotional states, (3) to verbal conceptions of emotional states, (4) to verbal concep conceptions of of dispositions, (5) to a metatheory of self similar to what is held by older of developmental changes in the structure of children and adults. An appreciation of children's self-concepts self-concepts is informative of of what children of of a infants' and young children'S developmental period have in common. common. However, a full account of of early particular developmental development also requires an understanding understanding of of how the self-conceptions self-conceptions personality development of same-aged children differ differ from child to child. We proposed that individual differ differof self-concept reflect differences differences in underlying emotionality and ences in the self-concept and emerge as aa result result of of aa complex complex interaction interaction of of variables variables such such as as parental parental personality, personality, parents' parents' as of the child, children's own behavior and/or temperament, and the perceptions of infant-caregiver attachment attachment relationship. relationship. We We suspect suspect that that the the emotionality emotionality inherent inherent infant-caregiver in one's self-concept provides provides aa common common core core between between early early self-conceptions self-conceptions and and one's self-concept in those held held later later in in life. life. Focusing Focusing on on the the emotional emotional basis basis of of personality personality should should those allow for for aa better correspondence between between research research on on child child and and adult adult personality, personality, allow better correspondence eventually leading to to a a better better understanding understanding of of the the course course of of personality personality develop developeventually leading ment. If this this proves proves to to be be the the case, case, then then it it will will eventually eventually be be possible possible to to predict predict the the ment. If course of of individual individual self-conceptions self-conceptions across across the the life life span. span. course

ACKNOWLEDGMENTS We thank Lila Braine, Susan Crockenberg, Alan Elms, Elms, Bob Emmons, Eva Schepeler, and David Watson for their insightful comments on earlier versions of this paper. Nancy Kohn's help with references is also appreciated.

REFERENCES infant-mother attachment. In B. M. CaldAinsworth, M. D. S. (1973). The development of infant-mother Cald Review of of child childdevelopment developmentresearch research (Vol. 3, pp. 1-94). (Vol. 3, pp. 1-94). well & H. N. Ricciuti (Eds.), Review Chicago: University of Chicago Press. Chicago: Ainsworth, M. D. S. S. (1983). Patterns of infant-mother attachment as related to maternal development:An An interactional interactional & V. Allen (Eds.), Human development: care. In D. Magnusson & (Eds.), Human perspective, (pp. 35-55). New York: York: Academic Press. perspective, & Wall, S. (1978). Patterns Patternsof ofattachment: attachment: Ainsworth, M. D. S., Blehar, M., Waters, E., E., & AA psychologicalstudy studyof ofthe the strange strangesituation. situation, Hillsdale, Erlbaum. Hillsdale, NJ:NJ: Erlbaum. psychological

CHAPTER 9

EARLY PERSONALITY DEVELOPMENT EARLY

235 23S

Diagnosticand andstatistical statisticalmanual manual mental disorAmerican Psychiatric Association. (1980). Diagnostic ofof mental disor DC: Author. ders (3rd ed.). Washington, DC: Arend, R, R., Gove, F., & Sroufe, L. A. (1979). Continuity of individual adaptation from infancy to kindergarten: A predictive study of ego-resiliency and curiosity in preschoolers. Child Development, Development,50, 50,950-959. 950-959. Child Bannister, D., & & Agnew, J. (1976). The child's construing of self. In J. Cole (Ed.), Nebraska Nebraska onMotivation Motivation(pp. (pp.100-125). 100-125).Lincoln: Lincoln:University UniversityofofNebraska NebraskaPress. Press. Symposium on Symposium Handbook of infant Bates, J. E. (1987). Temperament in infancy. In J. D. Osofsky (Ed.), Handbook infant development (pp. (pp.1101-1149). 1101-1149).New NewYork: York:Wiley. Wiley. development Bates, J. E., & Bayles, Bayles, K. K. (1984). (1984). Objective Objective and and subjective subjectivecomponents components in in mothers' mothers'perception perception 30(2), 111-130. of their children childrenfrom fromage age 66months monthsto to33years. years.Merrill-Palmer Merrill-PalmerQuarterly, Quarterly, 30(2), 111-130. Belsky, J., & & Isabella, R R. (1988). Maternal, infant, and social-contextual determinants of attachment security. In J. Belsky & & T. Nezworski (Eds.), Clinical Clinical implications implicationsofof attachment,(pp. (pp.41-94). 41-94).Hillsdale, Hillsdale,NJ: NJ:Erlbaum. Erlbaum. attachment, & Taylor, D. (1984). The origins of individual differences in infant infantBelsky, J., Rovine, M., & Development, 55, 55, mother attachment: Maternal and infant contributions. Child Development, 718-728. Block, J. (1989). Critique of the the act frequency approach to personality. Journal Journalof ofPersonality Personality Psychology,56, 56,234-245. 234-245. and Social Social Psychology, role of ego-control ego-control and and ego-resiliency in the the organization organization Block, J. H., & Block, J. (1980). The role of behavior. In W. A. Collins (Ed.), Minnesota MinnesotaSymposia Symposiaon onChild ChildPsychology Psychology(Vol. (Vol. 13, pp. 39-101). Hillsdale, NJ: NJ: Erlbaum. Bowlby, J. (1973). Attachment New York: Basic Books. Attachmentand andloss: loss:Vol. Vol.22Separation. Separation. New York: Basic Books. Bowlby, J. (1980). Attachment Attachment and andloss: loss:Vol. Vol.3.3. Loss, Loss,sadness sadnessand anddepression. depression.New NewYork: York: Basic Books. (2nd York: Basic Books. Bowlby, J. (1982). Attachment Attachmentand andloss: loss:Vol. Vol.1.1.Attachment Attachment (2nded.). ed.).New New York: Basic Books. (Original work published 1969). Bretherton, I. (1991). Pouring new wine into old bottles: The social self self as internal working model. In ofof Child Psychology In M. M. Gunnar & L. L. A. Sroufe Sroufe (Eds.), (Eds.),Minnesota MinnesotaSymposium Symposium Child Psychology (Vol. 23, pp. 1-41). Hillsdale, NJ: NJ: Erlbaum. Bretherton, I., & & Beeghly, M. (1982). Talking about internal states: states: The acquisition of an explicit theory of mind. mind. Development Development Psychology, Psychology,18,18,906-921. 906-921. Bretherton, I., Fritz, J., Zahn-Waxler, Zahn-Waxier, c., C, & & Ridgeway, D. (1986). Learning to talk about emotions: A functionalist perspective. Child Development, emotions: Development,55, 55,529-548. 529-548. Buck, R R. (1977). Nonverbal communication of affect in preschool children: children: Relationships Psychology, with personality and skin conductance. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 35, 35, 225-236. Buss, A. H., & Plomin, R developing personality traits. Hillsdale, R. (1984). Temperament: Temperament:Early Early developing personality traits. Hillsdale, NJ: NJ: Erlbaum. Campos, J. J., Barrett, K. C, C., Lamb, M. E., Goldsmith, H. H., & & Stenberg, C. (1983). Socioemotional development. In P. H. Mussen (Series Ed.) & & M. H. Haith & & J. J. ancy and Campos (Vol. Eds.), Handbook Handbook of ofchild childpsychology: psychology:Vol. Vol2.2.Inf Infancy anddevelopmental developmental psychobiology (pp. (pp. 783-915) New York: Wiley. Camras, L. L. A., A., Ribordy, Ribordy, S., S., Hill, Hill, J., J., Martino, Martino,S., S.,Spaccarelli, Spaccarelli,S., S.,&& Stefani, Stefani,R R.(1988). (1988).Recogni Recognition and posing of emotional expressions by abused children and their mothers. Devel Developmental opmental Psychology, Psychology,24, 24,776-781. 11^-1%\. Cassidy, J. (1988). Child-mother attachment and the self in in six-year-olds. Child ChildDevelopment, Development, 59, 121-134. 59,

236 236

EDER AND ANDMANGELSDORF MANGELSDORF EDn

R. (1988). Conceptions of states and traits: Chaplin, W. F., John, O. P., & Goldberg, L. R Dimensional attributes with ideals as prototypes. Journal of of Personality Personality and and Social Social Psychology, Psychology, 54, 54,541-557. 541-557. Cole, P. M. (1985). Display rules and the socialization of affective affective displays. In G. Zivin (Ed.), The Biology-environment interactions (pp. (pp. The development developmentof ofexpressive expressivebehavior: behavior: Biology-environment interactions 269-290). Orlando, FL: Academic Press. Child Crittenden, P. M. M. (1985). Maltreated infants: Vulnerability and resilience. Journal of of Child Psychology Psychology and andPsychiatry, Psychiatry,26, 26,85-96. 85-96. Crockenberg, S. B. B. (1981). Infant Infant irritability, mother responsiveness, and social support Development, 52, influences on the security of infant-mother infant-mother attachment. Child Development, 857-865. Daniels, D., Plomin, R, R., & Greenhalgh, J. J, (1984). Correlates of difficult difficult temperament temperament in infants. Child Development,55, 55,1184-1194. 1184-1194. Child Development, Dunn, J. (1988). The beginnings beginningsof ofsocial socialunderstanding. understanding.Oxford, Oxford,England: England:Basil BasilBlackwell. Blackwell. (1987). Conversations about feeling states between Dunn, J., Bretherton, I., & Munn, P. (1987). Psychology,23, 23,132-139. 132-139. mothers and their young children. Developmental Psychology, R. A A. (1988). The self conceived in memory. Early Child Development and Care, Eder, R Care, 40, 25-51. 40, Eder, R R. A A. (1989). The emergent personologist: The structure and content of of 3i-, Si-, 5i-, and 7i-year-olds' concepts of themselves and other persons. Child Development, Development,60, 60,121812181228. Eder, R R. A. (1990). Uncovering Uncovering children's children's psychological psychological selves: selves:Individual Individualand anddevelopmental developmental differences. Child Development, Development, 61, 61,849-863. 849-863. A, Gerlach, S. Eder, R R. A., S. G., & Perlmutter, M. (1987). (1987). In search of children's selves: Develop Development of the specific and general components of the self-concept. Child Development, Development, 58, 58, 1044-1050. The early earlyemergence emergence self-monitoring: Individual R. A., & Jones, M. (1989, April). The Eder, R A, & ofof self-monitoring: Individual PaperPaper presented differences and production ofemotional expressions. differencesininthe therecognition recognition and production of emotional expressions. presented at the biennial meeting of the Society for for Research in Child Development, Kansas City, MO. mothers' pereceptions Eder, R Mothers*self-perceptions, self-perceptions, mothers* pereceptions R. A., & & Mahmood, N. (1989, July). Mothers' Paper presented at the meeting of of of self-perceptions. of their theirchildren, children,children's children*s self-perceptions. Paper presented atbiennial the biennial meeting the International Society for the Study of Behavioral Development, Jyvaskyla, Finland. develop Egeland, B., B., & Farber, E. (1984). Infant-mother Infant-mother attachment: Factors related to its developtime. Child Development, Development,55, 55,753-771. 753-771. ment over time. Egeland, B., & Sroufe, L. L. A A. (1981). (1981).Attachment Attachmentand andearly earlymaltreatment. maltreatment.Child ChildDevelopment, Development, 52, 44-52. 52, Unmasking the theface. face.Englewood EnglewoodCliffs, Cliffs,NJ: NJ:Prentice PrenticeHall. Hall. Ekman, P., & Friesen, W. (1975). Unmasking R. N. (1983). The The prerepresentational prerepresentational self self and and its its affective affective core. core. Psychoanalytic PsychoanalyticStudy Study Emde, R of the the Child, Child,38, 38,165-192. 165''192. of R. N., Kligman, D., Reich, J., & Wade, T. (1978). Emotional expressions in infancy: Emde, R Initial studies of social signaling and an emergent model. In M. M. Lewis & & L. L. Rosenblum of affect (pp. 125-148). New York: Plenum Press. (Eds.), The development of S. (1973). The self-concept self-concept revisited. American Psychologist, Psychologist,28, 28,404-416. 404-416. Epstein, S. Field, T. M., & Walden, T. (1982). Perception and production of of facial expressions in infancy child develop developand early childhood. In H. W. Reese & L. P. Lipsett (Eds.), Advances in child ment and behavior (Vol. 16, pp. pp. 169-211). New York: Academic Press.

CHAPTER 99

EARLY PERSONALITY PERSONALITY DEVELOPMENT DEVELOPMEIST EARLY

237 237

Freud, S. (1950). Instincts and their vicissitudes. Collected Sigmund Freud. Collected papers papers of ofSigmund Freud. London: Hogarth Press. (Original (Original work published 1915) Freud, S. (1955). Beyond the pleasure principle. Collected Sigmund Freud. Freud. London: Collected papers papers of ofSigmund Hogarth Press. (Original work published 1920) Gaensbauer, T. J., Harmon, R. J., Cytryn, L., & McKnew, D. H. (1984). Social and affective affective development in infants with a manic-depressive parent. parent. American American Journal of of Psychiatry, Psychiatry, 141, 14h 223-229. Goldsmith, H. H., Bradshaw, D., & Riesser-Danner, L. (1986). Temperament as a potential development influence influence on attachment. In J. V. Lerner & R R. M. Lerner (Eds.), Tempera Temperament and social social interaction in infants and young young children. (New Directions in Child Development Development No. 31, pp. 5-34). San Francisco: Jossey-Bass. Goldsmith, H. H., Buss, A. H., Plomin, R., Rothbart, M. K., Thomas, A., Chess, S., Hinde, R R. A., & McCall, R R. B. (1987). Roundtable: What is temperament? Four approaches. Child Development, Development, 58, 505-529. Goldsmith H. H., & Campos, J. J. (1986). Fundamental issues in the study of early tempera temperament: The Denver twin temperament temperament study. In M. E. Lamb, A. L. Brown, & B. Rogoff Rogoff (Eds.), Advances Advances in developmental developmental psychology psychology (Vol. 4, pp. 231-283). Hillsdale, NJ: Erlbaum. Goldsmith, H. H., Losoya, S. H., Bradshaw, D. L., & Campos, J. J. (1994). Genetics of personality: A A twin study of of the five factor model and parental-offspring parental-offspring analyses. R. Martin, & G. Kohnstamm Kohnstamm (Eds.). The developing developing structure of of In C. Halverson, R temperament and and personality personality from from infancy to adulthood, adulthood, (pp. 241-265). Hillsdale, temperament NJ: Erlbaum. Gunnar, M. R., R, Mangelsdorf, Mangelsdorf, S., Larson, M., & Hertsgaard, L. (1989). Attachment, tempera temperament, and adrenocortical, activity in infancy: A A study of psychoendocrine regulation. Developmental Developmental Psychology, Psychology, 25, 355-363. Hart, D., & Damon, W. (1986). Developmental trends in self-understanding. Social Cognition, Cognition, 4, 388-407. Harter, S. (1982). The perceived competence scale for children. Child 87-97. Child Development, Development, 53, 53,87-97. Harter, S. (1983). Developmental Developmental perspectives on the self-system. In P. H. Mussen (Series Ed.), & E. M. Hetherington (Vol. Ed.), Ed.), Handbook Handbook of of child child psychology: psychology: Vol. 4. Socializa Socialization, personality, development (pp. 275-385). New York: Wiley. personality, social social development Hudson, J. A. (1986). Memories are made of this: General event knowledge and the develop developof autobiographical autobiographical memory. In K. Nelson (Ed.), Event Event knowledge: knowledge: Structure ment of and function function in development development (pp. 97-118). Hillsdale, NJ: Erlbaum. and R. (1993). Origins of of attachment: Maternal interactive behavior across the first year. Isabella, R Child Development, Development, 64, 605-621. Child R., & Belsky, J. (1991). International synchrony and the origins of infant-mother infant-mother Isabella, R, Child Development, Development, 62, 373-384. attachment: A A replication study. Child Izard, C. E. (1977). Human Human emotion. New York: Basic Books. R. R R. (1987). Infants' emotion expressions to acute Izard, C. E., Hembree, E. A., & Huebner, R pain: Developmental Developmental change and stability of of individual differences. differences. Developmental Developmental Psychology, 23, 105-113. Psychology, Izard, C. E., & Malatesta, C. Z. (1987). Perspectives on emotional development I: Differential Differential of early emotional emotional development. In J. Osofsky (Ed.), Handbook Handbook of of emotions theory of development (2nd ed., pp. 494-554). New York: Wiley. infant development James, W. (1892). The principles principles of of psychology. psychology. New York: Holt.

EDER AND MANGELSDORF EDERANDMANGELSDORF

238

Kagan, J. (1982). Psychological Psychologicalresearch researchonon human infant: An evaluative summary. thethe human infant: An evaluative summary. New New York: W. T. Grant Foundation. Kagan, J. (1984). The nature of of the child. child. New NewYork: York:Basic BasicBooks. Books. Kagan, J., Resnick, J. S., & Snidman, Snidman, N. N. (1988). (1988). Biological Biologicalbasis basisof ofchildhood childhood shyness. shyness.Science, Science, 240, 240, 167-171. A., Ford, L. M., & Meacham, 1. J. A. (1978). Dimensions of of self-concept self-concept in preschool Keller, A, 3-489. DevelopmentalPsychology, Psychology,14,14,48483-489. children. Developmental A, II. (1986). Thoughts on the intentional basis of early object word extension: Kuczaj, S. A., Evidence from from comprehension comprehension and production. In S. A. Kuczaj, II & M. D. Barrett Progress in cognitive development research (Eds.), The The development developmentof ofword wordmeaning: meaning: Progress in cognitive development research (pp. 99-120). New York: Springer-Verlag. A, II, & Maratsos, M. P. (1975). Merrill Kuczaj, S. A., (1975). What children can say before they will. will MerrillQuarterly,21, 21,89-11 89-111. Palmer Palmer Quarterly, 1. M.. E., Thompson, R R. A., Gardner, W., & Charnov, E E.. L. (1985). Infant-mother Lamb, M attachment. attachment Hillsdale, NJ: Erlbaum. Lazarus, R R, S. (1975). The self-regulation self-regulation of emotion. In L. Levi (Ed.), Emotions—Their Emotions-Their 47-67). parameters parameters and andmeasurements measurements(pp. (pp. Al-dl).New NewYork: York:Raven RavenPress. Press. infancy and early childhood. In J. Osofsky (Ed.), Lewis, M. (1987). Social development in infancy Handbook 19-493). New York: Wiley. Handbook of of infant development (pp. 4419-493). ZoU, D. (1984). (1984). Maternal Lyons-Ruth, K., Connell, D., Grunembaum, H., Botein, S., & Zoll, family history, maternal caretaking, and infant infant attachment attachment in multiproblem families. family Preventive Preventive Psychiatry, Psychiatry,2,2,403-425. 403-425. Mahler, M. (1967). On human symbiosis and the vicissitudes of individuation. Journal of of the 740-763. American 15,15, American Psychoanalytic PsychoanalyticAssociation, Association, 740-763. A. (1975). The psychological psychologicalbirth birthof ofthe infantNew NewYork: York: Mahler, M., Pine, F., & Bergman, A .theinfant. Basic Books. Malatesta, C. Z., & Wilson, A A. (1988). Emotion Emotion cognition cognition interaction interaction in in personality personality develop development: A decrete emotions functionalist functionalist analysis. British Journal Journalof of Social SocialPsychology, Psychology, 27, 27, 91-112. Mangelsdorf, S., Gunnar, M., Kestenbaum, R, R., Lang, S., & Andreas, D. (1990). Infant Infant proneness-to-distress temperament, maternal personality, and mother-infant mother-infant attach attachment: Associations and goodness of fit. Child Development, Development, 61, 61,820-831. 820-831. Mangelsdorf, S., Plunkett, J., Dedrick, c., C, Berlin, BerUn, M., Meisels, S., & McHale, J. (1996). Attachment security in extremely-low-birth weight infants. Developmental DevelopmentalPsychology, Psychology, 32, 914-920. 32, Markman, E. M., & Callanan, M. A A. (1984). (1984). An analysis of hierarchical classification. In R. J. Sternberg (Ed.), Advances Advances in inthe thepsychology psychologyofofhuman human intelligence 325-365). R intelligence (pp.(pp. 325-365). Hillsdale, NJ: Erlbaum. Markus, H., & Sentis, K. (1982). The self in social information information processing. In J. Suls (Ed.), Psychological (Vol. pp.pp. 41-70). NJ: Erlbaum. Psychologicalperspectives perspectivesononthe theself self (Vol1, 1, 41-70).Hillsdale, Hillsdale, NJ: Erlbaum. Matheny, A A. P., & Wilson, R R. S. (1981). Development tasks and rating scales for laboratory temperament. JSAS Documents in Psychology, assessment of infant temperament, JSASCatalog Catalogof ofSelected Selected Documents in Psychology, JI(MS. No. 2367). i7(MS. No. McCrae, R R. R, R., & Costa, P. T., Jr. (1985). Updating Updating Norman's Norman's "adequate "adequate taxonomy": taxonomy": Intelli IntelUgence and personality dimensions in natural language and in questionnaires. Journal 721 . of 49,49, 710710-721. of Personality Personalityand andSocial SocialPsychology, Psychology, R. R, R., & Costa, P. T., Jr. (1987). (1987). Validation of a five-factor five-factor model of personality McCrae, R 52,52, across instruments and observers. Journal Journalof of Personality Personalityand andSocial SocialPsychology, Psychology, 81-90. -

CHAPTER 99 EARLY EARLYPERSONALITY PERSONALITYDEVELOPMENT DEVELOPMENT

239 239

Miller, P. H., & Aloise, P. A. (1989). Young children's understanding of the psychological causes of behavior: A review. Child Development, Development,60, 60,257-285. 257-285. Mischel, W., Shoda, Y., & Rodriguez, M. L. (1989). (1989). Delay Delay of of gratification gratification in in children. children. Science, Science, 244, 244, 933-937. 933-937. New York: Rutledge && Kegan Paul. Piaget, J. (1952). The origins origins of of intelligence intelligenceininchildren. children. New York: Rutledge Kegan Paul. Radke-Yarrow, M., Cummings, E. M., Kuczynsky, L., & Chapman, M. (1985). Patterns of of attachment in two- and three-year-olds in normal families and families with parental depression. Child Child Development, Development,56, 56,884-893. 884-893. Rholes, W. W. S., S.,&&Ruble, Ruble,D. D.N. N.(1984). (1984).Children's Children'sunderstanding understandingofofdispositional dispositionalcharacteristics characteristics of others. Child Development, Development,55, 55,550-560. 550-560. Ridgeway, D., Waters, E., & Kuczaj, S. A., II. (1985). Acquisition of emotion-descriptive emotion-descriptive 18 months to 6 years. language: Receptive and productive vocabulary norms for ages 18 Developmental DevelopmentalPsychology, Psychology,21,21,901-908. 901-908. Rovee-CoUier, C. (1987). Learning and memory in infancy. In J. Osofsky (Ed.), Handbook Rovee-Collier, of infant development (pp. 98-148). New York: Wiley. of Ruble, D. N., & Rholes, W. S. (1981). The development of children's perceptions and attributions about their social world. In J. H. Harvey, W. Ickes, & R. F. Kidd (Eds.), directionsininattribution attributionresearch research (Vol. 3, pp. 3-36). Hillsdale, Erlbaum. New directions (Vol. 3, pp. 3-36). Hillsdale, NJ:NJ: Erlbaum. Schachtel, Psychiatry,10, 10,1-26. 1-26. 'Schachtel, E. G. (1947). On memory and childhood amnesia. Psychiatry, Schneider-Rosen, K., Braunwald, K. G., G., Carlson, Carlson, V., V.,& &Cicchetti, Cicchetti,D. D.(1985). (1985).Current Currentperspec perspectives in attachment theory: Illustration from from the study of maltreated infants. MonotheSociety Societyfor forResearch Researchin in Child Development, 194-210. graphs of of the 50, 50, 194-210. graphs Child Development, Shantz, C. V. (1983). Social cognition. In P. H. Mussen (Series Ed.) & 1. J. H. Flavell & E. M. M. Markman Markman (Vol. (Vol. Eds.), Handbook of ofchild childpsychology: psychology:Vol. Vol3.3.Cognitive Cognitive developE. Eds.), Handbook develop ment (pp. 495-555). New York: Wiley. self-monitoring of expression behavior. Journal of of Personality Personalityand and Snyder, M. (1974). The self-monitoring Social Psychology, Psychology,30, 30,526-537. 526-537. Social Sroufe, L. A. (1979). Socioemotional development. In J. Osofsky (Ed.), Handbook of of infant development (pp. 462-516). New York: Wiley. Sroufe, L. A. (1990). An self. In In D. D. Cicchetti Cicchetti & & M. M.Beeghly Beeghly An organizational organizational perspective perspective on on the the self. (Eds.), The to to childhood, (pp.(pp. 281-307). Chicago: University Theself selfin intransition: transition:Infancy Infancy childhood, 281-307). Chicago: University of Chicago Press. Sroufe, L. A., A., & &Waters, Waters,E. E.(1977). (1977).Attachment Attachmentasasananorganizational organizationalconstruct. construct.Child ChildDevelop Development, ment, 48, 48, 1184-1199. infant: A view from psychoanalysis and and Stern, D. N. (1985). The The interpersonal interpersonalworld worldofofthe the infant: A view from psychoanalysis developmental developmentalpsychology. psychology.New NewYork: York:Basic BasicBooks. Books. Sullivan, H. S. (1953). The interpersonal interpersonaltheory theoryof ofpsychiatry. psychiatry.New NewYork: York:Norton. Norton. Tellegen, A. (1985). Structures of mood and personality and their relevance to assessing anxiety, with an emphasis on self-report. In A. H. Tuma & J. D. Maser (Eds.), Anxiety and the anxiety disorders disorders (pp. (pp.681-706). 681-706).Hillsdale, Hillsdale,NJ: NJ:Erlbaum. Erlbaum. Tellegen, A. (1982). Manualfor Personality Questionnaire. Unpublished Manual forthe theMultidimensional Multidimensional Personality Questionnaire. Unpublished manuscript. University of Minnesota. Thomas, A., & Chess, S. (1977). Temperament Temperamentand anddevelopment. development.New NewYork: York:BrunnerIMazel. Brunner/Mazel. van den Boom, D. C. (1989). Neonatal irritability and the development of attachment. In G. A. Kohnstamm, J. E. Bates, & M. K. Rothbart Rothbart (Eds.), Temperament Temperamentin inchildhood childhood (pp. 299-318). Chichester, England: Wiley. Vaughn, B. B. E., Bradley, C. F., Joffe, L. S., Seifer, R., & Barglow, P. (1987). Maternal "difficharacteristics measured prenatally are predictive of ratings of temperamental "diffi-

240 240

EDn EDER AND ANDMANGELSDORF MANGELSDORF

culty" on the Carey Infant Infant Temperament Questionnaire. Developmental Developmental Psychology, Psychology, 23, 23, 152-161. Vaughn, B. E., Lefever, G. B., Seifer, R., & Barglow, P. (1989). Attachment behavior, 728-737. attachment security, and temperament during during infancy. Child ChildDevelopment, Development,60, 60,728-737. Waters, E. (1995). The attachment attachment Q-set. In E. Waters, B. E. Vaughn, G., Posada, & K. Kondo-Ikemura. Caregiving, cultural, and cognitive perspectives on secure-base behavior and working models: New growing points of attachment research. Mono Monographs of of the Society Society for for Research ResearchininChild ChildDevelopment, Development,60(2-3), 60(2-3),234-246. 234-246. Infant-mother attachment: Theories, models, recent data, Waters, E., & Deane, D. (1982). Infant-mother and some tasks for comparative developmental analysis. In L. W. Hoffman, Hoffman, R. A. Gandelman, & H. R. Schifman Schifman (Eds.), Parenting: Parenting:Its Itscauses causesand andconsequences consequences(pp. (pp. 19-54). Hillsdale, NJ: Erlbaum. & Deane, Deane,K. K.E. E.(1985). (1985).Defining Definingand andassessing assessingindividual individualdifferences differencesininattachment attachment Waters, E., & Q-methodology and the organization of behavior in infancy infancy and early relationships: Q-methodology of the Society Society for for Research ResearchininChild ChildDevelopment, Development,5050, 41-65. , 41-65. childhood. Monographs of A. (1984). (1984).Negative Negativeaffectivity: affectivity:The Thedisposition dispositionto toexperience experienceaversive aversive Watson, D., & Clark, L. A. emotional states. Psychological Bulletin,96, 96,465-490. 465-490. Psychological Bulletin, J., & & Clark, Clark, M. M. C. C.(1986). (1986).Individual Individualvariation variation in inattachment attachment security security Weber, R. A., Levitt, M. J., infant temperament. Child Child and strange situation behavior: The role of maternal and infant Development, 57, 57, 56-65. 56-65. Wellman, H. M., & Gelman, S. A. (1987). Children's understanding of the nonobvious. In of human intelligence intelligence(Vol. (Vol.4,4,pp. pp. R. J. Sternberg (Ed.), Advances in the psychology of 99-135). Hillsdale, NJ: Erlbaum.

CHAPTER CHAPTER 10 10

FAMILY INFLUENCES ON PERSONALITY DEVELOPMENT CHARLES F. JR. CHARLES F , HALVERSON, JR. UNIVERSITY UNIVERSITY OF OF GEORGIA GEORGIA

KAREN KAREN S. S. WAMPLER WAMPLER TEXAS UNIVERSITY TEXAS TECH TECH UNIVERSITY

I. INTRODUCTION INTRODUCTION I.

This essay is mostly about missing data. As we began the review of the voluminous literature on develop on parent-child parent-child relationships and and their effects effects on personality personaUty development, it became apparent to us that there is yet another another emerging crisis in the study of personality. In the past decade, there has been an increasing awareness that the data on personality development in the family context are seriously incomplete and problematic. This state of affairs all studies of the influence affairs exists because nearly all of the family on personality development have used a model that assumed that the influences of family family characteristics could adequately be assessed by by using mostly one-time assessments of some aspect of the family focused on one parent family that focused parent— usually the mother-and mother—and one child in in each family. And indeed, over the past 75 or 80 years we have amassed an enormous data base documenting how parents' child-rearing styles influence the child. Certainly influence the main-effect, single-child, single-parent models assessed in many studies have the main-effect, personalshown relations between parenting styles and children's temperament and personal 1949; Baum Baume.g., Baldwin, Kalhorn, & & Breese, 1949; ity. Beginning with the early studies ((e.g., 1967,1971,1983; & Moss, 1962; 1962; Sears, Maccoby, & & Levin, 1957; 1957; Sears, Sears, rind, 1967, 1971, 1983; Kagan & IlANDBOOK OFOF PERSONALITY PSYCHOLOGY HANDBOOK PERSONALITY PSYCHOLOGY241 241

COmlGIIT ACADEMIC PlESS COPYRIGHT C © 1997 1997 BY BY ACADEMIC PRESS.. ALL RIGHTS OF REPRODUCTION IN ANY FORM RESERVED.

ALL RIGHTS OF REPRODUCHON IN ANY FORM RESERVED.

242 242

HALVERSON AND ANDWAMPLER WAMPLER HALVERSON

Rau, & & Alpert, 1965; Yarrow, Campbell, & & Burton, 1968), we have indications of of family family influences influences on children. For For example, Baldwin's analyses of of the Fels Pels Study showed showed democratic parent parenting to be associated associated with assertiveness, vigor, high activity level, and sociability. In contrast, controlling parenting was associated with obedience, fearfulness, fearfulness, and found that that authori authoriwithdrawal (Baldwin, 1949). Baumrind's classic study (e.g., 1967) found tarian, controlling parents had children who were dependent and not particularly particularly sociable. In contrast, authoritative parents (rational, loving, firm parents) had had inde independent and socially competent competent children. Countless studies have subsequently subsequently replicated replicated in one form form or another another these general main-effect findings. Parenting has been studied in terms of main-effect findings. of control, warmth, combinations of of the three. Consistent, high levels of of involvement, and various combinations associated with children who are not aggressive and parental control have been associated generally well controlled controlled (J. H. Block & & Block, 1979; Patterson Patterson & & Bank, 1989), and of sociability and initiative. Low levels of of control (permissive (permissivewho show high levels of ness) have been associated with the opposite poles of of the above dimensions (e.g., high activity level; see Maccoby & & Martin, 1983; B. Martin, aggression, impulsivity, high & Thomas, 1979, for for excellent summaries of of this traditional literature literature). 1987; Rollins & ). affection in families-particularly families—particularly mothers-are mothers—are associated associated Similarly, warmth and affection of prosocial behaviors behaviors with attachment attachment security, compliance, altruism, and a range of of competence, including agreeableness and self-esteem for exam examindicative of self-esteem (see, for & Wall, 1978; Bretherton, 1985; Graziano & & Eisen Eisenple, Ainsworth, Blehar, Waters, & berg, this volume, chap. 30; Radke-Yarrow Radke-Yarrow & & Zahn-Waxler, Zahn-Waxier, 1984, 1984,1986). 1986).

n. FOR PERSONALITY PERSONALITY DEVELOPMENT DEVELOPMENT RESEARCH RESEARCH ll. THREE CRISES FOR

A. Direction of of Effects Effects The first first real crisis that influenced influenced the socialization literature could be called the bidirectional watershed. In 1968 R. Q. Bell published published his now classic review of of the literature demonstrating demonstrating the plausibility of of child effects effects on parents. Since that that time, it has become nearly pro forma forma to refer refer to correlational data obtained obtained in socialization studies as supporting interpretations interpretations of of either child effects effects on parents or vice versa (see R. Q. Bell & Harper, 1977; Harper, 1989). The crisis was that it was no longer clear who might be responsible for for children's personality development. Maybe, just parent-child correlations reflected reflected the action action of of constitutional characteris characterismaybe, parent-child tics of of the child on parenting and not the other other way around. B. Transactional Transactional Models The second second crisis that further further complicated complicated the picture of of the family family influence on personality personality development was the elaboration elaboration of of the transactional model of of develop development ment (A. Sameroff Sameroff & & Chandler, 1975). In Sameroff's Sameroff s original original formulation formulation of of the

CHAPTER 10

FAMILY INFLUENCES

243 243

environment were conceptualized as always in model, the child and the caretaking environment a state of mutual feedback; infant characteristics modified parental practices, which modified infant infant behavior that then further further modified in turn modified modified parental behavior. From this point of view, the partners in an intimate relationship over time could never be characterized without examining the reciprocal causality of one or the other shifting covariances between adults over time. Investigators need to examine the shifting offspring over time, paying close attention to the mechanisms underlying such and offspring 1992, especially chapters 7 and 8, for for an excellent discussion covariance (see Wachs, 1992, issues). Since the original formulation, there has been a veritable flood of of these issues). of examtheoretical and empirical articles from this transactional perspective (see, for exam ple, Belsky, 1981, 1981, 1984; 1984; Belsky, Rovine, & & Fish, 1989; 1989; Lerner, 1989; Lerner & & ple, 1978; R. P. Martin, 1983, Sameroff & & Seifer, 1983; 1983; Wachs, 1983; Spanier, 1978; 1983, A. J. Sameroff & French, 1971; Werner & & Smith, Smith, 1977, 1977,1982). These theorists theorists 1971; Werner 1982). These Werner, Bierman, & propose that neither the family context, nor the child, nor the physical environment significant determinant of any specific adaptive outcome. can be viewed as the sole significant A. Sameroff Sameroff and Chandler (1975) elegantly pointed out that when developdevelop ment is considered prospectively rather than retrospectively, there is a considerable effect of children or parents on development. In fact fact it is not lessening of the main effect main-effect characteristics that reside either in the possible to trace many simple main-effect organism or in the child-rearing environment (usually thought of as the parents). These authors proposed a continuum of caretaking causality. causality.AtAtone oneend, end,the the of caretaking sufficient to compensate for almost any developmental parental environment is sufficient difficulties. deviation so that it would not result in later behavioral or intellectual difficulties. On the other end of the continuum, the parenting environment lacks the resources from that which is normal in the child, thus to deal with even minor deviations from providing an environment that serves to maintain maladaptive behavior over time. identified three socialization models: models; (a) the single-factor model that that The authors identified single-factor model of development focuses on either constitutional or environmental determinants of the main-effects main-effects model, model, either parents parents or children), children), (b) the interactional interactionalmodel, model (e.g., the where child factors combine additively with parental characteristics so that child characteristics would be augmented or reduced by better or worse parenting envienvi ronments, and (c) the transactional transactionalmodel model inin which which development development isisproposed proposed toto result from the continual interplay between a changing organism and a changing family environment. Not only is it no longer clear who affects affects whom, but it is also not clear whether stability or change resides in organisms, environments, or some complex interplay between the two. C. Within-Family Within-Family Effects Effects c. The third crisis in socialization research on personality development is of of relatively is, in our opinion, the most far-reaching far-reaching in its implications. Nearly recent origin and is, all of the studies of family and parental influences on the personality of the child focus on the mother (as most are based on only one child per family. Whether they focus have), on the father father (a recent innovation), or on the family as a unit (an even more

244 244

HALVERSON HALVERSON AND AND WAMPLER WAMPLER

relationship is typically described for only recent innovation), innovation), the relationship only one child per family. There consideration of sibling and There has has been almost almost no consideration and parent similarities similarities and differences differences in terms of personalities personalities and and relationships relationships within within the same families. families. It is important remember that we have long made generalizations important to remember generalizations about about the effects effects of children and parents on each other other (generic children and parents at that), as if these relations hold within families these relations hold within families when when data have have been been derived derived from from between betweenfamily family analyses. For example, example, the oft-found oft-found relation relation of parental parental punitiveness punitiveness and child aggression interpreted as holding aggression is interpreted holding for parent-child parent-child relations in general, a effect as it were, for for all all children children in in families; families; if if one had had aa punitive, punitive, authoritarian authoritarian main effect parent, the outcome outcome was likely likely to be aggressive children (presumably (presumably all the children children family!). in the family!). About developmental behavior About 10 years ago, ago, the developmental behavior genetics genetics researchers researchers began began question the logic logic of this argument argument based based on their research research with twins and to question and adoptees (Plomin, DeFries, & & Loehlin, 1977; 1977; Scarr & & McCartney, McCartney, 1983; 1983; Scarr, adoptees Webber, Weinberg, Weinberg, & & Wittig, Wittig, 1981). In an early study of twin data (Loehlin (Loehlin & & Nichols, 1976), the startling conclusion conclusion was was that there appeared appeared to be no n o common common family variance variance that led to sibling sibling similarity similarity in personality. personality. From From the data investiga investigafamily tors concluded concluded that about about 50% of the variance variance in child personality personality was was genetic; the rest due to apparently apparently "nonshared" environmental environmental variance. In a scholarly scholarly review review of the existing data on twins, adoptees, and siblings, Plomin and Daniels (1987) of existing and Plomin and Daniels (1987) concluded that children in the same family do not resemble each other very much concluded that children same family resemble each other very much in personality personality development. In contrast, identical twins, even those raised apart, development. identical those were similar and had similar similar developmental developmental spurts spurts and andlags lags (see (see Matheney, Matheney, 1989).1 1989).^ were this "discovery" **discovery" that that environmental environmental factors factors are experienced experienced differently differently It is this by siblings growing growing up in the same same families families that is one of the major major findings findings in the by behavior genetic research. Siblings Siblings resemble resemble each each other other to a degree, presumably presumably behavior genetic because of shared D DNA, because N A , not shared shared experiences. The The behavior behavior geneticists geneticists designate designate this category '" "specific," or, more generally, category of environmental environmental influence influence as "E "Ei", or, more generally, l "nonshared." This index index of "nonshared" environment environment comes comes from from data that com compare sibling MZ sibling differences differences in personality personality to similarities, similarities, particularly particularly in studies studies of M Z twins twins raised raised together together and apart when when compared compared to siblings raised in biological biological and adoptive adoptive families. families. The The first first notice notice of the nonshared nonshared environment environment came came from from Loehlin Loehlin and Nichols' classic classic study study of personality personality in high school school twins twins (1976). They They concluded concluded was important important for determining personality—about half half the vari varithat environment environment was determining personality-about ance—but it was environment where twin pairs were not not correlated. Environmental Environmental ance-but environment where effects, instead instead of of being uniform uniform withm were random random and not not predictable predictable effects, within the family, were (see also Dunn Dunn & & Plomin, 1991; Rowe, Rowe, 1990). If there are are main effects effects of parenting parenting family variables variables on personality personality development documented in the large litera literaand family development (as documented ture on o n parental parental control, love, protectiveness, sensitivity, sensitivity, etc.), then why why don't don't children in the same family resemble each other more than they do? Plomin children in the same family resemble each other more than they do? Plomin and and

^But Butsee seeMcCartney, McCartney,Harris, Hams,and andBemieri Bernieri(1990) (1990)who whoshow showsome someadult adultdiminution diminutionofofidentical identical

1

twin twin similarities over age in those twinships raised apart compared to those raised in close contact.

CHAPTER CHAPTER 10 10 FAIIULY FAMILYINFLUENCES INFLUENCES

245 245

Daniels (1987) also summarized data from four recent adoption studies that show that the average adoptive parent-child personality correlations average about .05. The sibling sibUng correlations in such families are for the most part below .20 on personal personality measures (see also Scarr et al., 1981). Although it is obvious that at some some level, siblings sibUngs share the same family environment, it is also obvious obvious that important characteristics of that environment environment are largely nonshared for for personality devel development. A A corollary to the "nonshared" issue must be noted here as it also has profound profound imporeffects on traditional socialization research. In addition to discovering the impor of environment environment in child measures, the behavior geneticists geneticists have recently tance of documented the importance of genetically mediated effects effects on most of of our tradi tradidocumented & Bergeman, 1991, tional measures of "environment" (see Loehlin, 1992; Plomin & A number of of studies have shown that our environmental for extended discussions). A measures are themselves in part measures with a genetic influence. If our measures differences in parents, raters, and others who are really proxies for individual differences architecture can passively passively create correla correlapopulate family environments, then genetic architecture characteristics and our outcome tions between our predictor measures of family characteristics measures of child behavior. There may be some cause for pessimism here. Studies confounded even at the level of biological parents and their offspring may also be confounded of supposedly "independent", "environmental" assessments. Taken along with the of both parenting devastating critiques of self-report measures as valid indicators of functioning (Holden & & Edwards, 1989; Wampler & & Halver Halvertechniques and family functioning of genetic confounding confounding of of our measures must be carefully son, 1993), the issue of of personality development. examined in future research on family predictors of of this chapter that this essay was mostly about We asserted at the beginning of withinmissing data. Given the perspectives of child effects, transactionism, and low within family concordances on personality dimensions, most of of the literature appears to be only weakly applicable to discovering what it is in families that may contribute to personaUty development. development.^2 In fact, most of of the relations obtained in the socialization personality functioning may have really been mostly literature on parental impact on child functioning child main effects on parents. documenting the child literature—but should be cited as A study not often cited in the socialization literature-but of the paradigmatic studies in socialization-is socialization—is a model socialization experiment one of influenced by the work of of Scott and done by Freedman (1958). Freedman was influenced differences Fuller (1965) who showed that there were considerable temperamental differences of dogs, beagles, basenjis, wire-haired fox fox terriers, among four purebred breeds of of dogs know, dogs come in a remarkably and Shetland sheep dogs. As every owner of wide range of temperamental propensities. Freedman examined the interaction environbetween "inborn" "inborn" temperamental characteristics and a classic parenting environ ment: whether the "parent" was strict or permissive.

2^A Arecent recentarticle articlehas hasoffered offeredsome someinteresting interestingspeculations speculations on onwhich whichparent�child parent-childsystems systemsmay may influence (Cairns et at., al., 1990). be most susceptible to stability, change, and mutual inl1uence

246 246

HALVERSONAND ANDWAMPLER WAMPLER HALVERSON

Dogs from from each breed breed were socialized daily from from the 3rd week of of life through the 8th week. Dogs from from each breed breed were assigned to either a permissive-indulged permissive-indulged permissive-indulged regimen was one where the dogs or strict rearing regimen. The permissive-indulged encouragement of of play and aggression aggression via were never punished, and there was much encouragement human "parent." In the strict regimen, rough and tumble interactions with the human there was much emphasis on teaching to sit, stay, come, and so on (basically, the characterized by many dog schools). classic obedience training characterized After 5 weeks of of this regimen, a classic resistance-to-temptation resistance-to-temptation paradigm paradigm After was employed to test the effects effects of of the parenting regimen. The dogs were all hungry and were all placed in front front of of a bowl of of meat, and for for 3 min the caretaker caretaker prevented the animal from from eating by swatting the rump of of the dog with a rolled newspaper newspaper approached the food. The handler handler then and shouting "No!" every time the dog approached left the room room and the experimenter experimenter recorded recorded the elapsed time before before the dog ate left the meat. This test situation was repeated for for 8 days. What is so elegant about this model is that it examined examined breed breed (temperament) by rearing strategy strategy interactions interactions for important developmental outcome—the ability to resist temptation. Obviously, an important developmental outcome-the dogs who were not able to inhibit their impulse to eat in such a test would immedi immediately eat the food. If If parenting had a main effect, effect, then the disciplined dogs would indulged dogs would. If there were no important important breed breed differ differnot eat, whereas the indulged WOUld. If could attribute any systematic differences differences to the rearing environment. environment. ences, then we could results were indeed indeed curious. The Shetland Shetland sheep dogs never never ate the food food under The results either regimen. In contrast, all the basenjis basenjis ate the food food right away-it away—it did not either matter whether whether they were raised in a permissive or a strict environment. environment. matter interaction between breed and the rearing environment environment for for There was an interaction from what we would predict from from terriers and beagles: it was, however, the reverse from human parenting parenting theory. The indulged indulged terriers and beagles took took much longer to human eat the food food than the disciplined terriers and beagles. The importance of of this study eat that it underscores underscores a fairly fairly obvious point: If If the researchers had studied only is that found that the socialization socialization environment environment terriers and beagles they would have found interactively interactively determined determined self-control self-control (even if backward from from theory). If, If, however, however, determined that socializa socializathey had studied Shetlands or basenjis, they would have determined ineffective.^ tion was ineffective.3 D . Metaphors firom from Biology D. This study serves to illustrate two metaphors that are useful useful when discussing human human socialization socialization studies, reaction reaction range and canalization, canalization, both concepts introduced introduced by 3 ^Crawford Crawfordand andAnderson Anderson(1989) (1989)provide providefurther furtherexamples examplesfrom fromanimal animalresearch. research.They Theypropose propose gene-by-environthat the behavior genetic models emphasizing heritability coefficients may miss many gene-by-environ ment interactions that operate at the "breed" "breed" level and hence have zero heritabilities. For example, Sackett, Ruppenthal, Farenbruch, Holm, and Greenough (1981) found that rearing different breeds of of had differential differential effects effectsby bybreed, breed,with withsome somebreeds breedsbeing beingrelatively relativelyunaffected unaffected monkeys in social isolation had (see also Hinde & & Stevenson-Hinde, 1973, 1973, for more examples).

CHAPTER 10 CHAPTER

FAMILY INFLUENCES

247

Waddington (1957, 1962). 1962). Although neither of these concepts can be measured directly in human socialization studies, they serve to point to important possibilities in the research literature. The first of these, reaction reaction range, range,refers referstotothe thefact factthat that heredity is not rigidly related in any way to behavior, but instead establishes a range of possible responses to different different environments. In the dog example, the socialization environments provided were within the reaction range for for terriers and beagles, but were ineffective ineffective with the other two breeds. Closely related is the canalization, which whichrefers refers to tothe thefact fact that thatsome some genotypes genotypeswill willbe bemore more concept of canalization, difficult individudifficult to deflect from a maturational pattern of growth than others. Some individu als are difficult difficult to deflect from from a pattern while others are relatively easy. Within deflect while others will turn out to individuals some traits are relatively easy to deflect difficult (see Cairns, Gariepy, & & Hood, 1990, for for a different be relatively difficult different view).

E. E. Studies from the Parent-Child Domain For the parent-child & parent-child area, we use an example from from our own research (Halverson (Halverson & 1981). One way to begin to assess the relative contributions of the parent Martin, 1981). and the child to the developing relationship is by having data on parents before effects of and after they have children and data on infants before the possible effects of 1989; Cowan, sociaUzation have had much cumulative impact (see also Belsky et al, socialization aI., 1989; & Miller, 1991). 1991). We need to replace retrospective socialization Cowan, Heming, & accounts with actual event sequences over time (see as an example Jack Block's from childhood to adult adultexcellent study of the stability and change in personality from hood; J. Block, 1971). For this analysis we used data from the Bethesda Longitudinal Study (see, & Moss, 1976; Yang, Zweig, Douthitt, & & Federman, 1976). for example, Jacobs & data at five The sample included only those mothers and children with complete data five points in time-newlywed, time—newlywed, pregnancy (3rd trimester), birth, postpartum (3 months), and preschool (3 years). The analysis involved all five time periods and and included parent-infant influences, 28 boys, 28 girls, and their mothers. To study reciprocal parent-infant we devised a measure of maternal competence for the five time periods based on seven adjectives in a cluster score: and— score: calm, confident, organized, and thrifty, and negatively weighted-dependent, weighted—dependent, selfish, and temperamental. We divided the mothers into stable and unstable groups, and we used an iterative procedure that maximized the correlations between periods for for one group while minimizing the correlations between periods for the other. All correlations nonsignificant and differed differed statistically (ps<.001) in the unstable group were nonsignificant (ps<.OO1) from significant coefficients coefficients (average r = = .62). the stable group's significant factor analytically at each period. Specifi SpecifiData on the children were derived factor cally, the behavioral measures indicating high-magnitude behavior at each period princifactor analyses, and the first were subjected to separate principle components factor first princi ple component in each of these three analyses was chosen and represented an activity/irritability dimension at each period.

248 248

ANDWAMPLER WAMPLER HALVERSON AND

Our initial prediction was that mothers with stable attitudes (both high and low) would form form stable relationships with their infants infants and, therefore, therefore, the infants infants would show stability of of their behavior. The activity-irritability activity-irritability behavior in infants infants in terms of of mothers with stable attitudes did not fulfill fulfill our expectations. Indeed, the correlations among among the three periods periods were all essentially essentially zero. In contrast, contrast, for mothers unstable in competency, competency, we we found found that infants of of these mothers were highly stable over time (average r = = .54). Further, activity-irritability activity-irritability at 3 months predicted predicted maternal adequacy adequacy when the child was 3 years old. As A s 3-month irritability irritability increased, maternal maternal competence competence at 3 years decreased decreased (r[28] = = - . .65, 65, P p< < .(01). .001). No N o comparable comparable correlation correlation was obtained for for stable mothers and infants infants (r[28] = = .06). In addition, the contemporaneous contemporaneous correlation correlation between child and maternal scores at the 3-year period was highly significant = significant (r[28] = - . .71, 71, p p < < .(01), .001), with increases in activity activityirritabiUty again associated associated with decreases in maternal competence. No N o comparable comparable irritability correlations were obtained between child data and maternal data for for the stable group. Considering Considering the findings findings for for both stable and unstable groups, there was a infant effect effect on mothers mothers who who were predisposed predisposed to having having unstable unstable feelings feelings strong infant of competency. Infants Infants who showed showed strong strong continuities over time had mothers who who of showed discontinuities discontinuities over time. It appears possible that these infants' tempera temperashowed mental mental characteristics were influential influential in contributing contributing to maternal instability instability in the area of of competency competency as evidenced evidenced by the negative negative correlations correlations between between infant infant and maternal data over time. On the other hand, for stable mothers, it may have been been of competency competency that disposed them them to be relatively relatively immune to the their stability of influences of of infant temperamental temperamental characteristics. influences When we divided our stable and unstable samples into mothers who had boys boys or girls, we found found that instability instability in mothers was primarily primarily in the male sample. To T o explore explore this sex difference difference further, further, we analyzed analyzed the mean levels levels of of maternal competency competency over the four four periods as a function function of of having either a boy or a girl child. A A maternal maternal competence competence (stable, unstable) by sex of of child (male, female) female) by stage stage (marriage, pregnancy, postpartum, 3-year) repeated measures analysis of variance revealed a significant significant three-way interaction among stability of of maternal maternal competence, sex, and stage. unstable attitudes during pregnancy pregnancy and had had a boy boy showed showed Mothers who had unstable significant declines in competency competency for for both the postpartum and the 3-year periods. significant The other three groups groups were not significantly significantly different different from each other at any The period. These data are consistent consistent with our other findings findings in that it was the less lesscompetent mothers with irritable and fussy fussy males who found found that child-rearing child-rearing may competent been as difficult dif ficuh as they expected. In contrast, if mothers entered into pregnancy pregnancy have been feelings of of competency, competency, they were able to cope well with or childbirth with stable feelings active/irritable male child. A A most · interesting interesting finding finding was that those those having an active/irritable mothers who who tended to have unstable unstable feelings feelings of of competency competency but who had a female female mothers have discovered discovered child-rearing child-rearing was not so bad after after all (since (since their scores scores child may have did not decline over time).

CHAPTER 10 CHAmR

FAMILY INFLUENCES FAMILY

249 249

Our results also show that when mothers are stable across the four time periods. periods, infants show no stability of activity/irritability. In contrast, when mothers infant functioning are unstable, high stability of infant functioning was obtained. We interpret interpret these results as implicating a strong parent effect on infants infants when when maternal maternal behavior behavior was was stable. In contrast, a strong child effect effect on on mothers motherswas wasimplied impliedwhen whenmothers motherswere were unstable. Specifically, stable children seem to affect affect maternal instability whereas affect child instability. stable mothers seem to affect The points to be emphasized are (a) that we must characterize both the child and the family environment as differentially differentiallycanalized canalized and and susceptible susceptible to todirection direction effect (from child to family family or family family of effects, and (b) that the main direction of effect to child) will depend on important characteristics of both parenting and child further. systems. Two others examples from our research emphasize this point further.

F. Congenital Congenital Contributors Contributors to to Interactions Interactions F. Clearly, there is a need to examine both child and family effects effects in any study. There is also a requirement to identify identify and measure child and family characteristics independently of each other in order to examine reciprocal effects. As we have seen earlier in this paper, this is very hard (impossible?) to accomplish when children and parents share genes that relate (confound) (confound) "environments" and outcomes. One solution is the twin and/or adoption study. Another would be to use biological families but look for a possible early early environmental environmental contributor contributor to to personality personality that that would make children within families different different from each other very early on. If we could locate a nongenetic but congenital contributor contributor to individual differences, we could examine "how the twig is bent" within families of biologically biologically related individu individuals, searching for both "child-driven" and "family driven" effects over time. searching for both and "family driven" effects over als, investigated a congenital contributor contributor to the In our past research, we have investigated expression of sociability and impulsivity as related to activity level (R. Q. Bell & & 1973; Waldrop, Waldrop, 1982; 1982; Halverson & & Victor, 1976; 1976; Halverson & & Waldrop, 1973; 1976; Waldrop, Bell, McLaughlin, & & & Halverson, 1978; 1978; Waldrop & Bell, & & Goering, 1976; Halverson, 1971; 1971; Waldrop, Pedersen, & & Bell, 1968). These These characteristics characteristics of of impul impuloften been part of the formulations formulations of child temperament sivity and sociability have often personality (A. H. Buss, 1988; 1988; A. R. Buss & & Plomin, 1975; 1975; 1984; 1984; Eysenck, and personality 1970).

/. Child Child Minor Minor Physical Physical Anomalies Anomalies 1. The congenital contributor to these behaviors is indexed by the number of minor physical anomalies present at birth. There are 18 anomalies that can be assessed at any age and consist of minor growth abnormalities of the head, hands, and feet. for their occurrence in Down's Syndrome but The 18 anomalies are best known for present in the general population with an average of 2 to individual anomalies are present 4 anomalies per person (range 0-18). These minor developmental deviations most from either some form of genetic transmission or from from some event likely result from Renfield, occurring in embryogenesis that mimics genetic transmission (Quinn, Renfield,

HALVERSON HALVERSON AND ANDWAMPLER WAMPLER

250

& Rapoport, Rapoport, 1977; Rapoport, Quinn, & & Lamprecht, 1974; Waldrop Waldrop & & Halverson, Halverson, Burg, & 1971). The speculation is that the anomalies are the sequelae of of events occurring in the first trimester of of pregnancy pregnancy that also affect affect the physiology, structure, or biochemistry of of the central nervous system that can later lead to problems of of either under- or overcontrol. Stability Stability from from birth to 7 years is high ((rr = = .87; Waldrop & & Halverson, 1971). The importance of of the the anomaly index in predicting impulsive, aggressive behavior in boys aged 2 to 12 years has been established established in many studies (R. Q. & Waldrop, 1982; Burg, Hart, Quinn, & & Rapoport, 1978; Firestone, Peters, Bell & & Knights, 1978; Halverson, 1989b; Halverson Halverson & & Victor, 1976; Quinn et Riviere, & aI., al., 1977; Rapoport Rapoport et aI., al., 1974; von Hilsheimer Hilsheimer & & Kurko, 1979; Waldrop et aI., al., 1968, 1978; Waldrop & 1968,1978; & Halverson, 1971). Longitudinal data data have indicated that that the of anomalies at birth predicts, at 3 years of of age, the behavioral triad of of incidence of impulsivity, short attention aI., 1978) attention span, and high activity activity for for males (Waldrop et al., withdrawn and low activity behavior behavior for for girls (R. Q. Bell & & Waldrop, 1982; and withdrawn of this past past research points to a strong Halverson, 1989a; Waldrop et al., aI., 1976). All of effect for for the anomaly variable. main effect speculated that that the anomalies arise from from mild teratogenic teratogenic agents in We have speculated pregnancy (e.g., viruses, high blood pressure, stress) that tend to slow development development pregnancy differential exposure to circulating circulating androgens androgens necessary at a time when there is the differential for sex differentiation, differentiation, with males being exposed exposed to the male hormones and females females for preglacking such exposure. We suspect that there must be interactions between preg nancy stress agents and male hormone in the first trimester which produces this indexed later by high- and low-magnitude behaviors. dimorphic response as indexed 2. Study Siudy 1: Anomaly-Environment Anomaly-Environment Interactions Interactions 2. The anomaly index was done at birth, at 3 months, and at 3 years of of age in the Bethesda study. Anomalies of of the head include circumference circumference out of of normal range, low-set or malformed malformed ears, epicanthal folds, hyperteliorism, high steepled palate, and furrowed furrowed tongue. Anomalies Anomalies of of the hand include single transverse creases, short fifth finger, and clinodactyly of of the fifth finger. Anomalies of of the feet feet include wide spacing between between first and second toes, partial syndactyly syndactyly of of the second and third toes, and third third toe longer than than second second (see Halverson & & Shetterley, 1989, for for scoring details). For the present analysis we used the 3-year 3-year index because of of missing data at the earlier periods, although it should be noted noted that for for a subsample of of 23 a l , 1978), the index was stable from from birth to 3 years ((rr = = .86) boys (Waldrop et aI., interrater reliability ((rr = = .87). Children were divided at the median and had good interrater larger sample of of 199 into high- and low-anomaly low-anomaly children (scores based on the larger ranged from from 0 to 14, median median = = 4.2). Since there was a sex difference difference in the anomaly ranged of high- and low-anomaly low-anomaly boys score (boys had more), there were unequal subsets of and girls. from a larger longitudinal longitudinal sample of of 199 Research participants were drawn from ir parents. In tum, three-year-old children and the their turn, the parents were a subsample of of three-year-old Bethesda Longitudinal Study. Study. a larger cohort of parents who were participants in the Bethesda

CHAPTER 10 10 FAMILY FAMILYINFLUENCES INFLUENCES CHAPTER

251 251

Data were collected at five points in time: 3 months of marriage, during the first pregnancy, at the birth of of the child, 3 months postpartum, and when the child first was 3 years of of age. For the present analysis, we focus on 131 children (69 boys, 62 girls) with complete data at the marriage and preschool periods. As part of of an extensive test battery at 3 months of of marriage, each spouse responded to an exten extensive questionnaire on their marital relationship. Through cluster and factor analysis based on much larger samples of of around 2,000 couples, one prominent factor emerged that described couple couple complaints complaints (for example, see Ryder, Kafka, & & Olson, 1971). Based on 45 items for each spouse, a factor score was computed for each couple. The factor score was reliable and cross-validated across several subsamples of the larger sample. The score summarized complaints that each spouse had in of the area of of recreation, affection, friends, sexual satisfaction, money, housekeeping, and communication. For the present analysis, couples were split at the median (based on the number of of couples in the longitudinal sample, N N = = 199) into high highcomplaint couples or low-complaint couples at 3 months of of marriage. of age, children attended a 4-week mixed-sex nursery school. At 3 years of Comprehensive Comprehensive assessments of of their behavior were done by a team of three observ observfemale teacher who observed and rated each child's behavior ers and a male and female settings—indoor free play, rest time, and outdoor free play. daily in three behavior settings-indoor Analysis focused focused on behaviors that were reliable as well as stable across time of 27 items and setting. Data were summarized and condensed by factor analysis of including time-sampled observations of of play and social behavior, ratings of of child of behavior in experimental settings designed behavior by the teachers, and codings of to assess problem-solving ability and competence. Two varimax rotated factors emerged from the analysis: impulsivity and sociability. Data were cast into a 2 (complaints at early marriage-high, marriage—high, low) by 2 (males, females) x X 2 (anomalies-high, (anomaUes—high, low) design for each of of the two dependent variables, impulsivity impulsivity and sociability. Since Ns Ns were small and unequal from cell to cell, data were analyzed with planned comparisons for each dependent variable. For impulsivity, there was an additive effect effect of of anomalies and marital com complaints on the expression of of impulsive behavior for males. High-anomaly males with high-complaint high-complaint parents were significantly more impulsive than any of the other groups, whereas the low-anomaly low-anomaly boys with low-complaint low-complaint parents were less impulsive than any of of the other groups. When one factor was present, either anoma anomalies Ues or complaints, boys were moderately impulsive and not different different from each other. There were significant significant main effects for both anomalies and complaints. For girls, the data indicate that neither anomalies nor complaints moderated moderated the expres expression of impulsivity (or better, the lack of expression of impUlsivity). impulsivity). Girls were uniformly low and differed differed significantly significantly from all boys except those who were low on both anomalies and complaints. For sociability, sociability, the effect effect of newlywed complaints on 3-year-old children's sociability was moderated moderated by whether children were high or low anomaly. For both boys and girls, complaints affected affected low-anomaly children. In boys, the least sociable children were the low-anomaly boys with high-complaint parents, but for girls the

252 252

ANDWAMPLER WAMPLER HALVERSON AND

combination of of low anomalies and low complaints complaints led most sociable group led to the most of of girls. Between Between the sexes, low-anomaly, low-anomaly, high-complaint high-complaint boys boys were were significantly significantly there was a trend for social than the same subgroup of of girls, yet yet there for the low lowless social anomaly/low-complaint than their male counterparts. The anomaly/low-complaint girls to be more sociable than their male The effect for for anomalies anomalies was marginally marginally significant significant for for boys boys ((p p = = .06) and girls main effect ((p p = = .10) .10).. The main effect effect for complaints was significant significant for for boys and borderline borderline for girls (p = .10). for (p = We can see see how a child congenital congenital variable variable and a family together We family variable together contribute development. While contribute to improve improve our predictions predictions of of personality personality development. While there were main effects newlywed com effects on child behavior for for both child anomalies anomalies and newlywed complaints, we can see how in some of a child moderated some cases the constitutional constitutional variable variable of moderated the impact of the present of a preexisting preexisting family characteristic. For For the present sample, high highanomaly complaints than were anomaly children children were were less affected affected by variation variation in family family complaints were low-anomaly low-anomaly children. For the low-anomaly low-anomaly children, children, the the predominant predominant influence influence in this limited expected, the gender of hmited data set was family family complaints. Further, as expected, of the influenced the nature of of the problem behavior behavior when when it occurred, with child also influenced behavexternalizing behaviors, and girls more more internalizing internalizing behav boys exhibiting more externalizing Because of of the different different behavioral behavioral styles associated associated with imporiors. Because with gender, it is impor measure more than one one outcome outcome variable. For For example, example, impulsivity for girls tant to measure impulsivity for variable—in fact, most of children on was not a very important variable-in of the low-scoring low-scoring children dimension were were girls and there was a large sex for the total samthis dimension sex difference difference for total sam t[ 127] = = 4.07, P p < < .001). If we we had used only only impulsivity, impulsivity, we we would would have ple ((t[127] concluded concluded that there was a clear additive additive effect effect of of constitution constitution and families families for boys, but nothing for girls. With sociability dependent variable, however, we sociabiUty as the dependent observed differential impact observed both the congenital congenital and the family family variables variables having having a differential on boys and girls. We have completed analyses in a new new longitudinal longitudinal sample of young children and families above findings (Halverson & famihes that generally replicate replicate the above findings (Halverson & Wampler, 1993). Other results in the literature support the above transaction findings. findings. Matheney Matheney (1986) found especially when found stability in temperament temperament between between 12 and 24 months, especially when changes in sociability sociability (increases) mothers mothers were involved involved and expressive. expressive. Further Further changes (increases) over that same time period were associated associated with the same same maternal maternal personality personality characteristics. Engfer of German German mothers Engfer (1986) found found in a sample sample of mothers that infants infants when the parents were under increased in difficulty difficulty more from 4 to 18 months when stress than when they were not under stress. Belsky Belsky et aaI., Belsky (1984; Belsky l , 1989) found found similar results. There are characteristics of continuity, but of the child responsible responsible in part for for continuity, variables probably probably play a role as well. Patterson and Bank maternal and family variables transactional system system involves involves family (1989) proposed that the transactional family characteristics, marital forth that may at times "overshadow" the impact relationships, depression, and so forth of an aggressive child on the parent-child they tested and of parent-child relationships. Further, they found considerable considerable support for the parallel parallel continuities continuities hypothesis—that antihypothesis-that child anti found interval when when there was concomitant concomitant social behavior was stable over a 2-year interval continue to see see new new studies of development development stability in the parenting system. We continue studies of

CHAPTER CHAPTER 10 10

FAMILY FAMILYINRUENCES INFLUENCES

253 253

from the transactional perspectives that followed followed Belsky's (1984) important important theoriz theorizfrom ing about the determinants of that encompass more of parenting by using constructs that than one causal mechanism at more than one level in describing the child childenvironment. rearing environment.

G. The Family as a Unit factor many researchers do not include in their models is some conceptualiza conceptualizaOne factor of the family family as a unit and its impact on child behavior behavior and tion of and personality. Interest family therapy in this linkage has gained prominence with the family therapy movement movement where theorists view the functioning the primary primary locus of functioning of of the family family as a whole as the of health or pathology in children (see Gurman Gurman & & Kniskern, 1981, and Minuchin, for recent summaries of of theoretical and treatment treatment approaches within the 1988, for family therapy tradition). family family therapy approach has gained acceptance, the assump assumpEven though the family treatment of of problems in children children have not tions underlying this approach to the treatment tested systematically. Some evidence for for the impact of family family characteristics characteristics been tested impact of from positive results of of outcome studies on child characteristics can be construed from of family treatment treatment (Gurman & & Kniskern, 1981; 1981; Olson, Olson, Russell, Russell, & & Sprenkle, Sprenkle, 1989). 1989). of from studies comparing comparing families families with with a disturbed More direct evidence is available from families (Alexander, 1973; 1973; L. G. Bell & & Bell, 1992; Doane, 1978; child with control families 1963; Fischer, 1980; Garbarino, Sebes, & & Schellenbach, 1984; Jacob, 1975; Ferreira, 1963; & Phillips, 1976; Lewis, Rodnick, Michael, & & Goldstein, Lewis, Beavers, Gossett & 1981; Minuchin, Rosman, & & Baker, 1978; Mishler & & Waxler, Waxier, 1968; Riskin Riskin & & Faunce, Faunce, & Novak, 1974; Westley & & Epstein, 1969; 1972; Shepperson, 1982; Van der Veen & & AI-Khayyal, Al-Khayyal, 1982). of this research research has focused focused on Wynne, Jones, & 1982). Almost all of adolescents in treatment. See Minuchin et al. (1978) for for research with younger school-age children and L. G. Bell and Bell (1982), Lewis et et al. (1976), and Westley Epstein (1969) for for nonclinical studies. and Epstein between families found to differentiate differentiate between famihes of of Characteristics that have been found competent, well-socialized, and problem-free and multiple-problem problem-free adolescents and multiple-problem adolescents adolescents include positive affect affect (e.g., warmth, support, cohension), appropriate appropriate control (e.g., no parent-child coalition, parents in charge of of family), clear communi communication (e.g., low communication deviance, high agreement, low negative affective affective style), and problem-solving ability (e.g., flexibility, low conflict).

H. The Marital Relationship family therapists view the parents' marriage as the primary dysfunctional dysfunctional Most family subsystem in relation to negative child characteristics like aggression, impulsivity, effect of and poor ego control; that is to say, this relation is seen seen both both as a direct effect of disagreement about child-rearing and as an indirect effect of of parents who parental disagreement indirect effect avoid problems in their marriage by focusing focusing on problems in their their child. Most of of the research on the effects effects of of the marital relationship relationship on children has been done

254 254

HALVERSON AND ANDWAMPLER WAMPLER HALVERSON

by child behavior therapists who recognized that more than parental skill deficits deficits produced behavior problems in children (Cole & & Morrow, 1976; Patterson, 1976; Patterson & & Bank, 1989). Comparisons of of clinic and nonclinic groups of of children and adolescents adolescents have produced produced consistent consistent evidence that parents of of children with of marital distress (Fischer, negative personality traits also experience higher levels of 1980; Forehand, Griest, Wells, & McMahon, 1982; Griest, Forehand, Wells, & & McMahon, 1980; Johnson Johnson & Lobitz, 1974; Oltmanns, Broderick, & O'Leary, OTeary, 1977; Porter & O'Leary, O'Leary, 1980; Schwarz & Getter, Getter, 1980; Westley & Epstein, 1969). While While many investigators are moving toward using multilevel transactional family-child linkages, much of of it still remains ambiguous as to models in studying family-child of effects individual differences differences direction of effects issues. Many researchers do not include individual of children children like temperament temperament or personality. Some use retrospective retrospective measures of from only one parent, thus confounding confounding developmental, family, reports or reports from and child factors factors inherent in single-method approaches. Those studies in which the of child temperament temperament is appropriately appropriately examined rather consistently consistently find find impact of evidence for for a child effect effect (Hetherington, Stanley-Hagan, & & Anderson, Anderson, 1989; Lam Lamevidence for the bert, 1988; Powers, Hauser, Kilner, 1989). Elder makes a convincing case for impact of of child temperament temperament on later relationships, particularly marital and parent parentimpact & Elder, 1988; Elder, Caspi, & & Downey, 1986; Liker & & child relationships (Caspi & Even research based on a strong assumption of of a predominant predominant direction direction Elder, 1983). Even of effect effect from the family to the child tends to produce produce evidence evidence of of a child effect. effect. of series of of research studies, Patterson (1982; Patterson & & Bank, Bank, For example, in a series & Dishion, 1988) has produced evidence of of a strong relation between between 1989; Patterson & parenting practices and negative negative child characteristics like aggression and noncompli noncompliparenting negative ance. Yet, he also finds that parents are less adept at parenting the child with negative characteristics than they are at parenting more normal siblings, thus suggesting a effect. Patterson (1982) has also found that decreasing decreasing the child's antisocial antisocial child effect. behavior leads to decreases in the mother's depression. Our point here is not to reiterate simply the importance importance of of child effects. effects. Rather, we believe believe that much research is still guided by a unidirectional model of the family family influencing influencing the child child and that when when adequate measures of of child and parent are included, evidence evidence for for child effects effects is often more compelling than that for for family family effects. What What we are saying saying is that the magnitude of of the effects effects of of children children on parents parents may be greater greater than the magnitude of of effects effects of of parents on children. We We propose that investigators consider the magnitude of of effects effects issue seriously, seriously, represents a new new direction in the conceptualization of of the role of of parenting parenting since it represents in the family. These general points were considerably elaborated upon by Sandra Scarr presidential address to the Society Society for for Research Research in Child Development. Development. (1992) in her presidential Her thesis was that child-rearing environments within the normal species range sufficient for for normal development development but did not nor produce produce individual differences differences in were sufficient differences in genetics after the normal developmental developmental personality. Those arise from differences context applies to a great variety of of parenting parenting process were present. "Normal" in this context sufficient for for unremarkable unremarkable outcomes. When we we see differdifferenvironments that are sufficient

CHAPTER 10

INFLUENCES FAMILY INFLUENCES

255 255

reactions to different different ences in parental treatment, they may be due mostly to parental parental reactions children's personalities-a throughout this chapter. personalities—a point we have been making throughout This view has elicited much concern concern among socialization socialization researchers (see Baumrind, Baumrind, 1983; Jackson, 1993; Scarr, 1993). 1993). These These researchers researchers have been marshalling marshalling evidence evidence against against the position that being reared in one family family or or another another makes little or no difference Hoffman, 1991). Clearly, the difference in the development development of of personality (see also Hoffman, responses to the crises we have been writing about about here are beginning to appear. We have proposed that children (c. L. Martin & children also are active self-socializers self-socializers (C. & Halverson, 1981, 1983). Young children their roles vis-a-vis 1981,1983). children begin early to appraise appraise their behaviors considered considered appropriate appropriate (e.g., for for boys or others and learn and practice behaviors for persons like me). This activity may occur independently of of parenting girls, for of active gene-environment practices and is similar to the notions of gene-environment correlations from available environments (Scarr & & where children either create or pick niches from 1983; see the following following sections). McCartney, 1983;

I. New Models We summarize here models we have recently tested tested in our our new longitudinal longitudinal study of family family influences on child development development (The Georgia Georgia Longitudinal Halverof Longitudinal Study; Halver & Wampler, 1993; 1993; Wampler, Halverson, & & Deal, 1996). son & / . Study Study 2: 2: The Georgia Georgia Longitudinal Longitudinal Study Study 1. Briefly, our major major main-effect main-effect model involved involved seven latent variables derived derived for for 94 latent variables families seen in the first two waves of of our longitudinal longitudinal data set (the Georgia data Longitudinal Study). Six of of the constructs constructs were derived derived from from our our first first wave and were formed formed into a model to predict, first, child outcome in Year 2, and then then in a Year and second model, family function in Year 2. second family function predicted that child outcome in Year Year 2, operationalized operationalized as child externaliz externalizWe predicted behavior (active, impulsive, difficult difficult conduct), would be predicted predicted first first by a ing behavior direct path from from child minor physical anomalies (MPAs). This prediction prediction was based direct (summarized previously) where MPAs MP As have on our earlier work (summarized have been been consistently positively related to aggressive, impulsive, high active behavior. In addition addition to the MPAs would effect would have have an indirect indirect effect direct path to child behavior, we predicted that MPAs effects on both the constructs of on child externalizing through their effects of marital quality competent parenting. For the first time, we also employed employed parental MP As in and competent MPAs would have direct direct effects effects on on three three family our model, predicting that parental MPAs would family cohesion. The effects effects subsystems, marital quality, individual parent stress, and family of parent MPAs of As on on child aggressive, impulsive, externalizing externalizing behavior behavior could could be from marital quality, to parenting, to child outcome, and by a mediated by paths from from parent parent stress, to family cohesion, to child outcome. We predicted predicted these path from expectation that MPAs would related in adults personalwould be related adults to to parental parental personal paths in the expectation of negative emotionality, anxiety, and the tendency tendency to ity, most likely some aspect of become angry angry or or easily easily upset. upset. become

256 256

HALVERSON HALVERSON AND ANDWAMPLER WAMPLER

The second model employed the same latent variables but the focus was instead on family family cohesion cohesion in Year 2 as the outcome. We chose family family cohesion to represent represent competent competent family family functioning; functioning; our interest interest was in how aggressive, impul impulsive, externalizing child behavior behavior would have an impact impact on family family cohesion. We predicted As would would directly directly relate relate to their their externalizing externalizing behavior predicted again that child MPAs and that parental MPAs would would have a direct direct impact parent stress and and marital impact on parent family cohesion. quality, which would, in turn, relate to family Using a multimeasure (self-report (self-report and observation) definition definition of of our our constructs constructs 1993, for (see Halverson Halverson & & Wampler, 1993, for details) we found found that for for preschool boys and their families, famiUes, externalizing externalizing was mainly related to the congenital variable, MPAs. It It was not not predicted predicted by parenting or family family measures. For For girls, MP MPAs As were parenting or not related related to externalizing externalizing behavior behavior but were instead instead predicted predicted by parenting parenting quality quahty (i.e., self-ratings self-ratings of of competent competent behavior, observations of of parenting skill): the more (Le., competent the parent, the less externalizing. Using Using child externalizing as a predictor predictor competent of family cohesion, we found, for boys, a moderate effect of externalizing coheof family found, for moderate effect of externalizing on cohe stress—having difficult difficult boys had sion, as well as on marital quality and parental stress-having family systems. For girls, there was no effect effect of of externalizing externalizing an impact on all the family family cohesion. on family 2. The Revised Revised Model Model For expository purposes, we propose a match-mismatch match—mismatch model (or goodness of of & Chess, 1977) between families and children in order fit; Lerner, 1989; Thomas & importance of of aspects of of the family family environment environment that moder moderto examine the relative importance ate the direction direction of of influence influence between parents and children. The model predicts that certain temperamental temperamental or personality personality characteristics of of children associated associated with minor minor physical anomalies (e.g., impulsivity, aggression) will describe "difficult" "difficult" children. The goodness of of fit model predicts that when difficult difficult children are reared by problem problem families (those with limited resources, stress, etc.) both families families and children children will function function less well over time; the predominant predominant direction direction of of effects effects will be from to family (vulnerable to to child child effects). effects). This This could could be from child (high MPA) MPA) to family (vulnerable termed termed the "child-driven" model. When these same difficult difficult children are reared by competent, cohesive, effective effective parents, children will, over time, become less problematic problematic and the families families will continue continue to be competent over time. The direction of of effect effect in this case would be from from family family to child (the "family-driven" "family-driven" model). Rather Rather than than an effect effect of of one or or the other, our model focuses focuses on the question, under what what conditions is change predominantly predominantly child-driven and when is it predominantly predominantly family-driven? family-driven? From our revised revised transactional transactional perspective, different different child and adult domains domains may be relatively relatively important important at different different points in time. For example, during infancy, infancy, important factor factor in predicting marital the baby's personality may be the least important competent, peer-oriented peer-oriented couple, while during during later later childhood childhood relationships in a competent, difficult personality personality may assume an expanding expanding role in the the child's increasingly increasingly difficult adjustment of spouses to each other. other. adjustment of the spouses

CHAPTER 10 10 FAMILY FAMILYINFLUENCES INFLUENCES CHAPTER

257 257

In this transactional context we need to refer to yet another useful metaphor genotype—environment correla correlafrom the behavior genetics literature, that of the genotype-environment & McCartney, 1983). These authors have elegantly tion (Plomin et al, al., 1977; Scarr & biological families, parents' child-rearing and child character characterdiscussed how, within biological istics are correlated. By far the majority are positively correlated (e.g., active children with active parents) while some may be negatively correlated. Functionally, this concept is the same as the goodness of fit notion adapted from the temperament literature. The really interesting proposal about gene by environment correlations is the proposal that some environments and children passively resemble each other, while in others the child evokes evokes similarities in the environment (e.g., sociable chil children elicit sociable responses in others in many family forms), and at other times the child (usually older children, adolescents, and adults) pick niches that are congruent with their personalities (see Caspi & & Herbener, 1990, for an adult exam examcongruent ple). pie). From our research, and the model of influence we propose, the first two characconcepts, the passive and evocative correlations, need some qualifying. Some charac influential and others will not, depending on how vulnerable (or teristics will be influential buffered) the parenting system is to child effects. Further, congenital effects (i.e., gene-environment correlations into nongenetic) may lower or even turn positive gene-environment magnifying negative ones where parents will pressure children to change, possibly magnifying contrasts with siblings. While the research we have reviewed here (including our own) goes some direction-of-effects issues by focusing on transactional way toward acknowledging direction-of-effects of who will effect effect whom when, most studies are still, processes and moderators of of family effects that are shared or not by siblings unfortunately, silent on the issue of in the same family. As we reviewed earlier, most family effects effects on personality must be of of the unshared type (Plomin & & Daniels, 1987); siblings do not resemble each other to any great degree on personality measures. There are several possible responses to this very real dilemma. One possibility is that personality traits are simply not very responsive to what parents do: shaping, instruction, punishing, monitoring, discussion, or whatever. It is possible they are pretty much "given" in the biological template and that most of of the correspondences we track are either child effects on parents (idiosyncratic to each sibling within families) or shared passive genotype-environment genotype-environment correlations between children and parents. Another response is that the big differences differences among siblings in the same family may be due to unsystematic, idiosyncratic life events within each family. If, however, we were to entertain seriously this rather forbidding prospect, we would probably not do research on this issue-we issue—we would be left with essentially case studies of of life histories to account for each highly idiographic life course. In a recent study involving a sample of of Swedish twins and adoptees, Plomin, McClearn, Pedersen, Nesselroade, and Bergman (1988) found that using the Family retro Environment Scale (FES; Moos & & Moos, 1984) to obtain adult twin adoptee retrospective ratings of of their child and family environments produced puzzling results. Analyses revealed strong genetic influences on 9 of of the 10 subscales of the FES (for example, MZ twins raised apart were in substantial agreement about the

258 258

HALVERSON HALVERSON AND AND WAMPLER WAMPLER

similarity of their early environments in different different homes, homes,while whilesiblings siblingsin inthe the same same agree; (see also Pedersen, Plomin, McClearn, & & Frieberg, 1988; homes did not agree; Rowe, 1981, 1983). This approach seems limited at best. Family members do not 1981,1983). report differential differential treatment in very refined ways. We are not very confident from data in our study and others that children or parents accurately track what may be very subtle transactional processes. In our study, we used Block's Q-set items to measure parenting for two children in each of our families and are faced with little discrimination on parents' reports of practices. Parents told us that they treat all their children the same (i.e., they have global, trait-like self-perceptions). In contrast, children focus on differences between how each is treated. From their vantage point, everything is differential and state-like. We are not convinced, however, that young siblings of of differentially to personality personalitydifferent ages know that a parent is reacting differentially temperament differences in the sibling pair. Our present failure to describe the relevant constructs constructs responsible responsible for for these these sibling sibling variations variations within within families families is is due due in relevant part to the poverty of our attempts to develop decent measurement models (see & Halverson, 1993). Given our current data base, it is not possible even Wampler & to begin to document any one kind of systematic nonshared environmental effects We suspect suspect that direct observation observation combined combined with multisituation constructs constructs directly. We directly. that direct with multisituation is the the initial initial step step in in identifying identifying nonshared nonshared environmental influences. influences. is correlaWe propose a strategy that may facilitate the search. Recall that the correla tions between siblings is low (or even zero in some cases). What can be done is to select for sibling similarity and dissimilarities. With a zero correlation, one can form four equal-size groups from median splits of the variable: both siblings high on trait, both low, one high and one low, and vice versa. One can study, for example, the differences between families in those families who have siblings who are very similar on sociability and those who have siblings who are very different. If there are common family environments discriminating similar and different different siblings, we can begin to document the main important nonshared contributors to differences and similarities across families. Most of the nonshared nonsharedenvironment environmentatatthis thistime timeisisalso alsononvisible, nonvisible,arrived arrivedatat by default subtractions in in heritability equations (Wachs, 1983). We We need need hypotheses hypotheses about what the important processes might be. For example, example. Chess (1987) proposed that "poor" "poor" parenting parenting might might serve serve to to increase increase differences differences between between siblings. sibhngs. Parents Parents with limited or coercive styles that are not adaptable to differences in the children may may produce even larger largerdifferences differencesbetween betweensiblings siblingsthan thanthose thosethat thatshow showsensitive, sensitive, flexible flexibleparenting parentingmay may flexibleparenting. parenting.InInfact, fact,ititseems seemslikely likelythat thatsuch suchsensitive, sensitive,flexible serve to increase sibling similarities, similarities, not notdifferences, differences, particularly particularlywhen wheneffective effective elBfort on some valued outcome. We may not have parents expend considerable effort measured or even conceptualized well those things that parents do that enhance children (or (or siblings) siblings) their ability to influence even the most intractable, "canalized" children who start life with very different predispositions. Our discussion of the transactional perspective on personality development in goes beyond the outlines of similar models in the literature (e.g., Belsky, 1984) in

CHAPTER CHAPTER 10 10 FAMILY FAMILYINFLUENCES INFLUENCES

259 259

that some specification of the variables moderating the direction and magnitude of effects between domains over time has been attempted. We feel strongly that it is not particularly helpful to state that the child and the rearing environment are actively changing each other (Belsky, 1984; Plomin et aI., al., 1977; A. J. Sameroff, Sameroff, & & Seifer, 1983; Scarr & & Grajek, 1981). For example, Belsky's model does not really specify how domains jointly and separately contribute to either family or child development. Our model is more than a metamodel serving to explicate domains of potential relevance. What is needed is the difficult difficult theorizing and empirical work of of the specification specification of when and for what group group of of families families child characteristics would predominantly "drive" the parenting system, or when and for what group group of of families the parenting parenting system would "drive" behavior change in the child. Some characteristics take on importance only when considered in the light of of other dimen dimenof the contextual systems. sions of of the "individual differences" of of contexts contexts When the theoretical delineation of or domains domains is afforded afforded the same importance as individual differences differences in persons, persons, of relationship development. we will move closer to conceptualizing the complexity of difficult tasks for future research are specifying specifying the details of of not one "transac "transacThe difficult of such models that hold for certain families at certain tional model" but a series of times across development. Further, such models need to account not only for differential socialization sociaUzation of of parents and children variance between families but for differential within families. Such studies will need to be multicontext longitudinal longitudinal studies of of of functioning, including clinical, at-risk, multichildren families with a wide range of of families will need to be large and representative and healthy groups. Samples of so regression analyses can be cross-validated and strong external validation ensured. The task we have outlined is a major one-the one—the determination determination of of the direction and the amount of of variance that can be attributed to contexts controlling development of children and families. And it has only just begun. of summary, we suggest several guidelines to such a transactional transactionalBy way of ecological approach: 1. The nature of of family family and child adaptations should always be considered jointly. The adaptations are complex and multiply determined. The adaptations range from genetic and constitutional constitutional expression in children and parents to develop development histories and the personalities of parents and children, to the nature of of the interpersonal relationship within the family, to the social world outside the family. 2. Close attention must be paid to the shifting hierarchial nature of of the contex contextual inputs to functional functional outcomes (adaptations). Theory must be developed and tested regarding the particular structures of of these hierarchial models across across time and subgroups of of families. This process will require an interdisciplinary approach, using the theoretical insights of re of child developmentalists, marriage and family researchers, pediatricians, school personnel, and so forth. It is particularly important to assess the contributors to long-term stability from childhood into adulthood. Few studies deal with the thorny conceptual problems of of an early, relatively global of the "Big Five" personality dimensions personality developing into adult varieties of

260 260

HALVERSON HALVERSON AND ANDWAMPLER WAMPLER

(see A. H. Buss, 1988; A. H. Buss & & Finn, 1987). In fact, we we have puzzled over the almost structure beyond almost complete complete lack lack of of any research on personality personality traits traits and structure beyond the differences If we we must describe differences described described under the rubric of of temperament. temperament. If describe children, one Big Five one possible possible set of of dimensions dimensions might be the Big Five (John, 1990). In & Martin, 1994; Havill, Allen, Halverson, Halverson, & & our work (Halverson, Kohnstamm, & Kohnstamm, 1994) we have began began to develop develop measurement measurement instruments instruments for for children Kohnstamm, Five typology typology in order to begin begin to trace the continuities continuities to based on the Big Five these same constructs in adulthood. The findings findings emerging promising and we we emerging are promising elaboration of of the measurement measurement of of dimensions dimensions like Agreeableness, Agreeableness, anticipate the elaboration Conscientiousness, and Openness Openness to Experience Experience in children children and how how personality personality Conscientiousness, moderated by family family contexts contexts and vice versa. Until Until now, this area structure may be moderated empty set set in the developmental developmental area (see Halverson Halverson et al., for the research research is an empty aI., 1994, for interesting possibilities possibilities for for linking adult and agenda around these issues). Many interesting linking adult personality research are opened by having having a common overlapping set set of of common or overlapping child personality traits to trace over time. Only through the representative sampling of varieties of of cultural cultural contexts contexts 3. Only representative sampling of varieties families and children live can these models models be elaborated tested in which families elaborated and tested (Lancaster, Altmann, Rossi, & & Sherrod, 1987). Individual Individual differences differences and contextual contextual (Lancaster, differences must be coordinated coordinated in such models. differences placed on differential differential socialization socialization within within families families where where 4. Emphasis must be placed different different models may hold for different different children in the family family at the same same time time or overtime (Scarr & & Grajek, 1981).4 1981).^ overtime

ACKNOWLEDGMENTS The writing of this chapter was supported by NIMH Grants MH39899 and MH50302. The Bethesda Longitudinal Study was supported by the NIMH Intramural Program. The authors express their appreciation to William Graziano, Robert Hogan, John Johnson, Roy Martin, Hilary Rose, and James Walters for comments on earlier of this chapter. earlier versions of

REFERENCES C , Waters, E., & Walls, S. (1978). Patterns of of attachment. attachment Ainsworth, M. D. S., Blehar, M. C., NJ: Erlbaum. Hillsdale, NJ: Alexander, J. F. (1973). Defensive Defensive and supportive communications communications in normal and deviant families. Journal Journal of of Clinical and Consulting Consulting Psychology, Psychology, 40, 223-331. effect of of home environment environment on nursery school behavior. Child Baldwin, A. L. (1949). The effect Development, Development, 20, 20, 49-62. ^While Whilebeyond beyondthe thescope scopeofofthis thispaper, paper,we wemust mustconsider considerfamily familyrelationships relationshipsacross acrossthe thelifespan. lifespan. 4 We have have focused focused on on young young children, children, but but adult adult children children have have parents parents and and these these relationships relationships have have not not We 1990; been much much studied studied as as factors factors in in maintaining maintaining or or changing changing adult adult personality personality (Caspi (Caspi & & Herbener, Herbener, 1990; been Hagestad, 1990). Hagestad, 1990).

CHAPTER 10 10 FAMILY FAMILYINFLUENCES INFLUENCES

261 261

A. L., Kalhorn, J., & Breese, F. H. (1949). The appraisal of parent behavior. Baldwin, A PsychologicalMonographs, Monographs,63(Whole 6i(WholeNo. No. 299). Psychological 299). Baumrind, D. (1967). Child care practices anteceding three patterns of preschool behavior. GeneticPsychology PsychologyMonographs, Monographs,75,75, 43-88. Genetic 43-88. Baumrind, D. (1971). Current patterns of parental authority. Developmental Psychology Psychology Monographs, 4(No. 4(No. 1, 1, Part 2). Monographs, Baumrind, D. (1983). Rejoinder Rejoinder to Lewis' reinterpretation of parental firm control effects: effects: PsychologicalBulletin, Bulletin,94, 94,132-142. 132-142. Are authoritative families really harmonious? Psychological Bell, L. G., & Bell, D. C. (1982). Family climate and the role of the female adolescent: Family Relations, Relations, 31, 31,519-527. 519-527. Determinants of adolescent functioning. Family R. Q. (1968). A reinterpretation reinterpretation of the direction of effects effects in studies of socialization. Bell, R Psychological Review,75,75, Psychological Review, 81 81-95. -95. R. Q., & Harper, L. V. (1977). Child effects effects on on adults. adults.Hillsdale, Hillsdale, NJ: NJ: Erlbaum. Erlbaum. Bell, R R. Q., & «fe Waldrop, M. F. (1982). Temperament and minor physical anomalies. Ciha Bell, R Ciba Foundation Symposium,89,89, 206-220. Foundation Symposium, 206-220. Belsky, J. (1981). Early human experience: experience: A A family family perspective. perspective. Developmental DevelopmentalPsychology, Psychology, 17, 3-23. 1 7, 3-23. of parenting: A process model. Child Development, Belsky, J. (1984). The determinants of 55, 83-96. The developing developing family family system. system. In In M. M. R R. Gunnar Gunnar & & Belsky, J., Rovine, M., & Fish, M. (1989). (1989). The Minnesota Symposia Symposia on on Child Child Psychology Psychology (Vol. (Vol. 22, 22, pp. pp. 1119-166). E. Thelen (Eds.), Minnesota 19-166). Hillsdale, NJ: Erlbaum. through time. time. Berkeley, CA: Bancroft. Bancroft. Block, J. (1971). Lives through Block, J. H., & Block, J. (1979). The role role of of ego ego control control and and ego ego resiliency resiliency in in the the organization organization A. Collins (Ed.), Minnesota Minnesota Symposia Symposia on on Child ChildPsychology Psychology(Vol. (Vol. of behavior. In W. A 13, pp. 39-101). Hillsdale, NJ: Erlbaum. Societyfor for Research Researchinin Bretherton, I. (1985). Attachment theory: Retrospect and prospect. Society No.209, 209,50(1-2), 50(1-2),3-35. 3-35. Child Development, Development, No. C, Hart, D., Quinn, P., & Rapoport, J. (1978). Newborn minor physical anomalies Burg, c., and the the prediction prediction of of infant infantbehavior. behavior.Journal JournalofofAustism andChildhood Childhood Schizophrenia, Austism and Schizophrenia, 8, 8, 427-429. Personality: Evolutionary andand human distinctiveness. Hillsdale, Buss, A A. H. (1988). Personality: Evolutionaryheritage heritage human distinctiveness. Hillsdale, NJ: Erlbaum. Buss, A Personality A. H., & Finn, S. E. (1987). Classification Classification of personality traits. Journal of of Personality and Social Social Psychology, Psychology,52, 52,432-444. 432-444. New Buss, A. R., R, & & Plomin, R R. (1975). A temperament temperamenttheory theoryofofpersonality personalitydevelopment. development. New York: Wiley (Interscience). Early developing personality traits. Hillsdale, Buss, A A. R, R., & & Plomin, Plomin,R R. (1984). (1984). Temperament: Temperament: Early developing personality traits. Hillsdale, NJ: Erlbaum. R. B., (1990). Development, microevolution, and social Cairns, R B., Gariepy, J., & Hood, K. E. (1990). behavior. Psychological Review,97, 97,49-65. 49-65. Psychological Review, behavior, er, G. H. (1988). Emergent family Elder, family patterns: The intergenerational construc construcCaspi, A., A, & Eld R. A. Hinde & J. Stevenson-Hinde tion of problem behavior and relationships. In R Relationshipswith withfamilies: families: Mutual influences 218-240). Oxford, England: Mutual influences (pp. (pp. 218-240). Oxford, England: (Eds.), Relationships Oxford Oxford University Press. Caspi, A., A, & Herbener, E. S. (1990). Continuity and change: change: Assortative marriage and the Psychology, consistency of personality in adulthood. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 1990, 58, 58, 250-258.

262 262

HALVERSON AND ANDWAMPLER WAMPLER

temperament and fortuitous events. Brain and Chess, S. (1987). Let us consider the roles of temperament Behavior Behavior Sciences, 10, 21-22. Cole, C c.,, & Morrow, W. R R. (1976). Refractory parent behavior in behavior modification modification training groups. Psychotherapy, Psychotherapy, 13, 162-169. Cowan, C. P., Cowan, P. A., Heming, G., & Miller, N. B. (1991). Becoming a family: Marriage, parenting and child development. In P. A. Cowan & E. M. Hetherington (Eds.), Family transitions NJ: Erlbaum. transitions (pp. 126-141). Hillsdale, NJ: . Crawford, C. B., & Anderson, J. L. (1989). Sociobiology: An environmentalist discipline? American American Psychologist, Psychologist, 44, 1449-1459. A. (1978). Family interaction and communication deviance deviance in normal and disturbed Doane, J. A families: A A review of research. Family Process, 17, 357-376. R. (1991). Why are siblings so different? The significance of of differences differences Dunn, J., & Plomin, R in sibling experiences within the family. Family Process, 30, 271-284. Elder, G. H., Jr., Caspi, A., A, & Downey, G. (1986). Problem behavior and family relationships: A. B. Sorensen, F. E. Weinert, & L. R R. Life course and intergenerational themes. In A Human development development and the life course: Multidisciplinary Multidisciplinary perspectives perspectives Sherrod (Eds.). Human (pp. 293-340). Hillsdale, NJ: Erlbaum. A. (1986). Antecedents of behavior problems in infancy. infancy. In. A A. Kohnstamm (Ed.), Engfer, A Temperament discussed discussed (pp. 98-1 98-112). North America. Temperament 12). Berwyn, PA: Swets North Eysenck, H. J. (1970). The structure of of human personality personality (Rev. ed.). London: Methuen. A. (1963). Decision making in normal and pathologic pathologic families. Archives Archives of of General Ferreira, A Psychiatry, 8, 68-73. Psychiatry, R. M. (1978). Minor physical anomalies in Firestone, P., Peters, B., Riviere, M., & Knights, R Journal of of Child Child Psychol Psycholhyperactive, retarded, and normal children and their families. Journal ogy and Psychiatry, Psychiatry, 19, 155-160. ogy Fischer, J. L. (1980). Reciprocity, agreement, and family style in family systems with a and Adolescence, Adolescence, 9, 391-406. disturbed and nondisturbed adolescent. Journal Journal of of Youth and R, Griest, D. L., Wells, K. c., Forehand, R., C , & McMahon, R R. J. (1982). Side effects effects of of parent counseling on marital satisfaction. Journal Journal of of Counseling Counseling Psychology, Psychology, 29, 104-107. Freedman, D. G. (1958). Constitutional and environmental interactions in rearing four breeds of of dogs. Science, 127, 585-586. Garbarino, J., Sebes, J., & Schellenbach, C. (1984). Families at risk for for destructive parent parentchild relations in adolescence. Child Child Development, Development, 55, 174-183. Griest, D. L., Forehand, R, R., Wells, K. c., C , & McMahon, R R. J. (1980). An examination of of differences between nonclinic and behavior-problem-clinic-referred behavior-problem-clinic-referred children and their mothers. Journal Journal of of Abnormal Abnormal Psychology, Psychology, 89, 497-501. Gurman, A A. S., & Kniskern, D. P. (1981). Family therapy outcome research: Knowns and A. S. Gurman & D. P. Kniskern (Eds.), Handbook Handbook of of family family therapy therapy unknowns. In A (pp. 742-775). New York: BrunnerlMazel. Brunner/Mazel. Hagestad, G. O. (1990). Social perspectives on the life course. In R R. Binstock & L. George (Eds.), Handbook Handbook of of aging and and the social social processes processes (pp. 35-61). San Diego, CA: Academic Academic Press. Halverson, C. F. (1989a). Activity implications for Activity level: Measurement Measurement issues and and implications for personality personality structure in children. Unpublished manuscript. Minor physical physical anomalies anomalies and and sex differences behavior. Unpub UnpubHalverson, C. F. (1989b). Minor differences in behavior. lished manuscript. A., & Martin, R R. P. (1994). The developing developing structure structure of of Halverson, C. F., Kohnstamm, G. A, temperament from infancy to adulthood. temperament and personality personality from adulthood. Hillsdale, NJ: Erlbaum.

CHAPTER 10 10

FAMILY FAMILY INFLUENCES INFLUENCES

263 263

Halverson, C. F., & Martin, C. L. (1981, April). Parent-child over time. Paper Parent-child stability stability over presented at the meeting of the Society for Research in Child Development, Boston. Halverson, C. F., & Shetterly, for minor minor physical physical anomalies. Shetteriy, K. (1989). Revised Revised manual manual for anomalies. Unpublished manuscript. Halverson, C. F., & Victor, J. B. (1976). Minor physical anomalies and problem behavior Development, 47, 281-285. in elementary school children. Child Development, relations of of mechanically recorded activity Halverson, C. F., & Waldrop, M. F. (1973). The relations level to varieties of of preschool behavior. Child Child Development, Development, 44, 678-681. of child externalizing Halverson, C. F., & Wampler, K. S. (1993). The mutual influence influence of R. E. behavior and family functioning: The impact of a mild congenital risk factor. In R Cole & D. Reiss (Eds.), How amilies cope How do f families cope with chronic chronic illness? (pp. 120-147). Hillsdale, NJ: Erlbaum. Harper, L. V. (1989). The nurture of of human behavior. behavior. Norwood, Norwood, NJ: Ablex. v., Allen, K., Halverson, C. F., Kohnstamm, G. A. (1994). Parents use a Big 5 Havill, V., of children. In C. F. Halverson, G. A. categories in their natural language description of Kohnstamm, & R R. P. Martin (Eds.), The developing developing structure structure of temperament and and of temperament personality from from infancy to adulthood adulthood (pp. 371-386). Hillsdale, NJ: Erlbaum. personality Hetherington, E. M., Stanley-Hagan, M., & Anderson, E. R. A R (1989). Marital transitions: A American Psychologist, Psychologist, 44, 303-312. child's perspective. American R., & Stevenson-Hinde, J. (1973). Constraints on learning. New New York: Academic Press. Hinde, R, of the family environment on personality. Psychological Psychological Hoffman, L. W. (1991). The influence of Bulletin, 110, 187-203. Holden, G. W., & Edwards, L. A. (1989). Parental attitudes toward childrearing: Instruments, Psychological Bulletin, 106, 29-58. issues and implications. Psychological intervenJackson, J. F. (1993). Human behavioral genetics, Scarr's theory, and her views on interven tions: A critical review and commentary commentary on their implications for African African American Child Development, Development, 64, 1318-1332. children. Child methodological and Jacob, T. (1975). Family interaction in disturbed and normal families: A A methodological Psychological Bulletin, 82, 33-65. substantive review. Psychological of mother motherJacobs, B. S., & Moss, H. A. (1976). Birth order and sex of of sibling as determinants of Child Development, Development, 47, 315-322. infant interaction. Child of personality in the natural John, O. P. (1990). The "Big Five" factor taxonomy: Dimensions of Handbook of ofpersonality: personality: Theory Theory language and in questionnaires. In L. A. Pervin (Ed.), Handbook and research (pp. 66-100). New York: Guilford Press. of parents as Johnson, S. M., & Lobitz, G. K. (1974). The personal and marital adjustment of related to observed child deviance and parenting behaviors. Journal Journal of of Abnormal Abnormal Child Psychology, Psychology, 2, 193-207. Child Kagan, J., & Moss, H. (1962). Birth to maturity. New York: Wiley. American Psycholo PsycholoLambert, N. M. (1988). Adolescent outcomes for hyperactive children. American gist, 43, 786-799. gist, Parenting across the Lancaster, J. B., Altmann, J., Rossi, A. S., & Sherrod, L. R. R (1987). Parenting lifespan. New York: de Gruyter. R. M. (1989). Individual development and the family system: A life-span perspective. A life-span Lerner, R R. M. Lerner (Eds.), Family Family systems systems and and life-span development In K. Kreppner & R development (pp. 15-32). Hillsdale, NJ: Erlbaum. R. M., & Spanier, G. B. (1978). Child influences on marital Lerner, R marital and and family family interaction: A lifespan perspective. perspective. New New York: Academic Academic Press. A lifespan

264

HALVERSON AND ANDWAMPLER WAMPLER HALVERSON

Lewis, J. M., R., Gossett, 1. J. T., & Phillips, V. A. (1976). No single thread: thread: M., Beavers, W. R, New York: Brunner/Mazel. Psychological health Psychological healthininfamily familysystems. systems. New York: Brunner/Mazei. Lewis, J. M., M., Rodnick, E. H., Michael, J., & Goldstein, P. (1981). Intrafamilial Intrafamilial interactive behavior, parental communication deviance, and risk for of for schizophrenia. Journal of Psychology, 90, 90,448-457. 448-457. Abnormal Psychology, (1983). Economic hardship and marital relations in the Liker, J. K., & Elder, G. H., Jr. (1983). American Sociological SociologicalReview, Review, 343-359. 1930s. American 48,48, 343-359. environmentininpersonality personalitydevelopment. developmentNewbury NewburyPark, Park, C. (1992). Genes Genes and environment Loehlin, J. C. CA: Sage. C, & Nichols, R R. C. (1976). Heredity, Heredity, environment, environment, and personality. personality. Austin: Loehlin, J. c., University of Texas Press. (1983). Socialization in the context of of the family: Parent ParentMaccoby, E. E., & Martin, J. A. (1983). child interaction. In P. H. Mussen (Series Ed.) & E. M. Hetherington (Vol. Ed.), Handbook ofchildpsychology: Socialization, personality andand social development psychology:Vol. Vol.4.4. Socialization, personality social development (pp. 1-102). New York: Wiley. Martin, B. B. (1987). Developmental perspectives on family family theory and psychopathology. In T. Jacob (Ed.), Family psychopathology: Theory, methods and findings Familyinteraction interactionand and psychopathology: Theory, method^ and findings (pp. 163-202). New York: Plenum Press. Martin, C. L., & Halverson, C. F. (1981). A schematic processing model of sex typing and stereotyping in children. Child Development, 119-1134. Development, 52, 52,11119-1134. L.,& &Halverson, Halverson,C. C.F.F.(1983). (1983).The Theeffects effectsofofsex sextyping typingschemas schemason onyoung youngchildren's children's Martin, C. L., memories. Child Development,54, 54,563-574. 563-574. Child Development, R. P. (1983). Temperament: A review of research with implications for for the school Martin, R psychologist. School Psychology PsychologyReview, Review,12, 12,266-273. 266-273. P. (1986). (1986). Stability Stabilityand andchange changeof ofinfant infant temperament: temperament:Contributions Contributionsfrom frominfant, infant, Matheny, A. P. mother, and family family environment. In G. Kohnstamm (Ed.), Temperament Temperamentdiscussed discussed(pp. (pp. 81-97). Berwyn, PA: Swets North America. Matheney, A. P. (1989). Children's behavioral inhibition over age and across situations: Genetic similarity for for a trait during change. Journal of of Personality, Personality,57, 57,215-235. 215-235. McCartney, K., Harris, M. J., & Bernieri, F. (1990). Growing up and growing apart: A developmental meta-analysis of twin studies. Psychological PsychologicalBulletin, Bulletin,107, 107,226-237. 226-237. Minuchin, P. (1988). (1988). Relationships within the family: A systems perspective on development. In R R. A. Hinde & J. Stevenson-Hinde (Eds.), Relationships Relationships within families: families: Mutual influences influences (pp. (pp. 7-27). 7-27). Oxford Oxford England: England: Oxford Oxford University University Press. Press. Minuchin, S., Rosman, B B.. L., & Baker, L. (1978). (1978). Psychosomatic Psychosomaticfamilies. families.Cambridge, Cambridge,MA: MA: Harvard University Press. Mishler, E., & study of family Interactionininfamilies: families:An Anexperimental experimental study of family & Waxler, Waxier, N. (1986). Interaction processes processes and schizophrenia. schizophrenia.New NewYork: York:Wiley. Wiley. B. S. Moos, R R. H., & Moos, B. S. (1984). (1984). Family Family environment environment scale scalemanual manual(Rev. (Rev.ed.). ed.).Palo PaloAlto, Alto, CA: Consulting Psychologists Press. Olson, D. H., Russell, C. S., & Sprenkle, D. H. (1980). Marital and family family therapy with 42, 42, 973-995. children. Journal children. Journalof ofMarriage Marriageand andFamily FamilyTherapy, Therapy, 973-995. Oltmanns, T. T. F., F.,Broderick, Broderick,J.J.E., E.,&8LO'Leary, O'Leary,K.K.D.D.(1977). (1977).Marital Maritaladjustment adjustmentand andthe theefficacy efficacy Journalof ofConsulting Consultingand andClinical Clinical Psychology, Psychology, 45, 45, of behavior therapy with children. Journal 724-729. Patterson, G. R R. (1976). The aggressive child: Victim and architect architect of a coercive system. In L. A. Hamerlynk, L. C. Hardy, & E. J. Mash (Eds.), Behavioral modification modification and families: 1. I. Theory and research research(pp. (pp.189-216). 189-216).New NewYork: York:BrunnerlMazel. Brunner/Mazel.

CHAPTER CHAPTER 10

FAMILY INFLUENCES

26S 265

Patterson, G. R to family intervention: II. Coercive family R. (1982). A social sociallearning learningapproach approach to family intervention: IL Coercive family processes. processes. Eugene, Eugene,OR: OR:Castalia. Castalia. R, & Patterson, G. R., & Bank, L. (1989). Some amplifying mechanisms for pathologic processes R. Gunnar & & E. Thelen (Eds.), Systems Systems and development: development: The The in families. In M. R Minnesota SymposiumononChild Child Psychology (Vol. 167-209). Hillsdale, Minnesota Symposium Psychology (Vol. 22,22, pp.pp. 167-209). Hillsdale, NJ:NJ: ErErIbaum. lbaum. Patterson, G. R., R, & interac & Dishion, T. J. (1988). Multilevel family process models: models: Traits, interacR. A. Hinde & & J. Stevenson-Hinde (Eds.), Relationships Relationships tions, and relationships. In R within families: families:Mutual Mutualinfluences influences 283-310). Oxford, England: Oxford Univerwithin (pp.(pp. 283-310). Oxford, England: Oxford Univer sity Press. sity Press. R, McCleam, McClearn, G. E., & L., Plomin, R., Pedersen, N. L., & Frieberg, L. (1988). Neuroticism, extraver extraver55, 950-957. sion, and related traits in adult twins reared apart and reared together. SS, Plomin, R., R, & & Bergeman, C. S. (1991). The nature of nurture: nurture: Genetic Genetic influences influences on onenviron environBehavioraland andBrain BrainSciences, Sciences, 373-427. mental measures. Behavioral 14,14, 373-427. R., & & Daniels, D. (1987). Why are children in the same family so different from Plomin, R, andBrain BrainSciences, Sciences, 1-60. one another? Behavioral Behavioral and 10,10, 1-60. Plomin, R, R., DeFries, J. c., C, & & Loehlin, J. C. (1977). Genotype-environment interaction and correlation in in the the analysis analysis of of human human behavior. behavior. Psychological PsychologicalBulletin, Bulletin,84,84, 309-322. 309-322. correlation McCleam, G. E., Pedersen, N. L., L., Nesselroade, J. R, R., & & Bergeman, C. S. (1988). R, McClearn, Plomin, R., Genetic influence on childhood family environment perceived retrospectively from the last half of the life hfe span. Development DevelopmentPsychology, Psychology,24, 24,738-745. 738-745. Si O'Leary, K. D. (1980). Marital discord and childhood behavior problems. Porter, B., 8i Journalof ofAbnormal AbnormalChild ChildPsychology, Psychology,8, 8, 289-295. Journal 289-295. & Kilner, L. A. (1989). Adolescent mental health. American American Powers, S. I., Hauser, S. T., & Psychologist,44, 44,200-208. 200-208. Psychologist, Quinn, P., Renfield, M., Burg, C, & Rapoport, J. (1977). Minor physical anomalies: anomalies: A c., & Journalof ofChild ChildPsychiatry, Psychiatry, 662-669. newborn screening and a one-year follow-up. Journal 16,16, 662-669. & Zahn-Waxler, Zahn-Waxier, C. (1984). Roots, motives, and patterns in children's Radke-Yarrow, M., & & J. Reykowski (Eds.), prosocial behavior. In E. Staub, D. Bar-Tal, J. Karglowski, & Development and maintenance maintenanceofofprosocial prosocialbehavior: behavior: International perspectives Development International perspectives (pp.(pp. 81-99). New York: Plenum Press. Radke-Yarrow, M., & & Zahn-Waxier, Zahn-Waxler, C. (1986). The role of familial prosocial behavior: In D. D. Olweus, Olweus, J. J. Block, & & M. M. Radke-Yarrow Radke-Yarrow (Eds.) Researchfindings findings and questions. In Development Development of ofantisocial antisocialand andprosocial prosocialbehavior behavior(pp. (pp.207-233). 207-233).Orlando, Orlando,FL: FL:Aca Academic Press. Rapoport, J. L., Quinn, P. O., Lamprecht, F. (1974). Minor physical anomalies and plasma AmericanJournal JournalofofPsychiatry, Psychiatry, dopamine B-hydroxylase activity in hyperactive boys. American 131, 131, 387-390. Riskin, J., & & Faunce, E. (1972). An evaluative review of family interaction research. Family Family Process, 11,365-456. 365-456. Process, 11, C, & & Thomas, D. L. (1979). Parental support, power, and control techniques in Rollins, B. c., & I. L. Reiss (Eds.), the socialization of children. In W. R. Burr, R. Hill, F. I. Nye, & Contemporarytheories theoriesabout aboutthethe family: research-based theories pp. 317-364). Contemporary family: research-based theories (Vol.(Vol. 1 , pp.1,317-364). New York: York: Free Free Press. New Press. Rowe, D. C. (1981). Environmental and genetic influences on dimensions of perceived parenting: DevelopmentalPsychology, Psychology,1 7,17,203-208. 203-208. parenting: A twin study. Developmental Rowe, D. C. (1983). A biometrical analysis of perceptions of family environment: A study Development,54, 54,416-423. 416-423. of twin and singleton sibling kinships. Child Development,

WAMPLER HALVERSON AND WAMPLER

266

Rowe, D. C. (1990). As the twig is bent? The myth of of child rearing influences on personality development. Journal Journal of of Counseling Counseling and Development, Development, 68, 68, 606-611. R. G., Kafka, J. S., & Olson, D. H. (1971). Separating and joining influences in Ryder, R courtship and early marriage. American American Journal Journal of of Orthopsychiatry, Orthopsychiatry, 41, 450-464. Sackett, G., Ruppenthal, G., Farenbruch, C c.,, Holm, R., R, & Greenough, W. (1981). Social isolation rearing effects in monkeys vary with genotype. Developmental Developmental Psychology, Psychology, 117, 7, 313-318. A. J. & Chandler, M. J. (1975). Reproductive risk and the continuum of of caretaking Sameroff, A casualty. In F. D. Horowitz, E. M. Hetherington, S. Scarr, & G. M. Siegel (Eds.), Review Review of of child child development development research (Vol. 4, pp. 187-244). Chicago: University of of Chicago Press. A. J., & Seifer, R R. (1983). Familial risk and child competence. Child Development, Development, Sameroff, A 54, 1254-1268. Developmental theories for the 1990s: differScarr, S. (1992). Developmental 1990s: Development and individual differ Child Development, Development, 63, 1-9. 1-9. ences. Child Scarr, S. (1993). Biological and cultural diversity: The legacy of of Darwin for development. Child Child Development, Development, 63, 1333-1353. Scarr, S., & Grajek, S. (1981). Similarities and differences among siblings. In M. Lamb & B. Sutton-Smith (Eds.), Sibling Sibling relationships relationships (pp. 357-381). Hillsdale, NJ: Erlbaum. A theory of Scarr, S., & McCartney, K. (1983). How people make their own environments: A genotype-environment genotype-environment effects. Child Child Development, Development, 54, 424-435. R. A., & Wittig, M. A A. (1981). Personality resemblance Scarr, S., Webber, P. L., Weinberg, R among adolescents and their parents in biologially related and adoptive families. Jour Jourof Personality Personality and and Social Social Psychology, Psychology, 40, 885-898. nal of C , & Getter, H. (1980). Parental conflict and dominance in late adolescent adolescent Schwarz, J. c., maladjustment: A A triple interaction model. Journal Journal of of Abnormal Abnormal Psychology, Psychology, 89, 573-580. Scott, J. P., & Fuller, J. L. (1965). Genetics Genetics and the social social behavior behavior of of the dog. dog, Chicago: University of of Chicago Press. R, Maccoby, E. E., & Levin, H. (1957). Patterns Sears, R R. R., Patterns of of childrearing. childrearing, Evanston, IL: Row, Peterson. Sears, R R. R, R., Rau, L., & Alpert, R R. (1965). Identification and childrearing. Stanford, CA: Stanford University Press. Shepperson, V. L. (1982). Differences Differences in assertion and aggression between normal and neurotic family family triads. Journal Journal of of Personality Personality Assessment, Assessment, 46, 409-414. Thomas, A, A., & Chess, S. (1977). Temperament Temperament and and development. development. New York: BrunnerlMazel. Brunner/Mazel. Van der Veen, F., & Novak, A A. L. (1974). The family concept of the disturbed child: A A replication study. American American Journal Journal of of Orthopsychiatry, Orthopsychiatry, 44, 763-773. von Hilsheimer, G., & Kurko, V. (1979). Minor physical anomalies in exceptional exceptional children. Journal Journal of of Learning Learning Disabilities, Disabilities, 12, 462-469. Wachs, T. D. (1983). The use and abuse of of environment in behavior genetic research. Child Child Development, 54, 396-407. Development, nurture, Newbury Park, CA: Sage. of nurture. Wachs, T. D. (1992). The nature of strategy of of the genes. genes, London: Allen & Unwin. Waddington, C. H. (1957). The strategy New patterns and development. development. New York: Columbia Waddington, C. H. (1962). New patterns in genetics and University Press. Waldrop, M. F., Bell, R «fe Goering, Goering, J. J. D. D. (1976). (1976). Minor Minor physical physical anomalies anomalies and and inhibited inhibited R. Q., & Journal of of Child Child Psychology Psychology and and Psychiatry, Psychiatry, 117, behavior in elementary school girls. Journal 7,

113-122.

CHAPTER CHAPTER 10 10

FAMILY FAMILY INFLUENCES INFLUENCES

267 267

R. Q., McLaughlin, B., & & Halverson, C. F. (1978). Newborn minor Waldrop, M. F., Bell, R physical anomalies predict short attention span, peer aggression and impulsivity at age 3. Science, Science, 199, 199, 563-564. 563-564. & Halverson, C. F. (1971). Minor physical anomalies and hyperactive Waldrop, M. F., & exceptional infant infant (Vo\. (Vol. 2, 2, pp. pp. behavior in young children. In J. Helmuth (Ed.), The exceptional 343-380). New York: Brunner/Maze\. Brunner/Mazel. A., & & Bell, R R. Q. (1968). Minor physical anomalies and behavior Waldrop, M. F., Pedersen, F. A, in preschool children. Child Development, Development, 39, 39,391-400. 391-400. & Halverson, C. F. (1993). Quantitative measurement in family research. Wampler, K. S., & W. Doherty, Doherty, R R. Larossa, Larossa, W. W. Schumm, Schumm, & & S. S. Steinmetz Steinmetz (Eds.), (Eds.), Sourcebook Sourcebook In P. N. Boss, W. of family theories theories and methods: methods: A contextual contextual approach approach (pp. 181-194). New York: of Plenum Press. Wampler, K. S., Halverson, C. F., & & Deal, J. E. (1996). Risk and resiliency in nonc1inical nonclinical & E. Blechman young children: The Georgia Longitudinal Study. In M. Hetherington & (Eds.), Stress, inin children and thethe family (pp. 135-153). Hillsdale, Stress, coping copingand andresiliency resiliency children and family (pp. 135-153). Hillsdale, NJ: Erlbaum. Werner, E. E., Bierman, J., & & French, F. (1971). The children childrenof ofKauai. Kauai.Honolulu: Honolulu:University University of Hawaii Press. Werner, E. E., & & Smith, R R. S. (1977). Kauai's children children come of of age. Honolulu: University of Hawaii Press. & Smith, R R. S. (1982). Vulnerable Vulnerable but butinvincible. invincible.New NewYork: York:McGraw-Hili. McGraw-Hill. Werner, E. E., & majority. San Francisco: Jossey-Bass. A., & & Epstein, N. B. (1969). (1969). The silent majority. Westley, W. A, C, Jones, J. E., & & AI-Khayyal, Al-Khayyal, M. (1982). Healthy family family patterns: Observations Wynne, L. c., processes in families "at risk" for psychopathology. In F. Wash (Ed.), Normal family processes (pp. 142-167). New York: Guilford Guilford Press. (pp. Yang, R R. K., Zweig, A A. R., C, & & Federman, E. J. (1976). Successive relationship R, Douthitt, T. c., between maternal attitudes during pregnancy, analgesic medication during labor and Psychology, 12, 12,6-15. 6-15. delivery and newborn behavior. Developmental Psychology, R., Campbell, J. D., & & Burton, R R. V. (1968). Childrearing: Childrearing:An Aninquiry inquiryinto into research Yarow, M. R, research and methods. methods. San Francisco: Jossey-Bass.

This Page Intentionally Left Blank

CHAPTER 11 CHAPTER U

LONGITUDINAL STABILITY OF ADULT PERSONALITY T. COSTA, JR., AND ROBERT R. R. MCCRAE PAUL T. GERONTOLOGY RESEARCH RESEARCH CENTER, CENTER, NATIONAL NATIONAL INSTITUTE INSTITUTE ON ON AGING, AGING, NATIONAL GERONTOLOGY NATIONAL INSTITUTES OF OF HEALm, HEALTH, BALTIMORE, BALTIMORE, MARYLAND INSTITUTES MARYLAND

Borgatta and Lambert's 1968 handbook on personality included chapters on childchild hood and adolescent development and reviewed volumes of research conducted on college students, but said almost nothing about personality in adults over age 30. 30. Gerontologists and and life span developmentalists developmentalists (e.g., (e.g., Block, Block, 1971; 1971; Siegler, Siegler,George, George,&& Okun, 1979) studied personality and aging, but the relevance of theirfindings findings for personaUty was not widely appreciated. In the past decade, however, the field of personality the striking evidence of stability in adult personality traits has commanded more attention. Personality researchers and theorists are beginning to understand that the period of adulthood cannot be ignored, just as gerontologists have come to personaUty is central to the study of aging (McCrae & & Costa, 1990). recognize that personality This chapter reviews empirical evidence on the stability of personality and outlines some of the implications of stability for personality psychology.

I. I. THE THE DEFINITION DEFINITION AND AND ASSESSMENT ASSESSMENT OF OF PERSONALITY PERSONALITY An examination of stability or change in personality presupposes a definition of of and a method or or methods of assessing it. it. In personality and In principle, one could examine changes in personal constructs, instinctual impulses, cognitive styles, life structures, ego levels, or Jungian functions. In practice, most research has been conducted on HANDBOOK OF OF PERSONALITY PERSONAUTY PsYCHOLOGY PSYCHOLOGY IlANDBOOK

269

270 270

AND McCRAE MCCRAE COSTA AND

traits, dimensions of of individual differences differences in tendencies tendencies to show consistent consistent patterns traits, of of thoughts, feelings, and actions. This is not as serious a limitation as it might first appear, because traits are not the static, superficial, superficial, .o orr artifactual artifactual entities they are sometimes sometimes depicted as A s dispositions, traits are dynamic, in some respects equivalent equivalent to motives motives being. As and needs (Costa & & McCrae, 1988b); they are also inherently interactive, leading to the selection of of situations and the evocation evocation of of actions in others (Buss, 1992). Far from from being superficial, superficial, traits are among among the central central determinants determinants of of the life course (Conley, 1985b) and the sense of of identity identity (McCrae & & Costa, 1988), and traits familiar familiar from from questionnaires questionnaires and from English language adjectives are closely related to such theoretical constructs as Jungian attitudes and functions (McCrae & & related Costa, 1989b) and personality personality disorders disorders (Wiggins (Wiggins & & Pincus, 1989). And And traits are certainly not not artifacts of of person perception, as D' D'Andrade Andrade (1965) once argued. There is now now ample evidence evidence from from both observational observational studies (Moskowitz, 1988; Small, Zeldin, & & Savin-Williams, Savin-WiUiams, 1983) and cross-observer cross-observer rating rating studies (Funder (Funder & & Colvin, 1988; McCrae McCrae & & Costa, 1989a) to confirm confirm the objective reality reality of of personal personality traits. traits. ity It is sometimes sometimes objected that trait measures should not be used to study personality change, because because traits are, by definition, enduring dispositions. This is personality of days days a specious argument. Traits must persist across situations and over a period of we would not be able to distinguish them from from transient transient states. or weeks; otherwise, we But there is no requirement requirement that trait levels must remain stable over years or But decades. Similarly, the fact that trait measures must demonstrate short-term retest A reliable measure measure reliability does not ensure that they will show long-term stability. A show either stability or change; an unreliable measure would not be able to to can show show either. show Perhaps more than in any other approach to personality, trait theories theories have have been tied to measures. While While this has allowed allowed rigorous empirical tests of of the hypothe hypotheses of of trait theorists, it has also meant that critiques of of trait measures have clouded clouded the reputation of traits as constructs. It is doubtless true that many published scales of were poorly developed and validated, but as Block (1977) has argued, the value of from the best exemplars of of scale construction, not the trait theory must be judged from worst. Studies using instruments like the Personality Research Form (PRF; Jackson, 1984), the Interpersonal Interpersonal Adjective Adjective Scales (Wiggins, 1979), and the Eysenck Eysenck Person Personality Inventory Inventory (EPI; Eysenck Eysenck & Eysenck, 1964) have demonstrated their value in & Eysenck, demonstrated hundreds of applications. hundreds of recent years, two major concerns about trait measures have been largely In recent resolved: Social contaminant of of self-reports, self-reports, and self selfSocial desirability is not not a serious contaminant reports can be consensually consensually validated by observer observer ratings. Although Although researchers researchers and test constructors constructors have been preoccupied preoccupied by the specter of of social desirability since the 1950s, a solid body body of of evidence has demonstrated that individual differences evidence demonstrated individual differences in the tendency tendency to endorse endorse favorable favorable items are weak weak in comparison comparison with individual differences in substantive substantive traits (Block, 1965; Dicken, 1963; McCrae McCrae & & Costa, 1983; differences McCrae et al., 1989). Similarly, skepticism about agreement between self-reports McCrae aI., agreement between self-reports

CHAPTER CHAPTER 11 11

STABILITY OF PERSONALITY

271 271

and ratings (Fiske, 1978) has been answered by a number of studies which confirm that when knowledgeable raters use reliable instruments, substantial correlations self-reports and (in the range of .4 to .6) can be seen among raters and between self-reports & Colvin, 1988; 1988; McCrae & & Costa, 1989a; Woodruffe, Woodruffe, 1985). 1985). observer ratings (Funder & Thesefindings findings should bolster confidence in trait-based research on personality and aging. major drawbacks of the trait approach was the sheer Until recently, one of the major of words, words, hundreds hundredsof of published published number of trait measures and constructs. Thousands of scales, and dozens of of traits systems competed for for the researcher's or reviewer's attention. How could one make any generalization about the influence of age on personality traits when there appear to be an unlimited unUmited number of traits? Although it was widely understood that a few major major dimensions pervaded most of these traits, there was little agreement on what those dimensions were. In 1968, Wiggins concluded that Neuroticism and Extraversion (in some form form and 1968, under some label) were well established; in the 1980s a growing consensus (Bor (Borkenau, 1988; Digman & & Inouye, 1986; 1986; Goldberg, 1982; 1982; McCrae & & Costa, 1987) recognized the need to add the dimensions of Openness, Agreeableness, and and Consci Conscientiousness to complete what Norman (1963) called "an adequate taxonomy of per personality." Thefive-factor five-factor model provides a common basis for classifying natural language from a wide variety of personality inventories, and the items of trait terms, scales from of 1961; McCrae, Costa, & & Busch, 1986). 1986). It can therefore therefore California Q-Set (Block, 1961; the California framework for integrating research on aging and personality. Because it be used as a framework is comprehensive, conclusions about the five domains of Neuroticism, Extraversion, confidently be treated as Openness, Agreeableness, and Conscientiousness can confidently conclusions about the full range of personality traits.

II. II. STABILITY OF MEAN LEVELS The first of two major questions about personality stability or change concerns shifts in the mean levels of personality. Do individuals mature and grow normative shifts in wisdom as they age, or do they lose their youthful youthful exuberance and become depressed? Are old people better characterized as mellowed or cranky? cranky? Are Are there there stages or cycles of development, bringing predictable periods of crisis? Hypotheses might be based on either common stereotypes or on focused focused theories of aging (e.g., Erikson, 1950; 1950; Levinson, Darrow, Klein, Levinson, & & McKee, 1978), but in fact most research has been exploratory, seeking to document age differences differences or changes in various periods of adulthood. Research on aging faces two formidable problems rarely encountered in other branches of psychology. First, the independent variable, maturation, cannot be manipulated; in consequence, only quasi-experimental designs can be used, and these face well-known threats to their validity (Schaie, 1977). 1977). Second, aging is a process which requires decades to unfold; researchers who wish to study changes

AND McCRAE MCCRAE COSTA AND

171 272

differences) must study the life span in segments, or build (instead of simple age differences) upon the work of earlier researchers. A. Cross-Sectional Cross-Sectional Studies Studies A. The most convenient way to estimate age changes is by studying age differences differences in a cross-sectional design, and most research on age and personality has adopted 1977 review of these studies suggested that there were this strategy. Neugarten's 1977 very few consistent results, although there appeared to be evidence that older men and women scored lower on measures of Extraversion. Since then, a number of of large-scale studies have been published which included cross-sectional analyses (usually as as part part of of aa multiage-group multiage-group longitudinal design). Arenberg (usually design). Douglas and Arenberg (1978) examined age age differences differences on on the the Guilford-Zimmerman Guilford-Zimmerman Temperament Sur(1978) Sur vey (GZTS; (GZTS; Guilford, Guilford, Zimmerman, Zimmerman, & & Guilford, Guilford, 1976) 1976) in in two two samples samples of of men men aged vey 17 to 98 98 (Ns (Ns = = 605, 605, 310). 310). Siegler Siegler et et al. al. (1979) (1979) used used the Sixteen Sixteen Personality Factor Factor 17 to Questionnaire (16PF; (16PF; Cattell, Cattell, Eber, Eber, & & Tatsuoka, Tatsuoka, 1970) 1970) in a study of 331 men men and Questionnaire women aged aged 54 54 to to 70. 70. Costa Costa and and McCrae McCrae (1988a, (1988a, 1988b) age with women 1988b) correlated correlated age with MurMur ray's as measured by the the PRF FRF in in aa sample sample of of 296 aged 24 ray's needs needs as measured by 296 men men and and women women aged 24 to 81, 81, and with the thefive factors of of personality personality as as measured measured by by the the NEO NEO Personality Personality five factors Inventory (NEO-PI; Costa & & McCrae, McCrae, 1985b, 1985b, 1989a) in in aa different different sample sample of of 983 Inventory (NEO-PI; Costa men and women aged 21 21 to to 96. men and women aged 96. These studies can be compared if the scales are classified in terms of the five fivefactor model. model. Most Most of of the the Neuroticism Neuroticism (e.g., (e.g., 16PF 16PF Q4, Q4, Tension) Tension) and and Conscientious Conscientiousness (e.g., (e.g., PRF FRF Order) Order) scales scales showed showed no no age age differences. differences. Extraversion Extraversion appeared appeared to to decline, as seen in lower scores for older subjects on GZTS General Activity and Ascendance, PRF Dominance, Exhibition, and Play, and NEO-PI Extraversion. Older subjects also scored lower on PRF needs for for Change and Sentience and on NEO-PI Openness, suggesting a decline in Openness; however, there were crosscross sectional increases in PRF need for Understanding, another measure of Openness. Finally, Agreeableness appeared to be higher for older subjects, as seen in positive correlations between age and GZTS Friendliness, Friendhness, PRF Abasement, and NEO-PI Agreeableness, and a negative correlation between age and PRF Aggression. One problem with all of these studies is the possibility of sampling bias. Early Early cross-sectional studies compared college students with nursing home residents and inappropriately concluded that older people were were much muchlower lowerininintelligence, intelligence,mental mental health, and morale. morale. The studies cited above avoided such obvious confounds, but it is not clear how representative they were of the entire population. However, a recent follow-up study of a national sample of noninstitutionalized civiUans civilians (Cornoni-Huntley aI., 1983) 1983) can claim (Comoni-Huntley et et al, claim to tobe be representative. representative.Short Shortscales scalesmeasuring measuring Neuroticism, Extraversion, and Openness were administered to over 10,000 men and and women aged aged 32 32 to to 88. 88. As AsFigure Figure11shows, shows,plots plotsofofpersonality personalityscores scoresbybyage agegroup group showed little association; correlations correlationswith withage age ranged rangedfrom from -.12 -.12for forNeuroticism Neuroticismto to -.19 for Openness (Costa et al., 1986). 1986). Note that this pattern held for women as well as for men, and for blacks as well as for whites.

CHAPTER CHAPTER 11 11

273 273

STABILITY STABILITY OF OF PERSONALITY PERSONALITY 3 2 r 32

E CO

4 24 2 16

o—o White White Males Males • • ......•• White White Females Females cr'-'II Q-'—o Black Black Males Males •— Black Females Females . -- -•. Black

--

'•§ 16 CD Z

8

o

.« �---45-54 4 � 5� .5� 4--� 5� 5-� � '---� 65 �-� 74 '-�7 � 5-84 O �---3 � 5� 35-44 55-64 65-74 75-84

C c: 0

4r 2 24

o '(i) ...

16

ns ... X )( W 111

8 8

t'J.'-r-^s-w I

III. t —^''wjj^i:.r.iir.Ijti'i.|^

' ^ 16 Q) CD >

I

0

35-44

45-54

55-64

65-74

75-84

18 18

(/) CO

� S

c: C c: C

a

8.

o

12 12

l:::::::::::::t�:'�::�::::'::::::::::::iiil:''''';;;;'''' �

6 6h

35-44

45-54

55-64

65-74

J

75-84

Age Groups fiGVKE Mean levels Neuroticism, Extraversion,and and Opennesstoto Experience for 10-year age groups FIGURE 1 1 Mean levels ofof Neuroticism, Extraversion, Openness Experience for IO-year age groups of white men, black men, white women, and black women, aged 3S 35 to 84 years (from Costa et al., al, 1986). of

are Taken together, these cross-sectional studies suggest that older subjects are and open and and more agreeable than younger subjects. However, less extraverted and two qualifications are in order. First, the magnitude of the associations is uniformly small, especially considering the exceptionally wide age range and the near-perfect

274 274

COSTA COSTA AND ANDMCCRAE MCCRAE

reliability of 10% of of any of of measures of of age. Less than 10% of the variance of of the personal personality scales considered here is accounted cross-sectional stud accounted for for by age. Second, as cross-sectional studies, these findings are ambiguous as indicators of of maturation. In particular, cross crosssectional studies confound confound generational differences differences with maturational maturational changes. We do not know whether whether the oldest subjects in Figure 1 are lower in Openness because age leads to constriction constriction and conservativism, or because individuals born born in the brought up to be more closed to experience experience than first decades of this century were brought were subsequent generations.

B. Longitudinal and Sequential Studies deconfounding age and and birth cohort cohort is by conducting The most common approach to deconfounding a longitudinal study in which the same individuals are measured at two or more more times. In these repeated measures designs, subjects are matched matched on date of of birth, as well as on education, gender, and many other features; the longitudinal design is therefore Yet Siegler et therefore quite powerful powerful in detecting changes. Yet et a1. al. (1979) found found period of IntelUgence only one longitudinal change over the 8-year period of their their study: Intelligence repeated practice on the same items. scores increased slightly, perhaps due to repeated reported declines over over 7 years in General Activity, Douglas and Arenberg (1978) reported Thoughtfulness, Personal Relations, and Masculinity, but concluded Friendliness, Thoughtfulness, that only the changes in General Activity and Masculinity were likely to be true maturational changes. It was unlikely, for for example, that that the observed decline in maturational Friendliness was age related, because cross-sectional analyses had shown that that older friendly than younger. In a 6-year of NEO-PI scales, Costa people were more friendly 6-year study of found a significant significant decline in Neuroticism and McCrae (1988a) found Neuroticism but no change in Extraversion or Openness. Three-year Three-year longitudinal analyses of Extraversion of Agreeableness Agreeableness and Conscientiousness scales suggested a decline in both both these variables. Conscientiousness These longitudinal findings of personality findings show no consistent consistent picture of personality change, nor nor do they replicate the trends found found in cross-sectional cross-sectional studies. Neuroticism Neuroticism as measured by the NEO-PI showed a small decline, but but closely related scales in the 16PF and GZTS did not. Of Of the several scales related related to Extraversion, only one, one. General Activity, showed of the Openness scales (NEO showed longitudinal change. None of (NEOPI Openness; 16PF Tender-Mindedness, Imaginativeness, and Liberal Thinking; Thoughtfulness) declined. NEO-PI Agreeableness scores declined, in and GZTS Thoughtfulness) contradiction contradiction to cross-sectional findings of of increased Agreeableness in older sub subConscientiousness showed little or no change. These results jects. Scales measuring Conscientiousness strongly suggest that the small cross-sectional differences differences reported reported earlier eariier were due differences, not maturation. maturation. to generational differences, subject to confounding confounding influences. Readministra ReadministraLongitudinal studies are also subject of the test may itself itself affect affect scores, and changes between one administration administration tion of reflect social and historical change during the period, rather rather than and the next may reflect maturational shift. shift. Alternative analytic designs have been been devised devised to any universal maturational themselves subject subject to other confounds. tease out these confounds, although they are themselves other confounds. confounding, because time, date, and age are matheIt is impossible to remove all confounding,

11 CHAPTER 11

STABILITYOF OFPERSONALITY PERSONALITY STABILITY

275 275

interpreting results from from a variety of of designs, matically dependent. However, by interpreting reasonable inferences inferences can be made about maturational maturational effects (Costa & & McCrae, reasonable effects (Costa information from from cross- and time-sequential time-sequential analyses is added added to the 1982). When information conclusion that reinforces the conclusion that there is cross-sectional and longitudinal results, it reinforces little change during most of personality traits of adulthood adulthood in the mean mean levels of of personality & McCrae, 1988a; McCrae & & Costa, 1990). A A case could be made for for declines (Costa & in activity level, but these probably reflect reflect physical physical aging aging rather than psychological rather than psychological aging, and in any case, even these changes are small. Some 80-year-olds are more active than than many 30-year-olds.

c. C. Implications of of Mean Level Stability of evidence for Null findings rarely excite attention, and yet the pervasive pervasive lack of extraordinarily important. It flatly contradicts maturational change in personality is extraordinarily age stereotypes that portray older men and women as hypochondriacal, socially withdrawn, rigid in attitudes, irritable, and egocentric. These traits are no more (or less) common in older groups than in younger. Disease, cognitive and sensory and may help explain impairment, and societal neglect may mimic these traits and normative changes on these the stereotypes, but healthy individuals do not show normative personaUty dimensions simply as a result of of growing old. personality These findings should force force a reconsideration reconsideration of of some forms forms of of personality theory. They provide no evidence for for the discrete stages of of adult development theory; number of of theories depict (Erikson, 1950; Gould, 1978; Levinson Levinson et aI., al., which a number 1978)—a failure that that is also seen seen in more more focused focused investigations of of the midlife midUfe crisis 1978)-a & Rosenberg, 1981). 1981). Perhaps these theories should should (Costa & McCrae, 1980b; Farrell & or current current concerns; they be recast as developmental progressions in life structures or reflect changes in mood, social interaction, desire for for novelty, do not seem to reflect achievement strivings, because all these are stable across adulthood. adulthood. nurturance, or achievement construe personality personality in terms terms of of social roles Social psychologists sometimes construe personaUty (Veroff, (Veroff, 1983). and argue that role changes necessarily entail changes in personality Such theories might predict declines in nurturance nurturance in empty-nest or declines Such empty-nest women women or predictions would in achievement striving in retired men; these predictions would be wrong. Roles personality and apparently do not have clearly change with age, but roles are not personality and apparently influence on it. Indeed, it is more likely Ukely that that personality personality traits traits affect affect the roles much influence interpret both both chosen and assigned we choose to play and the ways in which we interpret & Costa, 1990). roles (McCrae &

ill ni. . STABILITY OF INDMDUAL INDIVIDUAL DIFFERENCES The fact fact that mean levels of of most personality variables show little or no change with age does not necessarily imply that individuals do not not change. Mean level had undergone stability could be observed even if all the individuals in the sample had of all personality personality variables—as dramatic changes in the level of variables-as long as the increases

1.76 276

COSTA AND MCCRAE MCCRAE COSTA AND

of some subjects were balanced by decreases of others. The second major question about stability thus concerns concernsthe themaintenance maintenanceof ofrank rankorder orderin inindividual individualdifferences. differences. This form form of stability stability isisusually usuallymeasured measuredasasaastability stabilitycoefficient, coefficient,aaretest retestcorrelation correlation based on readministration of a measure after after aa period period of of years. years.Note Notethat thatlongitudi longitudinal studies that follow the same individuals over time are required for research on stability in individual differences; there are no cross-sectional short-cuts.

A. Reports A. Retest Retest Stability Stability in in Self· Self-Reports Aside from the early studies of Strong (1951) on vocational interests and Kelly (1955) on personality from college ages to middle adulthood, virtually all the evievi dence on this question comes from studies published in the past 20 years. Table I summarizes studies using self-report methods. Across instruments, sexes, initial ages, and retest intervals, there is consistent evidence of substantial stability, with median correlations ranging from .34 to .77. .77. A careful examination of this table suggests that the magnitude of the stability coefficient depends depends more more on on the the instru instrusuggests that the magnitude of the stability coefficient ment than on initial age or retest interval. Instruments like the MMPI, which ment than on initial age or retest interval. Instruments like the MMPI, which is is better as aa measure of psychopathology psychopathology than an inventory inventory of of normal better regarded regarded as measure of than an normal personality, and the the Edwards Edwards Personal PersonalPreference Preference Schedule Schedule(EPPS; (EPPS;Edwards, Edwards,1959), 1959), personality, and which introduces distortion through a forced-choice format (Radcliffe, 1965), show which introduces distortion through a forced-choice format (Radcliffe, 1965), show relatively low stabiUty; standard personality questionnaires such as the California relatively low stability; standard personality questionnaires such as the California Psychological Inventory (CPI; Gough, 1987), 1987), GZTS, GZTS, and and NED-PI, NEO-PI, show show much Psychological Inventory (CPI; Gough, much higher values. higher values. There is some some evidence evidence that that retest retest correlations correlations are are lower lower for for subjects subjects initially under age 30. 30. Finn's Finn's (1986) (1986)comparison comparison of of young young and and middle-aged middle-aged adults adultson onMMPI MMPI factor factor scales over a 30-year interval showed substantially higher stability for the older subjects, and Block (1977) found higher stability on CPI scales for his over30 subjects than Helson and Moane (1987) reported for their under-30 subjects (although in this case initial age is confounded confounded with retest interval). Siegler et al. (1990) concluded that about half the variance in true scores for personality scales is stable between college and middle age.

B. Artifact Artifact and and Attenuation Attenuation in in Estimates Estimates of of Stability Stability B. The data in Table I are all taken from self-report instruments, and it might be hypothesized that artifacts of that method are responsible for the apparent stability. Individuals may wish to present a consistent image of themselves, recall how they answered previously, and duplicate their earlier answers. This argument is hardly plausible for long retest intervals, and one study that empirically examined it propro vided no support. Woodruff Woodruff (1983) asked middle-aged men and women to recall 25 years earlier when how they had answered the California Test of Personality 25 they had been in college; she also asked them to complete the test as it described them now. Subjects' recollections were poor; in fact, original scores were more

277 277

STABILITY STABILITYOF OFPERSONALITY PERSONALITY

11 CHAPTER 11

TABLE I

Stability Coefficients Coefficients for Recent Longitudinal Studies Studies Using Self-Report Instruments

Study Study

Instrument Instrument

A^ N

Sex

Initial age Initial 31-38 31-38 25-82 45-70 45-54 58-67 45-54 20-44 45-59 60-76 Seniors 18-35 Young managers

Block (1977) (1977) Block Costa and McCrae (1978) Siegler et al. al (1979) Leon et al. (1979)

CPI 16PF 16PF MMPI

219 139 139 331 331 71 71

M, F M,F M M M, F M,F M M

Costa et al. (1980)

GZTS

Mortimer et al. (1982) Conley (1985a) Howard and Bray" Stevens and Truss (1985)

Self-concept KLS factors EPPS GAMIN EPPS

Finn Finn (1986)

MMPI factors

Helson and Moane (1987)

CPI

60 87 32 368 378 378 266 264 85 92 96 78 81 81

M M M M M M M M M, F M,F M M M M M, F M,F M, F M,F M M M M F

Costa and McCrae (1988a)

ACL NEO-PI

78 234 164 140 140 133 133 101 101 96 273

F M M F F M M F F M, F M.F

Costa and McCrae (l992a) (1992a) Costa and McCrae (l992b) (1992b) Helson and Wink (1992) Costa et al. (in press)

GZTS GZTS CPI ACL CQS COS scales

College students 17-25 43-53 21 21 27 27 25-84 25-84 23-82 30-67 43 43 17-83

Retest interval 10 10 2 13 17 30 12 12 12 12 10 20 20 20 12 20 30 30 22 22 16 16 6 6 7 24 9 9 6

Correlations

Range Range .24-.64 .07-.82 .03-.76 .28-.74 .28-.74 .61-.85 .64-.85 .59-.87 .51-.63 .34-.57 .31-.54 .45-.61 .05-.58 -.05-.58 -.01-.79 -.14-.58 .10-.88 .21-.58 .40-.70 .49-.72 .60-.87 .60-.87 .58-.85 .65-.84 .65-.84 .61-.71 .66-.83

Median Median .71 .71 .50 .50 .50 .52 .52 .40 .72 .75 .73 .55 .46 .46

.42 .57 .34 .34 .44 .35 .56 .37 .51 .51 .61 .61 .74 .74 .73 .73 .65 .73 .73 .77

Note. Note. CPI, California Psychological Inventory; Inventory; 16PF, Sixteen Sixteen Personality Factor Questionnaire; MMPI, Minnesota Multiphasic Personality Personality Inventory; Inventory; GZTS, GZTS, Guilford-Zimmerman Guilford-Zimmerman Temperament Temperament Survey; Survey; KLS, KLS, Kelly Longitudinal Longitudinal Multiphasic Study; EPPS, EPPS,Edwards EdwardsPersonal Personal Preference Preference Schedule; Schedule;GAMIN, GAMIN,GuilfordlMartin Guilford/MartinInventory Inventoryof ofFactors; Factors;ACL, ACL,AdjecAdjecStudy; tive Check Check List: List: NEO-PI, NEO Personality Personality Inventory; Inventory; CQS, CQS, California California Q-Set. m part part from from Costa Costa and tive NEO-PI, NEO Q-Set. Adapted Adapted in and McCrae (1989b). McCrae (1989b). "Howard Howard and and D. D. W. W. Bray, Bray,personal personal communication, communication, May May 10, 10,1985. 1985. tI

strongly Strongly correlated with current scores than with recalled previous scores. Memory does not seem to inflate stability estimates. A second hypothesis concerns response sets: sets: To the extent that personality scores are determined by acquiescence, social desirability, or other response sets or styles, styles, the stability stability of personality may reflect stability of these artifacts. However, However, McCrae,and andArenberg Arenberg(1983) (1983)partialled partialledout outthe theeffects effectsofofacquiescence, acquiescence, when Costa, McCrae, coefficients extreme responding, naysaying, and falsification from 6-year stability coefficients

278

AND MCCRAE COSTA AND

in the GZTS, they found no appreciable decrease in stability. These data, again, suggest that the the stability seen in inTable Table IIreflects reflectssubstantive substantivecontinuity continuityof ofpersonality. personality. It is sometimes argued that even retest coefficients coefficients of .70 account for only half the variance in test scores and can be interpreted to mean that there is as as much change as stability in individual differences. These arguments overlook the fact that personality tests are fallible indicators of personality traits. True stability is always underestimated because of short-term unreliability. Costa, McCrae, McCrae, and Arenberg (1980) reported estimated 12-year stability coefficients on the GZTS of .80 .80 to 1.00, coefficients (1980) McCrae (1988a) found that Neuroticism, Extraversion, and Openness and Costa and McCrae (1988a) found that Neuroticism, Extraversion, and Openness and coefficients of .90 or above when corrected for unreliscales all showed stability coefficients unreli ability. ability. An alternative procedure for estimating true score stability utilizes utilizes an an external external criterion. When spouse ratings of Neuroticism, Extraversion, and Openness scales from the NEO-PI gathered in 1980 1980 were correlated with concurrent self-reports on the same scales, scales, the median cross-observer convergent correlation was .48. When were correlated correlatedwith withself-reports self-reportsobtained obtained66years yearslater, later,the themedian median the same ratings were .47 (Costa & & McCrae, 1988a). 1988a). Clearly, the passage of 66 years had correlation was .47 effect on the cross-observer validity of the spouse ratings, suggesting little little effect the cross-lagged correlation change in true personality. Mathematically, the ratio of the to the concurrent correlation estimates the stability of the true score. In this study, coefficient was therefore .98. the median estimated 6-year stability coefficient .98. c. Ratinls C. Stability Stability in in Personality Personality Ratings Although response biases and memory do not appear to inflate stability estimates, it can be argued that self-reports merely reflect the individual's self-image, and that these data speak only to the stability of the self-image. Rosenberg (1979) claimed that "people who have developed self-pictures early in life frequently frequently continue to hold to these self-views long after 58). If after the actual self self has changed radically" radically'' (p. 58). If so, so, stability coefficients coefficients may reflect only a crystallized self-concept. Personality ratings made by external observers provide a way to test this hypothesis. (CQS; Block, 1961, hypothesis. Studies of observer ratings on the California Q-Set (CQS; 1961, 1971) provided provided early early evidence evidence of of longitudinal longitudinal consistency consistency in in personality. personality. Block Block and his colleagues assessed men and women in junior high school, senior high school, and when subjects were were in intheir their 30s. 30s. Different Different panels panels of ofjudges judges rated rated the the individuals individuals at different periods. Correlations between senior high school and adulthood different time periods. for 90 CQS CQS items ranged from -.11 to .61; the median correlation was .26; .26; and % of the correlations were statistically significant. These correlations are, of of 61 61% course, much lower than those seen in Table I, but that is understandable: They are based on single items instead of scales, scales, confound changes due to aging with differences adoles differences between the two panels of judges, and trace personality from from adolescence, when it is not yet fully formed, into adulthood. direct evidence evidenceisisprovided providedby by aa6-year 6-yearstudy studyofofspouse spouseratings ratingsofofNeuroti NeurotiMore direct cism, Extraversion, Extraversion, and and Openness Opennessto to Experience Experience in in aa sample sampleof of167 167 men men and andwomen women cism,

11 STABILnY STABILITYOF OFPERSONALnY PERSONALITY CHAPTER 11

279 279

in the BLSA (Costa & & McCrae, 1988a). Ranging from .62 to .83 for men and from .63 to .86 for women, these correlations were very similar to those found in studies of self-reports. A A study of of peer ratings for all five major dimensions of of personality of & over a 7-year interval showed retest correlations ranging from .63 to .84 (Costa & McCrae, 1992b). Because unreliability affects observer ratings just as it does self-reports, there coefficients underestimate true stability of is again reason to suppose that these coefficients personality. Because they are based on observer ratings, it is also clear that the self-concept. high levels of stability are not artifacts of a crystallized self-concept.

IV. 20s rv. ADULT DEVELOPMENT: CHANGE IN THE 20S of stability during most of of the adult period do not necessitate a return The findings of developmental psychology ends with the 18th birthday. Although to the view that developmental Block demonstrated some continuity from high school into middle adulthood, the coefficients are not as high as those found between periods of adulthood, and coefficients development continues into several recent studies have suggested that personality development college and beyond. Haan, Millsap, and Hartka (1986) argued from their findings personality is not fully formed until it is shaped by the adult responsibilities that personality of work and marriage. Helson and Moane (1987) studied women between ages 21 and 27 and 27 and 43 and noted that stronger correlations were found found for the latter interval than for the former, despite the longer period of time covered. They concluded that "age 27 seems to have been at or near a watershed" (p. 179). Recent research research thus thus seems seems to to confirm confirm the the speculation speculation of of William William James James (1890) (1890) Recent that character is set at about age 30. that character is set at about age 30. differRelatively low retest correlations imply greater variability of individual differ ences over time. There is also evidence for systematic mean level changes during this period. Mortimer, Finch, and Kumka (1982) used self-concept self-concept scales in a 10of college seniors. They found increases in psychological year follow-up study of well-being and competence, and decreases in sociability and unconventionality. unconventionality. Bachman, O'Malley, and Johnston (1978) also reported increases in self-esteem self-esteem in late adolescence, and Jessor (1983) reported an increase in conventionality. In terms of of the five-factor model, these findings suggest decreases in Neuroticism (increased well-being and self-esteem), Extraversion (sociability), and Openness (unconven (unconvenA comparison of tionality), and an increase in Conscientiousness (competence). A of student and adult norms on the NEO-PI provides cross-sectional confirmation of these age changes: College students score higher in Neuroticism, Extraversion, and Openness, and lower in Agreeableness and Conscientiousness, than do adults (Costa & McCrae, 1989a). Data from the Revised NEO-PI continue to support these generalizations (Costa & & McCrae, 1994b). These trends do not appear to be limited to college students. Navy recruits in the age range from 17 to 21 show a very similar profile to college students, although the men are more closed to values than college men, and the women are

280 280

COSTA AND AND MCCRAE MCCRAE

higher in excitement seeking than college women (J. Holland, personal communica communication, February 15, 15, 1989). 1989). In both college students and Navy recruits, the most conspicious age difference is seen for Excitement Seeking, which is a full standard deviation above the mean of normative adult samples. Zuckerman (1979) has rere ported similar age trends for his measure of Sensation Seeking. Adults over the age of 30, regardless of educational background, appear to be less emotional and more settled and socialized than are individuals in their 20s. The effects of this change can be seen in the well-known decline in delinquency during the decade of of 20s, and historian Michael Rouche (1985/1987) even suggested that the violence, the 20s, intemperance, and rowdiness that characterized the Franks in the early Middle differences in personality: personahty: **A Ages might be attributable in part to such age differences "A society 60% of of the the population population under under age age twenty-five could not not avoid avoid being with more more than than 60% with youthful and and energetic, energetic, no no matter matter how how often often death death struck" youthful struck" (p. (p. 460). 460).

V. V. ALTERNATIVE ApPROACHES APPROACHES TO STABILITY OR CHANGE A. A. Altemative Altematiye Statistical Statistical Approaches Approaches

Most theory and research on personality in adulthood focuses on stability or change in mean levels or rank ordering, but some alternative statistical approaches have been advocated. A few studies have reported comparisons of factor structures in different age groups (e.g., Costa & & McCrae, 1976, 1976,1980b; & Aren1980b; McCrae, Costa, & Aren 1980; Monge, 1975). Changes in factor structure would suggest changes in the berg, 1980; relations among variables, variables, perhaps perhaps reflecting reflecting qualitative qualitative changes changes in in the the organization organization of personality. Such changes would, of course, complicate and qualify changes would, of course, complicate and qualifythe theinterpreta interpretation of mean level changes and stability coefficients. However, evidence to date points to the constancy of structure across age groups. In particular, the five-factor five-factor model of personality appears to apply equally well to school children (Digman & & Inouye, 1986), college students (Norman, 1963), and adults from the full age range (McCrae & & Costa, 1987). Gerontologists sometimes assert that variability increases with age (e.g., Fozard, Thomas, & & Waugh, 1976); 1976); apparently this is true for some cognitive and physiological variables. Personality variables, however, do not seem to show such of increases. We tested equality equaUty of variances for subjects over and under the age of 60 who completed the NEO-PI in 1986 (Costa & & McCrae, 1988a). Younger men showed significantly larger variance in in NEO-PI NEO-PI Openness and and Agreeableness scores than did older subjects (Fs [207, 293) 293] = = 1.41 and 1.38, respectively, p < .05), whereas younger women showed significantly larger variance than older women in Conscientiousness (F[268, 211) = .05). Neither sex showed age differences differences (F[268,211] = 1.36, p < .05). in the variance of Neuroticism or Extraversion scores. Even the significant differ differences seen here are quite small in magnitude; there is apparently little change in variance across the adult age span.

CHAmR CHAPTER 11 11

PERSONALITY STABILITY OF PERSONALITY

281 281

Some writers have called for an examination of ipsative changes among person personality variables over time (e.g., Mortimer et al, aI., 1982). Such analyses would seek to find increases characteristics.Bray Brayand andHoward Howard increasesininthe therelative relativesalience salienceof ofpersonality personalitycharacteristics. (1983) examined examined age-related age-related changes in a longitudinal study of AT&T managers who had taken the EPPS; the ipsative format format of this instrument instrument dictates that it can only assess the strength of motives relative to one another. Bray and Howard reported relative increases in the needs for Achievement and Autonomy and relative decreases in needs for Deference, Affiliation, Affiliation, and Intraception. Stevens and Truss (1985) also examined the EPPS over 12- and 20-year intervals, and interpreted the results to show maturational increases in Achievement, Autonomy, and Dominance, and decreases in Affiliation Affiliation and Abasement Abasement for for both sexes. Because both these 20s, the results may reflect reflect the same personality studies began with subjects in their 20s, meachanges reviewed in Section IV. However, when Murray's (1938) needs were mea from sured by a normative instrument, Jackson's PRF, in a sample ranging in age from 22 to 90 (Costa & McCrae, 1988a) none of these age relations was replicated cross sectionally. is also an ipsative measure of personality, and Q-correla Q-correlaCalifornia Q-Set is The California of tion across its 100 items on two occasions gives an estimate of the stability of rank order of personality attributes within an individual. Corrected for interrater unreliability. Block (1971) showed mean Q-correlations of .54 for for women and .56 unreliability, from senior high school to their mid-30s. Costa, McCrae, and Siegler (in for men from reported Q-correlations based on self-reports ranging from .12 to .86 over press) reported .71 for men. a 6-year interval; the median Q-correlations were .72 for women and .71 sahence of personality characteristics is stable, Both studies suggest that the relative salience after adolescence. particularly after

B. Stabllity Stability or or Change Change in in Other Other Personality Personality Variables Variables B. We have focused focused on studies of trait measures from self-reports or ratings because most aging research has employed these instruments and because they have shown themselves to be reliable and valid. However, it is of some interest to note that most studies using other theoretical perspectives and measurement approaches have also generally failed failed to find substantial age effects. A two-year longitudinal study of the perceptual variables measured by the Holtzman Inkblot Technique showed no evidence of maturational effects effects (Costa & McCrae, 1986a). Sentence Sentence completion completion measures of ego development have shown stability in adulthood (McCrae & Costa, 1980; 1980; Vaillant & McCullough, McCuUough, 1987), and although Vaillant (1977) has reported maturation of defenses between college age and early adulthood, most studies of of ways of coping in adults have found found that a wide variety of coping strategies are 1983; McCrae, 1982). used equally by adults of all ages (Lazarus & DeLongis, 1983; It is, of of course, possible to demonstrate age effects effects if instruments are tailored to reflect reflect social or psychological psychological characteristics that do change in adulthood. For example, there are regular changes in the proportion of individuals who consider Age identiidentithemselves middle-aged versus old or elderly (Bultena & Powers, 1978). Age

282 282

COSTA COSTA AND ANDMCCRAE MCCRAE

fication, fication, however, seems to be a relatively minor issue for for most adults, and is rarely mentioned self-concept (McCrae & Costa, 1988). mentioned as part of of the spontaneous self-concept

C. Subjective Personality Change The idea that personality is fundamentally fundamentally stable in adulthood adulthood is counterintuitive counterintuitive of change in their own personalities. to some people, who may perceive a great deal of It has also been challenged challenged by proponents of of a phenomenological phenomenological approach approach to the of life span development development (Ryff (Ryff & Heincke, 1983), who are concerned with study of subjective conceptions of of stability stabiUty and change. A A recent study by Krueger Krueger and subjective Heckhausen (1993) asked young, middle-aged, and older adults to describe the Heckhausen developmental course of of traits from from each of of the five personality personality domains. For all developmental respondents predicted increases in desirable desirable traits (e.g., age groups and all domains, respondents good-natured, purposeful, purposeful, realistic, intelligent) intelligent) up to age 60, with modest modest energetic, good-natured, dechnes thereafter. thereafter. Their Their subjective impressions of of personality personality change corres corresdeclines ponded to BUhler's Btihler's (1935) famous famous "curve of of life," but were inconsistent inconsistent with objec objecponded groups were asked to rate themselves themselves on the traits, tive data: When the three age groups differences. there were no age differences. Krueger and Heckhausen Heckhausen (1993) argued that these data may be interpreted interpreted in Krueger either as evidence that subjective subjective age changes are unfounded unfounded stereotypes, two ways, either or—more provocatively—as a demonstration demonstration of of the insensitivity insensitivity of of self-report self-report per peror-more provocatively-as sonality measures to real age changes. In particular, they suggest that individuals others their their own own age, effectively effectively eliminating any may rate themselves relative to others main effects effects for for age in both cross-sectional and longitudinal comparisons. If If valid, objection would would undermine undermine much much of of the psychometric research on personality this objection and aging reviewed reviewed in this chapter. But there are powerful powerful reasons to believe beUeve that it is not valid. vahd. First, the data show telltale signs of of stereotyping. Subjects appear appear to be much more sensitive to the desirability desirability of of the traits than to their specific specific content; their responses responses appear appear to reflect rather than an accurate reflect an evaluation evaluation of of aging aging rather accurate understanding understanding of of it. Second, this theory would make it difficult difficult to explain the age differences differences that are occasionally observed, such as the increase in conscientiousness after after adolescence and the decline decUne in activity among older individuals: Why are these differences differences not eliminated by the same process of of implicit age adjustment? adjustment? Third, the theory does not explain why age is singled out out by respondents as the basis for for comparison comparison instead of, of, for for example, race, social status, or gender. The fact fact that gender differences differences are routinely reported shows that that self-report self-report measures are sensitive to sex differences, differences, although a priori one might expect that people would use gender as often often as age as a basis of of social comparison. Fourth, McFarland, Ross, and Giltrow GiUrow (1992) explicitly tested this theory by asking older older subjects to explain the basis of of their self-ratings. In response to an open-ended open-ended question, none of of the subjects mentioned mentioned comparison with age peers, and and only 28% endorsed endorsed that that option when it was one of of two forced forced choices. McFarland McFarland et al. concluded concluded that "subjects "subjects do not appear appear to evaluate their current status by comparing comparing themselves with same-age peers" (p. 844). current

CHAPTER 11 11 CHAPTER

STABiLrrv OF OF PERSONALITY PERSONALITY STABILITY

283 283

A few longitudinal studies have contrasted subjective and objective changes. Woodruff Woodruff and Birren (1972) compared personality personality scores from from the California California Test of Personality taken in college with actual and recalled scores from an administration 25 years later. They found that objective changes in personality were small, but subjective changes were large. Most individuals "thought that their adolescent level 257). Krueger and of adjustment adjustment was much lower than it actually had been" (p. 257). Heckhausen (1993) pointed out that that study, too, could be explained study, too, could be explained by byaa changing changing basis of comparison at the two time points, but a later analysis of the data casts Woodruff considerable doubt on the accuracy of subjective perceptions of change: Woodruff Subjec(1983) showed that recalled scores were poorly correlated with initial scores. Subjec affect mean levels but not retest retest correlations, so these tive age adjustments would affect fallibility of memory and the need for data are best interpreted as evidence of the fallibility objective longitudinal measures of personality. When we asked subjects who completed the NEO-PI in 1986 if they had ^'changed a good deal," "changed a little," or "stayed pretty much the same" in "changed personality since 1980, 51% 51 % reported that they had stayed the same, and 35% said they had changed only a little (Costa & McCrae, 1989b). A substantial minority, 14%, felt felt that they had changed substantially. However, when 6-year retest correla correlations were examined for these three subgroups separately, the median values ranged from .79 to .82, .82, and repeated measures analyses showed little evidence of of mean from stability in their personalities, level changes. It appears that most people perceive stability contradicted by objective evidence. and the few who do not are contradicted VI. SOME IMPLICATIONS OF PERSONALITY STABILITY VI.

In the course of adult life individuals accumulate a lifetime of experience. They age biologically and face acute and chronic diseases. They pass through a variety from novice parents and workers to grandparents, widows, and of social roles, from retirees. They share with others the impact of great social and cultural changes, and face their own personal history of triumphs and tragedies. Yet all these events Uttle or no impact on basic personality traits. This fact fact should and experiences have little be the basis for a new perspective: Personality is not a product of the life course, an outcome or dependent variable, but a robust and resilient set of dispositions within the individual that themselves help shape the life course. People are not mere pawns of the environment, but active agents who steadfastly steadfastly pursue their own style of being throughout life.

A. A. Implications Implications for for Society Society and and the the Individual Individual Gerontologists interested in social planning have been concerned with normative personality change as a basis for policy. If people tended to withdraw and disengage from society with age, as Cumming and Henry (1961) proposed, then society should from leave them alone. If social activity is better for morale (Maddox, 1963), then society

COSTA AND MCCRAE

284

should develop community centers and provide transportation and opportunities for socializing. socializing. Clearly, our review supports neither of these hypotheses: We find little evidence for any kind. kind.IfIfwe we wish wishtotoserve servethe thepsychologi psychologifor normative changes of of any cal needs of older citizens, we should offer offer them a range of options as wide as the differences in personality. For the lifelong open introvert, a range of individual differences bookmobile may be more appropriate than a social center. the individual, stability stability in in personality personality isis aa source source of ofidentity identityand andcontinuity, continuity, For the for future future planning. planning.We We must mustbe be able ableto toanticipate anticipateour ourneeds, needs,interests, interests, and is a basis for inteUigently for career or family or retirement. and attitudes if we wish to plan intelligently Ideally, we should also be able to forecast how the significant others in our lives behave and and respond respondin in years yearsto tocome, come,and andthe thestability stabilityofofpersonality personalitytraits traitsallows allows will behave predictions. Sadly, Sadly,marriage marriagecounselors counselorsmust mustoften often confront confront the theproblems problemsthat that such predictions. arise when one spouse assumes that he or she will be able to change the other. The stability of personality poses a challenge for psychotherapy (Costa & 1986b). The The fact fact that that most most people people do do not not change change as as aa result result of of aging aging and and McCrae, 1986b). given the experience it typically brings does not mean that people cannot change, given the right right therapeutic therapeutic conditions. conditions. But But itit does doessuggest suggestthat that people peoplewho who are aredissatisfied dissatisfied the with their personality should seek help, because spontaneous improvement is unun alsomeans meansthat thatexpectations expectationsfor fortherapy therapymay may need needtoto be be modified. modified.Perhaps Perhaps likely. ItIt also likely. we will will come come to to see see high high Neuroticism Neuroticism or or low low Conscientiousness Conscientiousness as as conditions conditions that we that people must learn to live with, rather than as diseases that can be cured. In In that people must learn to live with, rather than as diseases that can be cured. that case, psychotherapy psychotherapy might might well well be be viewed viewed as as an an attempt to teach skills and case, attempt to teach social social skills and ways of coping that help the individual compensate for and adapt to basic personalways of coping that help the individual compensate for and adapt to basic personal ity dispositions. ity dispositions.

B. B. ImpDeations Implications for for PersonaDty Personality Research Researcli and and Tbeory Tlieory Most research on the topic of personality and aging has asked what we have come to believe is the wrong question: How does personality change as a result of of maturation or the the formative experiences of the the life course? Evidence Evidence on on the the stability stability very different different question: question:How Howdo do stable stablepersonality personalitydisposi disposiof personality suggests aa very tions shape the life course and affect affect adjustment adjustment at all ages? An example of this new perspective is found in research on psychological An happiness or or unhappiness unhappiness as a response to life events well-being. Instead of viewing happiness and circumstances, or to stages of development that yield a crisis at midlife and depression in in old old age, age, we we have have come come to to see see happiness happiness as as the the expression expression of of enduring enduring depression dispositions (Costa & & McCrae, 1980a; Emmons & & Diener, 1985; Watson & Clark, 1985; 1984). Individuals high in Neuroticism are prone to experience dysphoric affect, 1984). Individuals high in Neuroticism are prone to experience dysphoric affect, whereas those who are high in Extraversion are temperamentally high-spirited. whereas those who are high in Extraversion are temperamentally high-spirited. Love and and work workalso alsocontribute contributetotohappiness, happiness,particularly particularlyfor forthose thosehigh higjiininAgreeable AgreeableLove ness and Conscientiousness (McCrae & Costa, 1991). Because personality ness and Conscientiousness (McCrae & Costa, 1991). Because personality disposidisposi tions endure, endure, morale, morale, well-being, well-being, and and life life satisfaction satisfaction can can be 10 to to 20 tions be predicted predicted 10 20 we can can reinterpret reinterpret the the "mid-life **mid-life crisis" crisis" as as the the form form dysphoria dysphoria years in in advance, advance, and and we years

11 CHAPTER 11

STABILITY OF PERSONALITY

285 285

takes during middle age in individuals chronically high in Neuroticism (Costa & McCrae, 1980a, 1984). High levels of stability are generally assumed by researchers in psychosomatic medicine who attempt to link disease outcomes with psychological dispositions al., 1980). 1980). It is far more plausible to suggest that a lifetime of (e.g., Williams et aI., of hostility may predispose to coronary disease than to imagine that a transient phase of hostility in young adulthood will have lasting physiological consequences. How However, the recent demonstrations of continued personality development up to age 30 suggest that psychosomatic researchers may want to begin their longitudinal studies after after that age rather than in college, as has commonly been done (Siegler et aI., al, 1990). Longitudinal studies were essential for establishing establishing the stability of personality; & McCrae, for stability, they become useful useful in other ways (Costa & given the evidence for 1992a). 1992a). One particularly valuable feature feature of longitudinal research is the ability to accumulate data on a single sample. Every psychological variable is best evaluated its relation to other variables, but there is a practical limit to the number in terms of its of questionnaires that subjects can be asked to complete at a single session. Over the course of decades, however, periodic assessments can yield increasingly rich archives of data (e.g., Eichorn, Clausen, Haan, Honzik, & Mussen, 1981). If the subjects were over age 30 when the data were collected, the assessments may for many purposes be considered concurrent (Costa & McCrae, 1985a). Finally, the remarkable robustness of personality traits has clear implications for personality theory (Costa & McCrae, 1994a). Many different different theories, from for evolutionary to psychoanalytic, are consistent with the data on personality stability, but many others are not. Behaviorist, social learning, and social role theories in particular—theories that emphasize the shaping forces of the environment-cannot environment—cannot particular-theories proportioneasily account for dispositions that endure over long periods of time and proportion ately varied circumstances. This does not, of course, mean that the processes of of learning or of enacting roles are not important in determining behavior. They do not, however, appear to affect affect the basic tendencies of of the individual. Personality theories of the future future will need to consider how processes of of learning and adaptation allow individuals to cope with changing life circumstances while preserving their inner dispositions.

REFERENCES Bachman, J. J. G., O'Malley, P. P. M., Johnston, J. J. (1978). (1978). Adolescence Adolescence adulthood: Change Bachman, M., & Johnston, to toadulthood: Change Arbor, MI:MI: Institute forfor Social Research. and ofof young men. and stability stabilityininthe thelives lives young men.Ann Ann Arbor, Institute Social Research. Block, J. (1961). (1961).The The Q-sort Q-sort method in personality assessment and psychiatric method in personality assessment and psychiatric research.research, Spring SpringBlock, J. IL:Charles ChariesC.C.Thomas. Thomas. field, field, IL: Block, J. J. (1965). (1965). The The challenge challengeofofresponse response York: Appleton-Century-Crofts. Block, sets.sets. NewNew York: Appleton-Century-Crofts. Livesthrough throughtime. time. Berkeley, Bancroft Books. Block, J. J. (1971). (1971). Lives Berkeley, CA:CA: Bancroft Books.

286 286

COSTA COSTA AND AND McCRAE MCCRAE

Block, J. (1977). Advancing the psychology psychology of of personality: Paradigmatic shift shift or improving improving the quality of of research? In D. Magnusson & N. S. Endler Endler (Eds.), Personality Personality at the cross crossroads: Current interactional psychology Current issues in interactional psychology (pp. 37-64). Hillsdale, NJ: Erlbaum. Erlbaum. Borgatta, E. F., & Lambert, W. W. (Eds.). (1968). Handbook Handbook of of personality personality theory theory and and research. Chicago: Rand McNally. Borkenau, P. (1988). The multiple classification classification of of acts and the Big Five factors of of personality. personality. Journal Journal of of Research Research in Personality, Personality, 22, 337-352. Howard, A. (1983). The The AT&T longitudinal studies of of managers. In K. W. Bray, D. W., & Howard, Schaie (Ed.), Longitudinal Longitudinal studies studies of of adult adult psychological psychological development development (pp. 266-312). New New York: Guilford Guilford Press. BUhler, Btihler, C. (1935). The curve of of life as studies in biographies. Journal Journal of of Applied Applied Psychology, Psychology, 19, 405-409. Bultena, G. L., & Powers, E. A. (1978). Denial of of aging: Age identification identification and reference reference Journal of of Gerontology, Gerontology, 33, 748-754. group orientations. Journal Manipulation in close relationships: Five personality factors factors in interac interacBuss, D. M. (1992). Manipulation Journal of of Personality, Personality, 60, 477-499. tional context. Journal Cattell, R. B., Eber, H. W., & Tatsuoka, M. M. (1970). The handbook handbook for for the the Sixteen Sixteen Personality Factor Factor Questionnaire. Questionnaire, Champaign, IL: Institute for for Personality and Abil AbilPersonality ity Testing. stabiUty of of personality traits: A A multitrait-multimethod multitrait-multimethodConley, J. J. (1985a). Longitudinal stability multioccasion Journal of of Personality Personality and and Social Social Psychology, Psychology, 49, 1266-1282. multi occasion analysis. Journal Conley, J. J. (1985b). A A personality personality theory theory of of adulthood adulthood and aging. In R. Hogan & W. H. H. Jones (Eds.), Perspectives Perspectives in personality CT: JAI JAI Press. Press. Jones personality (Vol. 1, pp. 81-115). Greenwich, CT: Kleinman, Comoni-Huntley, J., Barbano, H. E., Brody, J. A., Cohen, B., Feldman, J. J., Kleinman, SL Madans, J. (1983). National Health and Nutrition Nutrition Examination Examination 1I— c.,, & J. C FoUowup Survey. Public Public Health Health Reports, Reports, 98, 245-251. Epidemiologic Followup differences in personality personality structure: A A cluster Costa, P. T., Jr., & McCrae, R. R. (1976). Age differences analytic approach. Journal Journal of of Gerontology, Gerontology, 31, 564-570. Costa, P. T., Jr., & McCrae, R. R. (1978). Objective Objective personality personality assessment. In M. Storandt, Storandt, I. C. Siegler, & M. F. Elias Ehas (Eds.), The The clinical clinical psychology psychology of of aging (pp. 119-143). New New York: Plenum Plenum Press. Costa, P. T., Jr., & McCrae, R. R. (1980a). Influence of of extraversion extraversion and neuroticism on subjective subjective well-being: Happy Happy and unhappy people. people. Journal Journal of of Personality Personality and and Social Social Psychology, Psychology, 38, 668-678. Costa, P. T., Jr., & McCrae, R. R. (1980b). Still stable after after all these years: Personality as a key to some issues in adulthood adulthood and old age. In P. B. Baltes & O. G. Brim, Jr. (Eds.), Life span span development development and and behavior behavior (Vol. 3, pp. 65-102). New New York: Academic Press. Life of age, period, period, Costa, P. T., Jr., & McCrae, R. R. (1982). An approach to the attribution of cohort effects. Psychological Psychological Bulletin, Bulletin, 92, 238-250. and cohort Personality as a lifelong determinant determinant of of well-being. Costa, P. T., Jr., & McCrae, R. R. (1984). Personality C. Izard (Eds.), Affective Affective processes processes in adult adult development development and and aging aging In G. Malatesta & c. (pp. 141-157). Beverly Hills, CA: Sage. Concurrent validation validation after after 20 years: Implications Implications Costa, P. T., Jr., & McCrae, R. R. (1985a). Concurrent of personality personality stability stability for for its assessment. In J. N. Butcher & c. C. D. Spielberger Spielberger (Eds.), of Advances in personality personality assessment assessment (Vol. 4, pp. 31-54). Hillsdale, NJ: Erlbaum. Erlbaum. Advances NEO Personality Personality Inventory Inventory manual. manual. Odessa, Costa, P. T., Jr., & McCrae, R. R. (1985b). The NED Assessment Resources. FL: Psychological Assessment

CHAPTER 11 U

STABILITYOF OFPERSONALID PERSONALITY STABILID

287 287

Costa, P. T., Jr., & McCrae, R R. R R. (1986a). Age, personality, and the Holtzman Inkblot and Human Development, 23, 23, 115-125. Technique. International InternationalJournal JournalofofAging Aging and Human Development, 115-125. R. R R. (1986b). Personality stability and its implications for Costa, P. T., Jr., & McCrae, R Clinical Psychology PsychologyReview, Review,6,6,407-423. 407-423. clinical psychology. Clinical R. R R. (1988a). Personality in adulthood: A six-year longitudinal Costa, P. T., Jr., & McCrae, R study of self-reports self-reports and spouse ratings on the NEO Personality Inventory. Journal of Personality Personalityand andSocial SocialPsychology, Psychology, 853-863. of 54,54, 853-863. Costa, P. T., Jr., & McCrae, R R. R R. (1988b). From catalog to classification: Murray's needs Journalof ofPersonality Personalityand andSocial Social Psychology, 258-265. and thefive-factor five-factor model. Journal Psychology, 55, 55, 258-265. Costa, P. T., Jr., & McCrae, R R. R R. (1989a). The NEO-PJINEO-FFI NEOPI/NEO-FFI manual supplement. Odessa, FL: Psychological Assessment Resources. Costa, P. T., Jr., & McCrae, R R. R R. (1989b). Personality continuity and the changes of adult life. In M. Storandt & G. VandenBos (Eds.), The adult years: years: Continuity Continuityand andchange change (pp. 45-77). Washington, DC: American Psychological Association. Costa, P. T., Jr., & McCrae, R R. R R. (1992a). Multiple uses for for longitudinal personality data. EuropeanJournal JournalofofPersonality, Personality, 6, 85-102. European 6, 85-102. T., Jr., Jr.,& &McCrae, McCrae,R R.RR.(1992b). (1992b).Trait Traitpsychology psychologycomes comesofofage. age.InIn Sonderegger Costa, P. T., T.T. B.B. Sonderegger NebraskaSymposium SymposiumononMotivation: Motivation: Psychology aging 169-204). Psychology andand aging (pp.(pp. 169-204). LinLin(Ed.), Nebraska coln: University of Nebraska Press. coln: R. R R. (1994a). "Set like plaster"? Evidence for for the stability of of Costa, P. T., Jr., & McCrae, R Can Personality PersonalityChange? Change? adult personality. In T. Heatherton & J. Weinberger (Eds.), Can (pp. 21-40). Washington, D.C.: American Psychological Association. Costa, P. T., Jr., & McCrae, R R. R. (1994b). Stability and change in personality from from adoles adolesF. Halverson, Halverson, G. G. A. A. Kohnstamm, Kohnstamm, & &R R.P. P.Martin Martin (Eds.), (Eds.), cence through adulthood. In C. F. Thedeveloping developingstructure structure temperament personality from infancy to adulthood The ofof temperament andand personality from infancy to adulthood (pp. (pp. 139-150). Hillsdale, NJ: Erlbaum. Costa, P. T., Jr., McCrae, R R. R., & Arenberg, D. (1980). (1980). Enduring dispositions in adult Journalof ofPersonality Personalityand andSocial Social Psychology, 793-800. males. Journal Psychology, 38, 38, 793-800. R, & Arenberg, D. (1983). Costa, P. T., Jr., McCrae, R R. R., (1983). Recent longitudinal research on aging. In In K. K.W. W.Schaie Schaie(Ed.), (Ed.),Longitudinal Longitudinal studies of adult psychological personality and aging. studies ofadult psychological development (pp. development (pp. 222-265). New York: Guilford Press. Costa, P. T., Jr., McCrae, R. R, R., & Siegler, I. C. (in press). Continuity and change over the adult life cycle: cycle: Personality and personality disorders. In C. R. Cloninger (Eds.), Personality Personalityand andpsychopathology. psychopathology.Washington, Washington,DC: DC:American AmericanPsychiatric PsychiatricPress. Press. R, Zonderman, A. B., Barbano, Costa, P. T., Jr., McCrae, R R. R., Barbano, H. H. E., E., Lebowitz, Lebowitz, B., B., & & Larson, Larson, sample: 2. Stability D. M. (1986). Cross-sectional studies of personality in a national sample: Stabihty in neuroticism, extraversion, and openness. Psychology Psychology and and Aging, Aging, 1,1, 144-149. 144-149. Cumming, E., & Henry, W. (1961). Growing old. old. New NewYork: York:Basic BasicBooks. Books. D' Andrade, R D'Andrade, R. B. (1965). Trait Trait psychology psychology and and componential componential analysis. analysis.American AmericanAnthropolo Anthropologist, gist, 67, 67,215-228. 215-228. Dicken, C. (1963). Good impression, social desirability, and acquiescence as suppressor variables. Educational Educationaland andPsychological PsychologicalMeasurement, Measurement, 699-720. 23, 23, 699-720. J. M., M.,& &Inouye, Inouye,J.J.(1986). (1986).Further Furtherspecification specificationofofthe thefive fiverobust robustfactors factorsofofpersonal personalDigman, J. Journalof ofPersonality Personalityand andSocial Social Psychology, Psychology, 50,50,116-123. 116-123. ity. Journal Arenberg, D. D. (1978). (1978).Age Agechanges, changes,cohort cohortdifferences, differences,and andcultural culturalchanges changeson on Douglas, K., & Arenberg, the Guilford-Zimmerman Guilford-Zimmerman Temperament Survey. Journal Journal of of Gerontology, Gerontology,33, 33,737-747. 737-747. Edwards Personal Personal Pref Preference Schedulemanual. manual. New York: Edwards, A. L. (1959). Edwards erence Schedule New York: TheThe Psychological Corporation.

288

COSTA COSTA AND ANDMCCRAE MCCRAE

Eichorn, D. H., Clausen, J. A., Haan, N., Honzik, M. P., & Mussen, P. H. (Eds.), (1981). Present Present and and past past in middle middle life. New York: Academic Press. Emmons, R. A., & Diener, E. (1985). Personality correlates of of subjective well-being. Personal Personality and and Social Social Psychology Psychology Bulletin, 11, 89-97. Erikson, E. H. (1950). Childhood Childhood and and society. New York: Norton. Eysenck, Eysenck, H. J., & Eysenck, S. B. G. (1964). Manual Manual of of the Eysenck Eysenck Personality Personality Inventory. Inventory, London: University Press. London: midlife, Boston: Auburn House. Farrell, M. P., & Rosenberg, S. D. (1981). Men at midlife. of personality self-ratings over 30 years: Evidence for an agel age/ Finn, S. E. (1986). Stability of cohort interaction. Journal Journal of of Personality Personality and Social Social Psychology, Psychology, 50, 813-818. Fiske, D. W. (1978). Strategies of of personality personality research. San Francisco: Jossey-Bass. of age and frequency frequency of of stimulus Fozard, J. L., Thomas, J. C C.,, & Waugh, N. C. (1976). Effects of repetitions on two-choice two-choice reaction times. Journal Journal of of Gerontology, Gerontology, 31, 556-563. Funder, D. C., C , & Colvin, C. R. (1988). Friends and strangers: Acquaintanceship, agreement, of personality judgment. Journal Journal of of Personality Personality and Social Social Psychology, Psychology, and the accuracy of 55, 149-158. of personality. Goldberg, L. R. (1982). From ace to zombie: Some explorations in the language of Advances in personality personality assessment assessment (Vol. In C. D. Spielberger & J. N. Butcher (Eds.), Advances 1, pp. 203-234). Hillsdale, NJ: Erlbaum. California Psychological Inventory Inventory administrator's administrator's guide. guide, Palo Alto, Gough, H. G. (1987). Calif ornia Psychological CA: Consulting Psychologists Press. Gould, R. L. (1978). Transformations. Transformations, New York: Simon & Schuster. Guilford-Zimmerman Guilford, J. S., Zimmerman, W. S., & Guilford, J. P. (1976). The Guilford-Zimmerman Temperament Survey Survey Handbook: Handbook: Twenty-five Twenty-five years years of of research and and application. application, San Temperament Diego, CA: EdITS. Haan, N., Millsap, R., & Hartka, E. (1986). As time goes by: Change and stability in Psychology and and Aging, Aging, 1, 220-232. personality over fifty years. Psychology Helson, R., & Moane, G. (1987). Personality change in women from college to midlife. Journal Psychology, 53, 176-186. Journal of of Personality Personality and and Social Social Psychology, Helson, R., & Wink, P. (1992). Personality change in women from the early 40s 40s to the early 50s. Psychology Psychology and and Aging, 7, 7, 46-55� 46-55. Jackson, D. N. (1984). Personality Personality Research Research Form manual manual (3rd ed.), ed.). Port Huron, MI: Research Psychologists Press. James, W. (1890). Principles Principles of of psychology. psychology. New York: Henry Holt. Jessor, R. (1983). The stability of of change: Psychosocial development from adolescence to young adulthood. In D. Magnusson & V. L. Allen (Eds.), Human Human development: development: An An interactional interactional perspective perspective (pp. 321-341). New York: Academic Press. 659-681. Kelly, E. L. (1955). Consistency of of the adult personality. American American Psychologist, Psychologist, 10, 10,659-681. Krueger, J., & Heckhausen, J. (1993). Personality development across the adult life span: Subjective conceptions vs cross-sectional contrasts. Journal Journal of of Gerontology: Gerontology: Psychologi Psychological Sciences, 48, P100-P108. American Lazarus, R. S., & DeLongis, A. (1983). Psychological stress and coping in aging. American Psychologist, Psychologist, 38, 245-254. Leon, G. R., Gillum, B., Gillum, R., & Gouze, M. (1979). Personality stability and change period—middle age to old age. Journal Jourrml of of Consulting Consulting and and Clinical over a 30 year period-middle Psychology, Psychology, 47, 517-524. C N., Klein, E. B., Levinson, M. L., & McKee, B. (1978). The Levinson, D. J., Darrow, C. seasons seasons of of a man's man's life. New York: Knopf.

CHAPTER 11 CHAmR

STABILITY OF OFPERSONALITY PERSONALITY STABILITY

289 289

(1963). Activity and morale: A longitudinal study of selected selected elderly subjects. Maddox, G. L. (1963). Social Forces, Forces,42,42,195-204. 195-204. Social mechanisms.Journal Journalof of Gerontol Gerontol(1982). Age Age differences differences in in the the use use of of coping coping mechanisms. McCrae, R. R. (1982). ogy, 37, 37, 454-460. 454-460. ogy, Costa, P. P. T., T.,Jr. Jr.(1980). (1980).Openness Opennesstotoexperience experienceand andego egolevel levelininLoevinger's Loevinger*s McCrae, R. R., & Costa, sentence completion test: Dispositional contributions to developmental models of of personality. Journal Psychology, 39,39, 1 179-1190. Journalof ofPersonality Personalityand andSocial Social Psychology, 1179-1190. McCrae, R. R., & Costa, P. T., Jr. (1983). Social desirability scales: More substance than Journalof ofConsulting Consultingand andClinical Clinical Psychology, 882-888. style. Journal Psychology, 51, 51, 882-888. of the five-factor model of personality McCrae, R. R., & Costa, P. T., Jr. (1987). Validation of Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, 52,52, across instruments and observers. Journal 81-90. McCrae, R. R., & Costa, P. T., Jr. Jr. (1988). (1988). Age, Age,personality, personality, and and the the spontaneous spontaneous self-concept. self-concept. Journalof ofGerontology: Gerontology:Social Social Sciences, S177-S185. Journal Sciences, 43, 43, S177-S185. McCrae, R. R., & Costa, P. T., Jr. (1989a). Different Different points of of view: Self-reports and ratings in the assessment of personality. In J. P. Forgas & M. J. Innes (Eds.), Recent advances advances socialpsychology: psychology:An Aninternational international perspective 429-439). Amsterdam: Elsevier. in social (pp.(pp. 429-439). Amsterdam: Elsevier. in perspective McCrae, R. R., & Costa, P. T., Jr. (1989b). Reinterpreting the Myers-Briggs Type Indicator Personality, from the perspective of the five-factor model of of personality. Journal of of Personality, 57, 57, 17-40. Costa, P. P. T., T., Jr. Jr. (1990). (1990).Personality Personalityininadulthood. adulthood. New York: Guilford Press. New York: Guilford Press. McCrae, R. R., & Costa, Arbeit: The full full five-factor McCrae, R. R., & Costa, P. T., Jr. (1991). Adding Liebe und Arbeit: 17,17, 227-232. model and well-being. Personality Personalityand andSocial SocialPsychology PsychologyBulletin, Bulletin, 227-232. of adult personality McCrae, R. R., Costa, P. T., Jr., & Arenberg, D. (1980). Constancy of structure in adult males: Longitudinal, Longitudinal, cross-sectional cross-sectional and times of of measurement measurement analy analyJournalof ofGerontology, Gerontology,35,35,877-883. 877-883. ses. Journal McCrae, R. R., Costa, P. T., Jr., & Busch, C. M. (1986). Evaluating comprehensiveness systems: The California California Q-Set and the five-factor model. Journal of of in personality systems: Personality,54, 54,430-446. 430-446. Personality, McCrae, R. R., Costa, P. T., Jr., Dahlstrom, W. G., Barefoot, J. C, C., Siegler, I. l. C, C., & Williams, research on B., Jr. (1989). A caution on the use of the MMPI K-correction in research R. B., psychosomatic medicine. Psychosomatic PsychosomaticMedicine, Medicine,51, 51,58-65. 58-65. McFarland, C., C, Ross, M., & Giltrow, M. (1992). (1992). Biased Biased recollections recollections in in older older adults: adults:The The role role of implicit theories of aging. Journal Psychology, 62, 62, 837-850. Journalof ofPersonality Personalityand andSocial Social Psychology, 837-850. Monge, R. H. H. (1975). (1975). Structure Structure of of the theself-concept self-concept from from adolescence adolescence through through old old age. age. Experi Experimental Aging Aging Research, Research,1,1,281-291. 281-291. Mortimer, J. T., Finch, M. D., & Kumka, D. (1982). (1982). Persistence and change in development: B. Baltes & O. G. Brim, Jr. (Eds.), Life-span The multidimensional self-concept. In P. B. development and and behavior behavior (Vol. (Vol. 4, 4, pp. pp. 264-315). 264-315). New New York: York: Academic Academic Press. Press. S. (1988). Cross-situational generality in the laboratory: Dominance and Moskowitz, D. S. Journalof ofPersonality Personalityand andSocial Social Psychology, 829-839. friendliness. Journal Psychology, 54,54, 829-839. Murray, H. A. (1938). Explorations Explorations in personality. personality. New NewYork: York:Oxford Oxford University University Press. Press. Neugarten, B. B. L. (1977). Personality and aging. In J. E. Birren & K. W. Schaie (Eds.), of the psychology of of aging aging (1st ed., pp. 626-649). New York: Van Handbook of Nostrand-Reinhold. of personality attributes: Replicated Norman, W. T. (1963). Toward an adequate taxonomy of factor structure in peer nomination personality ratings. Journalof of Abnormal Abnormal and andSocial Social Psychology,66, 66,574-583. 574-583. Psychology,

290 290

COSTA AND COSTA AND MCCRAE MCCRAE

Radcliffe, J. A. (1965). [Review of of Edwards Edwards Personal Personal Preference Preference Schedule]. Schedule]. In O. K. Buros Radcliffe, (Ed.), The The sixth sixth mental mental measurements measurements yearbook yearbook (pp. 195-2(0). 195-200). Highland Highland Park, NJ: Gryphon Press. Rosenberg, M. (1979). Conceiving self. New York: Basic Books. Conceiving the self A history history of of Rouche, M. (1987). The early Middle Ages in the West. In P. Veyne (Ed.), A private life: I.L From pagan pagan Rome Rome to Byzantium Byzantium (A. (A. Goldhammer, Goldhammer, Trans., Trans., pp. pp. 411-549). 411-549). private Cambridge, Cambridge, MA: Harvard Harvard University Press. (Original work work published 1985). of personality in adulthood adulthood Ryff, C. D., & Heincke, S. G. (1983). Subjective organization of and aging. Journal Journal of of Personality Personality and and Social Social Psychology, Psychology^ 44, 807-816. Schaie, K. W. (1977). Quasi-experimental Quasi-experimental research designs in the psychology of of aging. In J. E. Birren & K. W. Schaie Schaie (Eds.), Handbook Handbook of of the the psychology psychology of of aging aging (1st ed., pp. 39-69). New York: Van Nostrand-Reinhold. Nostrand-Reinhold. «fe Okun, M. A. (1979). Cross-sequential analysis of of adult c.,, George, L. K., & Siegler, I. C personality. Developmental Developmental Psychology, personahty. Psychology, 15, 350-351. Siegler, I. c., C , Zonderman, A. B., Barefoot, Barefoot, J. c., C , Williams, R. B., Jr., Costa, P. T., Jr., & McCrae, R. R. (1990). Predicting personality in adulthood adulthood from from college MMPI scores: Implications for for follow-up follow-up studies in psychosomatic medicine. Psychosomatic Psychosomatic Medicine, Medicine, 52, 644-652. Small, S. A., Zeldin, R. S., & Savin-Williams, R. C. (1983). In search of of personality traits: A multimethod multimethod analysis of of naturally occurring occurring prosocial and dominance behavior. behavior. A Journal Journal of of Personality, Personality, 51, 1-16. 1-16. and change in adult personality over 12 and Stevens, D. P., & Truss, C. V. (1985). Stability and 20 years. Developmental Developmental Psychology, Psychology, 21, 568-584. Strong, E. K., Jr. (1951). Permanence of of interest scores over 22 years. Journal Journal of of Applied Applied Psychology, Psychology, 35, 89-91. Vaillant, G. E. (1977). Adaptation Adaptation to life. life, Boston: Little, Brown. McCuUough, L. (1987). The Washington University University Sentence Completion Completion Vaillant, G. E., & McCullough, Test compared compared with other measures of of adult ego development. American American Journal Journal of of Test Psychiatry, 189-1194. Psychiatry, 144, 11189-1194. Veroff, J. (1983). Contextual determinants of of personality. Personality Personality and and Social Social Psychology Psychology Bulletin, 9, 331-343. Watson, D., & Clark, L. A. (1984). Negative affectivity: affectivity: The disposition disposition to experience aversive emotional states. Psychological Psychological Bulletin, 96, 465-490. Wiggins, J. S. (1968). Personality structure. Annual Annual Review Review of of Psychology, Psychology, 19, 293-350. A psychological taxonomy taxonomy of of trait-descriptive terms: The interpersonal interpersonal Wiggins, J. S. (1979). A domain. Journal Journal of of Personality Personality and and Social Social Psychology, Psychology, 37, 395-412. of personality disorders and dimensions dimensions Wiggins, J. S., & Pincus, A. L. (1989). Conceptions of of personality. Psychological Psychological Assessment: Assessment: A A Journal Journal of of Consulting Consulting and and Clinical Clinical Psychol Psycholof 1, 305-316. ogy, I, Type A A Williams, R., Haney, T., Lee, K., Kong, Y., Blumenthal, J., & Wahlen, R. (1980). Type behavior, behavior, hostility and coronary coronary heart heart disease. Psychosomatic Psychosomatic Medicine, Medicine, 42, 539-549. Woodruff, Woodruff, D. S. (1983). The role of of memory memory in personality continuity: A 25 year follow-up. follow-up. Experimental Experimental Aging Aging Research, 9, 31-34. Woodruff, D. S., & Birren, J. E. (1972). Age changes and cohort differences differences in personality. personality. Woodruff, Developmental Developmental Psychology, Psychology, 6, 252-259. Woodruffe, C. (1985). Consensual Consensual validation validation of of personality traits: Additional evidence and Woodruffe, Journal of of Personality Personality and and Social Social Psychology, Psychology, 48, 1240-1252. individual differences. Journal Sensation seeking: Beyond Beyond the optimal optimal level level of of arousaL arousal. Hillsdale, Zuckerman, M. (1979). Sensation NJ: Erlbaum.

CHAPTER 12 CHAPTER 12

Is THERE ADULT Is DEVELOPMENT DISTINCTIVE TO WOMEN? RAVENNA HELSON, JENNIFER PALS, AND MARJORIE SOLOMON INSTITUTE INSTITUTE OF OF PERSONALITY PERSONALITY AND AND SOCIAL SOCIAL RESEARCH, RESEARCH, UNIVERSITY OF OF CALIFORNIA, CALIFORNIA, BERKELEY UNIVERSITY BERKELEY

Adult development usually refers to positive personality change after late adolesadoles cence, such as increased competence, perspective, maturity, and understanding 1977; White, 1966). It It may may be be conceptualized conceptualized as taking place place in in the the process process (Vaillant, 1977; of meeting the psychosocial needs of identity, intimacy, generativity, and integrity (Erikson, 1963), or through developmental tasks such as rearing children or assum assuming responsibilities in work (Havighurst, 1948). It can refer to the actualization of of 1971; Jung, 1931/1960). Some would take adult one's individual potential (BUhler, (Btihler, 1971; development to be evidenced in the sequence of events that make a life story, or perhaps in the differentiation and coherence of one's life story (McAdams, 1993). Though attention tends to be focused on change in the positive direction, change may also be retrogressive (Baltes, 1987). 1987). Gains in self-control, for example, may be made at the cost of a loss in spontaneity. There is disagreement about whether features of adult development can be demonstrated in most people or in some people in some circumstances, or whether patterns of variation are so great and depend on so many factors as to render the construct of adult development of doubtful scientific use. use. The disagreement depends in part part on on diverse diverse conceptualizations conceptualizations of of what what development development consists of, of, on on the the way in and personality personality change change are are to to be be measured, measured, and and on on difficulties difficulties inherent inherent in in personality and personality Helson & & making comparisons comparisons across across cultures cultures or or historical historical periods periods (Helson, making (Helson, 1993a; 1993a; Helson H ANDBOOK OF HANDBOOK OF PERSONALrrY PERSONALITY PSYCHOLOGY PSYCHOLOGY

291 291

COPYRIGHT CI © 1997 BY ACADEMIC PlESS PRESS.. CoPYRIGHT ALL RJGHTS OF REPRODUCTION IN ANY FORM RESERVED.

ALL RIGHTS OF REPRODUCTION IN ANY FORM RESERVED.

292 292

HEUON, PALS, HELSON, PALS,AND ANDSOLOMON SOLOMON

of the question of Stewart, 1994). Whether there is adult development development is a version of of the meaning of life and the extent to which we can control our destinies. People are interested in such questions, but they are not easy to answer. To ask whether there is adult development distinctive to women involves all of the above issues along with considerations of of of what is special about women. Gender differences differences seem very small in some contexts contexts and very large in others Maccoby & Jacklin, 1974), and policies to either ((J. J. H. Block, 1976; Eagley, 1995; Maccoby differences have perplexing perplexing consequences (Hare (Hareminimize or maximize these differences emphasizing differences differences between men and Mustin & Marecek, 1990). For example, emphasizing women is often said to justify and perpetuate inequalities. However, minimizing differences may support the failure of of society to take into account women's gender differences special needs, such as those associated with childbirth. Research designs present numerous dilemmas and agendas. If one compares women with men, one neglects aspects of development unique to women or of of special interest to women. If one studies women alone, one leaves unclear the extent of of difference difference between men and women. If one studies middle class white women, results attributed women. If one studies middle class white women, results cannot cannot safely safely be be attributed to women in general (Yoder & Kahn, 1993). Research agendas are not neutral. to women in general (Yoder & Kahn, 1993). Research agendas are not neutral. If If one emphasizes biological differences, there is the implication of of universality and enduringness. If If one one emphasizes emphasizes how how social social institutions institutions subjugate there is is enduringness. subjugate women, women, there the implication that any picture one gets of women's adult development is incomthe implication that any picture one gets of women's adult development is incom plete or distorted. distorted. If If one one shows shows how how sex sex differences differences are from plete or are construed construed differently differently from one cultural group, society, or period of history to another, it would appear that one cultural group, society, or period of history to another, it would appear that women's adult development must be studied in context. women's adult development must be studied in context. The purpose of this chapter is not to conclude that there is or is not a pattern of adult development development distinctive among women, but to explore explore this very large of question. Whether there are general features in the way women change over the course of of any single investigation. of adulthood cannot be decided on the basis of Although cross-sectional studies (in which individuals of of different different ages are studied at the same time) may sometimes be useful, what appears to be an age difference difference in such studies may actually be attributable to differences differences in cohort cohort experience. (A cohort consists of people born about the same time who experience the same historical events at the same stage of of life.) One needs longitudinal studies (that is, studies of the same individuals followed over time), conducted in a variety of of historical and cultural contexts. Not only is no one study sufficient sufficient to show whether there is adult development of view is sufficient. sufficient. distinctive to women, but also it seems likely that no one point of Uke searchlights that clarify Theoretical perspectives are like clarify certain areas but throw others into obscurity. Therefore, we will discuss several theoretical approaches and use them all. The organization of this chapter is as follows:first first we state several assumptions assumptions about the question of of women's adult development. Then we consider four theoreti theoretical approaches to it, present two hypotheses or guiding ideas, and, after a brief brief review of social history affecting affecting women since the 1920s, we evaluate evidence for the hypotheses in studies of of American women born at various periods of of the century.

CHAPTER 12 12 CHAPTER

DISTINCHVE 1'0 TO WOMEN DISTINCI1VE

293 293

I. SOME SOME ORGANIZING ASSUMPTIONS Knowing how different people are, and the great variety of conditions under which Uve, is it reasonable to expect that their adult development could show importhey live, impor tant common features? Tooby and Cosmides (1990) make the case that human nature, which they define as a species-typical collection of complex psychological adaptations, is the same across races, ethnic groups, and classes (because of the evolutionary genetics of sexual recombination), but that males and females do "morphs." They believe beUeve that a complex coordination between constitute different "morphs." the physiological and psychological systems of males and females has evolved to support sexual reproduction (see also Buss, 1989). interOur assumption in this chapter is that biological differences as socially inter preted and elaborated and as experienced by individuals produce lives with content differences that is substantially different for men and women. There are universal differences in the early socialization of male and female offspring (Chodorow, 1978). In most cultures there there is is aa gender-based division of of labor. labor. Women Women have have almost almost always always cultures gender-based division occupied a subordinate position in society: they have been directed by men, occupied a subordinate position in society: they have been directed by men, pospos sessed less less income income or or property, property, and and held held fewer fewer positions positions of of high high status. status. Women's Women's sessed sexual attractiveness attractiveness to to men men and and their their childbearing childbearing function function have have always always played played sexual key roles roles in in the the social social construction construction of of women's women's lives. key lives. different A second assumption is that the lives of men and women are more different in some cultures and in some periods of history than in others, and that they are never entirely different. In the United States in the 1950s, many men went to work in the city while their wives maintained their suburban homes. Today both men do paid paid work. work. Even Even when when gender gender differences are are pronounced, pronounced, however, and women do men and women share experiences that give an important shape to adult developdevelop ment: both women and men see their children mature and their parents die. ment: A third assumption is that there are not only main themes but also main variations in in women's life life stories (Lott, 1987). 1987). For For example, example,there thereare areusually usuallywomen women who follow the pattern expected of them in their society along with others whose attitudes and values are more like those of men. men. There are women who emphasize mothering, those who emphasize the role of intimate companion, and those who emphasize work and and career. Other variations variations may may be be related relatedto toculture, culture,class, class,cohort, cohort, or psychological pattern. In sum, it is reasonable to look for general features in women's adult develop development as long as one realizes the complexity of the search.

n. n. SOME IDEAS AND THEORIES ABOUT WOMEN'S ADULT DEVELOPMENT

Ideas about women's adult development may be grouped for the purposes of this chapter into four categories: functionalist, relational, conflict, and normative. Each

294 294

BELSON, HELSON, PALS, PALS, AND AND SOLOMON SOLOMON

category includes a mixture of theories and what are more properly considered as interpretive frameworks frameworks or perspectives.

A. A. Functionalist Functionalist Perspectives Perspectives Functionalists interpret personality differences differences between men and women at different different stages of life as manifesting manifesting and as having the function function of supporting the biological or social system out of which they are said to have arisen. For example, Buss (1989) industriousshowed that across many cultures young women valued ambition and industrious ness in a mate more than young men did, whereas young men valued physical interpreted these differ differattractiveness in a mate more than young women did. He interpreted ences in terms of evolutionary selection pressures which led to different different reproductive strategies in males and females. differences might be given an alternative functionalist functionalist interpre interpreSimilar gender differences affective tation in terms of the division of labor in society between instrumental and affective tasks (Parsons & Bales, 1955). Perhaps it is found that young women are more interested in finding a marital partner than young men, who are more interested in vocational goals; or perhaps young women are less planful planful and more adaptable than young men. These differences of differences may be construed as serving the function function of preparing for adult roles: for for example, not having plans of their own, women can adapt more easily to the husband's way of life. Erikson (1963, 1968) characterizes the developmental task of adolescence and (1963,1968) the transition into adulthood adulthood as the formation formation of identity, with the resulting capacity to give fidelity fidelity to work, values, and other people. His treatment of differences differences between males and females in the development of identity is functionalist functionaUst in that it assumes the importance of of biology in determining determining gender roles. In discussing the development men, or in a general way, way, he emphasizes vocational development of identity in young men, choice. For a young woman, he says, identity formation formation involves the recognition of her "inner space" (reproductive capacity) and its integration into other aspects of **inner of her self. The choice of a husband, he says, is both the expression of a young of woman's identity and an important determinant of it. Thus, he links the process of of identity formation in a young woman to anatomy, potential for parenting, and formation for partner. The functionalist functionalist theory with the broadest life-span reach on the topic of sex differences differences is that of of Gutmann (1987). The personalities of young adults show a gender specialization, he says, that is conducive to the propagation of the species. Young men accentuate accentuate personality personality characteristics appropriate for for earning a living and protecting the family, and young women accentuate characteristics appropriate for helping the husband and caring for children. In adapting to young adult roles, each sex has to suppress characteristics construed as typical of the other; but in the second half half of life, men and women each begin to relax this suppression and show more of the characteristics that had been attributed to the other.

12 CHAPTER 12

DisTiNcnvE TO WOMEN DISTINrnVE

295 295

B. Women's Development in in Relationships Relationships B. Women's Development affiliations with object relations A second framework framework may be considered to have affiliations major modern development within psychoanalysis. A central idea is theory, the major that the self self in all human beings develops from early childhood throughout life in the context of important relationships. The influence of relationships is particularly evident in women, because men tend to resist what they construe as dependence and emphasize autonomy ((Jordan, Jordan, Kaplan, Miller, Stiver, & Surrey, 1991). Whereas Erikson considers intimacy a developmental task that requires a relational theorists theorists maintain maintain that issues issues of identity identity previous development of identity, relational Unked in female development, if not in the develop developand intimacy are inextricably linked of both males and females (Franz & White, 1985). Josselson (1987) describes ment of women's lives as unfolding in the context of organizing relationships that she calls "anchors." While each woman constructs her own sense of personal identity, the anchoring process by which this sense of self-definition self-definition comes about demonstrates the relational dimension of development characteristic of women. In Josselson's study, women successful successful in careers had anchoring relationships with mentors. Gilligan (1982) emphasized relational issues in her three-stage model for for the self self to ensure women's moral growth. Initially there is a focus on caring for connection between survival. The next phase is one in which the woman develops a connection self and others that is articulated by the concept of responsibility. The notion of of confused with self-sacrifice, self-sacrifice, so that women at this stage have responsible care is confused limited awareness of their own needs and agency. The transition to the third stage of the "illogic of the inequality between the other and is preceded by a recognition of the self" (p. 74). 74). Responsible care then becomes a self-chosen value which takes into account both the possibilities and the limitations of one's actions in the lives of others as well as one's responsibility for self-development. self-development. Feminists who hold the relational perspective believe that women's interest gift for for sharing and mutuality are downgraded downgraded and misinterpreted in others and their gift as dependence and triviality by a patriarchal society (GiUigan, (Gilligan, 1982; 1982; Jordan Jordan et et aI., al., 1991; Miller, 1976). These theorists say that Erikson (1963), Kohlberg (1973), and 1991; Levinson (1978) wrongly assume the central importance of autonomy in their acac counts of development. Women may appear inferior by these male standards, but inferior should be assessed in terms of of their own. C. C. Conflict Perspectives Conflict theorists, of whom Karl Marx is the best known example, study processes of power in society and the consequences of of this distri distrimaintaining the distribution of bution. Most feminists hold a conflict perspective. Without disputing the existence functionalists and rela relaof the differences differences between men and women described by functionalists differences tionalists, those with a conflict perspective attribute the origin of these differences to the subordinate position of women in society. For example, if women are more

296 296

HEI30N, HELSON, PALS, PALS, AND AND SoLOMON SOLOMON

adaptable than men, they would say that it is because society makes makes women women depen dependent on men. "What peculiarly signalizes the situation of woman," woman,*' wrote Simone de Beau Beauvoir (1953) more than 40 years ago, "is that she-a she—a free and autonomous being like all human creatures-nevertheless finds herself creatures—nevertheless finds herself living in a world where men compel her to assume the status of the Other" (p. (p. xxviii). xxviii). She analyzed how this affected young women, married women, aging women, and women in situation affected special categories, such as lesbians, independent independent women, women, and and prostitutes. prostitutes. "The "The first first twenty years of [a] [a] woman's life are extraordinarily rich rich.. .. .. .. [But] [But] at twenty or thereabouts mistress of a home, bound permanently to a man, a child in her arms, (p. 451). 451). Of Of old old women women she she said, said, she stands with her life virtuallyfinished finished forever" (p. "Old women women take take pride pride in in their their independence; independence; they they begin begin at at last last to to view view the the world through their their own own eyes eyes.. .. .. [but] [but] the the highest highest form form of of liberty liberty available available to to the the woman woman through parasite is stoical defiance or skeptical irony. At no time in her life does she succeed parasite is stoical defiance or skeptical irony. At no time in her life does she succeed in being being at at once once effective effective and independent" (p. (p. 561). 561). in and independent" Women's lives in the West have changed a great deal since Beauvoir wrote The Second SecondSex. Sex. However, However,American Americanfeminists feminists ofofthe the1970s 1970scontinued continued totoprovide provide The evidence that women's subordinate status caused them pain and restricted their development. Bernard (1972) wrote on the inequities of "his and her" marriage, and Bart (1971) described depression in "empty-nest" women whose whole adult lives had consisted of mothering their children. Contemporary feminists, among other holders of the contlict conflict perspective, often hold a constructivist position, according to which we do not discover reality reahty (K. J. J. Gergen, but actively construct the meanings that shape our experience (K. 1985; Hare-Mustin & & Marecek, 1990). 1985; 1990). They argue that power elites assert their meanings. For example, the idea that the authority through control over these meanings. functionalmenopause is devastating for women is attributed to patriarchal bias in functional ist constructions, which reduce women to their biological roles and disregard the fact that most women do not want to continue having children (Barnett & Baruch, 1978; 1978; Oatan, Datan, 1986). Some feminists work toward new constructions of women's lives (M. (M. M. Gergen, 1990; 1990; Heilbrun, 1988), 1988), including including constructions constructions that that bring bring to to light light the the intlu influence on women of subordinate status and the resourcefulness resourcefulness of individual women & Stewart, 1994; 1994; Stewart, Stewart, 1994). 1994). despite low power (Franz & D. D. Normative Nonnative Perspectives Perspectives The normative perspective is concerned with how adult development takes place influence of social norms. norms. Neugarten (1977) described the under the structuring intluence timing norms or "social clock" that regulates the age-appropriateness of various endeavors for men and women, in part through feelings of anxiety or self-esteem that one experiences in recurrent appraisals of how how well one is doing doing for for one's one's age. age. Helson, Mitchell, and Moane (1984) developed the concept of "social clock projects" related to the major commitments of adult life: family and work. They

CHAPTER 12 12

DISTINC11 VE TO DISTINCTIVE TOWOMEN WOMEN

297 297

of college women women who illustrated the idea in a sample of who grew up in an era (the late 1950s) when they were expected Most women expected to marry in their early 20s. Most women started out on the "feminine ''feminine social clock" pattern of of marriage, child-rearing, and homemak homemaking, but some Some women some followed followed the "masculine occupational occupational clock." Some women with a feminine social clock project were "late" in finding husband or becoming becoming a mother; finding a husband others experienced the disruption of of their project through through divorce. Personality characteristics influenced tum her experience influenced the path a woman followed, followed, and in turn in the different for example) influenced different phases of of her project (becoming a mother, for influenced subsequent et aaI., subsequent personality change. According According to Helson Helson et l , historical historical periods vary in the structure and strictness of of timing norms about about the age at which various advances in social clock projects should be accomplished, how different the projects how different assigned to men and women are, and how projects evolve evolve or are combined over time. In their view, adult development is related to the social social clock clock projects available differences in the ability to make commitments and to change and to individual differences different phases of of the project or with changing changing times. with different clock projects are influenced not only only by norms also by the occur occurSocial clock norms but but also of major social events or changes in social rence and timing of social climate. Stewart and Healy (1989) believe that events that occur in late adolescence or or early in young influence on the formation formation of adulthood have the biggest influence of identity, because at this time the individual is uncommitted and looking for guidelines. Duncan and Agronick found that women who had recently graduated from from college during the (1995) found of the women's movement were more likely at midlife height of midlife to report report this event having been a major influence influence on their lives than women as having women who who were were older (and already committed committed to wife-mother wife-mother roles) when when the women's movement was at already women's movement its height.

ni. WOMEN'S WOMEN'S LIVES LIVESFROM FROMmE THE 1920s 1920STO TOmE THE 199 1990S m 0s Before turning to our hypotheses and to empirical studies of of women's adult develop development, let us review briefly some social history affecting affecting American women women since ferment that included a break break with World War I. The 1920s brought strong cultural ferment Victorian conceptions of the Great Depression of of gender roles. However, both the of the 1930s, when the male role of of provider was threatened (Elder, 1974), and then World War II, when the male role of of soldier and protector needed support, reinforced the traditional notions. The psychosocial effects effects of of the Depression and World War II seem to have fed an exaggerated cult of of women's women's domestic role in the late 1940s force and 1950s. Even though women had in fact participated heavily in the labor force during World War II, they were "helping out." Their place was at home. During educathis time they married at a younger age, had more children, and dropped in educa of gender roles remained unusually firm and tional level. Thus, the conception of complex of newly distinct for three decades (Skolnick, 1991). Then a complex of long-term and newly arising factors led to dramatic changes in gender roles in the 1960s and 1970s. These overpopulation, leading to changes in factors included increased longevity and overpopUlation,

298 298

HEL'lON, HELSON, PALS, AND SoLOMON SOLOMON

attitudes toward birth control, a reduction in birth rate, and an elongated stretch of "child-free" time for women; women; the need for women as service workers in the expanding postwar economy, and beginning in the early 1970s the need for two wage-earners in a family to combat inflation; the changing structure of the family, including a rising divorce rate; increasing increasing levels levels of of graduate graduate and and professional professional educa education among women; and the women's movement (Bianchi & & Spain, 1986; 1986; Chafe, 1972; Giele, 1993; 1993; Skolnick, 1991). Figure 11 shows replies from a questionnaire returned by more than 700 alumni of a west coast women's college. The question was, was. When you were in college, how long did you expect to do paid work? It is apparent that women who attended college in different decades from the 1920s to the 1970s had very different different life expectations. Central to this chapter is the fact that several longitudinal studies cover these eras of rapid change, and if it is possible to compare their information we may be able to obtain much of value on the generality of adult development in women.

50

|E4400

+

married" -trried" ;f\^ Until I m \\ ""Until \\:/.A / \ .1A\ \/ . I

,.+

�

I 30

\

I

&

I

� +

•

I

^ J-' \,/>--"

2 o L \\

A" \

10

0 '21-'25 •2l-*25

-^., N

"Alwoys"

\

/: \Ax;\ I-�\" /

I

: I

v

:t-" 'y",, ... .. .... .. /}" ""until h o d/ / +. until II had -+:

�l

children

./ /

^•^ ,�

'ai-'SS

'•4l-*45 41-�5

/

\.

,"

\ "My My ideos ideos about about were '.\ \\\ work work were " unclear / \ \\ " " \\cunclear" leor

.....

\

:t- - +"

61

''61-65 -'65 '5l-'56 '5IJ� Cia" Gloss Groupifl91 Groupings

..

�

-4

\ \ \ \ \ ..... �. \\ \\ +....... � JL •7I-75

FIGUREl1 How How long long after after college college did didyou youthink think that that you youwould would work? work? Replies Replies from from 700 700 Mills Mills College College FtGURl 12 age age groups. Figure is from R. Helson, T. T. Elliott, and J. J. Leigh, "Adolescent Personality alumni of 12 Work (p. 266), by D. D. Stem and D. D. Eichorn Eichora (Eds.), Adolescence and Work and Women's Work Patterns," in Adolescence 1989, Hillsdale, NJ: Erlbaum. Copyright 1989 by Lawrence Erlbaum Associates, Inc. 1989, Inc.

CHAPTER CHAPTER 12

DISTINCTTVE WOMEN DIS11N C'IlVE TO WOMEN

299 299

IV. SOME POSSIBLE GENERAL GENERAL FEATURES OF WOMEN'S IV. SOME ADULT DEVELOPMENT What shall we look for from these studies? We have chosen two hypotheses, both derived from from functionalist functionalist ideas, because the functionalist functionalist perspective has a longer history and a more central place in the field of psychology than the relational, conflict, and normative perspectives. However, we will draw upon all four four perspec perspectives in evaluating the evidence for for or against these hypotheses. The evidence comes from from four four longitudinal studies a�d and one cross-sectional study, each representing a different different cohort. The first hypothesis is that in late adolescence women are primarily interested functionalist family and do not have strong vocational goals. This is a functionalist in marriage and family hypothesis because of its supposition that young women prepare themselves for from both the biological and social roles of mothering. When data are available from men and women, young men are expected to emphasize vocational goals over manifested by its frequency frequency of mention marriage. The strength of a goal may be manifested demonstrain response to a question about goals, its being given a high priority, or a demonstra tion that it predicted subsequent behavior. ILB) The second hypothesis tests Gutmann's conception (1987, see section II.B) of of a biosocial parental imperative that relaxes with age. Young adult women are expected to show nurturance and suppression of their own assertiveness as part of of a division of labor conducive to child-rearing, but to increase in independence and planfulness in later life. Men are expected to show independence and planfulness planfulness planfulness in the period of of early parenting but to become less achievement oriented and more concerned with relationships in later life. To evaluate these hypotheses we will examine longitudinal studies in which was about 25 25 years old in the middle of the Great Depression the average participant was following World War II (IHD studies), or in the (the Terman Study), in the era following early 1960s (Mills Study), and then turn to samples of younger males and females who have been followed followed over shorter time spans or in preadolescence and adoles adolescence alone. V. RELATIONAL AND VOCATIONAL GOALS OF YOUNG WOMEN V.

A. The The Tennan Terman Study A. Study The Terman sample consisted of men and women first identified as having high IQs lOs as school children (see the Appendix for details). Both men and women were 67% of the highly educated for the times, but it is particularly impressive that 67% from college and 24% completed graduate degrees (usually women graduated from California population was M.A.'s) during the 1930s, when less than 8% of the California graduating from from college (Tomlinson-Keasey & Keasey, 1993).

300 300

HEI.'iON, HELSON,PAI.'i, PALS,AND ANDSoLOMON SOLOMON

Most Terman women were between the ages of 23 23 and 27 in 1936. 1936. At At this this time the sample was was asked, "Have you definitely definitely chosen your life work?" and "Describe ''Describe your ultimate goals as fully fully as you can at this time." time.'* Nearly all of the men answered the question about goals with the name of one occupation, but many of the women gave at least two responses. Of codable replies from from 290 women women who of participated in both 1936 1972, half half mentioned mentioned homemaking and 61% 61% had participated 1936 and 1972, mentioned occupational occupational goals (Holahan, 1994). This rather rather surprising amount amount of mentioned of orientation may be attributable to the fact fact that they were adolescents in the work orientation was questioning questioning of prevailing gender roles, combined with the 1920s, when there was fact that they had been identified gifted and were highly educated. Also, during fact identified as gifted women often often needed to bring in income, though they were subject subject the Depression, women to the prejudice prejudice that women women who who worked were taking jobs away from from men men and were expected to give up their jobs when they married. married. "The careers of women are often often determined determined by extraneous circumstances circumstances rather than by training, talent, or vocational interest" (Terman (Terman & Oden, 1959, p. that 88% 88% of of the women women married, married, and and that that from from then then on on they they generally generally 144). It is true that from their husbands and their social status from from his achieve achievederived their identity from divorced or being (Tomlinson-Keasey & Keasey, 1993). ments (Tomlinson-Keasey 1993). Being single or divorced married with no children was highly related related to employment, employment, and Terman Terman and Oden married (see above) saw saw these features features of life style as "extraneous "extraneous factors" rather rather than as expressions of the women's own desires. Nevertheless, the Terman women were entirely husband husband oriented oriented and malleable: there was a significant significant relation relation between between not entirely homemaking or career during young adulthood adulthood and the predominant predominant life life plans for homemaking pattern as described described by the women many years later (Holahan, 1994). Though the largest categories of paid employment employment for these gifted, college-educated college-educated women largest office worker worker and school teacher, about 60% of the women in 1972 (average were office classified themselves as career or income workers and 40% as homemakers. age 61) classified Terman women were interested interested in marriage and In sum, as young adults the Terman accommodate the needs of the husband husband and family. However, many expected to accommodate family priorities, those who who did did also expressed occupational interests, and despite family tended tended eventually eventually to go on to stable work patterns.

B. The IHD rao Studies Women in the IHD samples (see the Appendix for details) were studied as high 1940s, a time of strongly differentiated differentiated school students in the late 1930s or early 194Os, gender roles. IHD investigators agree that the women were socialized to want a traditional gender role (Stroud, 1981) and that they were programmed programmed for for marriage traditional They were were not not given given the the encouragement encouragement that that their their brothers brothers received received (Clausen, 1993). 1993). They to go to college, and if they worked, they did not expect to stay in the job. According to Clausen (1993), young women gave more thought than young men to the kind of person they wanted to marry. planful competence in adolescents, Clausen (1993) measured In a study of planful investment. competence as the sum of three components: dependability, intellectual investment,

CRAPfER CHAPTER 12

DISTINCTIVE TO WOMEN

301 301

and self-confidence. In males the three components were correlated correlated and predicted vocational success and much else in their lives. For female adolescents competence was a much less powerful of powerful variable. It predicted status of husband and number of children, but little else. For females, the component component of self-confidence self-confidence was strongly related to physical attractiveness, and it predicted both marital satisfaction satisfaction and life satisfaction satisfaction more strongly than the total competence score. According to Stroud's study (1981) of college-educated women in the more Uttle evidence for career motivation advantaged of the two IHD samples, there was little among these women as adolescents. "The distinctive characteristics of adolescent work-committed women was their exclusion from. girls who became work-committed from . . . traditional pathways. . . . Viewed as the least attractive physically and the least sex-typed selfin feminine style and behavior, they were self-conscious, basically anxious, self defensive, and and had bodily concern" concern" (p. (p. 373). 373). Only Only much much later later did did these these women women defensive, had bodily come to to differ differ from from their their peers peers in in intellectual intellectual orientation orientation and and assertiveness. assertiveness. come from studies of samples in the IHD studies corroborate the These findings from functionalist conception that late adolescent women are primarily interested in a functionalist good marriage and that physical attractiveness is an important resource in attaining (planful competence) were, as this goal. Traits conducive to vocational success (planful hypothesized, more more important important in in young young men men than than in in young young women. women. hypothesized,

C. The The Mills Mills Study Study c. After World War II and throughout the 19508 1950s there was much emphasis on early After marriage and a big family. The Mills women were studied first as college seniors for details). The world was changing, but 88% in 1958 1958 or 1960 (see the Appendix for 88% of the Mills women had plans or hopes to marry, and 75% 75% of them wanted to have 3 or more children (Helson, 1993b). 1993b). (They actually had an average of 2.2 children, as compared with 3.4 for the IHD women.) qualified or Some 20% had long-term career aspirations, and many more had qualified difficult, they expected to work contingent career interests. Though planning was difficult, after their children were older, if their husbands were until they had children, after affected (Helson, cooperative, or as long as family interests were not adversely affected 1993b). As might be expected, women with traditional life plans were clearest in their goals. A personality index of the characteristics of purposiveness, independence, and ambition (labeled agency) at age 21 did not predict work status at age 27, 27, when most women were having or expecting children, but it did predict work status at age 43 (Roberts, 1994). Physical attractiveness in college (rated from yearbook powerful an advantage in this sample as in in the IHD sample: pictures) was not so powerful beautiful women married earlier but were also more likely to divorce (Kaner, 1994). beautiful In the early adult years, the marriage and family orientation of the Mills women fit fairly well the functionalist functionalist framework. However, the women did not "feminine lack the capacity to develop vocational goals; rather, the nature of the "feminine social clock clock project" project" they expected to to carry carry out, out, especially especially the the idea that their their lives lives social they expected idea that

302 302

HELSON, PAU, PALS,AND ANDSOLOMON SOLOMON HEISON,

would be shaped shaped by the man they married, precluded precluded differentiated differentiated planning and attached only marginal relevance to their college-age vocational goals. Over the attached next 15 years the women completed completed the most intensive phase of of child-rearing and the social world changed changed considerably, providing opportunities for for the expression expression of vocational goals. of D. Illinois Valedictorian Project As detailed in the Appendix, this study began with high achievers, more than half half of of them women, from from the graduating graduating classes of of Illinois lUinois high schools in 1981. 1981. The The followed through the next decade (Arnold, 1993). Although Although there were sample was followed differences in high school, the intellectual self-esteem self-esteem of of the women women no gender differences dropped sharply by the sophomore year of of college. It rose somewhat by senior dropped but remained remained significantly significantly lower lower than that of of the men. A persistent persistent theme theme year, but women was concern concern about about combining combining a career career and family. As among the young women seniors in college, their their professional professional expectations tended tended to be vaguer that those those of the men in the sample. of found also in the Mills women, can be interpreted interpreted in several several This vagueness, found ways. Arnold Arnold argues that that the realization realization of of early promise is more difficult difficult and complex for for women women than for for men, at least in part because they lack "tacit knowl knowledge," defined defined as career-related career-related understanding understanding that is rarely verbalized and not explicitly taught. Women, Arnold Arnold says, are constrained constrained in the development development of of their career career identity by their lack of of support support from from models and their inability inabihty to see themselves in high-level career career settings. She gives two case studies. One of of them woman with many self-doubts self-doubts and uncertainties who became describes a young woman successful in her profession profession through excellent support support and role modeling. highly successful other describes describes a young woman who had not developed developed a career career identity. In The other factor was that she had not found found sufficient sufficient support. support. Arnold's view, at least one factor The Valedictorian Valedictorian study gives evidence that despite radical changes in oppor opportunities for for women women and in attitudes toward toward their career career achievement, young women continued continued to suffer suffer uncertainty uncertainty about investment investment in careers.

E. Identity Formation in High School Males and Females Archer (1985) studied the process of identity formation formation in a sample of of 6th, 8th, Archer order to test the idea that that many many adolescent adolescent women engage 10th, and 12th graders in order "pseudoexploration" of of occupational identity in order to fill time until marriage. in a "pseudoexploration" She interviewed interviewed male and female female students to assess their identity development development relevant domains including occupational choice and career-family career-family overall and in relevant For each domain, the state of of identity formation formation was assessed as a function function priorities. For of two major major dimensions of of self-definition, self-definition, the exploration of of possibilities and the of commitment to a self-defining self-defining choice (Marcia, 1980). domain of of occupational choice, choice. Archer Archer found found no gender gender differences differences in In the domain of identity formation. In the domain domain of career-family career-family priorities, howthe process of

CHAPTER 12 CHAPTER

DISTINCTIVE TO WOMEN DlmNC'I1VE

303 303

ever, girls were more likely than boys to demonstrate evidence for both exploration and commitment. Among the 12th graders, girls were more likely than boys to anticipate conflict between occupational and family family roles, and of those who antici anticipated conflict, girls were more than three times more likely than boys to express concern about the resolution of this confl ict. Finally, 12th-grade girls were more conflict. likely than boys to be in a state of active exploration with regard to the overall sense of personal identity. These findings suggest that contemporary contemporary young women actively explore fu future occupational possibilities to the same degree as young men, but that their difficult to resolve because career exploration is more complex and ultimately more difficult family and career. of their greater concern and questioning about having both family of identity, particularly Furthermore, this conflict appears to be central to their sense of concludes: "Exploration girls is is genuine; in the the later later teenage teenage years. years. Archer Archer concludes: in "Exploration for for these these girls genuine; it is is not not pseudoexploration. pseudoexploration. This This is is in in the the face face of of aa society society that that offers offers little little support support it to these girls about about to to be be torn torn between between two two greatly greatly valued goals in in their their lives" lives" to these girls valued goals (p. 311). 311). These These were were high high school school students. students. We We have have already already seen seen from from the the Illinois Illinois (p. Valedictorian Study Study (Arnold, (Arnold, 1993) how the the identity identity exploration exploration of capable young young Valedictorian 1993) how of capable women becomes becomes more more confused confused as as the the realities realities of of adult life draw near. women adult life draw near.

F. Summary Summary and and Discussion Discussion F. All studies show that most late adolescent women have considerable interest in marriage and family. In the older studies family and career were considered separate tracks, so that serious investment in one tended to preclude the other. Nevertheless, a high proportion of Terman women (61%) were interested in vocational goals. goals. confined their Most IHD women, representatives of the most conservative cohort, confined ambitions to the roles of wife and mother. From the Mills sample to the most recent samples, many young women hoped to combine family and work. However, their vocational planning was found to be more diffuse diffuse and contingent than that of young men. This diffuseness may be functional for a sex-based division of labor, in that diffuseness functional willingness to relinquish or compromise vocational goals facili the young woman's wilUngness facilitated male careers, marriage, and the rearing of children. From the conflict point of view, however, our findings show that women have been disadvantaged by lack of opportunities in the work world, and they continue to be handicapped by difficulties difficulties in developing clear goals. Aspiring to and actively working toward a clear career goal allows one to incorporate these future future role expectations into the current self-concept self-concept (Heise, 1990; Markus & Nurius, 1986). Thus, self-esteem suffers suffers when one lacks self-enhancing self-enhancing goals. This may be the reason that the Terman women had a persisting lack of confidence and guilt that they had not lived up to their potential (Tomlinson-Keasey, 1990). In the studies of the younger women, both Arnold and Archer took a conflict conflict perspective in attributing the women's unclear life plans to lack of social support and an unfair share of of responsibility. responsibility. share

304 304

HEUON, HELSON, PALS, PALS, AND AND SOLOMON SOLOMON

Though young women's persistent concern for for marriage and family family may be counted counted as support support for for the functionalist functionalist perspective, it may also be construed as an indication of of their relational nature. Holders of of the relational point of of view would call attention to the social value of this concern and question its construction attention of concern question construction in terms of of deficiencies-"lack" deficiencies—"lack" of of planfulness planfulness or the "failure" of of women to under undertake and persevere in demanding careers. From From the normative point of of view, the demonstration demonstration of of differences differences in career career interest in different different cohorts indicates the importance of of the social world for for women's women's motivational structure. The fact fact that that the IHD sample showed showed more exclusive orien orientation toward toward marriage and family family than the Terman and Mills women is consistent consistent born in the 1920s grew up in a world structured structured by a with the idea that people born sequence sequence of of conditions-depression, conditions—depression, war, affluence, affluence, and stability-that stability—that produced produced maintained particularly particularly specialized gender roles (Skolnick, 1991). and maintained GENDER-RELATED PERSONALITY CHANGE VI. GENDER.RELATED VI.

Now we consider the second second hypothesis, that men and women change in personality personality reUnquish the responsibilities of of parenting, or more generally, as they assume and relinquish from young adulthood adulthood to later adulthood. In this section we will as they develop from from the three longitudinal longitudinal studies that cover the entire period period review findings only from from young adulthood adulthood to later later life. from

A. The Terman Study Terman Study has very few personality measures. However, there is information information The Terman about interests and values of of the men and women in the sample. Comparing a Terman Terman subsample of of college-educated college-educated men and women, Holahan (1984b) found found that the men emphasized emphasized work work values more than the women at both 30 and 70, but but men's occupational occupational concerns decreased more than women's by age 70, so that that the gap narrowed narrowed considerably. Even Even more striking was the change in endorsement endorsement of of home and family family as life life goals: in 1940 (age 30) only 7% of the men mentioned mentioned home and family as a life 1981 (age family life goal, in comparison comparison with 64% of of the women. In 1981 70), 40% of the men and 47% of the women mentioned home and family of of mentioned family as their life goal. The change in values of Terman men had begun before age 70: at age 62 of before Terman men derived even more satisfaction from their family life than from Terman derived satisfaction from family from their occupational involvements (Sears, 1978). of longitudinal longitudinal and cohort cohort factors in marital attitudes, Holahan Holahan In a study of (1984a) showed showed that a college-educated college-educated sample of of Terman Terman men expressed expressed more more agreement agreement at age 70 than at age 30 with items such as "The ' T h e husband husband and wife wife agree should express their love in words." Women were more inclined than men to agree not changed significantly significantly from from ages 30 to 70. Recruiting Recruiting with this item, but they had not younger a younger sample of 30-year-old men and women, she found that the younger cohort of of men agreed agreed significantly significantly more with this item than the Terman men had cohort

CHAPTER 12 12 CHAPTER

DISTINCDVE DISTINCTIVE TO TO WOMEN WOMEN

305 305

30, and did not differ differ from the younger women or from the Terman women. at age 30, (The young men and women were less satisfied with their marriages, however, than Tomlinsonthe Terman men and women had been at the same age [see also Tomlinson freer in Keasey & Blurton, 1992]. 1992]. Whether the younger men and women were freer being critical, had higher standards for for relationship, or were attesting to frustrations frustrations in modern marriage is not clear.) One concludes that the change in attitude on the part of the Terman men may be due in part to greater affiliativeness with age but that it may be due also to a change in social climate with a particular particular influence influence on men.

B. B. IUD IHD Studies Studies

Haan, Millsap, and Hartka (1986) conducted a study of personality change in men and women that combined the Oakland (OG) and Berkeley (BG) samples (see the for our purposes, Appendix). The combination of these two samples is not ideal for because they they differ differ in in age age by by about about 77 years. years. Nevertheless, we we shall shall assume assume that that the the because participants at at age age 30 30 (BG) and and 37 37 (OG) (OG) were were more more actively actively engaged engaged in in parenting parenting participants 61. From From than at at ages ages 40 40 (BG) and 47 47 (OG), and more more so so still still than than at at ages ages 53 53 and and 61. than (BG) and (OG), and ages 30 30 and and 37 37 to to ages ages 40 40 and and 47, 47, men men and and women women both both changed changed in in the the same same ages directions, but but only only for for women women were were the the changes changes significant: significant: women women increased increased on on directions, Q-sort factors labeled self-confidence, cognitive commitment, outgoingness, and Q-sort factors labeled self-confidence, cognitive commitment, outgoingness, and warmth. From From ages ages 40 40 and and 47 47 to to ages ages 53 53 and and 61 61 women women decreased decreased in in assertiveness assertiveness warmth. (a scale with connotations of undercontrol and hostility) and both men and women women (a scale with connotations of undercontrol and hostility) and both men and increased in warmth. increased in warmth. These findings are most consistent with an idea that has not been mentioned because it is not gender specific: that middle-aged adults show superior cognitive 1977; Stevens-Long, 1990). and executive abilities and relational skills (Neugarten, 1977; found in the fact that only In addition, support for the Gutmann hypothesis may be found significantly in self-confidence, outgoingness, and cognitive the women increased significantly significantly in warmth (though the women commitment, and that the men increased significantly did also). Though the women's decline on assertiveness between the middle and be taken as as contrary to the Gutmann hypothesis, it may indicate indicate the late periods may be was sufficiently sufficiently secure to reduce rebellious undercontrol and self-assertion was that their self-assertion hostility. The Haan et al. study is valuable, but it was not intended to test the Gutmann hypothesis, and the evidence from it for this purpose remains fuzzy. fuzzy. c. C. The The Mills Mills Study

Comparing the Mills women and their male partners at the early parental and periods. Wink and Helson (1993) found that the husbands described postparental periods, themselves on the Adjective Adjective Check List (ACL) (Gough & & Heilbrun, 1983) as less affiliative and as more competent and independent than their wives at the earlier affiliative time but not at the second. Looking at change in those couples who were the same individuals at both times of testing, they found that the men did not decline in

306 306

HELSON, PAlS, PALS,AND ANDSOLOMON SOLOMON HElSON,

competence, independence, and self-confidence self-confidence but the women showed sharp in increases. Husbands increased increased significantly significantly in affiliation, affiliation, whereas wives did not. Overall, the findings findings were in the direction direction predicted by Gutmann. However, However, women without without children changed in the same ways as women with children, so that literal parenting did not explain the findings. Furthermore, data available from from a subsample of of the parents of of the Mills women did not give the same results. The parents had filled out the ACL ACL only once but at the same postparental postparental ages as their daughters and sons-in-law. Even at age 52, the mothers scored lower on competence competence and independence independence than their husbands. Very few of of the mothers did paid work, and they had received less education than their spouses and daughters. D. Summary and Discussion longitudinal studies that that we have examined used different different kinds of of mea meaThe three longitudinal different age spans and periods of of history. Nevertheless, all show sures and covered different support for for the idea that gender gender differences differences in interests and personality personality some support fact that Mills women with and without children changed decrease with age. The fact in the same ways indicates that literal parenting was not the critical factor. However, from a functionalist functionalist perspective, a broad biosocial explanation explanation (perhaps in terms from of shifting shifting hormonal balances) is not ruled out. of From the relational point of of view, evidence that women become more indepen indepenFrom consistent with Gilligan's idea that it is a moral moral dent and assertive with age is consistent advance for for women to develop develop the ability abiUty to take care of of themselves as well as Increased affiliativeness affiliativeness in men is similarly consistent with the idea that it others. Increased is desirable for for men to develop communal skills and integrate them into their agentic orientation orientation (Bakan, 1966). From From the conflict conflict point point of of view, an increase in independence independence and assertiveness assertiveness adulthood can be interpreted in women from from young to later adulthood interpreted as a result of of the greater greater social control of of women women during their years of of sexual attractiveness, fertility, fertility, and active mothering. When this control lessens, women may have more autonomy, at least if if they have prospects for for interpersonal interpersonal power. In both the United States and Kenya, middle-aged women showed more power motivation than younger women when participants had high social status, but among women of of low social middle-aged and young adult women did not differ differ (Todd, Friedman, & status the middle-aged Researchers with a conflict conflict perspective point out that the clearest Kariuki, 1990). Researchers for increases in independence and assertiveness in women come from from evidence for opportunities for for women increased increased between their young adulthood adulthood samples in which opport�nities and late middle age. Thus, the Vaillants (1990) report report the appearance appearance of of late-life late-life accomplishments among creative Terman women that they attribute to increased accomplishments among creative Terman women attribute increased opportunities for women by the 1970s and 1980s. women who opportunities for women 19808. In the Mills Study, women remained in the traditional traditional role of of nonworking nonworking wife wife and mother mother did not increase remained significantly in independence and assertiveness between ages 21 and 43 as other significantly independence and between women did (Helson (Helson & Picano, 1990).

CHAPTER CHAPTER U 12

DISTINCTIVE TO TO WOMEN DISTINCOVE

307 307

The importance of social opportunity for for women's increases in independence and self-confidence self-confidence would also be emphasized by exponents of the normative point fact that the mothers of the Mills women did not seem to have shown of view. The fact the change that their daughters did seems to indicate that social resources and expectations moderate any biosocial factor factor (Gutmann, 1987) that may be involved. Holahan's study of marital attitudes points out the danger of attributing universality to changes that are attributable to social climate or cohort experience.

VII Vn. . OVERVIEW AND AND IMPLICATIONS Traditional science sets up a competitive relation among theories: the strongest wins. The conflict point of view often often attacks traditional science as narrow and biased. The relational view advocates connection over competition. In bringing together multiple frameworks frameworks for for a larger understanding of the complexities of together of adult development, we have tried to achieve a research strategy that maintains for scientific scientific procedures while asserting the value of recognizing and "con "conrespect for necting" a variety of perspectives. A second distinctive feature of this chapter is its use of longitudinal studies. major longitudinal studies began in the 1920s as a promising new approach Several major to child development. Seventy years later we see that there are studies of lives over several generations, and by comparing them we hope to address previously developinaccessible questions about social and historical influences on personality develop ment. It is not easy to make these comparisons, however. Different Different studies provide different kinds of information. Participants have been studied at different different ages and different differed in background background and resources. If the comparative study of longitudinal differed findings becomes recognized as a valuable endeavor, researchers can build on findings measures they have in common across studies to address particular questions (Helson, Stewart, & Ostrove, 1995) and develop techniques to increase and improve the knowledge obtained (Helson, 1993a). For example, with particular questions in mind, researchers can recruit new samples comparable (in intelligence, for exam exampie) ple) to the older ones, using some of the same questions or instruments (Holahan, 1984a; Mitchell & Helson, 1990; 1990; Tomlinson-Keasey Tomlinson-Keasey & Blurton, 1992). Has our attempt to study complex questions with a new strategy and longitudi longitudinal data from from several generations of men and women been successful? Within the limits of of our study, we believe that we have presented information information instructive for evaluating our two hypotheses. According to the first, young women were said to be primarily interested in marriage and family and to lack clear vocational goals. The evidence examined shows that from from the oldest to the most recent samples, young women did give higher priority to marriage and family than young men. Only in the IHD samples did young women show almost exclusive interest in marriage and family, but in the other samples vocational goals tended to be unstable for this hypothesis. or unclear. Thus, there seems to be considerable support for

308

BElSON, HELSON, PALS, AND SoLOMON SOLOMON

From the normative perspective, however, the large differences among the studies reviewed give ample evidence that changing social norms and expectations affected the planning of young women. From the conflict perspective, women are not by nature less able than men to set vocational goals, but the web of life circumstances associated with their subordinate status as helpers has made the process difficult. Even in the younger samples the complexity of the responsibility of planning for both family and career and insufficient insufficient social support for women's pursuit of careers contribute to ambivalent or tentative vocational plans. From the relational perspective, the importance of long-range vocational goals for most men and women has been overestimated by achievement-oriented academics. Women are are experts at at contingency planning, reconciliation of of conflicting interests, interests, and and taking taking The processes processes of of "connecting," "connecting," to towork work advantage of opportunities when they arise. The advantage issues as well as to people, need to be studied and supported in adult development. The second hypothesis was that men and women's personalities change be between young adulthood and and later life, in part because because of of aa biosocial biosocial parental parental impera imperative that relaxes over time, so that each sex shifts in the direction of interests or characteristics associated with the other. This is a very broad hypothesis, and it is not surprising that the research literature shows conflictingfindings, findings, depending in not part on the age span tested, characteristics of the sample, and effects of cohort and part social climate that vary with the personality measures used (Bengston, Reedy, & Gordon, 1985; 1985; Parker & Aldwin, in press). In all three longitudinal studies that we examined, however, the findings findings showed at least part of the hypothesized pattern. From youth to old age, Terman men men became less interested in occupation and and more more interested interestedin in family, family, and andTerman Terman women changed in the opposite direction. IHD women became more confident and intellectually invested as they moved from early to late adulthood; IHD men did not. Husbands of Mills women were more competent and independent and less affiliative than their wives in young adulthood, but not by their 50s. Though there was considerable support for the hypothesized pattern, the findings findings were weresometimes sometimes mixed mixedor oruneven. uneven.For Forexample, example,both boththe the IHD IHD and andthe the Mills studies found more change for women than for men, and the hypothesis was supported by data from the Mills women and their partners, but not from their parents, so that cohort influences would seem to be involved. Longitudinal findings findings are not yet reported from a cohort in which there was not a distinct gender-related division of labor in young adulthood. Thus, Thus, the thefunctionalist functionalist perspective perspective has has vitality, vitality, up to the present, but other perspectives pose important questions and alternative explanations for thefindings, findings, as reviewed in Section VI.D. Each new generation wants to know the story of its own adult development. At the same time, many people find find comfort and meaning in images that convey a common human wisdom, or a common feminine wisdom. Because many factors, interacting in numerous ways, shape the course of gendered lives, no life plot is inevitable, but cultural patterns recur. The answer to the question of an adult development distinctive to women may be imagined as a complex complex structure of more or less common patterns, forever interpretable from different different vantage points. The

12 CHAPTER 12

DISTINCTIVE TO TO WOMEN WOMEN DISTINCI1VE

309 309

factors that shape attempt to trace out this structure will teach us much about about the factors prepare us for for finding out whether whether young people under under modern modern our lives and prepare direction. conditions of life are moving it in a new direction. Vni.l. APPENDIX Vll ApPENDIX

A. The Terman Terman Study Study Terman Study began, participants (672 girls and and 856 boys) were California California When the Terman school children with a minimum IQ of 135. between 1907 and 135. Most Most were born born between Information was obtained from from parents and teachers as well as from from the 1915. Information children themselves. Questionnaires were sent to participants 10 times over their covered academic, family, and employment reports of of adult years and covered employment histories, reports few personality personality trait ratings. There There were also leisure interests, and at some ages a few of marital satisfaction. Some representative representative publications publications on the Terman Terman measures of adulthood are those of Holahan and Sears (1995); Sears (1978), Terman Terman sample in adulthood Tomlinson-Keasey and Keasey Keasey (1993), and Vaillant and Vail Vailand Oden (1959), Tomlinson-Keasey lant (1990).

B. IHD Studies Studies conducted at the Institute of DevelopThe two main longitudinal studies conducted of Human Human Develop ment (IHD) at the University of included males and females of California, Berkeley, included females Growth Study) and and in the late 1920s born in the early 1920s in Oakland (Oakland Growth represenintended to be represen in Berkeley (Berkeley Guidance Study). The samples were intended of the community, though there was some bias toward toward the middle class. The The tative of participants were studied in childhood or adolescence adolescence and and as adults in 1958-1959, participants 30,41, of the Berkeley Berkeley group were about about 30, 1969-1970, and 1982, when members of 41, and 53, and members of of the Oakland group about 37, 37,48, 1982,233 48, and 61. In 1982, 233 men and from the combined studies were interviewed. Studies of development women from of adult development in the IHD samples depend particularly on demographics, interviews, and measures from the California California O-Sort Q-Sort ((J. of 100 statements about derived from J. Block, 1978), a set of personality that the rater uses to describe an individual by assigning them to one personality of nine categories according to their salience in his or her personality. Four Four books of or her Clausen (1993), Eichom, Eichorn, Clausen, on the IHD samples are by J. Block (1971), Clausen Elder (1974). Haan, Honzik, and Mussen (1981), and Elder

C. The Mills Study The Mills College Longitudinal Study began in the late 1950s, recruiting 140 women from the senior classes of of 1958 and 1960. followed up by mail at from 1960. Participants were followed 1981, 1989). At At all times of of testing the study ages 26-27, 43, and 52 (1963-1964, 1981, included personality inventories, demographic data, and ratings and open-ended

310 310

nELSON, HELSON) PALS, PALS, AND AND SOWMON SOLOMON

Inventory was questions about about various areas of of life. The T h e California California Psychological Psychological Inventory administered Check List List was was obtained obtained from administered at each each testing, and and the t h e Adjective Adjective Check from both both the women and their partners at two ages. The California Q-Sort is available at t h e w o m e n and their at T h e California Q-Sort age age 443. 3 . Representative Representative articles are those by by Helson, Helson, Mitchell, and and Moane M o a n e (1984), and John Mitchell and a n d Helson Helson (1990), Wink Wink and and Helson Helson (1993), and and York Y o r k and J o h n (1992). Mitchell D. D . The The Illinois Valedictorian Study This study began with 46 w women who study began o m e n and and 35 men men w h o were were ttop o p achievers in Illinois high schools in 1981. 1981. Participants Participants were interviewed interviewed five five tto o six times times over over their their high and in the 10th year year of of the study, 1990-1991. Questionnaire Questionnaire data, data, college years and including a measure m e a s u r e of of intellectual intellectual self-esteem, self-esteem, were w e r e collected four occasions, and and including collected on on four m e a s u r e s of of gender gender roles, achievement achievement motivation, motivation, and and work were obtained obtained in measures work plans plans were T h e study study was begun begun by by Terry Terry Denny D e n n y and and is described described by by A r n o l d (1993, (1993,1996). 1985. The Arnold 1996).

ACKNOWLEDGMENT of this chapter chapter was supported supported by a grant from from the National Institute Institute of of Health Health The writing of (MH 43948).

REFERENCES Archer, S. L. (1985). Career Youth & Career and/or and/or family: The identity identity process for for adolescent adolescent girls. Youth & Society, Society, 16, 16, 289-314. Arnold, K. D. (1993). Academically talented talented women women in the 1980s: The The Illinois valedictorian project. In K. D. Hulbert Hulbert & D. T. Schuster (Eds.), Women's Women*s lives lives through through time: time: Educated Educated American American women women of of the twentieth twentieth century century (pp. 393-414). San Francisco: Jossey-Bass. school valedictorians. Arnold, K. D. (1996). Lives Lives of of promise: promise: What What becomes becomes of of high high school valedictorians, San Francisco: Jossey-Bass. Beacon Press. Bakan, D. (1966). The The duality duality of of human human existence. Boston: Beacon Baltes, P. B. (1987). Theoretical propositions of of life-span life-span developmental developmental psychology: On On the dynamics between growth and decline. Developmental Developmental Psychology, Psychology, 23, 611-626. Barnett, R. C c.,, & Baruch, G. K. (1978). Women Bamett, Women in the middle years: A A critique critique of of research research and theory. Psychology Psychology of of Women Women Quarterly, Quarterly, 3, 187-198. Gomick & B. K. Moran (Eds.), Bart, P. (1971). Depression in middle-aged women. In V. Gornick Women in sexist sexist society society (pp. 163-186). New New York: Basic Books. Women second sex. New York: Knopf. Beauvoir, Simone de. (1953) The second and self-conceptions: Personality Bengston, V. L., Reedy, M. N., & Gordon, C. (1985). Aging and processes and social contexts. In J. E. £ . Birren & K. W. Schaie (Eds.), Handbook Handbook of of psychology of of aging aging (pp. 544-593). New York: Van Van Nostrand-Reinhold. Nostrand-Reinhold. the psychology Bernard, J. (1972). The future future of of marriage. New New York: World-Time. Bernard. Bianchi, S. M M., & Spain, American women women in transition. New York: Russell transition. New Spain. D. (1986). American Bianchi. Sage Foundation. Foundation. Block, J. (1971). Lives Lives through through time. Berkeley. Berkeley, CA: Bancroft. Bancroft. Block. .•

CHAPTER 12

DISTINCTIVE TO WOMEN DISTJNCIlVE

311 311

Block, J. J. (1978). The The Q-sort Q-sortmethod methodin inpersonality personalityassessment assessment andpsychiatric psychiatric research, and research. PaloPalo Alto, CA: CA: Consulting Psychologists Press. Contemporary Psychology, Block, J. H. (1976). Debatable conclusions about sex differences. Contemporary Psychology, 22, 517-522. Biihler, C. C. (1971). Basic theoretical concepts of humanistic psychology. American AmericanPsycholo PsycholoBUhler, gist, gist,26, 26,378-386. 378-386. Buss, D. M. (1989). Sex differences in human mate preferences: Evolutionary hypothesis tested in 37 cultures. Behavioral 12,12, 1 -49. Behavioraland andBrain BrainSciences, Sciences, 1-49. economic, andand political Chafe, W. H. (1972). The American American woman: woman: Her Herchanging changingsocial, social, economic, political roles, roles, 1920-1970. 1920-1970.New NewYork: York:Oxford OxfordUniversity UniversityPress. Press. California Chodorow, N. (1978). The reproductionof ofmothering. mothering.Berkeley: Berkeley:University Universityofof California The reproduction Press. Clausen, J. A. (1993). American ofof thethe great depression. NewNew Clausen, Americanlives: lives:Looking Lookingback backatatchildren children great depression. York: York: Free Press. Datan, N. (1986). Corpses, lepers, and menstruating women: Tradition, transition, and the sociology of knowledge. Sex Sex Roles, Roles, 14, 14,693-703. 693-703. & Agronick, G. S. (1995). The intersection of life stage and social events: Duncan, L. E., & Journal of of Personality Personality and Social Personality and life outcomes. Journal Social Psychology Psychology 69, 69, 558-568. pohtics of comparing women and men. American American Eagley, A. H. (1995). The science and politics Psychologist, Psychologist,50, 50,145-158. 145-158. Eichorn, Eichom, D. H., Clausen, J. A., A., Haan, Haan, N., N., Honzik, Honzik, M. M. P., P., & Mussen, Mussen, P. P. (Eds.). (Eds.). (1981). (1981). Present Present and and past in middle middle life. life. San SanDiego, Diego,CA: CA:Academic Academic Press. Press. Elder, G. H., depression. Chicago: University of Chicago Press. H., Jr. (1974). (1974). Children Childrenof ofthe thegreat great depression. Chicago: University of Chicago Press. Erikson, E. E. (1963). Childhood Childhood and andsociety society(2nd (2nded.). ed.).New NewYork: York:Norton. Norton. New York: Norton. Erikson, E. (1968). Identity: Identity:Youth Youthand andcrisis. crisis. New York: Norton. creating lives. Franz, C. E., & & Stewart, A. J. (Eds.). (1994). Women Women creating lives.Boulder, Boulder,CO: CO:West Westview Press. & White, K. M. (1985). Individuation and attachment in personality developFranz, C. E., & develop Journalof ofPersonality, Personality,53,53,136,... 136-168. ment: Extending Erikson's theory. Journal ment: 168. Gergen, K. J. (1985). The social constructionist movement in modern modem psychology. American American Psychologist, Psychologist,40, 40,266-275. 266-275. Gergen, M. M. M. M. (1990). Finished at 40: Women's development within the patriarchy. Psychol Psychol471-493. ogy of of Women Women Quarterly, Quarterly,14,14, 471-493. Giele, J. Z. (1993). Women's role change and adaptation, 1920-1990. In K. D. Hulbert & D. T. Schuster (Eds.), Women Women*s livesthrough throughtime time(pp. (pp.32-60). 32-60).San SanFrancisco: Francisco: 's lives D. Jossey-Bass. theory and women's development. Gilligan, C. C. (1982). In a different different voice: voice:Psychological Psychological theory and women's development. Cambridge, MA: MA: Harvard University Press. Manualfor for the the Adjective AdjectiveCheck CheckList. List.Palo PaloAlto, Alto, Gough, H. G., & Heilbrun, A. B., Jr. (1983). Manual CA: Consulting Psychologists Press. CA: Gutmann, D. L. (1987). Reclaimed Reclaimedpowers: powers:Toward Towardaanew newpsychology psychologyof ofmen men and andwomen women in later later life. life.New NewYork: York:Basic BasicBooks. Books. in R., & & Hartka, E. (1986). As time goes by: Change and stability in Haan, N., Millsap, R, Psychologyand andAging, Aging,1,1,220-232. 220-232. personality over fifty years. Psychology R. T., & & Marecek, J. (1990). Gender and the meaning of of difference: PostmodHare-Mustin, R Postmod ernism and and psychology. In In R R. T. T. Hare-Mustin Hare-Mustin & & J. Marecek (Eds. (Eds.), Makingaadifference: difference: ernism ), Making

312 312

BELSON, HELSON, PALS, PALS, AND AND SoLOMON SOLOMON

Psychology and andthe theconstruction constructionof ofgender gender(pp. (pp.22-64). 22-64).New NewHaven, Haven,Cf: CT:Yale YaleUniver UniverPsychology sity Press. Havighurst, R. R. J. (1948). Developmental tasks and and education. education. New NewYork: York: David David McKay. McKay. Heilbrun, C. C. G. G. (1988). (1988). Writing Writing aa woman's woman's life. life.New NewYork: York: Ballantine. Ballantine. Heise, D. & D. R. (1990). Careers, career trajectories, and the self. In J. Rodin, C. C. Schooler, & Self-directedness:Cause Causeand and effects throughout course effects throughout thethe lifelife course (pp.(pp. K. W. Schaie (Eds.), (Eds.), Self-directedness: 59-84). Hillsdale, NJ: Erlbaum. Helson, R. (1993a). Comparing samples:Towards Towardsaaparadigm paradigm of oftension tension between between Comparing longitudinal longitudinal samples: stability and change. In D. D. C. C. Funder, R. D. D. Parke, C. C. Tomlinson-Keasey, & «& K. K. 19). Washington, Widaman (Eds.), Studying Studying lives livesthrough throughtime time(pp. (pp. 93-1 93-119). Washington,DC: DC:American American Psychological Association. Helson, R. (1993b). The fifties, young The Mills Mills classes classes of of 1958 1958 and and 1960: 1960: College College in the fifties, young adulthood adulthood in the sixties. In K. D. Hulbert & D. Women 's lives through time K. D. D. T. T. Schuster (Eds.), Women's (pp. (pp. 190-210). San San Francisco: Jossey-Bass. V., & & Moane, G. G. (1984). Personality and patterns of adherence and Helson, R., R., Mitchell, V., non-adherence to the social clock. Journal of of Personality Personality and andSocial SocialPsychology, Psychology,46, 46, 1079-1096. Helson, R., R., & & Picano, Picano, J.J. (1990). (1990).Is Isthe the traditional traditional role role bad bad for forwomen? women? Journal Journalof of Personality Personality and Social Psychology, Psychology, 59, 59,311-320. 311-320. Helson, R., & & Stewart, A. A. J. (1994). (1994). Personality change in adulthood. In T. Heatherton & J. Weinberger (Eds.), Can Canpersonality personalitychange? change?(pp. (pp.201-225). 201-225).Washington, Washington,DC: DC:American American Psychological Association. J., & & Ostrove, J. (1995). Identity in three cohorts of midlife women. Helson, R., Stewart, A. J., Journal of of Personality Personalityand andSocial SocialPsychology Psychology69, 69,544-557. 544-557. C. K. K. (1984a). Marital attitudes over 40 years: A longitudinal longitudinal and cohort analysis. Holahan, C. Journal of of Gerontology, Gerontology,39, 39,49-57. 49-57. Holahan, C. K. K. (1984b). The The relationship relationship between between life life goals goals at at thirty thirty and and perceptions perceptions of of goal goal attainment and life satisfaction satisfaction at seventy for gifted gifted men and women. International International Journal of Development, 20, 20, 21-31. of Aging and Human Development, Holahan, C. C. K. K. (1994). Women's goal orientations across the life cycle: Findings from from the Terman study of the gifted. In B. growing B. F. F. Turner & & L. L. E. Troll (Eds.), Women growing older (pp. 35-67). Newbury Park, CA.: CA.: Sage. Holahan, C. maturity. Stanford, C. K., K., & & Sears, R. R. (1995). The The gifted group in later later maturity, Stanford, CA: Stanford Stanford University Press. Jordan, J. V., G., Miller, J. B., P., & Surrey, J. L. V., Kaplan, A. A. G., B., Stiver, I.I. P., L. (1991). Women's growth in connection. connection. New NewYork: York:Guilford Guilford Press. Press. to to identity development in women. SanSan Fran Josselson, R. R. (1987). (1987). Finding Finding herself: herself:Pathways Pathways identity development in women. Francisco: Jossey-Bass. G. (1960). The The stages stages of of life. life. In In H. H. Reed, Reed, M. M. Fordham, Fordham, & & G. G. Adler (Eds.), (Eds.), Collected Collected Jung, C. G. works (Vol. 8, pp. pp. 387-403). Princeton, NJ: NJ: Princeton University Press. (Original work 1931) published 1931) 's lives: from Kaner, A. A. (1994). Physical attractiveness attractivenessand andwomen women's lives:Findings Findings froma longitudinal a longitudinal study. Unpublished doctoral dissertation, University of California, Berkeley. L. (1973). (1973). Continuities in childhood and adult moral development revisited. In Kohlberg, L. espan developmental P. B. Baltes & K. W. Schaie (Eds.), Lif Lifespan developmentalpsychology psychology (pp. (pp. 179-204). 179-204). New York: Academic Press. Levinson, D., D., with with Darrow, Darrow, C. C. N., N., Klein, Klein, E. E. B., B., Levinson, Levinson, M. M.H., H., & & McKee, McKee, B. B. (1978). (1978). The The seasons seasons of of a man's life. life. New New York: York: Knopf. Knopf.

12 CHAPTER 12

DISTINCTIVETO TOWOMEN WOMEN DIsnNCI1VE

313 313

Lx)tt, B. (1987). Women's Women*s lives: Themes and variations. Belmont, CA: Brooks/Cole. Lott, Maccoby, E. E. (1990). Gender Gender and relationships: A developmental account. American American Psychologist, Psychologist, 45, 513-520. erences. Stanford, CA: Maccoby, E. E., & Jacklin, C. N. (1974). The psychology psychology of of sex diff differences. University Press. Stanford University Marcia, J. E. (1980). Identity in adolescence. In J. Adelson (Ed.), Handbook Handbook of of adolescent adolescent psychology psychology (pp. 159-187). New York: Wiley. Markus, H., & Nurius, P. (1986). Possible selves. American American Psychologist, Psychologist, 41, 954-969. McAdams, D. P. (1993). The stories stories we live by: Personal Personal myths myths and and the making making of of the self self New York: Morrow. Miller, J. B. (1976). Toward Toward a new psychology psychology of of women. Boston: Beacon Press. V., & Helson, R Mitchell, v., R. (1990). Women's prime of life: Is it the fifties? Psychology Psychology of of Women Quarterly, Quarterly, 14, 451-470. Neugarten, B. L. (1977). Personality and aging. In J. E. Birren & K. W. Schaie (Eds.), Handbook of of the psychology psychology of of aging (pp. 626-649). New New York: Van Nostrand NostrandHandbook Reinhold. Parker, R, R., & Aldwin, C. C. M. (in press). Does sex-role identity change in adulthood? aduhhood? Differenti Differentiating age, cohort, and period effects. In M. E. Lachman & J. B. James (Eds.), Multiple Multiple paths paths of of midlife midlife development. development. Chicago: University of of Chicago Press. Parsons, T., & Bales, R R. F. (1955). Family, socialization socialization and and interaction process. process. Glencoe, IL: Free Press. Roberts, B. W. (1994). The reciprocal reciprocal relation between between personality personality and and work work experiences from from college to midlife. Unpublished doctoral dissertation, University of of California, Berkeley. Sears, R R. R R. (1978). Sources of life satisfaction of of the Terman gifted men. American American Psycholo Psychologist, 32, 119-128. Skolnick, A. (1991). Embattled Embattled paradise: paradise: The American American family family in an age of of uncertainty. New York: Basic Books. Stevens-Long, J. (1990). Adult development: Theories past and future. In R R. A. Nemiroff Nemiroff & C. A. Colarusso (Eds.), New New dimensions dimensions in adult adult development development (pp. 125-165). New York: Basic Sooks. Books. Stewart, A. J. (1994). Toward a feminist strategy for studying women's lives. In C. Franz & Women creating lives: Identities, resilience, resistance (pp. 111-35). 1 -35). A. J. Stewart (Eds.), Women Boulder, CO: Westview Press. Stewart, A. J., & Healy, J. M., Jr. (1989). Linking individual development and social changes. American American Psychologist, Psychologist, 44, 30-42. Stroud, J. G. (1981). Women's careers: Work, family, and personality. In D. H. Eichorn, Present and and past past in middle middle J. A. Clausen, N. Haan, M. P. Honzik, & P. Mussen (Eds.), Present life (pp. 353-389). New York: Academic Press. Genetic studies studies of of genius: Vol. 5. The gifted gifted group group at Terman, L. M., & Oden, M. H. (1959). Genetic Stanford University Press. mid-life. Stanford, CA: Stanford Todd, J. Friedman, A. & Kariuki P. W. (1990). Women growing stronger with age: The effect effect of of status in the United States and Kenya. Psychology Psychology of of Women Women Quarterly, Quarterly, 14, 567-578. TomUnson-Keasey, C. (1990). The working lives of of Terman's gifted women. In H. Y Gross GrossTomlinson-Keasey, experience and and meaning meaning of of work work in women's man & N. L. Chester (Eds.), The experience women's lives (pp. 213-239). Hillsdale, NJ: Erlbaum.

314 314

BELSON, HELSON, PALS, PALS,AND ANDSOLOMON SOLOMON

Tomlinson-Keasey, C, C., & Blurton, E. U. (1992). Gifted lives: Aspirations, achieve Gifted women's lives: achievements, and personal adjustment. In J. S. S. Carlson (Ed.), Advances in cognition cognition and and educational annual (Vol. 1, pp. 151-179). Greenwich, CT:CT: JAIJAI Press. research annual (Vol. 1, pp. 151-179). Greenwich, Press. educationalpractice: practice:AAresearch Tomlinson-Keasey, c., B. (1993). C, & Keasey, C. B. (1993). Graduating from from college in the 1930s: 1930s: The Terman genetic studies of genius. In K. D. D. Hulbert & D. T. Schuster (Eds.), Women's Women*s lives lives through through time (pp. 63-92). San Francisco: Jossey-Bass. Tooby, J., & Cosmides, L. (1990). On the universality of of human nature: The role of of genetics and adaptation. Journal of of Personality, Personality,58, 58,17-67. 17-67. Vaillant, G. E. (1977). Adaptation to life. life. Boston: Little-Brown. Vaillant, G. E., & Vaillant, C. O. (1990). (1990). Determinants and consequences consequences of creativity in a Quarterly, 14, cohort of gifted gifted women. Psychology Psychology of of Women Quarterly, 14,607-616. 607-616. progress (2nd ed.). New York: Holt, Rinehart, & Winston. White, R. W. (1966). Lives in progress Wink, P., P., & Helson, R. (1993). (1993). Personality change in women and their partners. Journal of of Personality and andSocial SocialPsychology, Psychology,65, 65,597-605. 597-605. Personality (1993). Working toward an inclusive psychology of women. Yoder, J. D., & Kahn, A. S. (1993). American Psychologist, Psychologist,48, 48,846-850. 846-850. (1992). The four four faces of Eve: A typological analysis of women's York, K. L., & John, O. P. (1992). Personalityand andSocial SocialPsychology, Psychology,63, 63,494-508. 494-508. of Personality personality at midlife. Journal of

PART IV PART IV

BIOLOGICAL DETERMINANTS OF PERSONALITY

This Page Intentionally Left Blank

CHAPTER 13 CHAPTER 13

EVOLUTIONARY FOUNDATIONS EVOLUTIONARY OF PERSONALITY DAVID M M.. Buss BUSS UNIVERSITY UNIVERSITY OF OF TEXAS, TEXAS, AUSTIN AUSTIN

In In the mid-1990s, after several hundred years yearsof of scientific scientificscrutiny, scrutiny,only onlyone onescientific scientific theory remains viable to account for all major characteristics of organic life, past present—the theory of evolution evolution by by natural selection (Darwin, (Darwin, 1859). 1859). Only Only two two and present-the "theories" have have received received even evenmarginal marginal scientific scientific attention-creationism attention—creationism competing "theories" (all forms of life created by a supreme being) and seeding theory (life planted on earth by extraterrestrial beings). The problems with creationism and seeding theory are (1) that multiple seedings or creations over time and space are required to bankaccount for the paleontological data, data, and (2) these theories are heuristically bank rupt—they lead to no important domains of inquiry and to no testable predictions rupt-they (Daly & Wilson, (Daly & Wilson, 1988). 1988). To remain as the sole scientific theory proposed to account for the major characteristics of all living forms surely qualifies evolutionary theory as a "great" Indeed, itit serves servesas as the themajor majormetatheory metatheoryin in the the biological biological theory (Alexander, 1979). Indeed, sciences. Although many details of the theory are vigorously debated (e.g., unit of selection, conceptualization conceptualization of of adaptation), adaptation), the the essentials essentials of of evolutionary evolutionary theory theory selection, are now now paradigmatic paradigmatic and and largely largely undisputed. undisputed. are It seems reasonable, therefore, to adopt an initial premise that millions of years of evolution might have left some mark on human psychology. Humans are part of the organic world, and there is no reason to assume that we are somehow exempt from from the the laws laws of of organic organiclife lifethat thatgovern governall all other otherliving livingforms. forms.Precisely Preciselywhat what exempt the consequences consequences of of this this premise premise are are for for psychology, psychology, however, however, remains remains stridently stridently the debated by aa few few and and largely largely ignored ignored by by many. debated by many. lIANoBOOK HANDBOOK OF OF PERSONALITY PERSONALITY PSYCHOLOGY PSYCHOLOGY

317 317

COPYRIGHT © 11997 CoPYII GHT C 997 BY ACADEMIC PRESS. ALL lIGHTS OF IEPRODUcnON IN ANY FOlM RESElVED.

ALL RIGHTS OF REPRODUOION m ANY FORM RESERVED.

318 318

DAVID DAVID M. M.Buss Buss

Part of of the problem problem is that it has not been immediately obvious what the implications of of adopting this premise are. Consider a noncontroversial noncontroversial example example— food the larynx (Symons, 1987). The larynx rises when we swallow, thus preventing food from getting into our lungs. One can grant that the larynx exists because it evolved from adaptive over time by natural selection, that it solved (and continues to solve) the adaptive problem of of getting food food down the right passage, and that those who had no larynx problem inefficient larynx were more likely to have choked choked to death in our evolutionary evolutionary or an inefficient past, and therefore therefore experienced experienced lower reproductive reproductive success than did those with an efficiently functioning functioning larynx. efficiently In this presumably noncontroversial example, it is not clear that the theory of evolution by natural selection has added materially to, or is even necessary for, for, of understanding of of the larynx. Human intuition has picked out the "prevent "prevent our understanding function, and studies of of the proximate proximate workings of of the larynx can proceed proceed choking" function, perfectly adequately adequately without being informed informed by explicit evolutionary evolutionary considerations. considerations. perfectly of psychology, where basic mechanisms that might have evolved In the domain of difficult to identify of human behavior, it has are difficult identify or even point to among the flux of remained even less clear what role evolutionary theory should or could serve. remained proceed with their research agendas to some some Psychologists, like anatomists, can proceed evolutionary forces that led to the origins of of our psychologi psychologiextent ignoring the evolutionary cal mechanisms. so—that anchoring anchoring I argue in this chapter, however, that it is a mistake to do so-that The personality psychology in evolutionary biology is essential and unavoidable. The problem is precisely how this can be done. The purpose of this chapter is to outline for doing so, and to point out the problems and some of the promising directions for pitfalls that have stymied previous attempts. pitfalls

AND HYPOTHESIS I. EVOLUTION: EVOLUTION: FACT, PATH, THEORY, AND

Much confusion confusion centers around the use of the term "evolution." Ruse (1986) distinguished distinguished among three different different uses-evolution uses—evolution as fact, fact, as path, and as theory. theory. refer to specific evolutionary hypotheses. I will add to these the use of evolution to refer

A. Evolution as Fact Evolution as fact fact refers refers to the natural unfolding unfolding or change of of organisms down Evolution from earlier forms. Evolution in this sense, referring referring to through the generations from known to occur long before before Darwin. The lines of of evidence change over time, was known fact of of evolution include biogeography biogeography (geographical distribution distribution of of supporting the fact organisms), comparative comparative anatomy anatomy (the study of of the relations between between the forms of of organisms, for for example, isomorphism isomorphism between the bones and arms of of humans and forelegs of of horses, the wings of of birds, and the paws of of mole), embryology embryology (the (the the forelegs study of of early forms forms of of organisms and and their their development; the embryos of of humans humans study

CHAPTER CHAPTER 13 13

EVOLUTIONARY EVOLUTIONARY FOUNDATIONS FOUNDATIONS

319 319

and those of dogs, for example, are indistinguishable due to common evolutionary origins), and paleontology (the study of the fossil record). from these multiple sources of of data The only conclusion that can be drawn from is that evolution in the sense of change in organic structure over time, although almost entirely unobservable directly, is a fact. fact.

B. Path Path of of Evolution Evolution B. A second meaning of "evolution" refers to the specific series of changes that led form—the evolutionary phylogeny, trajectory, or descent of a given to a current form-the species. The fossil record tells us much about the path of evolution. We know, for from past to example, that the path of evolution is not one uninterrupted drive from present, that it involves mUltiple multiple branchings rather than a continuous ladder, and that evolutionary dead ends are far more common than successes—most successes-most species that have evolved, 90% by some estimates, are extinct. We know from this record that life on earth emerged roughly 3,500 million years ago. It evolved beyond simple cells 500 million years ago. Mammals, of which we are one, emerged 200 million years ago. Dinosaurs reached their demise 60 million years ago. Primates made their first appearance on earth 55 million years ago, roughly 5 million years after after the dinosaur extinction. Human beings came on ago, the scene roughly 2-4 million years ago. And Neanderthal became extinct about 35,000 years ago, strangely coincidental with the emergence of ero Cro Magnan, our current form. Although many details of the path of human evolution remain un unknown, we do have the rough outlines of the specific evolutionary trajectory that can trajectory be traced from our mammalian and primate ancestors to our current modern forms. It is from the path of evolution, and in particular an examination of our own primate line, that a previous generation of psychologists attempted to glean insight into human functioning. There is some value to gauging trends in evolution through comparisons with other mammals and other primates (cf. A. Buss, this volume, chap. 14). 14). It is now generally recognized, however, that this form of "referential "referential modeling" (selecting this or that primate species for comparative analysis) is highly problematic, can be seriously misleading, and lacks the generativity typically associ associated with a good theory (Tooby & DeVore, 1987).

C. Theory Theory of of Evolution Evolution c. The theory of evolution, Darwin's contribution to the understanding of change in selection,the themechanism mechanismororprocess processby bywhich whichchange change naturalselection, organic form, refers to natural takes place. If variations in the organic world leading to greater survival and reproplace. repro frequently enough over generations, the cumulative effect effect will be evoduction recur frequently evo lutionary. Darwin did not simply want to account for for the evolution of organisms, he wanted to explain why they are the way they are. This implies a central role of of adaptation in his theory. We have fingers, hands, feet, eyes, keeness of sight, and adaptation

320 320

DAVID M.Buss BUSS DAVID M.

fleetness fleetness of of foot foot because those who had these mechanisms were better able to survive and reproduce than those lacking these mechanisms, or whose mechanisms mechanisms were different. different. Although natural selection selection is referred referred to as the theory theory of of evolution, it has received received overwhelming overwhelming empirical support. It has been observed observed in laboratory laboratory experi experiments and in the wild in thousands of of instances. Indeed, new species have been been "artificially" through through applying the principles of of selection in the laboratory. laboratory. evolved "artificially" In its essential forms, no one doubts that evolution occurs principally by natural evolved selection, and that human beings, as part of natural organic life, have also evolved What by natural selection. In this sense, we are all Darwinians (Symons, 1987). What doubt is the veracity veracity of of specific evolutionary evolutionary hypotheses hypotheses advanced advanced about is in doubt human behavior. human Specific Evolutionary Evolutionary Hypotheses Hypotlieses D. Specific refer to propositions about the function function of of a partic particSpecific evolutionary hypotheses refer feature of human behavior. One hypothesis, for for ular mechanism, structure, or feature offexample, is that male sexual jealousy evolved to ensure paternity in putative off spring by preventing encroachment by competing males. Males throughout our evolutionary past whose behavior behavior reflected reflected a mechanism of of sexual jealousy experi experievolutionary indifferent to the the enced greater reproductive success than did males who were indifferent considerable sexual contact that their mates had with other males. In this instance, considerable support this hypothesis (Buss, Larsen, Westen, & Semmel Semmelevidence has accrued to support roth, 1992; Daly, Wilson, & Weghorst, 1982). Another hypothesis is that that males have evolved a mechanism to detect when Another females females are ovulating. The ability to detect when women ovulate would enable a man to focus mating efforts efforts (generally (generally a costly enterprise) enterprise) more effectively, effectively, not bothering bothering to attempt attempt copulations with those females for for whom conception at that that moment moment is unlikely. In this case, the empirical data do not support support the hypothesis hypothesis— there is no evidence that human males can detect when females females are ovulating ovulating (Symons, 1987). Several crucial points can be made about specific evolutionary hypotheses. First, it is important important to evaluate each one separately. There is no such thing as "the" evolutionary " t h e " evolutionary hypothesis hypothesis about a given phenomenon. Indeed, for for any given feature or attribute, there are likely to be several different competing feature different competing evolutionary evolutionary hypotheses (cf. Buss, 1990a, 199Oa, 1990b; Rancour-Laferiere, Rancour-Laferiere, 1985). Second, evolutionary evolutionary hypotheses should be subjected subjected to the same empirical scrutiny as other other scientific scientific hypotheses. They have no special or elevated status by virtue of of their being evolutionary. evolutionary. Third, specific specific evolutionary evolutionary hypotheses differ differ in precision and testability. testability. Gould Gould and Lewontin Lewontin (1979), for for example, have marshaled the argument that many specific evolutionary evolutionary hypotheses hypotheses are "just so" stories, fictional fictional speculations that are are specific imprecise and incapable incapable of of being subjected subjected to the rigors of of empirical empirical scrutiny. There There imprecise indeed is some merit in their accusation. Many specific evolutionary hypotheses have indeed

CHAPTER 13

EVOLV110NARY EVOLUTIONARY FOUNDATIONS

321 321

vacuous. Perusing any recent copy of a been sloppy, imprecise, and empirically vacuous. journal in personality and social psychology, however, suggests that those with an evolutionary evolutionary perspective do not have a monopoly on sloppy, imprecise, and ungrounded hypotheses. Indeed, in any scientific discipline, hypotheses vary in ungrounded their precision, theoretical anchoring, and testability. testabihty. The key point is that useful useful evolutionary hypotheses are those that are specified sufficiently sufficiently precisely so that they can receive the same scientific scrutiny and be subjected to rigorous evidentiary GauUn, Agyei, & Gladue, in press; Buss et al, standards (see, e.g., Bailey, Gaulin, al., 1992; Cosmides, 1989; 1988; Gangestad 1989; Daly & Wilson, 1988; Gangestad and Simpson, 1990; 1990; Graziano & Eisenberg, this volume, volume. Chapter Chapter 30; 30; Kenrick, Groth, Trost, & Sadalla, 1993; Singh, 1993). II. BASICS OF EVOLUTIONARY PSYCHOLOGY n.

This section outlines the basics of current evolutionary psychology (Buss, 1991, 1991, in in press; Cosmides & Tooby, 1987, 1987, 1989; 1989; Daly & Wilson, 1988; 1992; 1988; Symons, 1987, 1987,1992; 1990b, 1992). This This provides provides aa foundation foundation for for the the analysis analysis Tooby & Cosmides, 1990a, 199Ob, of issues of central concern to the field of personality psychology.

A. Humans Humans as as Evolved Evolved Problem Problem Solvers Solvers

Humans, like other organisms, can be viewed as organized structures that exist in their present form because of a long history of natural selection, operating over miUions of years. Each one of us owes our existence to a long and unbroken line millions successfully solved problems posed by survival and reproduction in of ancestors who successfully our evolutionary past. Therefore, human structures, as well as human psychological description,can canbe beanalyzed analyzedininterms termsofof of description, mechanisms, at some fundamental level of the problems they solve. But "survival" and "reproduction" are broad categories, each subsuming a large and complex array of subproblems. To the extent that the evolutionary evolutionary psychologist can identify identify the nature of the specific specific problems that humans have psycholoevolved to solve, she or he has some advantage over the nonevolutionary psycholo gist in discovering the nature of human nature. identified many of the major major survival problems, subsumed by what Darwin identified food shortages, climate, he called "the hostile forces of nature." These include food surweather, diseases, parasites, predators, and natural hazards such as cliffs. But sur from an evolutionary perspective, is only a start. Natural selection operates vival, from on survival survival only only insofar insofar as as survival survival is is necessary necessary for for reproduction. reproduction. Thus, Thus, aa second second on major class class of of problems problems consists consists of of reproductive reproductive problems. problems. major At the broadest level, reproduction problems that must be solved are: intrasexual competition, competition, beating out members of your own sex to (1) successful intrasexual selection,choosing choosingfrom fromamong among obtain desirable mates of the opposite sex; (2) mate selection, the pool of potential mates those with the greatest value to one's reproductive

322 322

DAVID M.Buss Buss DAVID M.

success; (3) conception, conception, engaging in the necessary necessary social and sexual behaviors to fertilize, or to become fertilized; (4) mate mate retention, retention, preventing the encroachment encroachment of of intrasexual poachers as well as preventing one's mate from from defecting defecting (this problem problem must often often be solved in order order to actualize the promise of of reproductive reproductive effort); effort); (5) parental parental investment, investment, engaging in the necessary set set of of behaviors to ensure the survival and reproductive success of and (6) extraparental extraparental nepotistic of one's offspring; offspring; and nepotistic investment, incurring costs to self that benefit benefit nondescendant nondescendant genetic relatives. investment, Each of of these six problems, of of course, is extremely extremely broad broad and subsumes a host of subproblems. Successful Successful intrasexual competition, for for example, might require acquisition of of resources required required by a potential mate, appeasing, or not alienating, relatives and friends friends of of the potential mate, or (c) successfully successfully courting courting the potential potential mate, while simultaneously rendering intrasexual intrasexual competitors less attractive to the potential mate (e.g., derogating competitors). A A psychologist psychologist able to understand understand face in our environment environment of of evolutionary evolutionary adaptedness adaptedness these problems that humans face identify and explore the psychological (cognitive, emo emo(EEA) is in a position to identify evolved to solve them. tional, behavioral) solutions that have evolved of B. Other Humans Are the Most Important Source of Adaptive Problems and Adaptive Solutions of evolution evoke "nature "nature red red in tooth tooth and claw" Although traditional images of and "survival of of the fittest," fittest," sheer physical survival may not not have been the most important adaptive problem governing human evolutionary history. There appears human evolutionary important to be growing consensus that the most important problems were posed by members members of our own species. In other words, we are our own hostile force force of of nature. of from the prevalence of of group Findings that point to this conclusion come from warfare warfare (e.g., Alexander, 1987), the evolution of of language and nonverbal nonverbal signs for for manipulating others (Dawkins & 1978; Pinker, 1994; & Krebs, 1978; 1994; Pinker Pinker & & Bloom, Bloom, 1990), 1990), the intense sociality of of humans humans (Alexander, 1979), the extreme degree of of reciprocal altruism among humans (Trivers, 1971), the extreme degree of of parental investment investment among humans (Trivers, 1972), the degree of of sexual dimorphism dimorphism among humans, suggesting the importance of of intrasexual competition competition (Short, 1979), the extensive extensive and prolonged & Wilson, 1983), the existence prolonged mating rituals among humans (Daly & existence of of domain-specific domain-specific psychological mechanisms governing social exchange (Cosmides, 1989), and the intensity of of human effort effort devoted devoted to status, reputation, and hierarchy (Hogan, 1983). These suggest that the most important important problems that humans have had to face face in our EEA E E A for for survival and reproduction reproduction have been social problems. of subproblems. Forming Forming a Within each social adaptive problem lies dozens of successful dyadic alliance, for for example, may require require identi identifying posfying key resources pos successful assessing which individuals possess these resources, sessed by potential friends, assessing modeling the values of those individuals, gauging gauging potential sources of of strategic modeling interference, initiating initiating sequential and incremental chains of of reciprocity, and detecting detecting signs of "cheating" or nonreciprocity (see Cosmides & & Tooby, 1989). All of of these subproblems require solutions for for the formation formation of of a successful successful friendship. friendship.

CHAPTER 13 13

EvOWnONARY EVOLUTIONARY FOUNDATIONS FOUNDATIONS

323 323

Humans are probably unique in the duration duration and complexity complexity of of social relation relationships that they form. Humans sometimes form form lifelong lifelong mating relationships, develop friendships friendships that last for for decades, and maintain contact contact with their their brothers, sisters, of time and distance. Because social adaptive adaptive and other relatives over great expanses of problems were so crucial for for human survival and reproduction, many of of the most important important features features of of our evolved psychological psychological mechanisms will necessarily necessarily be social in nature. Social adaptive problems have been been so important important over human evolutionary history that many of of the dedicated dedicated psychological mechanisms currently studied by cognitive, personality, and developmental psychologists, in addition to inherently social. those studied by social psychologists, are inherently

c. C. Psychological Mechanisms as Evolved Strategies for Solving Problems A third essential feature feature of current evolutionary thought is the increasingly apparent apparent A psychological mechanisms as evolved solutions necessity to understand fundamental fundamental psychological to specific evolutionary problems (e.g., Barkow, 1989; Buss, 1989a; Cosmides & for example, Tooby, 1987, 1989; Shepard, 1987; Symons, 1987). Cosmides (1989), for developed a computational computational theory of of the psychological psychological mechanisms that must has developed for humans to solve problems associated associated with social exchange. have evolved in order for exchange must have a psychological mechanism mechanism that Organisms engaging in social exchange of cheaters-those cheaters—those who accept the benefit benefit of of a social exchange permits the detection of without paying the reciprocal reciprocal cost. cost. In In aa series series of of experiments, experiments, Cosmides Cosmides (1989) (1989) has has without paying the demonstrated that humans do indeed have such a mechanism, and that the mechademonstrated that humans do indeed have such a mechanism, and that the mecha nism is is domain-specific domain-specific in in the the sense sense that that it it operates operates particularly on problems problems struc strucnism particularly on tured as social exchanges. tured as social exchanges. another example, has studied studied preferences preferences as evolved Buss (1989a), to take another psychological mechanisms that solved certain problems associated associated with mate choice. psychological differ somewhat somewhat in the nature nature of of the reproductive reproductive Because human males and females differ problems they must solve, psychological mechanisms surrounding surrounding reproduction reproduction are expected to be somewhat sexually dimorphic (Symons, 1987). Sexually dimorphic dimorphic mate preference preference mechanisms for for age, physical appearance, and external resources have been documented documented across 37 cultures located in 33 countries on six continents and five islands (Buss, 1989a). Attempts to elucidate the nature of these psychological psychological mechanisms is just beginning. Undoubtedly, existing research in social, cognitive, personality, clinical, and developmental psychology will be invaluable when examined from an evolution examined from evolutionary perspective. Even given this incipient and somewhat inchoate stage in evolution evolutionary psychology, several important features features of evolved evolved psychological psychological mechanisms seem apparent. apparent.

D. Many Psychological Mechanisms Will Be Domain Specific Consider the human body. Although an integrated integrated whole, the body can to some Consider useful degree be studied as a set of interrelated interrelated parts, each serving specific specific functions functions useful

324 324

DAVID DAVID M. M.Buss Buss

or solving specific problems. We have sweat sweat mechanisms that solve the problem problem of of overheating, shivering mechanisms that solve the problem problem of of extreme extreme temperature temperature drop, a larynx that that solves the problem of of preventing chocking, a liver that solves the problem of filtering impurities, rods that solve the problem of filtering problem of of night vision, and a retina that solves the problem of depth perception. Because the problems that that require solutions are radically different different from from one another, there is no reason to assume that a mechanism that that evolved to solve one problem problem will be effective effective for solving a different different problem. To paraphrase von Neumann, when you are trying to maximize n functions, when n is equal to or greater than 2, you have to be very, very lucky. My sweat glands will not solve the problem problem of of getting my food food down the right passage, nor nor will my larynx solve the problem of of accurate visual perception. perception. Different Different problems require different different solutions. In this sense, a psychologist with an evolutionary evolutionary perspective expects that evolved evolved solutions are likely to be domain domainspecific. By analogy to the human body, the number number of of social problems posed by successful successful reproduction reproduction is also large. The psychological mechanisms needed needed to engage in successful successful social exchange (e.g., detect-cheater detect-cheater algorithm) are likely likely to engage be radically radically different different from from the preference preference mechanisms needed needed to choose a valuable of these mate. There is no reason to assume that psychological solutions to either of directly applicable to the psychological mechanisms that might problems will be directly of parental investment investment (e.g., empathy, nurturance, have evolved to solve problems of needed to retain a mate and fend fend off off poachers poachers (e.g., parental feeling) or those needed sexual jealousy). In this sense, an evolutionary perspective leads psychologists to suspect that many psychological mechanisms will be domain-specific. domain-specific. assumption made in behaviorism behaviorism that This view is at odds with the traditional assumption organisms possess only one or or a few few general general learning processes that operate operate in the organisms manner across content content domains and across species. But it is consonant consonant with same manner of the major major empirical advances that point to the existence of of a large number number of of many of the content-specificity content-specificity of of fears and phobias (Seligman & motives (Herrnstein, 1977), the specificity of of learning (e.g., Rozin & Schull, SchuU, 1988). Hagar, 1972), and the great specificity Indeed, without without domain-specific domain-specific psychological mechanisms, humans could not not Indeed, flexibility in behavior that is observed observed (Cosmides & Tooby, 1989). achieve the great flexibility E. Behavioral Flexibility Derives from Complex and Dedicated Psychological Psychological Mechanisms evolutionary psychological views of of James, McDougall, and even Lorenz The earlier evolutionary Tinbergen postulate instincts (or fixed action patterns) that connote rigidity and Tinbergen and intractability in their manifestation. Similarly, early sociobiological views that genes hold culture (and presumably presumably behavior) on a leash connote automaton-like automaton-like inflexibility of of human behavior. Few things are more obvious, however, than the inflexibility flexibility, discriminativeness, and context-dependency context-dependency of of human action extreme flexibility, (Cantor & Kihlstr6m, KihlstrOm, 1987).

CHAPTER CHAPTER 13 13

EVOLUTIONARY EVOLUTIONARY FOUNDATIONS FOUNDATIONS

325 325

But what psychological mechanisms could produce extreme behavioral flexi flexibility? biUty? Although some implicit thinking in psychology assumes that simple, amor amorflexibility, current evolutionary evolutionary phous psychological structures must underlie flexibility, thought stresses precisely the opposite: Extreme behavioral plasticity implies extreme mental complexity and stability; that is, an elaborate elaborate human nature. nature. Behavioral Behavioral plasticity plasticityfor for its itsown ownsake sakewould wouldbe be worse than useless, random variation suicide. suicide. During the course of evolutionary history the more plastic hominid behavior became the more complex the neural machinery must have become to channel this plasticity into adaptive action. (Symons, 1987, 1987,p.p.127) 127)

A carpenter's flexibility flexibility comes comes not not from from having having a single domain-general domain-general "all A carpenter's purpose tool" that is used to cut, saw, screw, twist, wrench, plane, balance, and purpose rather from from having many, more more specialized specialized tools, each each designed to hammer, but rather perform a particular particular function. function. It is the numerousness and specificity specificity of of the tools in perform that give the carpenter carpenter great flexibility, flexibility, not not a highly "plastic" the entire tool kit that single tool. of Identifying Identifying the Classes of of Acts to F. The Importance of Which Psychological Mechanisms Correspond A somewhat more controversial feature feature of current evolutionary thought is an em emA of acts to phasis not just on psychological mechanisms, but also on the classes of which they correspond, and without which they cannot be understood. Consider mechanism of of "male sexual jealousy." The most well-supported well-supported the psychological mechanism evolutionary hypothesis is that this mechanism evolved in men to increase confi confievolutionary dence in paternity (maternity never being in doubt) (Buss, 1988b; Buss et al., aI., 1992; Daly & Wilson, 1983; Daly et al., aI., 1982). human evolutionary history who were indifferent indifferent to the sexual contact Men in human that their mates had with other men presumably presumably experienced experienced lower reproductive reproductive success than did males who tended tended toward a jealous rage whenever they detected pair-bond defection defection or the presence of of potential intrasexual poach poachcues to potential pair-bond ers. Such a mechanism could not possibly evolve unless it produced produced behavior or action by the male that functioned functioned to prevent pair-bond pair-bond defection, interfere interfere with poaching males, and thus increase the probability of paternity. The psychological poaching probability of mechanism mechanism of of jealousy cannot be understood without identifying identifying this class of of "jeal "jealMechanisms can evolve only if they produce produce behavior that carries certain ous acts." Mechanisms consequences. Symons (1992) argues that "Darwin's theory of of natural selection sheds light affairs only to the extent that it sheds light on phenotypic design, and on human affairs manifested at the psychological rather rather than behavioral level." I design is usually manifested that Symons' dichotomy between between psychological and behavioral web would argue that strongly drawn. It is not just that psychological mechanisms mechanisms such as jealousy is too strongly

326 326

DAVID DAVID M. M.Buss Buss

cannot cannot be understood understood without without identifying identifying the corresponding corresponding class of of jealous acts, the class of of acts is part of of the scientific description of the nature of that mechanism. mechanism. Most single acts are ambiguous with respect to intention, motive, function, function, or design. Furthermore, single acts are invariably manifestations manifestations of of several psycho psychoConsider the following following acts: He called her at unexpected unexpected times logical mechanisms. Consider to see who she was with; He did not take her to the party where other males would break up if she ever cheated on him; He made her be present; He threatened to break feel guilty about talking to other men; He stared coldly at the guy who was looking considered alone is a product product of of several different different psychological psychological at her. Each act considered from competitors, another involves mechanisms. One involves concealing a mate from understanding of of the emotional states of of the partner and how those an empathic understanding manipulated (e.g., guilt induction), and still another involves knowledge states can be manipulated effects of of "cold stares" on an intrasexual competitor. competitor. about the effects What all of of these acts have in common, however, is that they were all presum presumproduced in part part by the operation of of a "sexual jealousy" mechanism. Jealousy Jealousy ably produced is the common common denominator denominator of of this class. The nature and function function of of jealousy is illuminated by considering these acts as a class rather than individually or in isola isolailluminated of acts constitutes constitutes the behavioral strategy strategy component component of of the psycho psychotion. This class of of male sexual jealousy. logical mechanism of understood without without Seen in this light, psychological mechanisms cannot be understood identifying identifying the class of of acts that composes the evolutionary evolutionary raison d'etre d'etre of of the mechanism. The acts are part of of the mechanism-part mechanism—part of of the phenotypic design that must be understood. There can be no science of evolutionary psychology from classes of of acts to which the psychological mechanisms correspond. correspond. divorced from Evolutionary Evolutionary thought in psychology has evolved dramatically dramatically over over the past century. The most compelling shifts shifts involve a change from from a view of of humans as possessing a rigid set of of instincts or fixed action patterns that are genetically deter determined and that invariably, intractably intractably manifest manifest themselves in behavior to a new view: humans are strategic problem solvers, whose evolved psychological psychological mechanisms mechanisms are at least somewhat somewhat domain-specific domain-specific and whose behavioral products products are highly dependent on eliciting context. dependent G. Specific Evolutionary Models Modern evolutionary psychology clearly recognizes that evolutionary evolutionary theory is a theory about the origins origins of of human human nature, but not not a theory theory of of that nature itself itself "That Darwin's theory theory of of adaptation adaptation (Cosmides & Tooby, 1989). As Symons says, "That constitutes the only scientifically scientifically tenable account of of the origin and maintenance of of organic design does not imply that this theory constitutes a Royal Road to the detection of of design" (Symons, 1989, p. 136). Thus, a clear limitation is a predictive predictive detection Evolutionary theory theory provides some broad guidelines about what is un unlimitation. Evolutionary likely to have evolved (e.g., adaptations that favor favor other species), but within those broad constraints, it cannot specify specify what must have or will have evolved. There for developing specific specific evolutionary evolutionary models models of of narrower sets of of is no substitute for

CHAPTER 13 13 CHAPTER

EVOLUTIONARY FOUNDATIONS EVOLUTIONARY

327 327

evolutionphenomena (e.g., a paternity confidence hypothesis of sexual jealousy, an evolution based theory of social exchange) and testing predictions based on these more circumscribed models. In this sense, current evolutionary psychology is consonant with existing practice within psychology, the elucidation psycho elucidation of of domain-specific psychological phenomena. The difficulties functionally signifi difficulties of cleaving psychological phenomena phenomena into functionally significant units beset all of psychology. A focus on function function simply makes explicit a treacherously difficult difficult task that has been largely bypassed in 20th century psychol psychology-a arbitrary description of basic psychological mechanisms. ogy—a non nonarbitrary III. III. FOUNDATIONAL IMPLICATIONS FOR PERSONALITY PSYCHOLOGY formulating the evolu evoluThis section draws out the most important implications for formulating specification of the core of tionary foundations of personality. These include a specification of nature—the major major directional tendencies toward which humans are moti motihuman nature-the vated—and identifying identifying important individual differences, differences, clarifying clarifying the concepts of vated-and of reframing the debate over personality consistency adaptation and adjustment, and reframing specificity. and behavioral specificity. A. Directional Directional Tendencies Tendencies of of Human Human Action: Action: Motives, Motives, Goals, A. Goals, and Desires Desires and Identifying the major major motives of humans has occupied personality psychologists Identifying force behind hu husince Freud postulated the "sex instinct" as the central energizing force major theories of personality contain assumptions, implicit or man action. All major selfexplicit, about what these major directional tendencies are. Maslow posited self actualization, Adler the striving for superiority, and Sullivan the negotiation of of interpersonal relations. A conception of major motives, desires, or directional tendencies must form the core of any major theory of personality. They specify what energizes the organ organism, what causes us to do something rather than nothing, and toward what goals human behavior is directed. Although individuals clearly differ differ in the particulars of the major major motives, all personality theories specify a core of species-typical motives around which individuals differ. differ. Modern evolutionary theory specifies specifies three broad broad classes of such directional tendencies: survival, tendencies: survival, reproduction, reproduction, and genetic genetic investment. investment. Each of these broad classes, in turn, may be partitioned into several important subclasses. Survival problems, for example, can be partitioned into striving to combat the "hostile forces of nature." Reproductive problems include selecting, attracting, and retaining a mate. Investment problems include deciding how to socialize children and which genetic relatives to channel resources toward. It is not by chance that preindustrial human groups devote a large portion of of their time to hunting and gathering; that individuals who sneeze or display signs

328 328

DAVID M.Buss Buss DAVID M.

of of ringworm are actively shunned; that we shiver, sweat, cover up, or strip down in response to extremes of of weather; that infants refuse to crawl over a "visual cliff"; and that snakes, spiders, darkness, and strangers are vigorously avoided. Individuals who failed to be motivated in these directions tended to become no one's ancestors. In sum, an evolutionary evolutionary psychologist expects that human action will be di diassociated with survival, growth, and rected, in part, toward solving the problems associated maintaining organismic integrity (Buss, 1986). But what is the conceptual status of of sociobiolthese directional tendencies? Contrary to accounts implicit and explicit in sociobiol ogy, there is no reason to expect that humans humans will will have have domain-general domain-general motives motives at level of of abstraction abstraction of of fitness fitness maximization. maximization. the level Evolutionary theory is a theory about the origins of of human nature, but is not itself conftated these, itself a theory of of that human nature. Many sociobiologists have conflated propose that humans have as and much conceptual confusion has resulted. To propose motives "survival" or "reproduction" or "fitness maximization" would be like A chess programming a computer chess strategy with the single instruction "win." A program with such a global instruction would have no possible means of of winning-it winning—it would not have the slightest clue about what moves to make in response to this or of that array of pieces. Computer chess programs instead contain many thousands of if-then statements. It is only through this detailed highly detailed, context-specific, if-then design that the program can solve the many problems that are correlated with, and sometimes necessary for, winning. Similarly, natural selection could not produce an organism with global motives like "survive," simply because the organism would have no way of of knowing how to behave when confronted with extreme cold, putrid meat, a thousand-foot thousand-foot cliff, or a poisonous snake. Instead, millions of of years of of natural selection have produced highly detailed, context-specific context-specific adaptations that solve specific specific problems that are associated with survival. Although maintaining and enhancing organismic integrity, at an abstract level, describes one central directional tendency of human nature, the goals, motives, preferences, fears, and desires that constitute part of of human nature are considerably more specific. It is in the details of of the specific survival problems that we can identify identify the end states toward which human action is directed. Human nature, therefore, in part consists of of the specific specific survival-related directional tendencies (problems that must be solved to survive) combined combined with the species-typical species-typical solutions or adaptations that have evolved in humans to accomplish these ends. Survival motives, however, may not be the most central to personality psychol psychologists. More important, perhaps, are those directional tendencies that involve social interaction. Group living confers reproductive advantage on those so doing to the extent that reproductively relevant resources are concentrated within groups. The fact that all known human groups are intensely social provides prima-facie pnma-/flde justifica justification for attempting to identify precisely what those resources are. Although the survival advantages of of group living, in the forms of of protection (e.g., through coalitional aggression and defense) and food provisions (e.g., through mating problems problems require close hunting), are fairly clear, the advantages for solving mating

CHAPTER CHAPTER 13 13 EVOLUTIONARY EVOLUTIONARYFOUNDATIONS FOUNDATIONS

329 329

attention..Humans successfully mated in our evolutionary past are our ances ancesattention .. Humans who successfully tors; those who failed to mate are no one's ancestors. Although reproduction provides a broad directional tendency toward which human action is directed, the "problem" at this level of abstraction is too broad of human nature. Like survival, successful reproduction to constitute a motive of specific problems. Three of the most requires the solution of a number of more specific important mating problems are: (1) successful intrasexual competition (Buss, 1988a), successful mate selection (Buss, 1989a), and (3) successful mate retention (Buss, (2) successful 1988b; Flinn, 1988). Although group living confers on members a concentrated pool of potential of intrasexual competitors. Since mates differ tremen tremenmates, it also imposes a field of dously in reproductively relevant ways, competition will be intense for those mates who are most desirable. Success at intrasexual competition constitutes a major directional tendency for human action. intrasexBut even this "problem" contains numerous subproblems. Successful intrasex often will involve (a) besting members of one's own sex in status, ual competition often rank, prowess, or resources, and (b) competing to embody those characteristics that members of the opposite sex desire. Thus evolutionary theory leads to the expectation that a central directional tendency of humans will be to accomplish these tasks. Given successful intrasexual competition, a second major problem of of mating is selecting a reproductively valuable mate. This problem differs in substance for human males and females. Although male and female mate preferences are highly similar for many characteristics (e.g., both sexes worldwide appear to value kindness, intelligence, stability, and dependability in potential mates), they differ in the nature of the problems that must be solved by the act of of mate selection. of For males more than for females, reproductive success is limited by ability to mate with fertile mates. Female fertility, however, is an internal characteristic that cannot be directly observed. Age and hence physical appearance correlates of age appear to provide the most powerful probabilistic cues to fertility. fertiUty. The hypothesis that males prefer youth and value physical appearance in potential mates more than do females has been confirmed in 37 cultures worldwide (Buss, 1989a). Females face a different different reproductive constraint. Relative to humans males, the reproductive potential of of human females is low and the variance in number of offspring offspring is low. Trivers' (1972) theory of of parental investment and sexual selection predicts that under certain ecological and mating conditions, females will select mates in part based on the external resources that a male can provide for her and her offspring. This is especially true where resources can be defended and accrued, n mating effort, and where where males are inclined to invest these resources iin males differ differ substantially in their ability and willingness to provide resources. The general hypothesis that women value resources, as well as cues to resource acquisition (e.g., ambition, industry), more than do men has been verified in 37 cultures (Buss, 1989a). In sum, although men and women are both faced with the

330 330

DAVID M. DAVID M.Bvss Buss

problem of of mate selection, the directional tendencies of men and women are ex exproblem pected to differ differ because the nature of of the problems that must be solved differs. differs. pected A A mate gained must be retained. This is true in contexts where long-term long-term mating bonds are formed, formed, and where reproductive success is impaired impaired by loss or defection of of a mate. Thus, a third third directional tendency tendency subsumed subsumed by mating is defection retention. mate retention. In a study of of mate retention retention tactics (Buss, 1988b), I found found a tremendous tremendous diversity of of acts and tactics devoted devoted to mate retention. These ranged from from vigilance vigilance diversity (e.g., (e.g.. He called her at unexpected unexpected times to see who she would be with; She had her friends friends check up on him; He snooped through her personal belongings; At At the her party, she did not of her sight) to violence violence (e.g., (e.g.. He hit the guy who not let him out of interested in her made a pass at her; She picked a fight with the woman who was interested mate; He vandalized vandalized the property property of the guy who had made a pass at her). Human Human effort is directed directed not simply toward competing for for initial access to mates mating effort and choosing among available alternatives-substantial alternatives—substantial effort effort appears to be directed directed toward retaining mates who have been acquired. acquired. toward In sum, evolutionary evolutionary theory theory provides a powerful powerful heuristic for for identifying identifying the ends toward which human action is directed. As such, it yields a model for for the **core" of of personality that is anchored not in arbitrary arbitrary speculations, but rather rather in "core" the biological biological process that governs governs all known forms of of organic organic life-evolution life—evolution by natural selection.

B. Identifying Identifying Important Individual Differences Differences Personality Personality psychologists have long been concerned concerned with identifying identifying the most im important ways in which individuals differ. differ. From among the thousands of of dimensions dimensions of of difference, difference, which ones should galvanize the attention of of personality personality theorists and researchers? As Goldberg (1972) succinctly phrased phrased the question: Why measure measure that that trait? Rationales for for designating some individual differences differences as particularly important have come from from folk folk psychology (Gough, 1968), factor factor analysis (Cattell, 1946), lexical analysis (Norman, 1963), and the act frequency frequency approach (Buss & Craik, 1985), to name a few. An important important alternative alternative rationale resides in evolution evolutionary criteria criteria for for importance importance (Buss, 1984). Individual differences differences that that are closely linked with components of natural natural selection, sexual selection, and life-history life-history reproductive strategies are crucial using an evolutionary evolutionary rationale rationale for for designating designating importance. Kenrick, Sadalla, Groth, and Trost Trost (1990), for for example, demonstrate that personality characteristics such as dominance, friendliness, and emotional emotional stability stability are intimately intimately linked with sexual sexual selection in that they are central central to mate choice (cf. Buss, 1989a). The possibility possibility differences have previously been, and may currently be, linked that these individual differences with evolution by selection grants them special importance from from an evolutionary evolutionary perspective, when contrasted contrasted with those individual differences differences not so linked. One could argue argue that personality psychologists have long studied studied the dimen dimenof dominance, friendliness, friendUness, and emotional stability, and do not need an evolusions of

CHAPTER 13 13 CHAFfER

EVOLUTIONARYFOUNDATIONS FOUNDATIONS EVOLUTIONARY

331 331

tionary perspective for for pointing them out. This argument argument is correct correct as far far as it goes-indeed, goes—indeed, these three dimensions are part of nearly every every taxonomy taxonomy in personal personality psychology (e.g., Goldberg, 1981; 1981; McCrae McCrae & Costa, 1989; Wiggins, 1979). But what an evolutionary evolutionary perspective does, in this case, is provide a further further rationale rationale useful supplement for why these dimensions are so important. In this sense, it adds a useful supplement for to existing taxonomic efforts of evolutionary evolutionary im efforts by anchoring anchoring dimensions in criteria criteria of importance. Another example, however, suggests that an evolutionary perspective can call Another evolutionary perspective attention to important important individual individual dif differences previously ignored, thought thought attention ferences that were previously subsumed within current current taxonomic taxonomic efforts. A A good to be unimportant, or not subsumed example of this comes from from the work of of Gangestad Gangestad and and Simpson (1990). Using example evolutionary perspective, they argued theoretically theoretically and and found found empirically that an evolutionary differ substantially in what they call "sociosexuality." At At one end of of individuals differ sociosexuality—they require this dimension are individuals who are "restricted" in sociosexuality-they commitment prior to entering a sexual relationship. more time, attachment, and commitment At the other other end are those who are "unrestricted" "unrestricted" in sociosexuality-they sociosexuality—they require require At commitment prior to sexual intercourse. less time, attachment, and commitment of the sociosexual dimension represent represent alternative strategies that These ends of reproductive success. Those who are restricted restricted tend to individuals use to pursue reproductive from their mates high levels of parental investment investment in their of parental provide and obtain from of men, tend to obtain a larger children. Those who are unrestricted, in the case of number of of matings with less parental investment devoted devoted to any particular offspring. number particular offspring. Unrestricted women, in contrast, tend to obtain temporary temporary mates who are highly Unrestricted presumably enables them to pass on to their their offspring offspring attractive to women. This presumably offspring who themselves will be highly attractive attractive to the genes that will lead to offspring achieved by two different different strategies, opposite sex. Thus, reproductive success is achieved important dimension of of individual difference difference that is not currently suggesting an important part of traditional taxonomies of individual differences. differences. part C. Environmentally Enyironmentally Induced Strategic Individual Differences Differences c. Evolutionary psychology provides a powerful powerful set of of conceptual tools for understandfor understand of individual differences. differences. The construction construction workers workers who are laboring ing the origins of on the building next door have thick callouses on their their hands. My academic col colleagues down the hall do not. These individual differences differences in callous thickness are highly stable over time. At At one level of of analysis, the variance can be traced traced solely to variance in the reliably recurring experiences of of the two groups. At At another level of of analysis, the existence of the species-typical callous-producing callous-producing mechanism is a explanation of of observed individual individual central and necessary element in the causal explanation Just as men men and and women women differ differ in the adaptive adaptive problems problems they they confront, confront, differences. Just different individuals individuals within within each sex sex face face different different adaptive adaptive problems over time. time. different problems over Some manifest individual differences differences are the strategic products of of species-typical recurrently different different adaptive problems across indi indimechanisms responding to recurrently viduals.

332 332

DAVID M. DAVID M.Buss Buss

In this callous example, the individual differences friction experiences differences in skin friction experiences are in some sense "environmental." If any academic colleagues colleagues were to trade places with the construction workers, then the manifest manifest individual differences differences would genotype-environment correlation correlation reverse. Nonetheless, we cannot rule out the genotype-environment proposed by Plomin, DeFries, and Loehlin (1977) and Scarr and McCart McCartprocesses proposed of heritable skills, interests, or proclivities ney (1983). Some individuals, because of preferentially select academic work or construction construction work work as occupations. These may preferentially friction-free versus friction frictionselections, in turn, may create repeated exposure to friction-free differentially activate activate the species-typical species-typical prevalent environments, which then differentially callous-producing mechanism. There are three central points: (1) stable manifest manifest individual differences differences can be caused by differences differences in the recurrent adaptive problems to which different different individuals are exposed; (2) the complex species-typical mechanisms are necessary of individual differences, differences, because and central ingredients in the causal explanation of without them the observed individual differences differences could not occur; and (3) the manifest individual differences differences are strategic outcomes of of recurrently recurrently different different input manifest into species-typical mechanisms. There are undoubtedly undoubtedly many recurrent recurrent environmental individual differences differences of precisely this sort. First-born First-bom children probably face face recurrently recurrently different different adaptive adaptive of compared with second-born children. These apparently apparently trigger in first firstproblems compared born children greater identification identification with the status quo, the parents, and the estab estabfor the fact fact that first firstlished power structure. This tendency may be responsible for revolutionary scientific theories (SuUoway, Laterborns typically oppose revolutionary (Sulloway, 1996). Later confront a niche already filled by an older sibling, tend more borns, who confront tend to rebel more greater identification identification with against established traditions, which may lead them to a greater revolutionary revolutionary scientific theories (Sulloway, (SuUoway, 1996). Individuals who grow up in environments where where resources are unpredictable, unpredictable, such as among Hungarian Gypsies, may adopt a more impulsive personality style, and even mating style, where it would be adaptively foolish foolish to delay gratification gratification (Bereczkei, 1993). In contrast, those growing up in environments where resources and future future prospects are more predictable may adopt adopt a personality strategy involving greater delay of gratification, including sexual gratification. The resulting individual differences differences represent represent strategic solutions to the different different adaptive adaptive problems problems encoun encountered. Recurrently Recurrently different different environmental input into species-typical mechanisms mechanisms can produce stable strategically patterned patterned individual differences. differences. Differences D. Heritably Induced Strategic Individual Differences Recurrently different different input into species-typical psychological mechanisms, of of course, from heritable individual differences, differences, whatever whatever their ultimate origin (i.e., may come from from selection for for alternative genetically genetically based strategies, whether they originated from frequency-dependent selection, genetic noise, pathogen-driven pathogen-driven selection for for genetic frequency-dependent uniqueness, or assortative mating). Individuals with an ectomorphic body type, for for confront different different adaptive problems than those who are mesomorphic. example, confront

CHAPTER CHAPTER 13 13

EVOLUTIONARY EVOLUTIONARYFOUNDATIONS FOUNDATIONS

333 333

Ectomorphs may risk being at the receiving end of of greater greater aggression than their more muscular peers, an adaptive problem that typically must be solved solved by means other than physical aggression. Genetic differences, differences, in other other words, pose different different other for different different individuals. adaptive problems for In addition addition to facing facing different different adaptive problems, some individuals experience experience rather than others. "Selection operates operates greater success at pursuing certain strategies rather through the achievement achievement of adaptive goal states, and any feature feature of of the world world— either of of the environment, or of of one's one's own own individual individual characteristics-that characteristics—that influences influences the achievement achievement of of the relevant goal state may be assessed assessed by an adaptively adaptively designed system" (Tooby (Tooby & Cosmides, 1990b, p. 59, emphasis added). Individuals who are mesomorphic, for for example, typically will experience far greater success at enacting far greater an aggressive strategy than individuals who are ectomorphic. Tooby Tooby and Cosmides (1990b) call this phenomenon "reactive heritability." (199Ob) phenomenon heritability." differences in physical attractiveness. There is evidence Consider individual differences successful at pursuing a "short-term" mating that physically attractive men are more successful strategy involving many sexual partners (Gangestad & Simpson, 1990). 1990). Physically of seeking and attractive women are better able to pursue a long-term strategy of partners (Taylor (Taylor & Glenn, actually obtaining higher-status, higher-income marriage partners differences in physical physical attractiveness affect affect the success of pursuing 1976). Heritable differences different mating strategies. The manifest manifest strategy strategy differences differences are in some sense different indirectly and reactively. Relative physical attractiveness func func"heritable," but only indirectly sex-typical psychological mechanisms, which tions as "input" into species-typical or sex-typical of different different individuals in different different directions. then canalize the strategies of of individuals-such individuals—such as differences keenHeritable dimensions of differences in body type, keen oratory skills, physical attractiveness, and spatial spatial ability-provide ability—provide ness of vision, oratory important input into species-typical mechanisms. These individually individually different different inputs important facing and the strategic solutions tell the organism about the adaptive problem it is facing successful The resulting product consists of individual differ differlikely to be successful. of strategic individual ences that are stable over time. The observed strategic differences differences are correlated correlated with genetic variance, but cannot be understood apart from from the central role played by our species-typical psychological mechanisms that that were "designed" to receive input, both environmentally environmentally and heritably based, about about the adaptive problems problems confronted confronted and the strategic solutions likely to be successful. successful.

E. The Five-Factor Model-Personality Model—Personality and the Adaptive Landscape One need not believe that there are only only five important important personality dimensions reach the conclusion conclusion that (see, e.g., de Raad & Hoskens, 1990; Tellegen, 1985) to reach discovered so repeatedly-(l) repeatedly—(1) Surgency (dominance, power, extraversion extraversion the five discovered versus submissiveness, weakness, and introversion), (2) Agreeableness (cooperative (cooperative Conscientiousness (industri (industriand trustworthy versus aggressive and suspicious), (3) Conscientiousness Emotional Stability (secure (secure ous, responsible versus lazy and irresponsible), (4) Emotional Intellect-Openness (intelligent, (intelligent. and stable versus insecure and anxious), and (5) Intellect-Openness

334 334

DAVID M. Buss BUSS DAVID M.

perspicacious, and creative versus stupid, boarish, bearish, and unimaginative)-must unimaginative)—must surely be included in some form within any major personality taxonomy. Descriptive work documents the robustness of these factors, but does not elucidate why why they are so frequently found. From an evolutionary perspective, there are three ways to approach this crucial question. The first, discussed above, is that these individual differences differences may represent strategic differences, based on either heritable or environmental environmental system—variations that differences. Second, they may signify mere "noise" in the system-variations were neutral with respect to natural and sexual selection. The third approach, entirely compatible with the first, is that these five dimensions of differof individual differ ences summarize summarize the most most important important features features of of the social social landscape that humans have had to adapt to (Buss, 1989b). From this second perspective, "to know others is an adaptive necessity" (Symons, 1979, p. 310). The core of of the "personality as adaptive landscape" view is that perceiving, attending to, and acting upon differences differences in others has been (and likely still is) crucial for solving adaptive problems. The first piece of of evidence in favor of of this view is that trait terms are inherently evaluative. Peabody (1985) found that less than 3% of trait terms were evaluatively neutral, the remaining 97% having definite evaluative (as well as descriptive) aspects (see also Hofstee, 1990). Hogan (1983) of others as potential contributors argues that trait terms reflect observer evaluations of of others to, or exploiters of, the group's resources. For example, the Intellect of (Factor 5) must be evaluated so that a person knows whom to go to for advice. Conscientiousness (Factor 3) must be evaluated to know who to trust with tasks. Borkenau argues that a selective advantage would accrue to those persons who differences in have the ability to perceive and act upon these major individual differences others. Graziano and Eisenberg (this volume, chap. 30) place Agreeableness (Factor 2) in evolutionary perspective. They argue that coordinated group action is best accomplished when individuals are willing to cooperate and conform conform to group norms, and suspend their individual concerns for the good of of the group (see also Wiggins, 1991, for a similar account). Humans are an intensely group-living species. Groups afford protection from predators, protection from other groups of aggressive males, the possibility of cooperative hunting of large game, and a population of of potential mates. But groups intensification of of competition, risks of of also carry costs. With group living comes an intensification communicable diseases, and aggression from other group members. Other humans can cripple our survival and reproductive success. Other humans are our primary of strategic interference. Other humans are our primary "hostile force of sources of nature" (Alexander, 1987). In a phrase, other humans define our primary adaptive landscape, and are capable of facilitating or interfering with our reproductive strate strategies (see also Byrne & Whiten, 1988). features I have argued that personality traits summarize the most important features of information for of that adaptive landscape (Buss, 1989b). They provide a source of answering important life questions: Who is high or low in the social hierarchy? of my coaliWho is likely to rise in the future? Who will make a good member of

CHAPTER CHAPTER 13 13

EVOLUTIONARY EVOLUTIONARY FOUNDATIONS FOUNDATIONS

335 335

tion? Who possesses the resources that I need? With whom whom should I share my re reWhom can I depend depend on when in resources with me? Whom sources? Who will share their resources whom should should I mate? Who should should I befriend? befriend? Who might do me harm? harm? need? With whom for sage advice? The hypothesis is that that people Who can I trust? Who can I go to for evolved psychological psychological mechanisms sensitive to individual differences differences in others have evolved that are relevant to answering these critical adaptive questions. of the most important important features features of of human human groups are: (1) they are intensely Two of hierarchical, hierarchical with important important reproductive resources closely linked with position in the hierarchy (e.g., Hogan, 1983; 1983; Lopreato, 1984), and (2) they are characterized characterized by elevated forms forms of cooperation cooperation and reciprocal reciprocal alliance alliance formation formation (Axelrod, 1984; Cosmides & Tooby, 1989; Trivers, 1971). The importance importance of of hierarchy suggests that location of of others in the hierarchy, as well as proclivities to ascend ascend in the hierarchy, are extremely important features features of of the human adaptive landscape. The prevalence prevalence of reciprocal alliance formation suggests that a second critical feature of the human of formation second feature of adaptive landscape is the differential proclivity of others to "cooperate" or to to adaptive landscape is the differential proclivity of others to "cooperate" or "defect" (see Jones, Couch, & Scott, this volume, chap. 19). "defect" (see Jones, Couch, & Scott, this volume, chap. 19). of Surgency (dominance-submissiveness) (dominance-submissiveness) and The persistent emergence of (cooperative-aggressive) as the two two major major axes in interpersonal interpersonal Agreeableness (cooperative-aggressive) iht first first two factors factors in personality-descriptive personality-descriptive taxonomies (e.g., Wiggins, 1979) and as the taxonomies (McCrae & Costa, 1989; Trapnell & Wiggins, 1990) represents the adaptive significance to all humans of discerning in others their their hierarchical position wiUingness to form form reciprocal reciprocal alliances (Buss, 1989b). In and proclivity and their willingness evolutionary history, those individuals who were able to accurately discern human evolutionary and act upon these individual differences differences likely enjoyed enjoyed a considerable reproductive advantage over those who were oblivious to these consequential consequential individual differ differences. Evidence from studies of competition and mating support specific predictions ences. Evidence from studies of competition and mating support specific predictions from an an evolutionary evolutionary analysis analysis of of these these features features of of the the human adaptive landscape. landscape. from human adaptive For example, kindness (Factor 2), dependability (3) emotional stability (4) and and For example, kindness (Factor 2), dependability (3) emotional stability (4) intelligence (5) are among the most valued characteristics in potential mates (Buss intelligence (5) are among the most valued characteristics in potential mates (Buss et aaI., l , 1990). 1990). et Historically, evolutionary approaches have ignored differences. The ignored individual differences. increased sophistication sophistication and precision recent work shows that models can gain increased of individual differences. differences. The integration of of basic by incorporating an analysis of psychological mechanisms with evolved strategic individual differences psychological differences may provide the most compelling theoretical bridge to close the current current chasm between the of psychology psychology that deal with typical human human mechanisms and those that branches of focus on individual differences. differences. focus Psychological Metatheory of of F. An Evolutionary Psychological Differences Sex Differences Evolutionary psychology, in addition to shedding shedding light on individual differences differences within sex, also provides a unique meta metatheory for understanding understanding differences between within theory for differences between the sexes: Men and women are expected to differ differ in their underlying psychology

336 336

DAVID DAVID M. M.Buss Buss

only in the delimited domains where they have faced recurrently different different adaptive faced recurrently problems. In domains where the sexes have faced faced the same adaptive adaptive problems, no sex differences differences are expected. Alternative theories of of sex differences differences typically cannot specify specify in advance advance the domains in which sex differences differences will be found, found, in which direc direction they will be found, and why they will be found found in these domains domains and directions. Men and women historically historically have faced faced many many adaptive adaptive problems problems that that are highly similar. Both sexes needed to maintain body temperature, so both both sexes Repeated friction friction to certain certain areas of of have sweat glands and shivering mechanisms. Repeated of nature" to both sexes in ancestral environments, the skin was a "hostile force of evolved callous-producing callous-producing mechanisms. Both sexes needed so men and women have evolved of identifying identifying a good cooperator for strategic conflu confluto solve the adaptive problem of cooperator for long-term mate, and this may be one reason why both sexes ence when seeking a long-term partner so highly across all cultures cultures whose partner preferences value **kindness" "kindness" in a partner partner preferences have been studied (Buss, 1989a). In several domains, however, the sexes have faced faced different different adaptive problems. For For 99% of of human evolutionary history history men faced faced the adaptive problem problem of of hunting of gathering, possible possible selective reasons for for men's greater greater upper upper body and women of strength and spatial spatial rotation rotation ability ability and for for women's greater greater spatial location memory strength location memory ability (Silverman (Silverman & Eals, 1992). Internal female female fertilization and gestation created ability fertilization and problem of of uncertainty of of parenthood for men, but for women. the adaptive problem parenthood for but not for for men Cryptic ovulation created the adaptive problem for men of of knowing knowing when a woman ovulating—a possible causal force force in the origins of of sexual activity through through the was ovulating-a entire cycle and the emergence of of long-term committed committed mating mating bonds between between a entire man and a woman (Alexander (Alexander & Noonan, 1979). The dual male mating mating strategy man of seeking both short-term short-term sexual partners and long-term long-term marriage marriage partners created of an adaptive problem for whether particular particular men saw for women of of having to discern discern whether temporary sex partners or as potential spouses (Buss & Schmitt, 1993). them as temporary Sex differences differences in mate preferences preferences (Buss, 1989a), courting strategies (Buss, 1988a; Tooke & Camire, 1991), jealousy (Buss et aI., al., 1992), mate guarding tactics (Buss, 1988b; Flinn, 1988), sexual fantasies fantasies (Ellis & Symons, 1990), and and sexual desires (Buss, 1994) correspond correspond remarkably remarkably well to these sex-linked sex-linked adaptive adaptive problems. Evolutionary Evolutionary psychology offers offers the promise promise of of providing providing a coherent coherent theory theory of of strategic strategic sexual differences differences as well as strategic sexual similarities. G. Anchoring Conceptions of of Adaptation Adaptation and Adjustment adaptation and adjustment adjustment have been frameConcepts such as adaptation been central central to nearly all frame works of signify an ability to deal of personality. As typically used, these concepts signify effectively with the varied varied demands of everyday living. The related related dimensions of of effectively neuroticism (Eysenck, 1981), and emotional instability instabiUty (Norman, 1963), which imply imply ineffective negotiation negotiation of of life's tasks, are central to nearly nearly every every taxonomic system ineffective of personality. These concepts represent a particular particular manner of manner of of coping, typically subjective distress, intrapsychic intrapsychic discomfort, discomfort, large fluctuations fluctuations one with high levels of subjective of subjective states. of mood or affect, affect, and a relative inability inability to terminate terminate negative subjective

13 CHAPTER 13

EVOLUTIONARYFOUNDATIONS FOUNDATIONS EVOLUTIONARY

337 337

of adaptation adaptation is also central to evolutionary biology, although although it is The concept of defined somewhat somewhat differently. differently. Adaptations in evolutionary evolutionary biology refer defined refer to evolved posed by the complex tasks of of survival and reproduction. reproduction. solutions to problems posed features of of behavior behavior or morphology morphology are considered to be adaptations, and Not all features the evidential standards that must must be met for for considering something something an adaptation adaptation are complex and often often difficult difficult to meet (Williams, (WiUiams, 1966). Nonetheless, the effective effectiveanchored ness with which reproductive problems are solved provides a biologically anchored of "adjustment." "adjustment." meaning of Draper and Belsky (1990) articulate an intriguing evolutionary evolutionary theory theory of of Draper alternative reproductive strategies based based on the environments environments that that humans encoun encounter in early childhood. They propose that in environments and cultures where where fathers are present present during early childhood, the reproductive strategy tends to involve fathers of sexual activity, stability of adult pair-bonds, and delayed puberty, delayed onset of a set of concomitant concomitant personality characteristics that facilitate this strategy such as self-monitoring and high cooperativeness. In environments and cultures cultures low self-monitoring fathers tend to be relatively absent, an alternative personality constellation where fathers and reproductive reproductive strategy is followed, followed, one involving early onset of of puberty and selfsexual activity, instability of of adult pair-bonds, low parental investment, high self monitoring, and high aggressiveness. intriguing about about this theory theory is that it posits that both both strategies strategies are are What is intriguing of our our species-typical repertoire-we repertoire—we all have the capacity to follow either either part of strategy. But which one we do follow follow depends depends on the environment environment that we encounter encounter Draper and Belsky emphasize that it is possible to rear rear children while growing up. Draper successfully successfully under under both both regimes, and that neither strategy is inherently superior or inferior to the other-they other—they are both "adaptive" in the environmental environmental contexts in inferior which which they occur. This implies that conceptual clarity might might be achieved from from anchoring definitions definitions of of "adjustment" "adjustment" and "adaptation" in the effectiveness effectiveness with particular environments. The definition definition which reproductive problems are solved in particular of adjustment adjustment would would shift shift from from the content content of of the strategy strategy (i.e., it is not necessarily necessarily of maladjusted maladjusted to be impulsive or aggressive) to the success success of of the strategy strategy in a specific specific environment. environment. In sum, evolutionary evolutionary thinking has implications for for how personality personality psycholo psychologists might conceptualize conceptualize adaptation adaptation and adjustment. adjustment. It suggests that the equation equation of of maladjustment maladjustment with strategies that might appear appear distasteful distasteful or repugnant (e.g., those that that are aggressive, impulsive, or wanton) is inappropriate. These strategies strategies may be functional functional in the particular particular environments in which they occur (e.g., where resources are unstable or unpredictable). The effectiveness effectiveness with which survival and reproductive problems are solved is one biological criterion by which the concepts of of "adapted" and "adjusted" "adjusted" can be anchored. anchored. Personality Stability and Change H. Implications for Personality

This framework framework provides a coherent theoretical rationale for for when when we will observe observe personality stability and change, and, perhaps even more importantly, an answer personality

338 338

DAVID DAVID M. M.Buss BUSS

to why why some aspects of personality remain stable while others change. Stability is expected at several levels. First, many of our evolved species-typical mechanisms will remain stable over time, even if they remain unactivated. All humans, including emics, retain their callous-producing mechanisms, even if they rarely cloistered acad academics, encounter the repeated skin friction necessary for their activation. This is stability in the species-typical species-typical sense. Second, stability in the individual differences differences sense can be produced by stable environmental recurrences in exposure to adaptive problems-stability problems—stability in manifest manifest individual differences. Just as differences differences in callous thickness between academics academics and construction construction workers are stable over time, so differences experiencing hostility from others may be stable over time (Dodge & Coie, 1987). Differences Differences in expres expressions of of jealousy, to use another example, may be stable over time due to being differences in married to a spouse who displays frequent cues to infidelity. Stable differences the adaptive problems to which one is exposed, of course, may be created by of individual actors, either heritably or environmentally based. The properties of recurrent barrage of sexual "come-ons" experienced by physically attractive women, for example, may be an adaptive problem that stems from heritable differences in of manifest personality is determined physiognomy. The key point is that stability of in part by the recurrences in the adaptive problems to which individuals are exposed. A third source of stability stems from the retention of successful or well wellA practiced problem solving strategies. Some strategies rely on the exploitation of quaUties. Those with many resources, for example, can cau retain their certain personal qualities. mates with lavish gifts that are inaccessible to the less endowed. Those lacking of any kind may be forced to resort to self-abasing tactics, positive inducements of subordinating oneself oneself to the goals of of one's mate, or cost-imposing tactics, such as subordinating Well-practiced strategies are such as threats and violence (Wilson & Daly, 1992). Well-practiced generally more effective effective than less-well-practiced ones, and so some stability occurs through the retention of well-practiced effective effective strategies. This account is similar to the fascinating theory recently recently proposed proposed by Caspi and Moffitt Moffitt (in press) that suggests that individual differences differences manifest manifest themselves most strongly in times of transition, in part because individuals deploy strategies to deal with those transitions that have worked for them in the past, can be enacted quickly, and cost little energy, presumably because they have been well practiced. This framework theoretical account of framework simultaneously provides a meta metatheoretical of when and why we will observe change. First, change in underlying mechanisms may occur over ontogeny with species-typical shifts in adaptive problems. The most obvious apparaexample is puberty. Women develop enlarged breasts and their reproductive appara tus becomes functional to solve adaptive problems that were irrelevant during childhood. Predictable psychological psychological shifts undoubtedly accompany these changes, such as increased interest in the opposite sex, a honing of one's mate preferences, of vivid sexual and increased attention to one's physical appearance, and the onset of romantic fantasies. Second, change can occur as a result of of developmental shunting of individuals down one path versus another. Those growing up in father-absent households, for

CHAPTER 13 13 CHAPTER

EVOLUTIONARYFOUNDATIONS FOUNDATIONS EVOLUTIONARY

339 339

example, may be shunted into a more promiscuous short-term mating strategy, whereas those growing up with an investing father may be shunted into a more monogamous long-term mating strategy (Belsky, Steinberg, & Draper, 1991). Pre Presumably, individuals have psychological mechanisms that determine whether secur securing a long-term investing mate is likely, or whether the individual would do better to extract a variety of of different different resources from different different shorter-term mates. Third, change in manifest manifest behavior can occur as result of change in the adaptive friction-prevalent occupa occupaproblems to which one is exposed. Just as shifting from a friction-prevalent tion to a friction-free friction-free one causes change in one's manifest callouses, so a shift from a low-Conscientious to a high-Conscientious mate may cause a change in manifest manifest jealousy. More transient shifts in the adaptive problems to which one is exposed may produce more transient shifts in manifest behavior. The fourth source of of change occurs when an old strategy for solving an adaptive problem is eliminated, or a new strategy is acquired or activated. Crying as a tactic for getting one's way becomes less effective as one moves from childhood to adulthood, prompting its diminution with development. Gaining a job promotion of resource-bestowal resource-bestowal as a strategy for attracting and retaining may permit the use of mates, a strategy that was previously inaccessible. New strategies are added and old strategies are jettisoned, in part based on changes in the assets one can exploit effectiveness with changing circumstances. and on shifts in effectiveness Thus, personality stability and personality change can both be understood within a single integrative conceptual framework. This framework framework provides a meta metatheoretical account of of why why we expect stability and change and under under what conditions conditions we expect stability and change. Perhaps through this integrative framework framework for examining stability and change, we can start bridging the traditions that historically have isolated the study of individual differences from the study of human nature.

IV. IV. CONCLUSIONS The evolutionary conception of personality proposed here starts with the premise that personality cannot be properly understood without articulating the problems history— that humans have had to solve over thousands of generations of human history problems that are ultimately related to survival and reproduction. Personality, from this perspective, consists centrally of the psychological psychological mechanisms mechanisms and behavioral behavioral strategies that humans have evolved for solving these problems. Because the problems that must be solved are numerous and complex, the psychological mecha mechanisms and strategies that have evolved are likely to be many and, at least in part, domain-specific. This view differs from previous "evolutionary" approaches to personality that attempt to trace "traits" in humans by comparing them with other primate species such as chimpanzees or baboons. These early approaches fail because: (1) all species, including humans, are unique in some respects, (2) selection of particular species for comparative analysis is often misleading (Tooby & DeVore, 1987), and

340 340

DAVID DAVID M. M.Buss Buss

(3) they ignore the core of Darwinian theory-the adaptation by natural theory—the evolution of of adaptation selection. Adaptation must be central to evolutionary personality personality psychology, and the adaptations of of greatest interest to personality personality psychologists occur in the form of of psychological mechanisms and behavioral strategies. psychology permits personality psychology psychology to escape from the Evolutionary psychology basic directional tendencies of of huendless anchorless speculations about what the basic hu man nature are. Since evolution by natural selection selection is the principle principle guiding guiding force force of organic life, there is no responsible for the creation of all known structures of exempt from this process. This suggests reason to believe that humans have been exempt that the numerous reproductive problems that humans have had to solve (e.g., competition for mates, selection of mates, ensuring paternity in children, rearing children in particular ways, aiding genetic relatives, forming reciprocal alliances, negotiating social hierarchies) must be the core directional tendencies toward which human action is directed. In addition to specifying the core directional tendencies, an evolutionary perspective clarifies several confusing debates that have dominated the field for identifying the past 20 years. years. Evolutionary psychology provides sound criteria for identifying important individual differences, and leads researchers researchers to dimensions previously not considered; anchors the concepts of adaptation and adjustment in ways that escape the previous arbitrary and value-laden definitions; and provides a powerful framework for resolving the debate over personality consistency and behavioral framework specificity. segIn a curious way, an evolutionary perspective also brings together two seg ments of the field that have been separated for decades-the decades—the segment that focuses focuses on grand theories of personality and the segment that focuses on understanding historically is that branch of of psychol psycholindividual differences. Personality psychology historically of all the subfields of ogy that is in principle the broadest and most integrative of framework within which its psychology. An evolutionary perspective provides the framework breadth and integration can be achieved.

ACKNOWLEDGMENT This chapter chapter was was supported supported in in part part by by aa fellowship fellowship from from the the Center Center for for Advanced Advanced Study Study in This the Behavioral Sciences and by by NIMH NIMH Grant Grant MH-44206. MH-44206. The The author author thanks thanks Leda Leda Cosmides, Cosmides, Martin Daly, Don Symons, John Tooby, and Margo Wilson for illuminating discussions of the issues in this chapter.

REFERENCES Alexander, R. D. D. (1979). Darwinism Darwinism and andhuman humanaffairs. affairs. Seattle: University Washington Alexander, Seattle: University ofof Washington Press. Alexander, R. D. (1987). The The biology biologyof ofmoral moralsystems. systems.New NewYork: York:dedeGruyter. Gruyter.

CHAPTER CHAPTER 13 13

EVOLUTIONARY EVOLUTIONARY FOUNDATIONS FOUNDATIONS

341 341

Alexander, R. D., & Noonan, K. M. (1979). (1979). Concealment of ovulation, parental care, and human social evolution. In N. A. Chagnon & W. Irons (Eds.), Evolutionary biology and human AnAn anthropoLogical perspective (pp.(pp. 402-435). North Scitu humansocial socialbehavior: behavior: anthropological perspective 402-435). North Scituate, MA: Duxbury Press. Axelrod, R. (1984). The evolution of of cooperation. cooperation. New NewYork: York:Basic BasicBooks. Books. Bailey, J. M., Gaulin, S., Agyei, Y., & Gladue, B. B. A. (1994). (1994). Effects Effects of gender and sexual orientation on evolutionarily relevant aspects of human mating psychology. Journal of Personality 66,66, 1074-1080. Personalityand andSocial SocialPsychoLogy, Psychology, 1074-1080. Barkow, J. (1989). Darwin, Darwin, sex, sex, and status. status. Toronto: Toronto:University University of ofToronto Toronto Press. Press. Belsky, J., Steinberg, L., & Draper, P. (1991). (1991). Childhood experience, interpersonal develop developDevelevolutionary theory of socialization. Child Devel ment, and reproductive strategy: An evolutionary opment, 62, 62,647-670. 647-670. opment, T. (1993). (1993). r-selected r-selected reproductive reproductive strategies strategiesamong among Hungarian Hungarian Gypsies: Gypsies:AAprelimi prelimiBereczkei, T. analysis. EthoLogy Ethology and andSociobiology, Sociobiology,14, 14,71-88. 71-88. nary analysis. universals.Philadelphia: Philadelphia:Temple TempleUniversity UniversityPress. Press. Brown, D. E. (1991). Human universals. Buss, A. H. H. (in press, this volume). volume). An An evolutionary evolutionary perspective on on personaUty In personality traits. traits. In Handbook of of Personality PersonalityPsychology. Psychology.New NewYork: York:Academic AcademicPress. Press. Hogan et al. (eds.), Handbook person-environment (PE) correlation: The role Buss, D. M. (1984). Toward a psychology of person-environment Journalof ofPersonality Personalityand andSociaL Social Psychology, 361-377. of spouse selection. JournaL PsychoLogy, 47,47, 361-377. M. (1986). (1986).Can Cansocial socialscience sciencebe beanchored anchoredininevolutionary evolutionarybiology? biology?Revue RevueEuropeene Europeene Buss, D. M. SciencesSociales, Sociales,24,24, 41-50. des Sciences 41-50. evolution of human intrasexual competition: Tactics of mate Buss, D. M. (1988a). The evolution Journal of ofPersonality Personalityand andSociaL Social Psychology, 661-628. attraction. JournaL PsychoLogy, 54,54, 661-628. Buss, D. M. (1988b). From vigilance to violence: Tactics of mate retention. Ethology and Sociobiology,9,9,291-317. 291-317. SociobioLogy, differences in human mate preferences: Evolutionary hypotheses Buss, D. M. (1989a). Sex differences Behavioraland andBrain BrainSciences, Sciences, 1-49. tested in 37 cultures. Behavioral 12,12, 1-49. theory of of strategic strategic trait traitusage: usage:Personality Personality and Buss, D. M. (1989b, June 26-30). A theory and thethe adaptiveLandscape. landscape.Paper Paperpresented presentedatatthe theInvited InvitedWorkshop WorkshopononPersonality PersonalityLanguage, Language, adaptive University of Groningen, The Netherlands. Buss, D. M. (1990a). Evolutionary social psychology: Prospects and pitfalls. Motivation Motivation and and Emotion, 14, 14,265-286. 265-286. Emotion, Toward a biologically informed informed psychology of personality. JournaL Journal of Buss, D. M. (1990b). Toward of Personality,58, 58,1-16. 1-16. Personality, Annual Review of of PsychoLogy Psychology (pp. (pp. Buss, D. M. (1991). Evolutionary personality psychology. AnnuaL 459-491). Palo Alto, CA: Annual Reviews, Inc. of human mating. NewNew York: BasicBasic Buss, D. M. (1994). The The evoLution evolutionof ofdesire: desire:Strategies Strategies of human mating. York: Books. Buss, D. M. (1995). Evolutionary psychology: A new paradigm for psychological science. PsychologicalInquiry, Inquiry,6, 6,1 -300. 1-300. PsychologicaL Buss, D. M., Abbott, M., Angleitner, A., Asherian, A., Biaggio, A. (plus 45 additional authors). (1990). International preferences in selecting mates: A study of 37 societies. Journalof of Cross-CuLtural Cross-CulturalPsychoLogy, Psychology, 5-47. JournaL 21,21, 5-47. Buss, D. M., & Craik, K. H. (1985). Why not measure that trait? Alternative criteria for of Personality Personalityand andSocial SocialPsychoLogy, Psychology, identifying important dispositions. Journal JournaL of 48,48, 934-946. Buss, D. M., Larsen, R., Westen, D., & Semmelroth, Senmielroth, J. (1992). (1992). Sex differences differences in jealousy: PsychologicalScience, Science,3,3,251-255. 251-255. Evolution, physiology, and psychology. Psychological

342 342

DAVID M. M. Buss BUSS DAVID

Buss, D. M., & Schmitt, D. P. (1993). Sexual Strategies Theory: A contextual evolutionary analysis of human mating. Psychological PsychologicalReview, Review,100, 100,204-232. 204-232. R, & Whiten, A. (Eds.). (1988). Machiavellian Clarendon Press. Machiavellianintelligence. intelligence.Oxford: Oxford: Clarendon Press. Byrne, R., Cantor, N., & Kihlstrom, Kihlstr5m, J. (1987). Social intelligence. intelligence.New NewYork: York:Academic AcademicPress. Press. Caspi, A., & Moffitt, Moffitt, T. (in press). Psychological PsychologicalInquiry. Inquiry. R. B. (1946). Description Description and andmeasurement measurementof ofpersonality. personality.Yonkers-on-Hudson, Yonkers-on-Hudson,NY: NY: Cattell, R World Book. The logic logic of of social social exchange: exchange: Has Has natural natural selection selection shaped shaped how how humans humans (1989). The Cosmides, L. (1989). Cognition, 31, 31,187-276. 187-276. reason? Cognition, Cosmides, L., & Tooby, J. (1987). From evolution to behavior: Evolutionary psychology as the missing link. In J. Dupre Dupr^ (Ed.), The The latest latest on on the the best: best:Essays Essaysononevolution evolutionand and optimality. optimality. Cambridge, MA: MIT Press. Cosmides, L., & Tooby, J. (1989). Evolutionary psychology and the generation of culture. A computational computational theory theory of of social social exchange. exchange.Ethology Ethologyand andSociobiol SociobiolPart II: Case study: A ogy, 10, 10, 51-98. ogy, behavior. Boston: Boston:Willard Willard Grant Grant Press. Press. evolution, and behavior. Daly, M., & Wilson, M. (1983). Sex, evolution, Daly, M., & Wilson, M. (1988). Homicide. Homicide. New York: de Gruyter. 982). Male M., & &Weghorst, Weghorst,S.S.J.J.(1(1982). Malesexual sexualjealousy. jealousy.Ethology Ethologyand andSociobiol SociobioU Daly, M., Wilson, M., ogy, 3, 11-27. ogy, originof ofthe thespecies speciesbyby means natural selection, or. Preservation Darwin, C. (1859). On the origin means ofof natural selection, or, Preservation of favoured races races ininthe thestruggle strugglefor forlife. life.London: London: Murray. of Murray. Dawkins, R, R., & Krebs, J. R R. (1978). Animal signals: information or manipulation. In J. R R. signals: information evolutionary approach approach(pp. (pp. & N. B. Davies (Eds.), Behavioral Behavioral ecology: ecology: An evolutionary Krebs & 282-309). Oxford: Blackwell. European Journal of (1990). Personality descriptive nouns. European of de Raad, B., & Hoskens, M. (1990). Personality, Personality,4,4,131-146. 131-146. Social-information-processing factors in reactive and Dodge, K. A., & Coie, J. D. (1987). Social-information-processing proactive aggression in children's peer groups. Journal of of Personality Personalityand andSocial SocialPsy Psychology, 53, 53, 1146-1158. chology, (1990). Personality Personality development development in in evolutionary evolutionary perspective. perspective.Journal Journal Draper, P., & Belsky, J. (1990). of of Personality, Personality,58, 58,141-163. 141-163. J., & &Symons, Symons,D. D.(1990). (1990).Sex Sexdifferences differencesininsexual sexualfantasy: fantasy:An Anevolutionary evolutionarypsychologpsychologEllis, B. J., Journalof ofSex SexResearch, Research,27,27, 527-556. 527-556. ical approach. Journal A model model for for personality. personality. Berlin: Berlin:Springer-Verlag. Springer-Verlag. Eysenck, H. J. (1981). A Sociobiology,9,9,1-28. 1-28. Flinn, M. (1988). Mate guarding in a Caribbean village. Ethology and Sociobiology, Gangestad, S. W., & Simpson, J. A. (1990). Toward an evolutionary history of of female of Personality, sociosexual variation. Journal of Personality,58, 58,69-96. 69-96. R. (1972). Some Some recent recent trends trends in in personality personality assessment. assessment. Journal Journalof of Personality Personality Goldberg, L. R Assessment, Assessment,36, 36,547-560. 547-560. R. (1981). Language and individual differences: The search for for universals in Goldberg, L. R personality lexicons. In L. Wheeler (Ed.), Review Review of ofpersonality personalityand andsocial socialpsychology psychology (pp. 141-165). Beverly Hills, CA: CA: Sage. interpreter's syllabus for the California California Psychological Inventory. Gough, H. G. (1968). An interpreter's , pp. In P. McReynolds (Ed.), Advances in psychological psychological assessment assessment (Vol. (Vol. 11, pp. 55-79). 55-79). Palo Alto, CA: Science & Behavior Books. R. C. (1979). (1979). The spandrels of San Marco and the Panglossian Gould, S. J., & Lewontin, R Proceedingsof ofthe theRoyal RoyalSociety Society paradigm: A critique of the adaptationist programme. Proceedings of of London, London, Series SeriesB, B,205, 205,581-598. 581-598.

13 CHAPTER 13

EVOLUTIONARY FOUNDATIONS EVOWnONARY

343 343

Herrnstein, R. (1977). The evolution of of behaviorism. American American Psychologist, Psychologist, 32, 593-603. Hofstee, W. K. B. (1990). The use of everyday personality scientific purposes. personaUty language for scientific European Journal Journal of of Personality, Personality, 4, 77-88. Hogan, R. (1983). A A socioanalytic theory of personality. In M. M. Page (Ed.), Nebraska Nebraska Symposium Symposium on Motivation. Motivation, Lincoln: University of of Nebraska Press. Kenrick, D. T., Groth, G. E., Trost, M. R., & Sadalla, E. K. (1993). Integrating evolutionary and social exchange perspectives on relationships: Effects Effects of of gender, self-appraisal, self-appraisal, involvement on level of mate selection. Journal Journal of and Social Social Psychol Psycholof Personality Personality and and involvement 951-969. -969. ogy, 64, 951 Kenrick, D. T., Sadalla, E. K., Groth, G., & Trost, M. R. (1990). Evolution, traits, and the Qualifying the parental investment investment model. Journal Journal of of stages of human courtship: Qualifying Personality, 58, 97-1 16. 97-116. Lopreato, 1. '(1984). Human J.*(1984). Human nature nature and and biocultural biocultural evolution. evolution. Boston: Allen & Unwin. McCrae, R. R., & Costa, P. (1989). The structure of interpersonal traits: Wiggins circumplex Journal of of Personality Personality and and Social Social Psychology, and the five-factor model. Journal Psychology, 56, 586-595. Norman, W. T. (1963). Toward an adequate taxonomy of of personality attributes: Replicated Journal of of Abnormal Abnormal and Social Social factor structure in peer nomination personality ratings. Journal Psychology, 66, 574-583. Psychology, National Characteristics. Characteristics, Cambridge, England: Cambridge University Peabody, D. (1985). National Press. Pinker, S. (1994). The language instinct. New York: Wm. Morrow. Pinker, S., & Bloom, P. (1990). Natural language and natural selection. Behavioral Behavioral and and Brain Brain Sciences, 13, 707-784. Plomin, R., DeFries, J. c., C , & Loehlin, J. C. (1977). Genotype-environment Genotype-environment interaction and Psychological Bulletin, 84, 309-322. correlation in the analysis of human behavior. Psychological SchuU, J. (1988). The adaptive-evolutionary adaptive-evolutionary point of Rozin, P., & Schull, of view in experimental psychology. In R. C. Atkinson, R. J. Herrnstein, G. Lindzey, & R. D. Luce (Eds.), Steven 's handbook Steven's handbook of of experimental experimental psychology. psychology. New York: Wiley. Taking Darwin Darwin seriously: seriously: A naturalistic approach to philosophy, Ruse, M. (1986). Taking approach to philosophy. Oxford: Blackwell. people make their their own environments: A A theory of of Scarr, S., & McCartney, K. (1983). How people —> environment environment effects. effects. Child Child Development, Development, 54, 54, 424-435. 424-435. genotype -+ learning. New Biological boundaries of learning. boundaries of Seligman, M. E. P., & Hagar, J. L. (Eds.) (1972). Biological York: Appleton-Century-Crofts. Appleton-Century-Crofts. Shepard, R. N. (1987). Toward a universal law of of generalization for for psychological science. Science, 237, 1317-1323. component parts, somatic and gentical selection, Short, R. V. (1979). Sexual selection and its component Advances in the Study of Behavior, Behavior, 9, 131-158. Study of as illustrated by man and great apes. Advances Silverman, I., & Eals, M. (1992). Sex differences differences in spatial abilities: Evolutionary theory and adapted mind mind (pp. 533-549). data. In J. Barkow, L. Cosmides, & J. Tooby (Eds.), The adapted New York: Oxford University Press. Singh, D. (1993). Adaptive significance of waist-to-hip ratio (WHR) and female attractiveattractive Journal of of Personality Personality and Social Social Psychology, Psychology, 65, 293-307. ness. Journal Radical thinking thinking in in science science and social social thought. Cambridge, Cambridge, SuUoway, F. (1996). Born Born to rebel: Radical Sulloway, MA: MIT Press. evolution of of human sexuality. New York: Oxford Oxford University Press. Symons, D. (1979). The evolution Symons, D. (1987). If we're all Darwinians, what's the fuss about? In C. Crawford, M. Sociobiology and and psychology: psychology: Ideas, issues, and and applications applications Smith, & D. Krebs (Eds.), Sociobiology (pp. 121-146). Hillsdale, NJ: Erlbaum.

344 344

DAVID M. M.Bvss Buss DAVID

Ethology and and Sociobiology, Sociobiology,10, 10, Symons, D. (1989). A critique of Darwinian anthropology. Ethology 131-144. Symons, D. (1992). On the use and misuse of Darwinism in the study of human behavior. In J. The adapted adaptedmind mind(pp. (pp.137-159). 137-159).New New J. Barkow, L. Cosmides, & & J. Tooby (Eds.), The York: Oxford University Press. & Glenn, Glenn, N. N. D. D. (1976). (1976).The The utility utility of of education education and and attractiveness attractiveness for for females' females' Taylor, P. A., & American Sociological SociologicalReview, Review, 484-498. 41,41, 484-498. status attainment through marriage. American Tellegen, A. (1985). Structures of mood and personality and their relevance to assessing anxiety, with an emphasis on self-report. In A. H. Tuma & & J.J. D. D. Maser Maser (Eds.), (Eds.),Anxiety Anxiety and the and the anxiety anxietydisorders. disorders.Hillsdale, Hillsdale,NJ: NJ:Erlbaum. Erlbaum. Tooby, J., & & Cosmides, L. (1990a). The past explains the present: Emotional adaptations and the structure of ancestral environments. Ethology 375-424. Ethology and andSociobiology, Sociobiology,11,11, 375-424. & Cosmides, L. (1990b). On the universality of human nature and the uniqueness J., & Tooby, J., Therole roleof of genetics geneticsand andadaptation. adaptation.Journal JournalofofPersonality, Personality, of the individual: The 58,58,17-68. 17-68. Tooby, J., & & Cosmides, Cosmides, L. L. (1992). (1992). Psychological Psychological foundations foundations of of culture. culture. In In J. J. Barkow, L. Tooby, Cosmides, & & 1. J. Tooby (Eds.), The The adapted adapted mind (pp. 19-136). New York: Oxford University Press. & DeVore, DeVore, 1.I. (1987). (1987).The The reconstruction reconstruction of of hominid hominid behavioral behavioral evolution evolution through through Tooby, J., J., & Primate strategic modeling. In W. G. Kinzey (Ed.), The Theevolution evolutionofofhuman humanbehavior: behavior: Primate models (pp. (pp. 183-237). New York: York: State University of New York Press. models & Camire, Camire, L. L. (1991). (1991). Patterns Patterns of of deception deception in in intersexual intersexual and and intrasexual intrasexual mating mating Tooke, W., & Tooke, Ethology and andSociobiology, Sociobiology,12, 12,345-364. 345-364. strategies. Ethology & Wiggins, J. S. (1990). Extension of the Interpersonal Adjective Scales to Trapnell, P. D., D., & Journalof ofPersonality Personalityand and include the Big Five dimensions of personality (IASR-B5). Journal Social Psychology, Psychology,59, 59,781-790. 781-790. Social Quarterly Review Reviewof ofBiology, Biology,46, 46, Trivers, R. (1971). The evolution of reciprocal altruism. Quarterly 35-57. Trivers, R. (1972). Parental investment and sexual selection. In B. Campbell (Ed.), Sexual Sexual 1871-1971 (pp. 136-179). Chicago: Aldine. selection selectionand andthe thedescent descentofofman: man: 1871-1971 (pp. 136-179). Chicago: Aldine. Wiggins, J. S. (1979). A psychological taxonomy of trait descriptive terms: terms: The The interpersonal interpersonal Psychology, 37, 37, 395-412. domain. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 395-412. Wiggins, J. understanding J.S.S.(1991). (1991).Agency Agencyand andcommunion communionasasconceptual conceptualcoordinates coordinatesfor forthe the understanding and measurement of interpersonal behavior. In W. Grove & & D. Cicchetti (Eds.), Thinking psychology Thinkingclearly clearlyabout about psychology(Vol. (Vol.2).2).Minneapolis, MinneapoHs,MN: MN:University UniversityofofMinne Minnesota Press. selection: A critique ofsome current evolution Williams, WiUiams, G. G. C. C. (1966). (1966).Adaptation Adaptationand andnatural natural selection: A critique of some current evolutionary thought. thought.Princeton, Princeton,NJ: NJ:Princeton PrincetonUniversity UniversityPress. Press. Wilson, M., & & Daly, M. (1992). The man who mistook his wife for a chattel. In J. Barkow, L. Cosmides, & & J.J.Tooby Tooby (Eds.), (Eds.), The Theadapted adaptedmind mind(pp. (pp.289-322). 289-322).New NewYork: York:Oxford Oxford University Press.

CHAPTER 14 14

EVOLUTIONARY PERSPECTIVES EVOLUTIONARY ON PERSONALITY TRAITS H . Buss BUSS ARNOLD H. UNIVERSITY OF OF TEXAS TEXAS UNIVERSITY

Three evolutionary perspectives may help us to understand human personality traits. The The first first emphasizes emphasizes similarities similarities between between humans humans and and the the animals animals with with whom whom traits. we share share an an evolutionary evolutionaryheritage heritageininbehavior, behavior,personality, personality,and andadaptations adaptationsrelated related we to personality personality traits. traits. Like Like the the sociobiological sociobiological approach, approach, this this perspective perspective focuses focuses to on similarities similarities between our species and others, but it it differs differs in in two two ways. on between our species and others, but ways. First, First, the similarities similarities are are limited limited to to mammals and and primates. primates. Second, Second, the the focus focus is is mainly mainly the personahty traits, not on principles involving the propagation of on behavior and personality genes. evolutionary trends trends in the line that led The second perspective is broader: evolutionary to our species. An example is the waning of instinct as a determinant of behavior. Thus in lower mammals, stimuli often serve as releasers of an entire program of innate behavior, whereas in our species, stimuli typically serve as conditioned stimuli, cues for instrumental behavior, or information to be cognitively propro cessed. cessed. Like other species, we differ anatomically and psychologically from all other species, and and we we posses posses personality personality traits traits unique unique to to us. us. We We study study them, them, and and not not the the species, other way way around. around. Clearly, Clearly, aa third third perspective perspective is is needed, needed, one one that that examines examines how how other we are are distinctively distinctivelyhuman. human. we The chapter is organized around these perspectives, which are discussed in primate, and and human humanbehavior behaviorand and person personsequence. I cite literature on mammalian, primate, ality, but but also also attempt attempt aa conceptual conceptual integration integration of of this this knowledge. knowledge. ality, IlANDBOOK HANDBOOK OF OF PERSONAL1I'Y PERSONALITY PSYCHOLOGY PSYCHOLOGY

345 345

COPYRIGHT © 1997 BY ACADEMIC PlESS. PRESS. CoPYII GHT C ALL liGHTS OF REPIODUcnON IN ANY FORM RESERVED.

ALL RIGHTS OF REPRODUCHON IN ANY FORM RESERVED.

346 346

ARNOLD ARNOLD H. H. Buss BUSS

I. THE COMMON HERITAGE The animals closest to us in evolutionary terms are the primates. Primates have of course been shaped by evolutionary processes and therefore share behavior with other mammals and with highly social mammals, but some features are distinctively primate. In what follows, imagine that we are nothing more than primates, especially the great apes, and even more specifically, the species closest to us, us, chimpanzees. chimpanzees. to

A. Primates as Mammals

activity, which descends Like other mammals, primates follow a daily rhythm of activity, to a low ebb at night during sleep and reaches various peaks during the day, usually before feeding. During the day there are also intense bursts of energy in aggression, escape, sexual behavior, or just frisky play. Within any particular species, there are marked individual differences in the level of energy expenditure, which are well known to owners of pets and have been reported in chimpanzees by de Waal (1982). Many mammals range over a territory, sniffing and searching for any changes exploreisisespecially especiallystrong strongininthe theline linefrom fromwhich which The tendency tendency to to explore that might occur. The primates evolved. Their curiosity can lead to contact with unfamiliar animals of and members of other species. species. Then Then they theyare are wary waryand andless lessmobile. mobile. species and their own species Thus when the chimpanzees at Gombe first saw Jane Goodall, they retreated and kept their distance, though her behavior objectively posed no threat (Goodall, 1986). Exploratory behavior temporarily diminished but but slowly returned to its its previ previous level. Only after many months of her continued presence were they able to overcome their shyness and allow her to approach closely and occasionally interact with them. Such timidity may be one component of a more general tendency to be fearful. fearful When a young primate is exposed to excessive social novelty, temporary abandonment, or threat, it reacts by howling, shrieking, clutching, and kicking. Puppies and kittens tend to wail and whine. Despite species differences among mammals in the expressive display of fear, there is a common set of reactions sysstem: increased increased rate rate of of breath breathtriggered by activation of the sympathetic nervous sy�tem: ing, higher higher blood blood pressure, pressure, and and shunting shunting of of blood blood from from the the digestive digestive system system to to the the ing, skeletal muscles-all muscles—all preparations preparations for for flight. flight. skeletal Primates share with other mammals a behavioral repertoire of aggressive aggressive and dogs is well known, known, against both both their their own kind responses. Thefighting fighting of cats and and members members of of other other species. species. Earlier Earlier field studies of of chimpanzees chimpanzees found found them them to to and field studies be gentle gentle and and peaceful, peaceful, but but later later research research has has revealed revealed clear clear patterns patterns of of aggression aggression be (Goodall, 1986). (Goodall, 1986).

CHAPTER 14 14 CHAFfER

PERSPECTIVES EVOLUTIONARY PERSPECDVES

347 347

B. Primates as Sociable Sociable Mammals Mammals

Primates are part of a group of mammals that are highly sociable, some of the best examples being elephants, dolphins, and dogs. They prefer prefer to associate with one off from from companionship. Dogs mourn for for their another and become upset when cut off masters, and chimpanzees become depressed by the absence of their companions (H. (H. F. Harlow & Harlow, 1962). After After social isolation, reunion elicits not just relief relief but also elation. Dogs greet comrades by nuzzling and licking each other; primates seem to enjoy grooming others and being groomed. All mammalian mothers form an attachment bond with their newborn off offspring. In less sociable mammals, this bond wanes and gradually disappears as the offspring offspring become capable of self-help. But primates and other highly social mammals form enduring bonds of attachment, which which may may continue continue throughout throughout life. life. Adult Adult of attachment, primates, especially the great apes, show affection affection to friends by grooming, hugging, observed among humans, we call patting, or even kissing. When such attachment is observed friendship or love. it deep friendship Primate infants may may become accustomed to being the sole focus of the mother's affection and refuse refuse to share this affection affection with others. When the mother gives affection jealous. Jealousy is birth to the next child, the older sibling may become intensely jealous. not limited to the young, however, and is a potential consequence of any close bond of affection. Pet dogs have been known to display jealousy when another pet after the birth of a child. is introduced into the household or after

c. C. Two Kinds of Emotions

otions. I group jealousy, love, elation, and grief under the heading of relationship relationshipem emotions. They are absent in nonsocial mammals (cougars, leopards) and occur in weak form among moderately social mammals such as the common cat. It is clear that love is part of attachment and that jealousy can arise when exclusive love is demanded. The origins of elation and grief grief were mentioned earlier. Grief so Grief occurs when a social animal is isolated from from an attached other, and joy is elicited by acceptance and affection affection from from others. The primordial reaction of a social animal to separation from its fellows is melancholy, just as the primordial reaction to being reunited is elation. Three other emotions-fear, emotions—fear, rage, and sexual arousal-are arousal—are grouped under the heading of arousal emotions. All three involve a massive emotions. All three involve a massivereaction reactionof ofthe thesympathetic sympathetic division of the autonomic nervous system. Fear and rage are part of the preparation of the body for fight or flight (Cannon, 1927). 1927). The The physiological physiological reactions reactions in in sexual sexual arousal are more complex, but in the late stages of sex at least, there is the same kind of bodily arousal as in fear and anger (Masters & Johnson, 1966). Uttle or no autonomic arousal. In the relationship emotions, however, there is little Consider the affection affection of friends friends or the love of a mother for her children. These

348 348

ARNOLD ARNOLDH. H.Buss Buss

feelings may be experienced as are physiologically placid emotions, though the feelings intense.l intense.^ Similarly, there is no autonomic arousal in grief grief or depression, and there may even be a lowered arousal. There may be autonomic arousal in jealousy but only when there is a strong anger component. Subtract the anger component from jealousy, and autonomic arousal is absent. Given that the arousal emotions occur in all mammals and the relationship emotions occur in all highly social mammals, they are likely to be adaptive. The arousal emotions can be seen as aiding survival. Fear and rage are part of of bodily preparation for the massive muscular exertion of fight or flight in the face of of possible injury or death. Sexual arousal enhances the sexual motivation required for the continuity of of the genes. Thus the arousal emotions are crucial in survival of of the of the individual (fear and rage). genetic material (sex) or in survival of doubtful whether whether the relationship emotions play a similar direct role in It is doubtful survival. If they have an indirect adaptive function, it may be to facilitate group cohesiveness and cooperation. The joy of of being accepted and welcomed by others affiliate with them. The grief of isolation in the group reinforces the tendency to affiliate grief of sufficiently aversive to motivate affiliative affiliative behavior, which ends the from others is sufficiently melancholy. And an individual's depression spurs others to offer offer the help that might lift the individual's mood. Thus the emotions of of love, joy, and grief grief may serve to lift enhance sociality, which is known to be adaptive (Wilson, 1975). Jealousy may be of love in children. However, in adults jealousy merely a nonadaptive consequence of of a demand for exclusiveness is a relationship. Women require is often the result of of love, and men of of sex (see David Buss, this volume, chap. 13). exclusivity of In brief, when emotions are examined from an evolutionary perspective, two different kinds can be discerned. They differ in arousal, prevalence among mammals, and adaptive functions (see Table I). The potential for aggressive behavior may be one reason that affiliative affiliative animals tend to have a social organization that minimizes aggression, usually one based on dominance. dominance. Larger, stronger animals use aggression or threats to achieve superior status over smaller, weaker members of the group. Being dominant usually means having better access to food, to attention from others, and often often to females. Domi-

Theassertion assertionthat thatelation elationinvolves involvesno noautonomic autonomicarousal arousalmay maycontradict contradictpersonal personalexperience. experience. 1^The I suggest three reasons why we should not rely on such experience. First, it confuses the feeling of being person may may dance dance or or leap leap in in the the air air or on top of the world with physiological arousal. Second, a joyful person sufficient muscular muscularexertion exertionto toelevate elevateheart heartrate rateand andbreathing breathingrate. rate.Subtract Subtractthe the muscular otherwise have sufficient muscular exertion, and there is no particular physiological arousal in elation. Third, what is the base rate against which autonomic arousal should be compared? If we compare the physiological state that accompanies might be regarded as as aa mild mild state state joy with that accompanying complete quiescence or NREM sleep, joy might of arousal. But But quiescence is is an aninappropriate inappropriatebaseline baselinecondition. condition.Instead, Instead,we weshould shoulduse usethe thephysiological physiological arousal that occurs in such nonemotional states as problem-solving (Lacey, 1956) 1956) or the orienting reflex to novel stimuli (Lynn, 1966). 1966). When elation, unaccompanied by exercise, is compared with the one*s own thoughts or to external stimuli), it is clear that elation nonemotional state of attention (to one's has none of the physiological arousal of fear, rage, or sexual arousal.

CHAPTER CHAPTER 14 14

349 349

EVOLUTIONARY EvoLimoNARY PERSPECflVES PERSPECTIVES TABLE I TABLE Arousal and Relationship Emotions

Emotions Autonomic arousal Reaction Adaptive function function Occurrence

Arousal

Relationship

Fear, rage, sex High To threat or a sexual stimulus Survival of individual/genes In all mammals

Love, jealousy, joy, grief grief Low To acceptance or rejection rejection Group cohesion In very social mammals

nant primates tend to be groomed by subordinate ones. And those who associate with a dominant animal, either as companions or as sex partners, may share some of the ascendant animal's status. The concepts of dominance and status seem necessary to account for the refers social organization of primates and other highly social mammals. Dominance refers from its own strength and to the animal's position in the group that derives mainly from ferocity, but also from from cleverness and guile. Status refers to the position and privi privifrom dominance but also from age (older leges of the animal, which derive not only from animals have higher status), reproductive status (ovulating females have higher status), and even genealogy (rank of the mother). Genealogy is more important for females, whereas dominance through fighting is more important for males. D. D, Distinctively Distinctively Primate Primate Features Features The primate behavioral tendencies relevant to personality represent extensions of features seen in lower mammals. The mammalian tendency to explore contains an element of curiosity, but primates are even more curious. And primates, especially the great apes, are distinctive in displaying cognitive curiosity. curiosity. In the laboratory, they will work to receive stimulation (Butler, 1957). 1957). Chimpanzees are known to solve puzzles and other problems in the absence of any extrinsic reward, just to satisfy an insatiable curiosity. There is an anecdote about an experimenter who used banana slices to reward a chimpanzee for solving problems. The animal was not hungry and just lined Hned up the banana slices as he solved the problems, but puz eventually the experimenter's supply was exhausted. He continued to present puzzles and was astonished to receive a banana slice after after each puzzle was solved. The chimpanzee obviously was rewarding the experimenter with banana slices sHces and continued to do so until the experimenter regained all of them. More generally, the curiosity of primates, especially chimpanzees, extends beyond the seeking of of tangible rewards to to an intrinsic interest in solving problems and seeking novel stimuli. Their curiosity is abetted abetted by a capacity for for manipulation manipulation that is far beyond that of other mammals. Chimpanzees have been taught such fine motor acts as striking aa match, lighting a cigarette, and wielding a paint brush. Given their manipu-

350 350

ARNOLD H. ARNOLD H.Buss Buss

lative ability, primates' curiosity often often takes the form form of of poking, polcing, prying, squeezing, and generally handling objects and generally objects in the environment. And like any young human human child left left unsupervised unsupervised in a home, they open every every unlocked unlocked cabinet, closet, and around the room. drawer, strewing the contents around Primates are exceedingly exceedingly imitative. A A Japanese's macaque monkey discovered discovered that sweet potatoes tasted tasted better better when she washed off off the dirt in water. Later, she ventured to cast wheat on shallow ocean water, where the clinging sand sank to of the bottom and the wheat could easily be scooped up and eaten. The cleaning of potatoes and wheat quickly spread throughout throughout the troop until virtually all the adopted it. monkeys in the troop had adopted childhood of of primate young is longer than that of of most mammalian mammaUan young, The childhood relation to the life span of of the species. This extended childhood may especially in relation for young primates have much to learn. In the potato-washing incident, be necessary, for the novel behavior behavior was adopted adopted first younger monkeys and only later by their first by younger elders, a reminder that the young tend to be the most imitative, perhaps because because they have so much to learn. This incident also suggests that primates are especially educable, and the longer longer childhood childhood provides ample opportunity opportunity to learn the ways of their group. of of primate behavioral features features are summarized summarized in Table II. The The variety of underlying assumption assumption is that in the line that led to our species there is a progression progression from mammals to social mammals to primates. Species higher in this progression progression from retain features features seen in species lower in the progression but have added behavioral behavioral tendencies. Thus primates possess features features seen in mammals and social mammals, specific to primates. as well as specific 1. /• Development Development

affect personality personality traits. In primates, The focus here is on developmental developmental issues that affect only two developmental developmental eras are necessary: infancy infancy and the juvenile period. period.

Infancy. Like all mammalian mammalian infants, newborn newborn primates primates are completely completely a. Infancy, helpless and would perish without without adult assistance. The mother mother supplies food, food, protection, and love. Infants Infants respond to this maternal care, and a close bond of of affection affection develops.

TABLE n TABLE n Evolution and Behaviorrrraits Behavior/Traits Mammals

Social mammals

Primates

Activity Exploration Fear Aggression

Sociability Enduring attachment attachment Jealousy Dominance Dominance

Manipulativeness Manipulativeness Cognitive curiosity Close imitation Greater educability

CHAPTER CHAPTER 14 14

EVOWnONARY EVOLUTIONARYPERSPEcnvES PERSPECTIVES

351 351

The period of attachment in primates has been divided into three phases (H. F. Harlow, Harlow, & Hansen, 1963). At first the mother is solicitous and completely accepting of the infant, and she is a haven of safety and nurturance. The infant's feeling of of security depends in large part on the mother. If she is sufficiently sufficiently protective and available, the infant will be secure enough to venture out in the wider environment. Primate infants appear to be motivated by two opposing tendencies: the need to seek novelty and stimulation versus the need for security and protection (Mason, 1970). An insecure infant remains close to the A secure infant infant tends to be low in mother, too scared to explore the environment. A fear and can venture away from the mother so long as she is in sight. diminishIn the second phase of attachment, the mother withdraws affection, diminish ing attention to the infant, and starts to punish the infant. The latter may react with withdrawal, anger, resistance, or negativism. These first signs of of independence are typically met by even more irritability irritabihty and punishment by the mother. In the third phase, the mother is often occupied with the birth of the next offspring therefore is even more rejecting of her older child. The presence of offspring and therefore this new infant is likely to elicit jealously and temper tantrums by the displaced sibling. If the mother can spare some attention and affection affection for her older sibHng, sibling, the latter's jealousy and annoyance should gradually wane. of the attachment period may be expected to affect affect personality The events of traits. The mother's behavior should be regarded as only one determinant, albeit an important one, of her youngster's personality. If she is not sufficiently sufficiently protective and a haven of of security, her infant may become fearful fearful and inhibited. If she fails to provide enough attention and social stimulation, her infant may become with withdrawn and less sociable. And if she cannot share at least some affect with her older offspring after after the birth of a new one, the older one may become intensely jealous. offspring In brief, the events of of attachment are assumed to affect the personality traits of fearfulness, impulsivity (the opposite of of inhibition), sociability, and the anger of (jealousy). component of aggressiveness Uealousy). Juvenile Period. Period, Like mammalian young, juvenile primates love to play. b. Juvenile They especially prefer rough-and-tumble rough-and-tumble play fighting, chasing one another, and follow-the-Ieader. foUow-the-leader. While their muscles and coordination are developing, so are bonds of of friendship and alliance. Now peers and unrelated animals can substitute for the mother in offering offering attention and social stimulation. Gender affects the social bonds that develop. Females tend to stay closer to their mothers and are intensely interested in infants. While still in the juvenile period, they start to practice mothering or attempts to do so with available infants. After becoming mothers, they form the nucleus of a group that includes their own offspring and other mothers and their offspring. Thus the sociability of females is enhanced by close interpersonal relationships. Male juveniles tend to separate more from their mothers and are less interested in infants. The social bonds established by males during play are more important, especially the bonds of friendship needed for alliances (Goodall, 1986).

352 352

ARNOLD H. H. Buss BUSS ARNOLD

The other side of of friendship friendship in social interaction interaction is power power relationships. relationships. Domi Dominance is ordinarily achieved through through aggression. It starts with the mock aggression aggression of of juvenile play, and often often becomes real as males struggle for for ascendance. Size and stronger males tend to win fights and become dominant. dominant. strength are crucial. Larger, stronger tend to lose fights and become become submissive. Some males, Smaller, weaker males tend tend to be ferocious, overcoming physical liabilities with sheer determina determinahowever, tend assume leadership through guile. But size and strength strength tion to win fights, and others assume domiusually win. Thus older males tend to dominate juveniles, and males tend to domi females. nate females. Gradually, a dominance dominance hierarchy develops, and each primate primate knows his or Enduring status in the hierarchy diminishes aggression, except except for for in inher place. Enduring stances when older primates become weaker and younger ones attain adult adult size of power struggles is a major major determinant determinant of of and strength. Clearly, the outcome of of aggressiveness and dominance. the traits of 2. Primate Primate Personality Personality Traits Traits

that in this examination examination of the common heritage, humans are being regarded regarded Recall that personaUty traits have been discussed on the as just primates. Issues relevant to personality assumption that that a particular set of of traits is common to primates, especially the great our species. The seven traits listed below have already been mentioned mentioned apes and our divided into two groups. in previous sections. They may be divided The first involves activation, activation, which is defined defined as involving various kinds of of arousal (here defined defined broadly): Activity, the total energy energy output output as observed in rate of movements and 11.. Activity, their vigor Fearfulness, wariness, running away, cowering, and the concomitant concomitant physio physio2. Fearfulness, logical arousal 3. Impulsivity, Impulsivity, acting suddenly and on the spur of of the moment; the opposite is the tendency to inhibit behavior of personality traits are all social: The second set of Sociability, preferring preferring being with others (though primates are a highly social 4. Sociability, differences in sociability within each species) group, there are still individual differences 5. Nurturance, Nurturance, helping others, especially those who need help, even at a cost to the helper helper (altruism) Aggressiveness, attacking attacking or threatening others 6. Aggressiveness, 7. Dominance, Dominance, seeking and maintaining superior status over others versus the opposite pole, submissiveness submissiveness Though aU all four four traits are in the realm of of social behavior, they differ differ markedly. The The first two can be characterized characterized as prosocial in that they involve seeking the presence presence of others (sociabiUty) conflict of (sociability) or helping them (nurturance). The second two involve conflict for power (dominance). As a means of of attaining (aggressiveness) or a struggle for superior status, aggression is closely linked to dominance. However, it is possible superior

CHAPTER CHAPTER 14 14

EVOLUTIONARY EVOLUTIONARYPERSPECflVES PERSPECTIVES

353 353

to be irritable and strike at others without seeking power, just as it is possible to attain superior status vicariously (through association with a dominant individual), distinby trickery, or even by leadership ability. Thus it is worthwhile to distin guish between aggression and dominance, though the two are closely linked in primates. Keen observers of primate behavior have described individual differences in six of of the seven traits just described, the two most prominent observers being Goodall (1986) in the wild and de Waal (1982) in a more restricted environment. Goodall Concerning activity, de Waal (1982) described two extreme chimpanzees. One moved slowly, became fatigued easily, and generally was lethargic. The other was "the moving force behind all developments. His boundless energy and boisterous, provocative effect of of a catalyst" (p. 70). No one would be provocative behavior has had the effect surprised to see such extremes among domestic pets, and marked differences differences in activity have been systematically recorded in domestic cats (Feaver, Mendl, & activity have been systematically recorded in domestic cats (Feaver, Mendl, Bateson, 1986). Bateson, 1986). differences in fearfulness Individual differences fearfulness have repeatedly been documented. Goodall (1986) described a chimpanzee male who went to pieces in tense social Goodall situations, rushing to others for reassurance or clutching his own genitals. De Waal (1982) observed a female chimpanzee who was always the first to raise the alarm and who became so scared that she shook or vomited. . The personality personality trait of of sociability is also manifest manifest in chimpanzees: "Luis is much more sociable than Yeroen. He has an open and friendly character and sets great store by company" (de Waal, 1982, p. 63). And cats vary considerably in gregariousness with each other and with humans (Feaver et al., 1986). differences in nurturance have been largely ignored by observers Individual differences of differences among mothers of animal behavior, but Goodall (1986) reported sharp differences offspring could not obtain. Some were completely in whether they shared nuts their offspring selfish, whereas others shared the food with even their 6-year-old offspring. offspring. differences in these behav behavConcerning aggression and dominance, individual differences iors have been the focus focus of of many observers. No one seriously questions the existence of of these traits in primates. In addition to observations of of individual primates, there have also been system systemof monkeys, complete with factor analyses. Chamove, Eysenck, and atic studies of Harlow (1972) coded the behavior of rhesus monkeys and came up with three factors: fear, hostility (aggressiveness), and affiliativeness affiliativeness (sociability). A A 4-year study by Stevenson-Hinde, Stillwell-Bames, Stillwell-Barnes, and Zung (1980) yielded three bipolar effective, and confi confidimensions: fearful, tense, and subordinate versus aggressive, effective, dent; slow and equable versus active and excitable; and solitary versus sociable. These studies of of monkeys have provided provided evidence evidence for six of of the seven primate of it in the term traits. The only one missing is impulsiveness, though there is a hint of eariier descriptions of the excitable. Notice that impulsivity is also missing from the earlier of individual animals. Most observers of animal behavior tend to ignore behavior of personality traits, and the handful who do observe individual differences differences may not have attended to behaviors indicating impulsiveness. I predict that when observers

354 354

ARNOLD H. H.Buss BUSS ARNOLD

of impulsiveness in primates, they will find it. Meanwhile, its search for evidence of status as a primate trait must be considered as tentative. II. EVOLUTIONARY TRENDS IL Our common heritage includes features we share with mammals, highly social mammals, and primates (see Table II). The evolutionary sequence just mentioned mentioned implies evolutionary evolutionary trends in the features that affect behavior and personality. These trends may be divided into those that occurred during mammalian evolution in the line that led to primates and those that occurred during primate evolution in the line that led to our species.

A. Mammals to Primates Anatomical adaptations are important throughout the animal kingdom, but they Anatomical are especially important among lower mammals. Thus some animals have claws which are anatomically specialized for gripping and ripping. But other animals, primates especially, have hands, which are open up a wide range of of behavioral behavioral adap adaptations. offspring, many have litters. All Though some mammals give birth to a single of/spring, primates give birth to a single offspring, which means that the mother-infant mother-infant bond affectional tie is strengthened by the longer longer childchild is closer to that individual. This affectional hood hood of of primates in relation to total life span, another trend in mammalian evolution. A A long childhood offers youngsters more time to learn the ways of their group and their species. Linked to the stretching out of childhood is a corresponding waning of of innate behavioral tendencies, a shift from instinct to learning. Primates have fewer instincts, and they tend to be more generalized and diffuse than the innate tendencies of lower mammals. And as might be expected, the relationship between hormones and behavior is weaker in primates than in lower mammals. Innate tendencies are crucial for animals that mature quickly (and therefore therefore have little time to learn). But primates care for their young for years, which dimin diminishes the need for having innate tendencies immediately immediately and automatically available for the contingencies of of everyday life. For lower mammals the gap between adaptive needs and behavior that satisfies these needs is filled largely by instinct. For primates, especially the great apes, it is filled largely by learning. As learning became a more prominent mechanism during mammalian evolu evolution, the role of of stimuli changed. No longer mainly releasers of of innate behavioral programs, stimuli came to serve as cues providing information about how to act rather than simply triggering acts. And learning became more complex in a variety of of ways: delays, detours, and chaining of of responses. Greater complexity of learning required a larger brain, and during mammalian evolution the brainlbody brainA)ody ratio increased markedly. A A larger, better brain continued the cycle by making possible

14 CHAPTER 14

355 355

PERSPECTIVES EVOLUTIONARY PERSPEcnvES TABLE III Trends: Mammals Mammals to to Primates Primates Evolutionary Trends:

From From

To To

adaptations Anatomical adaptations childhood Shorter childhood Instincts, releasers Conditioning Small brain Less individuality

Behavioral adaptations adaptations Behavioral Longer Longer childhood childhood Learning, cues cues Cognitive learning Large brain More individuality individuality

even higher forms of learning: learning: learning to learn, learning sets, and matching-to matching-to(H. F. F. Harlow, 1949). Such advanced kinds of learning were, essentially, sample (H. cognitivelearning. learning. cognitive The manipulative ability of primates is the result of another evolutionary trend: trend: the the freeing freeing of of the the hands handsto to hold, hold,carry, carry,or or grasp grasp objects. objects. This Thisability abilitycombined combined with a larger, better brain, which delivered at least primitive cognition, could then lead to tool use. The use of tools, thereby extending what can be done only with the limbs, limbs, is is the the most most conspicuous conspicuous example example of of aa behavioral behavioral adaptation, adaptation, in in contrast contrast the to an an anatomical anatomical adaptation that uses uses only only the the body. These evolutionary evolutionary trends trends are to adaptation that body. These are summarized in in Table Table III. summarized III. As the right-hand column of this table suggests, primates (especially the great apes) are similar to humans in behavioral adaptations. More than lower mammals, the lives of primates are shaped by their everyday experiences, which differ from one animal to the next. Having fewer innate tendencies, a longer childhood, and greater range range of of life Ufe experiences, experiences, primates primates are are more more likely than lower lower mammals mammals aa greater to be different different from one another in behavior. Thus the inevitable outcome of of these evolutionary trends is greater individuality, as reflected in the seven personality traits that humans share with primates. m. DISTINCTIVELY m. DISTINCTIVELY HUMAN TENDENCIES

Most of these evolutionary trends continued during the evolution of humans: humans: the predominance of behavioral adaptations, a longer childhood, more advanced cognicogni tions, a larger and more complete brain, and greater individuality. These trends, especially the psychological psychological adaptation of of aa high level of of cognitive ability, produced a species distinct from all others. Let me offer an example of how a small change, working over time, can produce a qualitative difference. Consider two 5-year-old boys, one of whom learns to read and the other who does not. At first first there will be Uttle difference in their cognitions. But within a few years the nonreader will little difference few have to drop out of school, or if he remains, he will learn little there. He will be shut out of the world of the written word, and his intellectual growth will be

356 356

ARNOLD ARNOLD H. H.Buss Buss

stunted. The reader will be exposed exposed to a wide array of of information information not available to nonreaders, and he will proceed normally into advanced advanced cognitions. The small quantitative quantitative difference difference of of early childhood childhood will inevitably inevitably widen into a gulf gulf large enough for for an observer observer to infer infer qualitative differences differences in cognitive ability and from that of of other information. Similarly, we were set on a path which diverged from difference between between us and them them is sufficiently sufficiently large to primates so that now the difference yield psychological tendencies unique to our species. A. Tools Primates move about mainly on four four limbs, which means that the hands are only partially free for manipulation. The freeing free for freeing of of the hands was completed completed with the evolution of bipedalism in our species. The human hand offers a of offers paradoxical paradoxical exam example of of an anatomical anatomical and behavioral adaptation. It is a fairly fairly typical typical primate primate hand, the only difference difference being an opposable thumb, which which means means the human human hand is anatomically unspecialized. But when we consider consider the potential potential uses of of the hand, especially when a tool is placed in it, we can see that it is essentially a behavioral ad adaptation. The occasional occasional use of of tools by primates primates was inevitable once they possessed the requisite manipulative ability and a brain capable of of integrating a task with a means of of performing performing the task. But two points need need to be made. First, primates primates use tools only occasionally; they do not need tools and can get along well without without them. of the time; we need tools and and cannot We use tools most of cannot survive without without them. Second, after after we marvel at the inventiveness inventiveness of of chimpanzees who fashion fashion sticks to poke into termite mounds mounds or use leaves as sponges, we must must bear bear in mind that these devices are so primitive that they barely qualify as tools. Though barely qualify Though the great great apes possess rudimentary rudimentary imagery and planning ability, these two cognitive processes are far far advanced advanced in our species. When When such cognitions are harnessed harnessed to the superior brain-hand connections superior brain-hand connections of of our species, we can produce produce sophisticated sophisticated tools that multiply our native capabilities: engines to multiply power of multiply the power of muscles, telescopes and radar to amplify the senses, and computers to enhance our amplify computers our cognitions. B. Childhood The second trend that continued was the elongation of of childhood, childhood, of the period of which occupies a portion of of our life life cycle larger larger than than that of any other species. This that of other extension of of childhood has given rise to the concept concept of neoteny: the persistence of of neoteny: persistence of infantile or juvenile characteristics into adulthood. The concept may be appreciated infantile juvenile concept appreciated when a more advanced species is compared with a less advanced advanced species. In the less advanced advanced species, most childish features features disappear disappear in adulthood, and adults look and act entirely different from the young. In more advanced look different from advanced species, childish features are retained. An infant chimpanzee, for instance, has these human-like human-like features infant for features: a small, flat fiat face; a relatively relatively large brain brain and and brain brain case; small teeth; thin nails; and sparse body hair. It more closely resembles resembles a human human than than does an adult

CHAnER CHAPTER 14 14

EVOWnONARY EVOLUTIONARY PERSPECDVES PERSPECTIVES

357 357

chimpanzee. As Gould (1977) has written, "this associated of characters associated complex of characters— neoteny, large brains . . . , slow development, small litters, intense parental care, large body sizesize— must have suggested a look in the mirror" (p. 351). He is more specific about about the sequence of of events: specific major human adaptations adaptations acted acted synergistically throughout their I assume that major gradual development. The interacting system of delayed development-upright development-upright posture-large brain is such a complex: delayed development has produced a large brain by prolonging fetal growth rates and has supplied a set of cranial freed the hand for for tool proportions adapted to upright posture. Upright posture freed use and set selection pressures for an expanded brain, brain. (p. 339) of neoteny may be seen in our MammaTwo psychological examples of our species. Mamma lian and primate primate young are extremely extremely playful playful but but their their playfulness playfulness declines with playful throughout throughout life. Primate young are intensely curious. We maturity. We are playful remain curious all our days.

C. Cognitions started during mamma mammaThe evolutionary trend toward greater cognitive capability started continued during the evolution evolution of lian evolution and continued of humans, and it is cognitive from primates and and all other other animals. capability that most clearly distinguishes us from Thus we can originate complex and abstract abstract concepts, examples of of which are re revealed in science, religion, and the arts. Our Our imagery enables us to envision worlds for we alone are capable of of conceptions such as heaven and hell, that do not exist, for black holes. We are as curious as any animal but but go beyond curiosity ghosts, and black for understanding, so that we seek causes and make attributions. to the search for Adult humans can leave behind the egocentricity egocentricity of of children and animals to view the world world from from perspectives other other than our own. Thus an adult adult may adopt adopt a of his prey, and and a teacher child's perspective, a hunter the perspective of teacher the perspective of a learner. Chimpanzees are capable of of learning sign language (Gardner & Gard Gardof ner, 1969), but it is impoverished compared compared to human adults' language. When teaching is connected connected to language, we can render knowledge cumulative cumulative by passing it on to future other mammals play, future generations. And we make rules; primates and and other but we are the only species to have games, which are distinguished from other distinguished from other play by the presence of (arbitrary) (arbitrary) rules. Our social cognitions are unique. We make social comparisons; for for example, by engaging in downward downward comparison of of others less fortunate fortunate in order order to make ourselves feel better. We derive considerable pleasure from from the achievements of of sharing the joy of of their their successes. And And our personal those close to us, vicariously sharing for example, repression and projection projection as mechanisms of of cognitions are unique, for defense against anxiety.

D. Self Advanced Advanced cognitions also establish establish a potential for for self-consciousness, self-consciousness, which may form in the great apes but is fully fully developed only in humans. occur in rudimentary form

358 358

ARNOLD H. ARNOLD H.Buss Buss

This statement requires elaboration, for research has established estabUshed that primates display behavior that can be interpreted as awareness of self, for example, recogni recognition of of self self in a mirror. When a chimpanzee has a red dot surreptitiously placed on its head and looks in a mirror, it tries to touch the marked area on its head (Gallup, 1970). This finding and related results have led Gallup (1977) to conclude that "the extent that self self recognition implies a rudimentary concept of of self, these data show that contrary to popular opinion and preconceived ideas, man may not have a monopoly on the self-concept. Man may not be evolution's only experiment self-recognition is restricted in self-awareness" (p. 333). As to the question of why self-recognition inabihty to the great apes and humans, Gallup (1977) suggests that *'the "the monkey's inability to recognize himself himself may be due to the absence of of a sufficiently sufficiently well-integrated well-integrated self-concept" (p. 334). Notice the conceptual leap from recognition of self self in a mirror to a well-integrated self-concept. self-concept. of mirror-image recognition, which is absent in Adult humans are capable of infants and develops slowly during the second year of life as part of more general trends in cognitive development. By the age of 2 years most infants possess this capacity (Amsterdam, 1972; 1972; Schulman & Kaplowitz, 1977). Does this mean that of 2 years have a self-concept self-concept and the same kind of self-awareness as older children of children and adults? There are five cognitive attributes present in older children that are absent in 2-year-olds, which suggests that the answer is no. The first is self-esteem. The basis for later self-esteem self-esteem may be laid down in 2year-olds, but children of of this age do not show behavior that allows us to infer the general self-evaluation self-evaluation called self-esteem. This diffuse feeling of self-worth develops gradually and can be measured perhaps by the age of 4 years. Nor are infants clearly aware of of the difference difference between their private feelings and public behavior. It is still too early for the sense of of covertness covertness and an awareness that private thoughts and feelings cannot be observed. Infants and primates lack the sense of covertness that can be inferred in children of of 4 years. Infants are still egocentric and do not know that others view the world from different different perspectives. Even children of several years of age are limited Umited in social perspective-taking. In one study children were asked to select gifts for their parents, teacher, brother, sister, and self self (Flavell, 1968). Most 3-year-olds selected the same gifts for others as for themselves. Some 4-year-olds selected gifts appropriate for others, half half the 5-year-olds did, and all the 6-year-olds did. Social perspective-taking perspective-taking evidently emerges during the fifth year of life. Linked to perspective-taking perspective-taking is the abihty to view oneself oneself as a social object. Such public public self-awareness, as seen in the ability of embarrassment, does not occur until the fifth year of life (Buss, Iscoe, & reaction of Buss, 1979). The last facet of the advanced self self to develop is identity. It may be a personal of being different different from everyone else in appearance, behavior, identity, the sense of oneself to be character, or personal history, or it may be social identity, knowing oneself other group that that offers a a member of a nation, religion, race, vocation, or any other something larger than oneself. oneself. And most of us have a sense sense of belonging to something

14 CHAPTER 14

EVOLUTIONARY PERSPECOVES PERSPECTIVES EVOLUTIONARY

359 359

of continuity, identifying identifying ourselves as the same person across decades of of time or of across diverse social roles. of the self self are absent in 2-year-old 2-year-old human selfThus five aspects of human children: self idenesteem, a sense of covertness, perspective-taking, public self-awareness, and iden advanced or cognitive self, which tity. These may be regarded as evidence for an advanced conspicuously absent in human infants and the great apes. They They do appear appear to is conspicuously self—an awareness of where the body ends and not-me have a primitive, sensory self-an begins, and mirror-image recognition (Buss, 1980). But But they lack lack the advanced self that is implicit in constructs such as self-concept, self-concept, self-esteem, self selfcognitive self consciousness, and identity, constructs easily applied applied to older older human children and adults.

E. Socialization Advanced cognitions also affect affect the way we train our our offspring offspring to function function in the gender and dominance, world. Primates are capable of primitive social roles based based on gender masculine and feminine feminine but human roles are more complex. Every society defines defines masculine complicated rules involving status; when property property is involved, social roles, and has complicated And every every society develops religion classes based on ownership are sure to follow. And and a code of morality. On the basis of of a survey of literature, Dubin Dubin and and Dubin Dubin (1963) found found that these minimal goals were common to most cultures: 1. 1. 2. 3. 4.

of food food Control of elimination and of grabbing of of thumb sucking and masturbation masturbation Control of Learning how and what can be touched touched Learning how to relate to others, including control of of aggression

Another Another survey, this one of the files on 50 different different countries, revealed revealed that starting in the 5- to 7-year period, children are inculcated inculcated in the traditions, beliefs, of their society (Rogoff, Sellers, Perrotta, Fox, & White, 1975). and values of 1975). They for adulthood by practicing must adhere to the rules and prepare for practicing roles they will for divergent assume later in life. Boys and girls tend to be separated and trained for sex roles. Distinctiveness F. Human Distinctiveness

These singular human tendencies are summarized summarized in Table IV. It is not much of a stretch stretch to conclude that superior cognitive ability is the engine of of all the attributes in the table. This ability has set us on a path different different from from all other animals, culminating in two revolutions in tools: substituting substituting machines for for muscle and substi substituting machines (computers) for for some of of our cognitive functions. functions. Our advanced of two evolutionary trends advanced intelligence rests on the continuation continuation of mentioned earlier. The first is the trend toward a larger brain in relation to body size. The ratio of of brain to body size, already large in primates, continued to increase,

360 360

ARNOLD ARNOLD H. H. Buss Buss

TABLE IV IV Distinctively Human Tendencies Tools Cognitions

Self Socialization

Sophisticated and necessary Abstract concepts; concepts; search for understanding; sociocentric perspective; advanced language; language; rules Self-esteem; sense of covertness; public self-awareness; identity Kinship; Kinship; advanced social rules; rules; culture; culture; religion; morality

ours being four times that Jerison, 1976). The second trend was an tliat of primates ((Jerison, increase in cerebral cortex and frontal frontal lobes, anatomical adaptations that underlie underUe our advanced cognitions. These evolutionary traits have strongly shaped human personality. personaUty. A longer childhood, strong socialization pressure, true social roles, and a variety of work and social contexts combine to widen the environments in which humans exist. Advanced cognitions and language extend the human response repertoire. These distinctively human features have produced two effects. First, the traits shared with primates have been elaborated in ways to be described shortly. Second, there are uniquely human personality personaUty traits. G. G. Elaborated Elaborated Primate Primate Traits Traits

activity continued to consist of tempo and vigor of response, but now In humans activity manifested in the rate and loudness of speech. And the complexity of human it also manifested environments has added contexts in which the expenditure of energy may be seen. Thus physical activity occurs in a variety of sports and types of exercise. The current fitness fad may appeal to those high in activity but not those at the other end ("When I have the urge to exercise, I lie down until it passes"). passes")- Many factory factory and farm jobs require considerable vigor, and the assembly line demands a speedy rate of response. Just living living in in aacity city like like New New York York may may require require aa fast fast pace pace of of responding. responding. The major major addition to fear is cognitions. As far far as we know, we are the only species that worries. The down side of our ability to imagine the future future is a fear fear of dying. We can teach great apes to use sign language, but would they then be subject to existential anxiety? The range of of possible apprehensions has been opened up in the modem world: fear flying, fear fear of offlying, fear of a nuclear nuclear attack, and fear fear of c�mtracting cpntracting AIDS, to name just a few. And our advanced cognitions have probably contributed to two problems of adjustment adjustment unique to our species: panic and agoraphobia. The trait of impulsivity has been strongly affected affected by socialization practices. All societies require some inhibition of behavior, and some push hard for of for delay of gratification. As a result, the trait dimension of impulsivity has been extended at the inhibition end. Impulsivity has also been elaborated by the human propensity for planning and by the need to make decisions, which means that the trait is much more complex in our species.

CHAPTER CHAPTER 14 14

EVOLUTIONARY EVOLUTIONARYPERSPEcnVES PERSPECTIVES

361 361

For those who are sociable sociable there is an added human outlet of formal groups of of people: work, sports, clubs, politics, or organized spectator activities. The low end of of sociability sociability is extended extended by the possibility of privacy, which is available to humans but not to primates. But because humans may become isolated, the issue of of loneliness now becomes important, especially for those high in sociability. To the trait of of nurturance is added the imagery that allows us to put ourselves in the place of of another person (empathy), which facilitates nurturant behavior (Davis, 1983). Socialization practices strongly push the helping of of others. More than any other animal, we are likely to save women and children first and to render aid to the old and infirm. Furthermore, there are charitable organizations, religious orders, and occupations that by their nature enhance nurturance tendencies or provide an outlet for them (e.g., medicine, nursing, and child care). Aggressiveness Aggressiveness is expanded by the availability availabiUty of of language. We are the only species to curse, derogate, maliciously tease, verbally threaten, and spread nasty of advanced cognitions and language enables us to make gossip. The combination of negative attributions to others, engage in prejudice, and generally to hate others. Our advanced tools also allow us to inflict greater harm on others. dominance is elaborated in several ways. The presence of organized The trait of dominance competition in our species allows an individual to attain dominance by means of competition—politics, for example. Closely allied to competitiveness competitiveness success sin competition-politics, of achievement (McClelland, Atkinson, Clark, & Lowell, 1953), power are the traits of Machiavellianism (Christie & Geis, 1970), all of which are (Winter, 1973), and Machiavellianism Unked to dominance. More than any other species, humans can attain dominance linked by means of of nonaggressive leadership through personality traits (charisma) or quali qualities the group seeks in a leader. At the opposite end of the dominance dimension, submissiveness is also deference not just to a leader but elaborated. Submissive humans have the outlet of deference to generalized authority: the trait of of authoritarianism (Adorno, (Adomo, Frenkel-Brunswik, Frenkel-Brunswik, Levinson, & Sanford, 1950). In brief, evolutionary trends in the line that led to our species have resulted in the elaboration of of primate personality traits. Specifically, three human features features are important here. Cognitions and tools have added to our response repertoire, socialization socialization has emphasized one or another end of of trait dimensions, and a broader range of of environments has opened up possibilities for personality personality traits that are not present in other species. How important are these seven traits in personality research and theory? In one form or another they are listed in virtually all classifications classifications of of personality (see Buss & Finn, 1987). Eysenck's (1970) superfactor of extraversion extraversion consists mainly of of sociability and impulsivity, though activity also appears as a minor aspect. His neuroticism neuroticism appears to be mainly fearfulness. The four social traits have been been emphasized in one way or another in circumplex models of of interpersonal behavior. To cite just one example, Kiesler (1983) uses two axes: dominant versus submissive and sociable versus aggressive.

362 362

ARNOLD ARNOLD H. H. Buss Buss

H. H. Uniquely Human Traits Traits

1. Self Self As the only species with advanced, cognitive self-consciousness, humans are oneself unique in possessing traits that concern the reactions to or awareness of oneself Private self-consciousness self-consciousnessrefers referstotoa afocus focuson on (Fenigstein, Scheier, & Buss, 1975). Private as may be seen in a typical item: "I'm "I'm generally generally the nonobservable aspects of oneself, as attentive to my inner feelings." Public self-consciousness self-consciousnessinvolves involvesoneself oneselfasasaasocial social object and being concerned with appearance or behavior that can be observed by self-consciousnessinvolves involves others: "I'm self-conscious about the way I look." Body Body self-consciousness sensations: "I can often feel my heart beating" (Miller, a focus on internal bodily sensations: Murphy, & Buss, 1981). anxiety, which is related to acute public self-consciousness self-consciousness as well as Social anxiety, to fear, has been assessed as a global trait (Fenigstein et aI., al., 1975). But two compo component traits have been isolated. One is speech anxiety: "I always avoid speaking in 1984). The other trait is shyness, which also public if possible" (Slivken & Buss, 1984). fear and public self-consciousness self-consciousness (with correlations appears to be a combination of fear of .50 and .26, .26, respectively; Cheek & Buss, Buss, 1981). Self-esteem is well known and apparently such a generalized, pervasive trait Self-esteem that all measures of it intercorrelate. Finally, humans are capable of putting on an act to please other people, behavior that has been called self-presentation. self-presentation.The Thebest best self-report self-report measure is called ability to modify modify self-presentation: "I *'I have the ability to control how I come across to people, depending on the impression I want to give them" (Lennox & Wolfe, 1984). 2. 2. Other Cognitive Traits Humans make attributions, but there are individual differences of differences in the kind of attributions made and the extent to which people make them. The most prominent attributional trait is locus of of control control(Rotter, (Rotter, 1960), 1960),which whichconsists consistsof ofseveral severalcompo compo(Deci && nents (Collins, 1974). A A more more recent recent trait, trait, called calledgeneral generalcausality causalityorientation orientation (Deci Ryan, 1985), involves whether people seek autonomy or control or whether they feel controlled. Individual differences differences in imagery are tapped by two related traits. Absorption Absorption is the tendency to have intense subjective experiences: **If "If I wish, I can imagine (or daydream) some things so vividly that they hold my attention as a good movie or story does" (Tellegen & Atkinson, 1974). of 1974). Openness Openness to experience experience consists of being interested in novelty, the arts, and, especially, one's fantasies and feelings feelings (McCrae & Costa, 1983). There are also several cognitive styles which deal with the way information information 1978, for a review). The best known is processed (see Goldstein & Blackman, 1978, trait is field dependence-independence, dependence-independence, also known as psychological differentiation differentiation (Witkin, Dyk, Faterson, Goodenough, & Karp, 1962).

CHAPTER CHAPTER 14 14

EVOLU11 0NARY PERSPEcnVES EVOLUTIONARY PERSPECTIVES

363

IV. THREE PERSPECTIVES IV.

Thefirst first perspective deals with the animal heritage of human personality. It starts w^ith this this question: question: What What ifif we we are are regarded regarded as as nothing nothing but but primates? primates? If If so, so, we we can can with examine personality relevant features that primates share with other animals. Thus primates have arousal and and relationship emotions and and a long childhood, childhood, and and they theytend tend to to be be cognitively curious, manipulative, and and especially educable. educable. We We can can examine examine the the events of infancy, especially attachment, and the long juvenile period and see how they affect personality. The focus on our primate heritage culminates with a list of of seven personality traits present in primates and humans. The second perspective alerts us to evolutionary trends in the line that led to our species. There are the broad trends of mammalian evolution toward behavioral adaptations, longer childhood, learning, and the beginning of cognitions. A crucial trend concerns concerns the the innate, innate, species-wide species-wide behavioral behavioral tendencies tendencies called called instincts. instincts. Intrend In stincts wane wane as as aa determinant determinant of of behavior behavior and and so so do do hormones. hormones. Instinctive Instinctive behavior behavior stincts persists in in humans, humans, but but instinct instinct is is often often overridden overridden by Hormones can persists by learning. learning. Hormones can affect human human behavior-the behavior—the sexual sexual motivation motivation of of women, for instance. But unlike affect women, for instance. But unlike other mammalian mammaUan females, females, women women do do not not come come into into heat, their sexual sexual behavior behavior other heat, and and their is determined determined less less by by hormonal hormonal changes changes than than by by socialization, and cogniis socialization, learning, learning, and cogni tion. These These biological biological determinants determinants still still affect affect our our behavior behavior but but much much less less so so than tion. in other other animals. in animals. from There are also narrower trends that occurred in the evolution of humans from ancestral primates such as advanced cognitions, which led to tool use and its many consequences, especially true socialization. Infants do not possess the advanced significognitions they will later have, and of course they have not been exposed to signifi cant socialization. It follows that with respect to personality, human infants should be regarded as nothing more than primates. The only personality traits seen in infants are are some some of of those those we we share share with with primates. primates. infants differThese trends lead directly to the third perspective: humans as distinctly differ ent. There are novel personality traits in humans not seen in other species: species: self selfrelated traits and cognitive traits. traits. These uniquely human traits differ in two ways cogni from the traits we share with primates. First, they derive from our advanced cognitions and socialization practices. Second, all the primate traits are known to be inherited in humans (see Buss, 1988, for a review), but the inheritance of the uniquely human traits has yet to be established. The evolutionary perspective also shows how the traits humans share with primates become more nlore elaborate in our species. We have a wider range of response options, especially those of cognitive nature. Thus humans have a larger repertoire differof aggressive and dominant behavior, which means that these traits are more differ entiated in our species. And we live in a wider range of social environments: work, home, school, and play. Consequently, an active child might be energetic at home or on the playground but relatively still in the classroom. As this example illustrates, the traits we we share share with with primates primates are are likely likely to to show show less less consistency consistency in in humans. the traits humans.

364 364

ARNOLD ARNOLD H. H.Buss Buss

A final comment concerns the practices of of those who study personality. The The A seven traits we share with primates have been the focus of interest by personality focus of personality psychologists for for many decades. As mentioned mentioned earlier, these traits are found found in virtually all classifications uniquely human classifications of of personality traits. In contrast, the uniquely human traits, with the exception of of self-esteem, have received less attention. And these distinctly human traits tend to be omitted from from most classifications classifications of of personality. Why have the seven traits we share with primates been examined examined more? Perhaps they are more observable. Or perhaps these traits are regarded regarded as more important because, as derivatives of they are more' closely linked to of our evolutionary past, they adaptive needs in the line that led to our species.

REFERENCES Adorno, T. W., W.,Frenkel-Brunswik, Frenkel-Brunswik, E., E.,Levinson, Levinson,D. D.J., J.,&&Sanford, Sanford,R. R.N. N.(1950). (1950).Theauthoritar The authoritarian personality. personality.New NewYork: York:Praeger. Praeger. ian B. (1972). (1972). Mirror self-image self-image reactions before the age of Developmental Amsterdam, B. of two. Developmental Psychology, 5,5,297-305. 297-305. Psychology, Buss, A. H. (1980). (1980). Self-consciousness Self-consciousnessand andsocial socialanxiety. anxiety,San SanFrancisco: Francisco:Freeman. Freeman. Buss, A. H. (1988). Personality: Personality:Evolutionary Evolutionaryheritage heritage human distinctiveness. Hillsdale, andand human distinctiveness. Hillsdale, NJ: Erlbaum. Buss, A. H., & Finn, S. E. (1987). Classification Classification of personality traits. Journal of of Personality Personality Social Psychology, Psychology,52, 52,432-444. 432-444. and Social Buss, E. (1979). (1979). The development of embarrassment. Journal of of Buss, A. H., Iscoe, I., & Buss, Psychology, 103, 103,227-230. 227-230. Psychology, Butler, R. A. (1957). The effect effect of deprivation of visual incentives on visual exploration in monkeys. Journal Psychology, 50, 50, 177-179. Journalof of Comparative Comparativeand andPhysiological Physiological Psychology, 177-179. Cannon, W. B. (1927). Bodily changes fear and rage. New York: Appleton. changesininpain, pain,hunger, hunger, fear and rage. New York: Appleton. Chamove, A. S., Eysenck, H. J., & Harlow, H. F. (1972). Personality Personality in monkeys: Factor Psychology, analyses of rhesus social: social: Behavior. Quarterly Quarterly Journal JournalofofExperimental Experimental Psychology, 24, 24, 496-504. Cheek, J. M., & Buss, A. H. (1981). Shyness and sociability. Journal of of Personality Personalityand andSocial Social Psychology, Psychology, 41, 41,330-339. 330-339. Machiavellianism. New Christie, R., & Geis, F. L. (Eds.). (1970). (1970). Studies in Machiavellianism. NewYork: York:Aca Academic Press. Collins, B. B. E. (1974). Four components of the Rotter Internal-External Internal-External scale: Belief in a difficult world, a just world, a predictable world, and a politically responsive world. difficult Journalof ofPersonality Personalityand andSocial Social Psychology, 381-391. Journal Psychology, 29,29, 381-391. differences in empathy: Evidence for a multidimen multidimenDavis, H. (1983). Measuring individual differences Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, 113-136. 44,44, 113-136. sional approach. Journal &. Ryan, Ryan,R. R. M. M. (1985). (1985). Intrinsic Intrinsicmotivation motivationand andself-determination self-determination in human Deci, E. L., & in human behavior. New York: Plenum Press. behavior. de Waal, F. (1982). Chimpanzee Chimpanzee politics. politics.New NewYork: York:Harper Harper&&Row. Row. Dubin, E. R., & Dubin, R. (1963). (1963). The authority inception period in socialization. Child Child Development,34, 34,885-898. 885-898. Development, structure of of human humanpersonality personality(Rev. (Rev. ed.). ed.).London: London: Methuen. Methuen. Eysenck, H. J. (1970). The structure

CHAPTER 14 14 CHAFfER

36S 365

EVOLUTIONARY PERSPECTIVES EvoLUnONARY

Feaver, J., Mendl, M., & Bateson, P. (1986). A method for rating individual distinctiveness distinctiveness in cats. Animal Animal Behaviour, 34, 1016-1025. Fenigstein, A, A., Scheier, M. F., & Buss, A A. H. (1975). Public and private self-consciousness: self-consciousness: Assessment and theory. Journal Journal of of Consulting Consulting and Clinical Clinical Psychology, Psychology, 42, 523-527. Flavell, J. (1968). The development development of of role-taking role-taking and and communication communication skills skills in children. New York: Wiley. Gallup, G. G., Jr. (1970). Chimpanzee: Self-recognition. Self-recognition. Science, 167, 86-87. Gallup, G. G., Jr. (1977). Self-recognition in primates: A comparative approach to bi bidirectional properties of of consciousness. American American Psychologist, Psychologist, 32, 329-338. Gardner, R R. A., & Gardner, B. T. (1969). Teaching sign language to a chimpanzee. Science, 165, 664-672. Goldstein, K. M., & Blackman, S. (1978). Cognitive Cognitive style. New York: Wiley (Interscience). Goodall, J. (1986). The chimpanzees Gombe. Cambridge, MA: Harvard University chimpanzees of ofGombe. LFniversity Press. Gould, S. J. (1977). Ontogeny Ontogeny and and phylogeny. phylogeny. Cambridge, MA: Harvard University Press. Harlow, H. F. (1949). The formation formation of learning sets. Psychological Psychological Review, Review, 56, 51-65. Harlow, H. F., & Harlow, M. K. (1962). Social deprivation in monkeys. Scientific American, American, 07(5) , 136-146. 2207(5), Harlow, H. F., Harlow, Hariow, M. K., & Hansen, E. W. (1963). The maternal affectional affectional system of monkeys. In H. L. Rheingold (Ed.), Maternal behavior behavior in mammals mammals (pp. 254-281). New York: Wiley. Jerison, H. (1976). Paleoneurology and the evolution of mind. Scientific American, American, 234, 90-101. Kiesler, D. J. (1983). The 1982 interpersonal circle: A taxonomy for complementarily complementarity in human transactions. Psychological Psychological Review, Review, 90, 90, 185-214. Lacey, J. 1. I. (1956). The evaluation of of autonomic responses: Toward a general solution. Annals Annals of of the New New York York Academy Academy of of Sciences, 67, 123-163. Lennox, R R. D., & Wolfe, R R. N. (1984). Revision of of the Self-Monitoring Self-Monitoring Scale. Journal Journal of of Personality Personality and Social Social Psychology, Psychology, 46, 1349-1364. Lynn, R R. (1966). Attention, Attention, arousal arousal and and the orientation orientation reaction. Oxford, England; Perga Pergamon Press. Mason, W. A A. (1970). Motivational factors in psychosocial development. In U. J. Arnold & M. M. Page (Eds), Nebraska Nebraska Symposium Symposium on Motivation Motivation (pp. 335-364). Lincoln: Lincoln: Univer University of of Nebraska Press. Masters, W. H., & Johnson, V. E. (1966). Human Human sexual sexual response. Boston: Little, Brown. McClelland, S. c., C , Atkinson, J. W., Clark, R R. A., & Lowell, E. L. (1953). The achievement achievement motive. Englewood Cliffs, NJ: Prentice-Hall. McCrae, R R. R, R., & Costa, P. T., Jr. (1983). Joint factors in self-reports and ratings: Neuroticism, extraversion, and openness to experience. Personality Personality and and Individual Individual Differences, 4, 245-255. C , Murphy, R, R., & Buss, A. H. (1981). Consciousness of body: Private and public. Miller, L. C., Journal of of Personality Personality and and Social Social Psychology, Psychology, 41 41, 397-406. Journal Rogoff, B., Sellers, J., Perrotta, S., Fox, N., & White, S. (1975). Age of of assignment assignment of of roles and responsibilities 8, 353-369. responsibiUties in children. Human Development, Development, 118, for internal internal versus external external control of reinforce reinforceRotter, J. B. (1960). Generalized expectancies for ment. Psychological Psychological Monographs, Monographs, 80(Sehal 80(Serial No. 609). Schulman, A. H., & Kaplowitz, L. (1977). Mirror-image response during the first two years of life. Developmental Developmental Psychology, Psychology, 10, 133-142. of ,

366

ARNOLD ARNOLDH. H.Buss Buss

Slivken, K. E., & & Buss, Buss,A A.H. H.(1984). (1984).Misattributions Misattributionsand andspeech speechanxiety. anxiety.Journal JournalofofPersonal Personality and Social Social Psychology, Psychology,47, 47,396-402. 396-402. Stevenson-Hinde, J., J., Stillwell-Barnes, Stillwell-Bames, R, R., & &Zung, Zung,M. M.(1980). (1980).Subjective Subjectiveassessment assessmentof ofrhesus rhesus monkeys over four successive years. Primates, Primates,21, 21,66-82. 66-82. A, & Tellegen, A., & Atkinson, G. (1974). Openness to absorbing and self-altering experiences ("absorption"), a trait trait related related to to hypnotic hypnotic susceptibility. susceptibility.Journal Journalof ofAbnormal AbnormalPsychol Psychology, 43, 43, 111-122. 1 1 1-122. Wilson, E. O. (1975). Sociobiology. Sociobiology, Cambridge, Cambridge,MA: MA:Harvard HarvardUniversity UniversityPress. Press. Winter, D. G. (1973). The The power power motive. motive.New New York: York:Free Free Press. Press. R, & Witkin, H. A., Dyk, R R. B., Faterson, H. F., Goodenough, D. R., & Karp, S. A A. (1962). Psychological New York: Wiley. Psychologicaldifferentiation. differentiation. New York: Wiley.

CHAPTER 15 CHAPTER 15

GENETICS, TEMPERAMENT, AND PERSONALITY AND DAVID C. C. ROWE UNIVERSITY UNIVERSITY OF OF ARIZONA ARIZONA

One founder of the theory of evolution, Charles Darwin, theorized that emotions and their accompanying facial expressions had a natural origin. In his book The The Expressionsof ofEmotions EmotionsininMan Manand andAnimals Animals(187211965), (1872/1965),ofofthe theemotions emotionsofofjoy joy Expressions and pleasure, pleasure, he he observed observed that that "with "with all all the the races races of of man man the the expression expression of of good good and spirit appears appears to to be be the the same, same, and and is is easily easily recognized" recognized" (p. (p. 211). 211). He He further further analyzed analyzed spirit the specific specific facial facial features features involved: involved:". when in in good good spirits spirits.. .. .. the the eyes eyes sparkle, the ". .. .. when sparkle, with the the skin skin aa little little wrinkled wrinkled round round and and under under them, them, and and the the mouth mouth aa little little drawn with drawn back at at the the corners" corners" (p. (p. 16). 16). Darwin Darwin queried queried "missionaries "missionaries and and protectors" protectors" of back of native peoples peoples around around the the world world about about the the facial facial expressions expressions showing showing emotions. emotions. native The descriptions descriptions they they returned returned matched matched his his own own observations observations of of western westernEuropeans. Europeans. The From the the expression expression of of emotion's emotion's universality universality in in all all races, races, its its early early appearance appearance in From infancy, and and its its similarity similarity in in many many species species (including (including primates primates and and dogs), dogs), Darwin Darwin infancy, decided that that the the emotions emotions were were biologically biologically inherited. inherited. According According to to his his description description decided of the the young young woman woman pictured pictured in in Figure Figure 1, she was was tearing tearing up up aa photograph photograph of of 1, she despised lover. lover. Who Who would would mistake mistake her her disdain disdain for for the the feelings feelings of of fear, fear, joy, joy, aa despised or surprise? surprise? or In cross-cultural field studies, Eibl-Eibesfeldt Eibl-Eibesfeldt (1975) collected evidence on the universality of emotional expression more objective than Darwin's. He toured diverse, nonwestern societies with a special camera, onefitted fitted with a false forward looking lens and a hidden, side viewing one. From the latter lens, he took motion pictures of of people people who who were were unaware unaware of of the the camera's camera's eye. eye. As As had had Darwin, Darwin, Eiblpictures Eibl Eibesfeldt concluded concluded that that emotions emotions and and their their accompanying accompanying facial facial expressions expressions were Eibesfeldt were HANDBOOK HANDBOOK OF OF PERSONAUTY PERSONAUTY PSYCHOLOGY PSYCHOLOGY

367 367

CoPYRIGHT COPYRIGHT C © 1997 1997 BY BY ACADEMIC ACADEMIC PRESS PRESS.. Au RIGJrni OF REPRODUCOON IN ANY FORM RESERVED.

ALL RIGKTS OF REPRODUCTION IN ANY FORM RESERVED.

368 368

DAVID DAVID C. C. ROWE ROWE

FIGURE 11 A A woman woman showing showing disdain disdain from from Darwin's Darwin*s The The Expressions Expressions of of Emotions Emotions FlGUO

in Man Man and and Animals. Animals. in

human universals based upon biological instincts. For instance, in in all human groups, groups, gestures: first aa smile at at her partner and and aa brief a flirting girl made similar facial gestures: lift of the eyebrows, and then a turning away from the person of her affection. She may cover her face with a hand and may laugh or smile in embarrassment. Thus everywhere people can communicate in universal "language" "language" of emotional ex expression. This chapter examines the genetic bases of temperament and personality. Undoubtedly, our ability to infer subtle emotions from facial appearance, to use facial expressions to deceive others, and to use them to conform to social norms is more elaborate than in other animal species. Nonetheless, it is good to consider in western Europe the above evidence on the biological bases of emotion because in and the United States a cultural belief that all human behavior is mainly shaped by culture and experience prevails. This chapter first discusses methods used in the field of behavior genetics for the estimation of genetic and environmental components of personality variation (Plomin, 1990). Second, it surveys evidence of genetic variation in temperamental and personality traits. The third section of the chapter considers several special topics, namely, shared environmental influence on personality variation, genetic variation in environmental measures, and a theory of personality development.

CHAPTER CHAPTER 15 15

GENETICS, GENETICS, TEMPERAMENt, TEMPERAMENT, AND AND PERSONALITY PERSONALnr

369 369

I. SEPARATING SEPARATING NATURE AND NURTURE The idea of separating nature and nurture is sometimes greeted with deep skepticism by the lay-public and by social scientists alike. The social critics of behavior genetic results may ask, "How *'How can a scientist separate nature and nurture, when people need both genes (in the DNA from from parents) and an environment environment in which to develop physically and psychologically?" These concerns are understandable, but they are also misdirected. Although always a part of behavioral development, genetic and environmental influences are separable as components of trait variation. In the context of trait variation, nature can be clearly more influential than nurture, and vice versa. For instance, in spoken accents, variation may depend on the environmental factor factor of geographic locale, whereas in physical height or weight, variation may depend on which genes were inherited from one's parents. That is, as established by research, most physical traits possess a strong genetic component in their variation. In this regard, the reader may consider the striking physical resemblance of identical (also called monozygotic, MZ) twins, whether they were reared apart or or together (Bouchard, (Bouchard, Lykken, McGue, McGue, What degree degree of of genetic geneticinfluence influence exists exists in in behavioral behavioral trait trait Segal, & Tellegen, 1990). What careful empirical studies variation is thus a scientific question, to be settled through careful differences. of individual differences.

A. Heritability Heritability A. The heritability coefficient of coefficient is a numerical estimator of the genetic component of trait variation. The equation that gives its mathematical definition is h2 _ Genetic variation in a trait . '' entabIllty h heritability = /i^ = . . . . . , Ph enotyplc vanatlon Phenotypic variation 10 in a traIt trait

where "phenotypic" is the measured trait variation across individuals and "genetic variation" is that part attributable If attributable to the substitution of genes for for one another. If all variation in a trait were attributable to genetic variation, then the heritability coefficient 100%. At the other extreme, when it takes a value near coefficient would be 1.0, or 100%. zero, gene substitutions would lack effects effects upon trait variation (e.g., as possibly illustrated illustrated by speech accents). A true example corresponding to the former former case-nearly case—nearly 100% 100% heritabil heritability-is ity—is a genetic disease named Huntington's chorea. This disease is caused by a single gene mutation that is dominantly inherited. When only one copy of of the defective gene is received, all affected of affected people (who live long enough) will die of progressive neuronal degeneration. When two normal genes are inherited, no one will die of of the disease. Of Of course, except for such devastating single-gene diseases, human traits typically lack heritabilities near 1.0. Bouchard et al's al.'s (1990) study of of separated separated MZ twins allows the estimation of heritabilities of several physical and from highest to lowest: psychological traits. Consider these heritabilities ordered from (IQ), .69; .69; systolic blood pressure, fingerprint ridge count, .97; .97; intellectual ability (10),

370 370

DAVID C. CRowE DAVlD RoWE

.64; heart rate, .49; .48; and Jackson .64; .49; California Psychological Inventory Traits, .48; Vocational Interests, .43. .43.^1 Except for fingerprint ridge counts (i.e., a quantitative measure of the fingerprint), no heritability coefficient coefficient approached 1.0. of heritability can take one of of two meanings. The first meaning is Estimates of narrow-sense heritability. It represents genetic effects transmissible in families. A narrow-sense A parent-child correlation on a trait is calculated as one-half one-half the narrow-sense herita heritaparent-child trait correlation would be bility (e.g., if the former were .30, then the parent-child .15). The narrow-sense heritability has one limitation, however; it fails to account for possible interactions among genes, which may be located separately in the DNA. Epistasis Epistasis is the interaction of of genes on different different chromosomes, whereas dominance dominance is an interaction of genes at a single genetic locus. Broad-sense Broad-sense heritability removes this deficiency; it is used to estimate all genetic variation contributing to trait variation, including both types of of gene-gene gene-gene understanding (Lykken, McGue, interactions. An analogy may aid the reader's understanding Tellegen, & Bouchard, 1992). Suppose that we regard parents as giving their children cards from separate decks. If the father gives a daughter a 10 and a 9, whereas her mother gives her a 7, a 5, and a Queen, then the girl has the cards with which to hand—the high card would be a Queen. But suppose construct an unexciting poker hand-the now that the father gives his daughter an Ace and a King, whereas her mother provides a Queen, a Jack, and a 10. With these cards, the daughter would have the of makings of an extraordinary poker hand—a hand-a royal flush. Specific combinations of transmissigenes may thus construct rare and extreme traits (ones that would not be transmissi combination of genes ble from parent to child, because, except very rarely, this combination al. named traits in which the gene-gene interaction would fail to reoccur). Lykken et a1. component is large "emergenic" traits. They observed that combinations of rare of "a Ramanujan genes (i.e., "cards") might produce a child who achieves the status of record—or a True Crime miniseries for [a mathematical genius], a new Olympic record-or television" (p. 1575). of h^ not always always identified, identified, h2 isis not Although in behavior genetic studies the meaning of study research the reader should be aware that estimates derived from the twin study design—either MZ twins raised apart or MZ and fraternal (also called dizygotic, design-either DZ) twins compared-would compared—would be heritabilities in the broad sense. Because MZ fertilized egg cell) the exact twins share (as they originate biologically from one fertilized

Becauseofofsampling samplingvariation, variation,heritability heritabilityestimates estimatesare areinexact. inexact.Falconer's Falconer's(1981) (1981)formula formulafor for 1^Because T^)V(N pairs paurs - 1)]°·5. 1)]^'^.The Thestandard standarderrors errorswere were the standard error error of of the correlation coefficient coefficient is is [(1 [(1 - r)2/(N the standard the correlation .01;systolic systolicblood bloodpressure, pressure,.08; .08;heart heartrate, rate,.11; .11; WAIS full scale IQ, .08; California fingerprintridge ridgecount, count,.01; fingerprint WAIS full scale IQ, .08; California andJackson JacksonVocational VocationalInterest InterestSurvey, Survey,.12. .12.Because Becauseof ofsampling samplingvariation variation Psychological Inventory, Psychological Inventory, .13; and and different different selections selections of of local local genotypes genotypes (Le., (i.e., peoples' peoples' genetic and local local environmental environmental and genetic compositions) compositions) and conditions, heritability heritability coefficients coefficients may may thus thus vary vary from one study study to to another. another. This This variation, variation, however, however, conditions, from one does not not mean mean that that heritability heritability coefficients coefficients are are nongeneralizable nongeneralizable from from one one population population to to another. another. Once does Once sampling variation has been excluded, heritability heritability coefficients may be generalizable generalizable across across social social class class variation has been excluded, coefficients may sampling contexts and other other environmental environmental clines-the clines—the degree degree of of generalizability generalizability is is an an empirical empirical issue issue that that can contexts and can be addressed addressed with with behavior behavior genetic (Rowe & & Waldman, Waldman, 1993). genetic research research designs designs (Rowe -

CHAPTER CHAPTER 15 15

GENETICS, TEMPERAMENT, AND PERSONAUTY PERSONALITY GENETIcs,

371 371

same combination of genes, their trait resemblance contains all genetic variation, attributed to the main effect effect of genes, as well as that which can that which can be attributed be attributed to the gene-gene interactions. Often, heritabilities from twin studies from other research designs, and epistatic effects effects are greater than those estimated from may be part of this.

B. Environmental Environmental Variation Variation B. found to be less than 1.0, As heritabilities for most personality traits have been found and often they were considerably less than this limit, environmental variation must also contribute to trait variation among individuals. In behavior genetic research designs, this environmental variation is also subject to an analytic decomposition. nonshared The two main subcomponents of environmental variation are called the nonshared and shared components. By definition, nonshared environmental variation refers to all environmental effects acting uniquely on individuals and not contributing to psychological resem resemeffects blance among family brother and sister). sister). To Tocalculate calculate family members (e.g., the alikeness of brother from the this component properly, measurement error variation must be removed from environmental component component of variation. Nonshared environmental environmental effects effects can make environmental genetically matched MZ twins dissimilar in behavior. The shared component of variation refers, by definition, to environmental effects effects correlated across family members and which operate to make them psycho psychologically alike. For instance, social class level is correlated across siblings and may make them alike in a psychological outcome (e.g., risk-taking). In summary, trait variation can be separated into three main variance compo components: (1) heredity, (2) shared environment, and (3) nonshared environment. The strength of each component, as a proportion of total trait variation, is represented by a proportion that can take any value between 0 and 1. The heritability coefficient coefficient is symbolized as h2• h^' the shared environment component as 2· c^- and the nonshared environment component as e2 e^ or u2 u^ (the latter being used to distinguish it from from measurement error, which is also sometimes symbolized as e^). e2). C Behavior Behavior Genetic Genetic Research Research Designs Designs c. All classic behavior genetic research designs were first used in the late 19th or early 20th centuries. They may involve the comparison of adoptive children with their biological or adoptive parents, the comparison of MZ twins raised apart, or the comparison of MZ and DZ twins.

/. Adoption Adoption Studies 1. In the adoption research design, trait variation is apportioned through the compari comparison of children raised apart from their biological parents. Adoptive children are different in social circumstances from from placed in adoptive families that are often different those of their biological parents. Hence, a type of "experiment" is created in which

372 372

DAVID C. ROWE DAVID CRowE

the genetic similarity of parent and child is eliminated. The adoptive parents raise children who have a genetic constitution that is different different from theirs in an environ environdifferent from from that which they would have received had their biological ment that is different coefficient is used to express parents not decided on adoption. The correlation coefficient family members' psychological resemblance. In the adoptive study, one-half family one-half the effect is realized in the trait correlation of the adoptive child and biological genetic effect tho genetic effect effect is expressed because first degree relatives (e.g., parent. One-half the one-half their (poly (polysiblings or parents-children) can be expected to possess only one-half 2) may morphic^) genes. genes. The The full full genetic genetic effect effect (h (h^) may be be calculated calculated as as twice twice the the correla correlamorphic2) tion coefficient coefficient of of adoptive adoptive child child and and biological biological parent. parent. tion effect is realized in the trait In an adoptive study, the family environmental effect family members correlation of adoptive parent and adoptive child. Because these family do not share genes (except by chance), their psychological similarity can be inferred inferred from exposure to common environmental influences. to arise from

MZ Twins Twins Raised Raised Apart Apart 2. MZ The study of MZ twins raised apart is a special type type of adoption adoption design. design. The The research research design falls into the adoption design category because at least one twin would be raised by adoptive parents. In the ideal separated twin design, the MZ twins would for adoption. Of course, real adoptions be separated and, by simple chance, assigned for can only approximate the requirements of the ideal design. For instance, one MZ twin may remain with his biological parents, whereas the other is adopted; or both different parents who are related by blood to the twins' biological may be raised by different parents; or, approaching an ideal design, each twin may be raised by separate adoptive parents, who lack any connection to the twins' biological parents. The two most recent studies of twins raised apart were the Minnesota Study of Twins Reared Apart (Bouchard et al, aI., 1990) and the Swedish AdoptionfI'win Adoption/Twin Study of Aging (Pedersen, Plomin, McClearn, McCleam, & Friberg, 1988). 1988). The latter study included about 350 pairs of twins raised apart, and so was larger than the entire world's previous samples of separated twins. 3. 3.

The Comparison f MZ and DZ Twins Comparison o ofMZ andDZ

Another research design is that comparing MZ and DZ twins. MZ' twins share Expressed as asaacorrelation correlation coefficient, coefficient, their their genetic geneticsimilarity similarity exactly the same genes. Expressed = 1.0. In contrast, as any first degree relative DZ twins share only would be r = = .50. 1/2 their genes (on average), yielding a genetic correlation of r = .50. Heritability is estimable from from this research design by a simple calculation, h DZ), h^2 = = 2(rMZ 2(rMz - r'•DZ), where h2 h^ is the heritability, rMZ TMZ is trait correlation in identical twin pairs, and rDZ rpz is that in fraternal fraternal twin pairs. ^Polymorphic Polymorphicgenes genesare arevariable. variable.For Forinstance, instance,the thegenes genesfor foreye eyecolor colorare arepolymorphic polymorphicbecause because 2 different different forms of these genes yield different eye pigment colors.

15 CHAPTER 15

GENETIcs, GENETICS,TEMPERAMENT, TEMPERAMENT,AND ANDPERSONALITY PERSONALITY

373 373

Among several several assumptions, this twin design design requires assumption of of equal equal Among requires the assumption treatments treatments (relevant to a particular trait) for MZ M Z and DZ D Z twins. Clearly, if if MZ MZ twins were treated more alike than DZ D Z twins, and this treatment treatment were were to affect affect a particular particular trait, then the attribution attribution of of MZ M Z twins' greater greater resemblance resemblance to genetic genetic variation variation may be flawed. flawed. Although Although the assumption assumption of of equal equal treatments treatments may be false false for for some traits, for for a majority of of behavioral traits it has received received considerable considerable support (Loehlin & Nichols, 1976). The The main reason that that the assumption assumption has been been not that the treatment treatment of of MZ M Z and DZ D Z twins was was exactly exactly alike in all cases. safe was not Rather, it survived because because many environmental environmental treatments treatments (contrary (contrary to widespread widespread cultural beliefs) beliefs) failed failed to influence influence the personality personaUty traits that the treatments treatments were were cultural expected to influence. influence. expected One particular particular concern about equal twin treatment treatment deserves deserves consideration. consideration. Many critics of of twin studies (Lewontin, Rose, Rose, & Kamin, Kamin, 1984) suppose suppose that MZ MZ Many would be especially alike alike in behavior behavior because because they they physically physically look look alike. This This twins would of the twin research design, though, is unconvincing. The The association association of criticism of generally weak weak (r < nonexisphysical, facial features with personality traits is generally < .25) or nonexis appearance-personality trait correlations, appearance appearance alone alone tent. Given low facial appearance-personality induce much of of a personality personality correlation correlation across twin pairs (the reader reader cannot induce M Z twins are not exaCtly exactly alike in their physiques physiques or facial facial should note also that MZ appearances). appearances ) .

4. Model-Fitting Model-Fitting Approaches Approaches 4. In model-fitting model-fitting research designs, studies can be extended extended to cover cover any combination combination of family family types. That is, they can be done with many many combinations combinations of of relatives relatives (e.g., of families of of MZ M Z twin aunts or uncles). In step-siblings, half-siblings, adoptees, and families model-fitting designs, equations are first written to express express the the correlational correlational data data model-fitting different family family types in terms of of the variance variance components components described described from the different mathematically for for them. above. The equations are then solved mathematically An A n example example should help to clarify clarify these ideas for for the the reader. Adoptive Adoptive siblings siblings from successive successive placements placements in a home home would be biologically biologically unrelated unrelated to one one from another. Similarly, a biological biological child and an adoptive adoptive child raised raised in the same same family family would be biologically unrelated because their biological parents parents would would be different. different. In both family family structures, because because the siblings have different different biological biological parents, hereditary constitution constitution would be dissimilar (i.e., uncorrelated). As A s the siblings' their hereditary trait similarity similarity must arise entirely from from common common exposures exposures to environmental environmental experi experiequation ences, the equation

r'"URT URT

= = c'(^

would hold, where rTURT unrelated children biologically unrelated URT is the trait correlation in biologically reared together and c'c^ is the shared environmental environmental effect. effect. Now one from from M MZ Now suppose that this correlation is combined with one Z twins raised together. The latter's trait resemblance can be attributed potentially potentially to a combination combination of of shared shared environmental environmental and hereditary hereditary influences. influences. Because Because MZ M Z twins equation also possess exactly the same genetic constitution, as mentioned above, the equation

374 374

DAVID C. C ROWE rMZT =

h2 + c'-

follows, where rTMZT MZ twins raised together. MZT is the trait correlation in MZ Using these equations together, one can solve for the variance components, h^2 and 2-. c^-For Forexample, example,ififthe theMZ MZtwin twincorrelation correlationwere were.53, .53,whereas whereasthat thatininadoptive adoptive h 2 == .36. siblings was .17, .17, then c'c^ = = .17 .17 and and hh^ .36. Of Of course, course, this this example example would would become become more complex if more traits or family types were added. The general point is that model-fitting gives a flexible flexible analysis of kinship data, one that allows these data to be fit and theoretical assumptions to be evaluated simultaneously. In this regard then, the classical research designs of MZ twins raised apart, MZ and DZ twins compared, and and adoptive adoptive families families are are but but important special special cases cases of of aa more more general general compared, of handling handling kinship kinship data. data. method method of

n VARIATION IN n. GENETIC GENETIC VARIATION IN TEMPERAMENTAL TEMPERAMENTAL AND AND PERSONALITY TRAITS TRAITS PERSONALITY A. The The Heritability Heritabiiity of of Childhood Childhood Temperament Temperament A. Temperament is the first expression of personality in the very young. As Darwin observed, infants reveal many rudiments of a complex emotional life:

see children, children, only only two two or or three three years years old, old, and and even even those those born born blind, blind, We may may see We of aa very very young young infant infant reddens reddens from from blushing from from shame; shame; and and the the naked naked scalp scalp of blushing passion. passion. Infants Infants scream scream from from pain pain directly directly after after birth, birth, and and all all their their features features then assume the same same form form as as during during subsequent subsequent years. years, (p. (p. 351) 351) assume the Modem of Modern temperament theorists also focus on infancy, on that first appearance of differences in personality (Goldsmith et ai., al., 1987). Theorists Theorists have have differed differed individual differences temperament and in their assessment of the main in their particular definitions of temperament temperamental traits. As shown in Table I, Goldsmith chose temperamental categotemperamental TABLE TABLEII Trait Categories in in Different Different Models of Temperament Buss & & Plomin

Darwin

Goldsmith

Low spirits High spirits Hatred & & anger Disdain, disgust Surprise, fear Shame, shyness

Sadness Joy, pleasure Anger Disgust Fear

Fear

Fear

Interest Activity

Activity

Orienting Activity Soothability

Sociability Anger

Note. Trait categories from Goldsmith et et a!. al. (1987).

Rothbart Distress Smiling

Thomas & & Chess Predominant mood Intensity of mood Approach/withdrawal Adaptability Sensory threshold Distractibility Persistence/attention Activity Rhythmicity

CHAPTER 15 15

375 375

GENETICS,'TEMPERAMENT, TEMPERAMENT,AND ANDPERSONALITY PERSONALITY GENETICS,

ries on the categories overlap considerably considerably with Dar Darthe basis of basic emotions. His categories win's. Buss and and Plomin (1984), on the other hand, identified identified just three three broad temperaments: emotionality subdivided into general general arousal, fear, and emotionality (which can be subdivided anger), activity, and and sociability. Activity Activity level refers refers to total motoric motoric activity-the activity—the amount amount of movement, such as pacing and and running. Sociability Sociability refers to the prefer preference for social interactions with others, as opposed to being alone. Rothbart's Rothbart's (1981) temperamental temperamental categories included Buss and Plomin's activity and several others with an emotional flavor flavor such as Darwin's and Goldsmith's. One of of Rothbart's Rothbart's temperaments referred to a more cognitive, attentional attentional aspect aspect of behavior, that is, duration of orientirig. orienting. Finally, Thomas and and Chess (1977) defined defined nine categories of temperament. They shared with the other theorists theorists a category category of activity. Approach/ may be related to both emotionality emotionality and sociability; sociabihty; it did not appear withdrawal may as a separate factor in a joint factor analysis of Thomas and Chess's and Buss and temperament items (Rowe (Rowe & Plomin, 1977). They They also added added two catego categoPlomin's temperament (i.e., distractibility, persistence/attention persistence/attention span). ries with a possible basis in cognition (Le., these proposed categories of temperament temperament were were different different among research Although these agreement also exists among the temperamental temperamental traits proposed. groups, some agreement differences in infancy infancy that Together, they adequately span the variety of individual differences first drew Darwin's attention. evidence has accumulated that basic In light of twin and adoption studies, evidence extensively explored explored the herita heritatemperaments are heritable. Buss and Plomin (1984) extensively Because bility of four temperaments: the three listed in Table I and impulsivity. Because they decided decided it was not a tempera temperathe evidence of impulsivity's heritability was poor, they temperament must be heri herimental trait by their (somewhat restrictive) criteria that a temperament table. emotionality on Table II presents heritabilities for sociability, activity, and emotionality and Plomin Plomin pooled pooled over over studies. These correlations correlations the basis of twin data that Buss and computed in three ways: (1) as the 'TMZ twice the difference difference of were computed MZ itself, (2) as twice ' ''MZ ~ ''"Dz, and (3) as twice the 'oz. TDZ. The first first estimate, of course, assumes assumes an OZ, and MZ absence absence of family family (shared) environmental environmental influences influences on temperament. temperament. The second second

II TABLE lI

Heritability Estimate from EAS Questionnaire Questionnaire Heritability Broad sense sense Broad h2 h'^Method Method 11

Activity Activity

.62 .62

.53 .53

Sociability Sociability

Emotionality Emotionality

.63 .63

Narrow sense sense

h^ Method Method 22 h2 1.50 1.50

1.12 1.12 11.02 .02

h2 h^ Method Method 3

< 00 <

< 00 <

.24 .24

Note. 2 (rMZ ); Note. Method Method 11isisthe theMZ MZtwin twincorrelation; correlation;Method Method2,2,2 (rMz-- roz '"DZ); Method Method 3, 3, 2roz. 2rDz. Data Data given given in in Buss Buss and and Plomin Plomm (1984, (1984, Table Table 9.2, p. 122). 122). N = 228 228 MZ MZ twin twin pairs pairs and and 172 172 DZ DZ twin twin pairs. pairs. p. =

376 376

DAVID C. CRowE DAVID ROWE

estimate-that estimate—that of the classical MZ versus DZ twin comparison-assumes comparison—assumes "equal" *'equar' twin environments relevant to the temperamental traits. Like the first, the last estimate assumes an absence of of family (shared) rearing influences. As noted above, because the DZ twins share only half half their genes (on average), their trait correlation must be doubled to estimate heritability. As shown in Table II, these heritability estimates appear to conflict with one another. The rMZ TMZ twin estimate was the most reasonable one-across one—across the three traits, a mean of 51 % of variation was attributable to genetic variation. Both the of 51% of classical twin and the roz-based estimates, however, were nonsensical, which was an outcome of of small DZ twin trait correlations. The former estimates were nearly 1.0; the latter were 0 or less than 1 (but a variance component component can never take a value less than 0.00 because variances are always positive numbers). Why were the DZ twin correlations so small? Buss and Plomin suggested suggested one possible reason for it: the data on twins' temperament depended almost entirely on parental ratings of behavior. If parents tend to contrast their fraternal twins with one another and so accentuate their differences, then this rating bias would make them appear more dissimilar in temperament than they really were. This rating artifact could create zero, or even a negative, DZ DZ twin correlation on the temperamental traits, yielding the unreasonable heritabilities (i.e., (i.e.. Methods 2 and 3 above). Data from twin studies using objective measures permit an evaluation of of Buss and Plomin's contrast effect conjecture. Consider the trait of activity. As shown in effect of Table II, parents rate the DZ DZ twins as no more alike in activity than randomly paired children-the children—the roz was about O. 0. Activity is clearly a temperamental trait exists—mechanically for which an observational alternative to parental ratings exists-mechanically recording the twins' bodily movements on an instrument that records them. In a study of of twin infants (mean age = = 8 months; Saudino & Eaton, 1991), actometers were attached to 78 MZ twins (in 39 pairs) and to 42 same-sex DZ twins. The activity data were collected over a 2-day period. The twin correlations for composite actometer activity were rMZ TMZ = = .76 '76 and rTDZ = .56. Heritability estimates by the three OZ = methods above are .76, .40, and 1.08, respectively. The high estimate from doubling the DZ twin correlation (1.08) may reflect reflect an influence of shared environment on infants' activity level. From the MZ versus DZ twin comparison, this estimate of shared environmental variation would be .36.^ .36.3 Adjusting the DZ correlation for effect results in a more consistent heritability estimate on the basis of Method this effect 2= DZ twin correlation; h h^ = .40). None of the heritability heritability 3 (i.e., doubling the adjusted DZ of course, because of the small number of of twin pairs. Nonethe Nonetheestimates is precise, of tendency towards 0 or negative DZ DZ twin correlations was clearly absent in less, a tendency objective measurements of activity. these objective 3 ^Shared Sharedenvironmental environmental variation variation (c2) (c^)was wasestimated estimated asas2roz Ir^z -- PMz. r^z-For Forfurther further details detailsof of the the basis of this derivation, see Rowe (1994). Because the twins could play together, shared environment here may be the result of the twins influencing one another.

CHAnER CHAPTER IS 15

GENETICS,TEMPERAMENT, TEMPERAMENT,AND ANDPERSONALITY PERSONALITY GENE11cs,

377 377

Another Another study leads to similar conclusions for other temperamental temperamental traits (Emde aI., 1992). As (Emde et al., A s shown in Table III, III, for parental ratings of temperament temperament the heritability heritability estimate estimate as derived derived from from the DZ D Z twin correlation alone was was always always inconsistent inconsistent with the other other estimates. In contrast, contrast, estimates estimates from from the observational observational measures varied varied widely, but except except for task orientation, orientation, were were never inconsistent inconsistent (i.e., 0 or less than 0 whereas whereas other estimates estimates were positive). The one exception, exception, task D Z (.22) than MZ M Z twin (.15) (.15) correlation. Hence, Hence, task orientation, yielded yielded a greater DZ the comparison comparison of MZ M Z with DZ D Z twins suggested suggested that this observed trait was not heritable heritable (one of the few such nonheritable nonheritable traits among among 23 traits examined examined by Emde Emde et a1.). a l ) . In summary, these these data data strongly strongly suggest that that contrast contrast rating effects effects led to inconsistent inconsistent heritability heritability estimates estimates when parental ratings were were used used for inhibition, inhibition, shyness, and activity.4 activity."* Moreover, the narrow-sense narrow-sense heritability heritability as estimated estimated from from the DZ D Z twin corre corresubstantial for temperamental temperamental traits. Averaged Averaged over inhibition, shyness, shyness. lations was substantial

in TABLE m

Heritabilities Based Based on on Observational Observational versus versus Parental Parental Report Report Heritabilities Measures Measures of of Infant Infant Temperament Temperament

Broad-sense Broad-sense

h^ Method Method 11 h2

Narrow Narrow sense sense

/i^Method Method22 h2

h^Method Method33 h2

Laboratory Laboratory observational observational measures measures

Shyness Shyness

.70 .70

.50 .50

.62 .62

.52 .52

Activity

.42

.58

.26

Inhibition

Task orientation*

.57

.15

.90

< < 00

.44

ratings on on the the CCTI Parental ratings Parental CCfI Emotionality Emotionality

.35 .35

.74 .74

.04

Shyness Shyness

.38 .38

.82 .82

< < 00

Activity

.50

1.50

< < 00

.38 .38

.84

< < 00

Sociability Sociability

Attention! Attention/

persistence persistence

.35

.64

.06

Note. Note. CCfI, CCn,Colorado ColoradoChildhood ChildhoodTemperament TemperamentInventory Inventory(Rowe (Rowe&i

Plomin, 1977). 1977). Data Data were were from from Emde Emde et et al. al. (1992). Plomin, (1992). * 'oz

> 'MZ

^In InEmde Emdeetetal. al.(1992), (1992),another anotherparental parentalrating ratingscale scaledid didnot notproduce producelow lowDZ DZtwin twincorrelations correlations

4

for the the traits traits of of negative negative and and positive positive emotionality. As not not all all parental parental ratings ratings produce produce contrast contrast effects, effects, for emotionality. As

more research research is is needed needed comparing comparing rating rating formats formats and and the the semantic semantic globality globality versus versus specificity specificity of of more

rating items. rating items.

378 378

DAVID C. ROWE C ROWE

and activity, the mean h2 /i^ was .56. .56. The broad-sense heritabilities (Le., (i.e., with MZ .56, respectively. Because twins) averaged over the same three traits were .57 .57 and .56, the narrow- and broad-sense heritabilities heritabiUties were close in magnitude, I believe that effects effects of gene-gene interactions on temperament variation were minor (at least for these temperaments). Although this brief brief review has neglected many excellent for the omitted omitted studies studies studies of the inheritance of temperament (see Goldsmith, 1983), the reinforce the conclusion stated here: one-third to one-half one-half of individual mainly reinforce differences in in temperamental temperamental traits traits can can be be attributed attributed to to genetic genetic variation variation among among differences children. children. B. B. The The Heritability Heritability of of Personality Personality in in Adulthood Adulthood The domain of adult aduU personality is characterized by an extensive behavior genetic self-report questionnaires. literature. Studies have been done using a great variety of self-report different theories of personality propose somewhat different different sets of adult traits, As different different self-report self-report scales, the domain of adult personality would and also use different seem confusing. Fortunately, personality personaUty theorists have reached a consensus that at least five independent personality factors exist (Le., (i.e., the "Big Five," John, 1990). different names than one of the Big Five Five trait trait Many personality traits, although given different factors, may be actually synonymous with one of them or represent a combination of of several of of them. them. The The Big Big Five Five trait trait factors factors are are listed listedin inTable TableIV, IV,along alongwith withadjectives adjectives representative of them. them. The The table table uses uses the the factor factor names names popularized popularized by by Norman Norman representative of (1963), except except that that the the first trait factor factor was was named named extraversion extraversion instead of Nor Norinstead of (1963), first trait man's surgency. man's Loehlin (1992) reanalyzed data on the Big Five factors from studies around the world that used behavior genetic research designs. He fit a variety of behavior IV TABLE IV The "Big Five" Trait Dimensions

Extraversion

I

II Agreeableness

Talkative Active Outgoing Dominant Enthusiastic

Sympathetic Appreciative Soft-hearted Generous Helpful

Organized Planful Responsible Dependable Precise

Retiring Withdrawn Silent Shy Reserved

Hard-hearted Quarrelsome Unfriendly Cold Fault-finding

Slipshod Irresponsible Frivolous Disorderly Careless

III Conscientiousness

IV Emotional stability Tense Nervous Worrying Fearful Self-pitying

V Culture Wide interests Intelligent Insightful Sophisticated Clever Unintelligent Shallow Simple Narrow interests Commonplace

Note. Note. Adapted from Loehlin (1992). (1992), with permission. Copyright 1992 by Sage Publications.

15 CHAPTER 15

379 379

GENETICS,TEMPERAMENT, TEMPERAMENT,AND ANDPERSONALITY PERSONALITY GENmcs,

parsimogenetic models to kinship correlations for Big Five factors, seeking the most parsimo theoretically convincing convincing explanation explanation for variation in each one. His results results nious and theoretically led to a remarkable finding finding that fairly fairly simple models can give an account account of personal personality variation. behavior genetic model model worked worked for the world's adoption data on extraver extraverOne behavior sion. Loehlin's model model postulated just two sources of trait variation: heredity and sion. nonshared environment. On the basis of the best statistical fit, the heritability of .35, and the nonshared environmental environmental contribution was was .65. The extraversion was .35, .65. The environmental part of variation was set to zero (2 (c^ = = .(0). .00). The reason for shared environmental eUminating the latter was that a model without this parameter fit statistically nearly eliminating environment was therefore therefore not costly costly as well one with it. The "loss" of the shared environment explanatory power. power. in terms of the model's explanatory Table V presents actual extraversion correlations correlations and and those model-fitted model-fitted to them. In the case of biologically biologically unrelated family members, members, the the model-fitted model-fitted correla correla0. In the case of biologically related family members, it was .18. The latter tion was O. one-half the model-estimated model-estimated heritability of extraversion extraversion (h2 (h^ = = .36). These These value is one-half model-fitted correlations matched the obtained correlations closely enough two model-fitted accepted statistically. Note that under this model's assumptions, that the model was accepted all family resemblance personality was was merely attributable to shared genes. all family resemblance for personality (i.e., via adoptive family family relation relationThat is, once the genetic similarity was removed (Le., ships), personality than persons reared in ships), family members were no more alike in personality different families. dif ferent families. Data from additional family family types on the Big Five personality factors, however, complex models than the heredity-nonshared heredity-nonshared environment environment one. required more complex one. found that two models gave equally good descriptions of the personality personality Loehlin found variation in the Big Five. One model required the variance components of M Z twins, (3) shared shared (1) transmissible heredity, (2) a special shared environment in MZ model had the same same environment, and (4) nonshared environment. The other model

TABLE V V TARLE Obtained and Model-Fitted Correlations on Extraversion Obtained

Family relationship Mother and biological child Father and biological child Mother and adoptive child Father and adoptive child Biologically related siblings Biologically unrelated siblings

r observed observed

N (pairs) N

.12 .21 -.01 .03 .20 07 -.07

621 547 571 522 570 258

r

-

.

r

r fitted

.18 18 18 .18 .00 .00 .18 .00 . .

N (pairs), total number of of pairs across three studies. Correla CorrelaNote. N tions weighted average correlations from Loehlin's (1992) Table 2.4, p. 32.

380 380

DAVID C. CRowE DAVID ROWE

variance variance components, except that that one for for gene-gene gene-gene interactions replaced replaced that of of special MZ MZ twin environments. Table VI shows the parameter parameter estimates for for this second second theoretical model. Over Over the Big Five trait factors, the mean narrow-sense heritability heritability was .30, whereas whereas mean estimate of of shared environmen environmenthe mean broad-sense was .42. In contrast, the mean effects was relatively small in magnitude (2 (c^ = = .08). The remainder remainder of of trait tal effects of course, would be due to nonshared environmental effects effects (e2 (e^ = = .50). variation, of About 40% of of this nonshared nonshared variation would would be attributable to measurement measurement error, About remainder to the lasting consequences of of experiences uniquely changing and the remainder for a special M Z twin twin each person. In Loehlin's second model—the model-the one allowing for MZ environment rather than emergenic gene action-the action—the variance component component estimate estimate environment for shared shared environment environment was also small (mean 2 c^ = = .05, not shown in Table VI). In for heritability and shared environmentability environmentability of major Big Five summary, the heritability of the major factors suggest that first degree relatives will correlate only about .15-.20 on them, factors shared genes, not shared shared experiences, mainly mainly determine the familial familial resem resemand shared of "blood" relatives. blance of in. fuRTHER FURTHER TOPICS OF INTEREST m. of Shared Rearing Effects Effects A. The Absence of A widespread widespread cultural belief belief in western European European countries and in the United United A parenting styles will create variations in children's personal personalStates is that variation in parenting behavior genetic findings findings reviewed above generally generally refute refute this cultural ity traits. The behavior belief. For example, none of of Loehlin's adoption models required required a parameter parameter representing an environmental environmental effect effect of of parental behavior behavior on the development development of of representing component of of environ environchildren's traits. Moreover, as noted above, the variance component 11% of trait variation in Table ments shared by siblings was always small, from from 2 to 11 % of currently held by behavior behavior geneticists is that family family environments environments VI. The consensus currently influence on personality personality development development (Rowe, 1994; in the normal range may lack influence

TABLE TABLE VI VI Estimates of Variance Components in in "Big Five" Personality Personality Traits Traits factor Big Five trait factor

Narrow-sense heritability h2 h^

Gene-gene interactions ;2 i^

Shared c^ environment c2

I. Extraversion Extraversion I. II. II. Agreeableness Agreeableness III. III. Conscientiousness Conscientiousness IV. Emotional Emotional stability stability IV. V. Culture/openness Culture/openness Mean

.32 .32 .24 .22 .27 .43 .30

.17 .11 .11 .16 .14 .02 .12

.02 .02 .11 .11

Note. Adapted from from Loehlin Note, Adapted Loehlin (1992, (1992, Table Table 3.2, 3.2, p. p. 67). 67).

.07 .07

.06

.08

CHAPIER CHAPTER 15 15

GENETICS, TEMPERAMENT, AND PERSONAUTY PERSONALITY GENETIcs,

381 381

Scarr, 1992, 1993). In other other words, although different different parents treat children in different different ways-for ways—for instance, showing affection affection demonstrably, covertly, or not at all—except for extremes (Le., (i.e., child abuse or neglect), these parental treatments all-except effects on child development. would be equivalent to one another in their effects The assertion of the essential equivalence equivalence of family family environments, which to families appear very different, different, has provoked provoked heated controversy many observers of families 1993; Jackson, 1993; 1993; Scarr, 1993). The The reason reason among social scientists (see Baumrind, 1993; for the controversy is easy to comprehend. The shared environment relates most for family social influences. directly to the relative malleability of traits in response to family environment of strong effects effects means that a trait would be highly malleable, A shared environment from one family rearing circumstance in the sense that if children were "swapped" from A small small shared component component means means to another their trait scores would greatly change. A the opposite, that in range of family environments studied, all of them would be effects on children's traits. If changing rearing circum circumroughly equivalent in their effects from one family environment stances has little consequence, then moving children from will not not change change them. them. The Thereader reader should should note notethat that the the shared shared environment environment to another another will to is more more relevant relevant to to the the social social malleability malleability of of traits traits than than is is heritability. heritabiUty. In In her her 1991 1991 is ResearchininChild ChildDevelopment, Development,Scarr Scarr(1992) (1992) for Research presidential address to the Society for explained these these implications implications of of behavior behavior genetic genetic results results for for personality personality malleability: malleability: explained

The flip flip side side of of this this message message is is that that itit is is not not easy easy to to intervene intervene deliberately deliberately in The in children's lives lives to to change change their their development, development, unless unless their their environments environments are are out outchildren's side side the the normal normal species species range. range. .. .. .. for for children children whose whose development development is is on on aa predictable predictable but but undesirable undesirabletrajectory trajectory and andwhose whoseparents parentsare are providing providingaasupport supportive environment, interventions interventions have ive environment, have only only temporary temporary and and limited limited effects. effects. . . .. .. Should we we be be surprised? surprised? Feeding Feeding aa well-nourished well-nourished but but short short child child more more and Should and more will will not not give give him him the the stature stature of of aa basketball basketball player. player. Feeding Feeding aa belowmore below average intellect intellect more moreand andmore moreinformation informationwill willnot notmake makeher herbrilliant. brilliant.Exposing Exposing average shy child child to to socially socially demanding demanding events events will will not not make make him him feel feel less less shy. shy. The The aa shy child with with aa below-average below-average intellect intellect and and the the shy shy child child may may gain gain some some specific specific child skills and and helpful knowledge of of how how to to behave behave in in specific specific situations, situations, but but their skills helpful knowledge their enduring intellectual and personality personality characteristics characteristics will will not not be be fundamentally fundamentally enduring intellectual and changed, changed. (pp. (pp. 16-17) 16-17) effects can be found in an exchange of Further details on the issue of rearing effects of environmenviews among Sandra Scarr (1993) and her two critics who adopt strong environmen talist positions, Baumrind (1993) and Jackson (1993). B. B. Genetic Genetic Variation Variation in in Environmental Environmental Measures Measures Social scientists may regard measures of environmental influences as independent environof genetic ones in the DNA. This view may ignore a major characteristic of environ mental measures widely used in child development research: that they summarize the behavior of individuals. For instance, consider the measure "number of books in the home," which has been related to children's IQ variation. For young children, parents must decide which books to purchase and how many; hence, "books in the

382 382

DAVID ROWE DAVID C. CRowE

home" may indirectly assess heritable personality traits such as general intelligence intelligence (10) (IQ) and the Big Five factor factor trait of of culture. In a review article on genetic variation in environmental environmental measures, Plomin and Bergeman Bergeman (1991) observed, "Environments ''Environments have no DNA DNA and can show no genetic influence. Measures Measures of of the environment environment perfused with characteristics of of individuals, individuals, however ., .. .. [they] [they] can can . .. .. may be perfused show genetic influence" (p. 374, italics in original). Plomin Plomin and Bergeman's review documents that that genetic variation occurs in a variety variety of of environmental measures. Social economic status (SES) is often often assessed by the measures of of "years of of education," incomes, or occupational prestige. Although Although parental SES is associated with children's current current "environments," these measures measures also distill in themselves the behavior behavior of a parent over many years, who either had educational and occupational success or who lacked them. the traits leading to educational Summarizing behavior behavior genetic studies of of SES, Plomin and Bergeman Bergeman concluded concluded Summarizing about .40. Hence, SES cannot be regarded regarded as capturing capturing that its heritability was about environmental variation. purely environmental interpretations of social class, a correlation of of an Therefore, contrary to most interpretations personality trait trait outcome is not not proof proof of of environmental environmental influ influSES measure with a personality Common genes may make parent and child alike for for dif different ence. Common ferent traits, so that the effects effects of of genes can be mistaken for for environmental ones when statistical data are interpreted. For example, in parents the genes may be expressed expressed as differences differences in SES; in their children, they may be expressed as another trait, for for instance, intelUgence (10). (IQ). Perhaps Perhaps social science editors should should adopt the policy general intelligence that both genetic and environmental alternatives be posed when an SES-child SES-child personality statistical relationship is found. Such hypotheses can be investigated in personality behavior genetic studies that that include measures measures of of social class (Rowe & Wald Waldbehavior man, 1993). Genetic variation also can can be found found in measures of of parenting styles. In my twin studies of of adolescent twins' reports of of parental behavior, I have found found evidence for for genetic variation in measures of of parental affection/warmth, affection/warmth, but not in those of of 1983). Plomin and his colleagues took the reverse parental parental control control (Rowe, 1981, 1981, 1983). approach. Genetic variation in parenting styles of of adult twins (who had families) families) was investigated. In the Swedish Adoption Adoption Twin Study of of Aging (Plomin McCleam, Pedersen, Nesselroade, & Bergeman, 1989), adult twins reported reported on their general general home environments environments on the Family Environments Scale questionnaire. Twin pair correlation correlation coefficients coefficients were then computed computed for for four four types of of pairs: MZ twins raised apart, MZ twins raised together, DZ D Z twins raised apart, and DZ D Z twins raised together. In the different different groups, the number of of twin pairs ranged from from 50 to 129. As shown shown in Table VII several scales produced statistically significant significant heritabilities: expressiveness, culture, organization, and control. Across the nine home environ environment scales, the average heritability .25. About heritabihty was .25. About one-quarter one-quarter of variation in home environment environment is thus attributable to genetic variation. variation. Other environmental environmental measures reviewed reviewed by Plomin and Bergeman Bergeman (1991) Other revealed a genetic component. Television viewing (total hours) yielded a herita heritaalso revealed of .44. Controllable Controllable life events had a heritability heritability of of .43, .43, and uncontrollable uncontrollable bility of

CHAPTER15 15 GENETICS, GENETICS,TEMPERAMENT, TEMPERAMENT,AND ANDPERSONALITY PERSONALITY CHAPTER

383 383

TABLEVU Vn TABLE

Heritability Estimates Estimates for for Home Home Environment Environment Heritability

Scales Scales

Home environment environment Home scale scale

Broad-sense Broad-sense

heritability heritability hhj2

Cohesion Cohesion

.19 .19

Conflict Conflict

.25 .25

Culture Culture

.40* .40*

Expressiveness Expressiveness Achievement Achievement Active Active

Organization Organization Control Control

Mean Mean

.27* .27*

.12 .12 .21 .21

.26* .26* .26* .26*

.25 .25

Note, Adapted Adapted from from Plomin Plomin and and Bergeman Bergeman (1991, (1991, Note.

p. 376). 376). p.

*p < < .05. *p .05.

life events, .18. .18. Perceived Perceived adequacy adequacy of social support support had a heritability heritability of .30. Adolescents' choice of different different peer groups groups (e.g., delinquent or nondelinquent) nondelinquent) was also heritable. It is no exaggeration to say that genetic variation suffuses that suffuses measures of the environment. Social scientists should be encouraged to consider encouraged whenever interpreting interpreting "effects" "effects" of of environmental environmental measures. this genetic variation whenever Furthermore, whether whether genetic variation in measures of of "environment" represents represents gene effects effects shared with known personality traits (i.e., the Big Five), or whether whether it is really of most most personality really something something unique and outside outside the domain domain of personality traits, should be investigated investigated (Chipuer, Plomin, Pedersen, McClearn, & Nesselroade Nesselroade 1993). C. A of Personality Development c. A Theory of

To advance findings in this chapter chapter must must be be placed To advance social science knowledge, the findings theory of of personality personality development. development. Behavior Behavior genetics genetics is is part part of of a into a general theory general theory of of personality development that that requires requires an an integration of behavior general theory personality development integration of other psychological psychological subdiscipUnes. subdisciplines. This This theory theory attributes attributes trait genetics with other trait variation to variation variation in the biology biology of of the the human human nervous nervous and and endocrine endocrine systems. An An analytic to layer between between the the genes genes and and measured measured traits traits is being being investigated investigated through through aa whole layer range of of new new brain brain imaging imaging systems. Hypotheses Hypotheses about about physiological physiological bases bases of of per range personality sonality variation variation are are currently currently becoming becoming more more refined refined (see (see Geen, Geen, this volume, chap. 16). Behavior Behavior genetics genetics also explains explains the the intergenerational intergenerational transmission transmission of of the absence of strong strong biological biological selection selection and and given given large large population population sizes sizes traits: in the absence of (certainly (certainly true true of of humans), genetic genetic variation variation in in one one generation generation will will be be reliably reliably transmitted transmitted to to the the next. Both Both within-population within-population variation variation in in personality, personality, and and some some

384 384

DAVID ROWE DAVID C. CRowE

population population mean differences differences (Kagan, Arcus, & Snidman, 1993), may thus possess possess a basis in genetic variation. variation. scientific progress, many many interesting and unresolved questions remain remain Despite scientific general theory of of personality. Some fundamental fundamental questions are, are. Why to complete a general do family family environments environments have so little effect effect on personality personality development? What variability in personality personality over many generations? Is genetic varia variamaintains genetic variability tion related to human human adaptive traits, or is it merely genetic "junk"? How How does environmental transmission transmission occur? Can the specific environmental causes of of "non "nonenvironmental shared" environmental environmental variation variation be found? found? shared" With regard regard to the first question question posed above, behavior behavior genetics gives the metaphor of of "niche picking" (Scarr (Scarr & McCartney, 1983). Just Just as a nonhuman nonhuman metaphor finding a local environment environment most most suited to its adaptive adaptive animal "niche picks" by finding strengths, so may people find the local social environments most suited to their reinforce them and would allow their full full heritable traits, which would tend to reinforce other animals, people may also change change local environ environexpression. Moreover, as do other opportunities have been ments to suit themselves. Hence, only when environmental opportunities constrained would a child fail to develop heritable personality personality traits to severely constrained inherent potentials. nearly their inherent questions asked in this chapter chapter should occupy occupy the next generation of of The questions social scientists concerned with trait variation. The field has come a long way since of genetic and environmental transmission. His Darwin's rudimentary rudimentary concepts of incorrect Lamarckian Lamarckian view that incorrect that traits may be inherited merely because they are abandoned in the 20th century. Yet many puzzles about tempera temperaexercised has been abandoned ment and personality personality remain to be solved. As Darwin wrote, "the language of of the ment of importance for for the welfare welfare emotions, as it has sometimes been called, is certainly of of mankind. To understand, as far far possible, the source or origin of of the various of of men around around us . . . ought expressions which may be hourly seen on the faces of interest for for us" (p. 366). to possess much interest

REFERENCES Baumrind, D. (1993). The average expectable environment is not good enough: A response to Scarr. Child Development, Development, 64, 64,1299-1317. 1299-1317. Bouchard, T. J., Jr., Lykken, D. T., McGue, M., Segal, N. L., & Tellegen, A. (1990). Sources Science, of human psychological differences: The Minnesota study of twins reared apart. Science,

250, 223-228. 223-228. 250, personality traits. Hillsdale, Buss, A. H., H., & &Plomin, Plomin, R. R. (1984). (1984). Temperament: Temperament:Early Earlydeveloping developing personality traits, Hillsdale, NJ: Erlbaum. Chipuer, H. M., Plomin, R., Pedersen, N. L., McClearn, G. E., & Nesselroade, J. R. (1993). Genetic influence on family environment: The role of personality. Developmental 10-118. Psychology, Psychology, 29, 29,1110-118. Darwin, C. (1872/1965). The expression animals. Chicago & expression of of the emotions in man and animals. London: University of Chicago Press.

CHAPTER CHAPTER 15 15

GENETICS,TEMPERAMENT, TEMPERAMENT,AND ANDPERSONALITY PERSONALITY GENETIcs,

38S 385

Eibl-Eibesfeldt, Eibl-Eibesfeldt, I. (1975). Ethology: Ethology: The science science of of behavior. behavior.New NewYork: York:Holt, Holt,Rinehart Rinehart&& Winston. Emde, R R. N., Plomin, R, R., Robinson, J., Corley, R, R., DeFries, J., Fulker, D. W., Reznick, J. S., Campos, J., Kagan, J., & Zahn-Waxler, C. (1992). (1992). Temperament, emotion, and fourteen months: The MacArthur longitudinal twin study. Child Develop Developcognition at fourteen ment, ment, 63, 63,1437-1455. 1437-1455. quantitativegenetics genetics(2nd (2nded.). ed.).New NewYork: York:Longman. Longman. Falconer, D. S. (1981). Introduction to quantitative Goldsmith, H. H. (1983). Genetic influences on personality personality from from infancy infancy to adulthood. Child Development, Development,54, 54,331-355. 331-355. Goldsmith, H. H., Buss, A. H., Plomin, R, R., Rothbart, M. K., Thomas, A., Chess, S., Hinde, R R. A., & McCall, R R. B. (1987). Roundtable: What is temperament? Four approaches. Child Child Development, Development,58, 58,505-529. 505-529. Jackson, J. F. (1993). Human behavioral genetics, Scarr's theory, and her views on interven intervenfor African African American tions: A critical review and commentary on their implications for Child Development, Development,64, 64,1318-1332. 1318-1332. children. Child The "Big "Big Five" Five" factor factor taxonomy: taxonomy: Dimensions Dimensions of of personality personality in in the the natural natural John, O. P. (1990). The John, languages and in questionnaires. In L. A. Pervin (Ed.), Handbook Handbook of of personality: personality: Theory and research research(pp. (pp.83-100). 83-100).Lisse, Lisse,The TheNetherlands: Netherlands:Swets Swets&&Zeitlinger. Zeitlinger. Theory (1993). The idea of temperament: Where do we go Kagan, J., Arcus, D., & Snidman, N. (1993). R. Plomin & G. E. McClearn McCleam (Eds.), Nature, Nature, nurture, nurture, and psychology from here? In R (pp. 197-210). Washington, DC: American Psychological Association. R. C, (1984). Biology, Biology, ideology, ideology, and human nature: nature: Not Not Lewontin, R c., Rose, S., & Kamin, L. (1984). genes. New York: Pantheon Books. in our genes. Loehlin, J. J. C. C. (1992). (1992). Genes Genes and environment environmentininpersonality personalitydevelopment. development. Newbury Park, Newbury Park, Loehlin, CA: Sage. & cfe personality: A study of850 C, & & Nichols, R R. C. (1976). Heredity, Heredity,environment, environment, personality: A study of 850 Loehlin, J. c., sets of of twins. twins. Austin & London: University of Texas Press. sets A., & & Bouchard, Bouchard, T. T. J., J., Jr. (1992). (1992). Emergenesis: Emergenesis:Genetic Genetic Lykken, D. T., McGue, M., Tellegen, A., Psychologist,47, 47,1565-1577. 1565-1577. traits that may not run in families. American Psychologist, Norman, W. T. (1963). Toward an adequate taxonomy of personality attributes: Replicated Journalof ofAbnormal Abnormal and andSocial Social factor structure in peer nomination personality ratings. Journal factor Psychology, Psychology,66, 66,574-583. 574-583. Pedersen, N. L., Plomin, R, R., McClearn, G. G. E., E., & &Friberg, Friberg,L. L.(1988). (1988).Neuroticism, Neuroticism,extraversion, extraversion, and related traits in adult twins reared apart and reared reared together. Journal Journalof of Personality Personality and Social Social Psychology, Psychology,55, 55,950-957. 950-957. Plomin, R to human behavioral genetics. Pacific R. (1990). Nature Natureand andnurture: nurture:An Anintroduction introduction to human behavioral genetics. Pacific Grove, CA: Brooks/Cole. Plomin, R R. & Bergeman, C. S. (1991). The The nature nature of of nurture: nurture: Genetic Genetic influence influence on on "environ "environ1 4,14, 373-386. mental" measures. Behavioral Behavioraland andBrain BrainSciences, Sciences, 373-386. Plomin, R, R, & Bergeman, C. S. (1989). R., McClearn, McCleam, G. E., Pedersen, N. L., Nesselroade, J. R., Genetic influence on adults' ratings of their current Marriage current environment. Journal of of Marriage Family,51, 51,791-803. 791-803. and the Family, Development, 52, 52, Rothbart, M. K. (1981). Measurement of temperament in infancy. Child Development, 569-578. Rowe, D. C. (1981). Environmental and genetic influences on dimensions of perceived parenting: A twin study. Developmental Developmental Psychology, Psychology,17, 17,203-208. 203-208. family environment: A study Rowe, D. C. (1983). A biometrical analysis of perceptions of family Development, 54, 54,416-423. 416-423. of twin and singleton sibling kinships. Child Development,

386 386

DAVID ROWE DAVID C. CRowE

of family family influence: influence: Genes, Genes,experience, experience,and andbehavior. behavior. New Rowe, D. C. (1994). The limits of New York: Guilford Guilford Press. c., & Plomin, R. (1977). Temperament in early childhood. Journal of Personality Rowe, D. C, of Personality Assessment, Assessment, 41, 41, 150-156. Rowe, D. c., C, & Waldman, I. D. (1993). The question "How?" reconsidered. In R. Plomin & G. E. McClearn (Eds.), Nature, Nurture, & Nature, Nurture, & Psychology (pp. 355-373). Washington, DC: American Psychological Association. Saudino, K. J., & Eaton, W. O. (1991). Infant Infant temperament temperament and genetics: An objective twin study of motor activity level. Child Development, 174. Development, 62, 62,1167-1 1167-1174. Scarr, S. (1992). Developmental theories for the 1990s: 1990s: Development and individual differ differDevelopment, 63, 63,1-19. 1-19. ences. Child Development, Scarr, S. (1993). Biological and cultural diversity: The legacy of Darwin for development. Child Development, 1333-1353. Development, 64, 64,1333-1353. & McCartney, K. K. (1983). How people make their own environments: A theory of of Scarr, S., & genotype -+ -* environment effects. Child Development, Development, 54, 54,424-435. 424-435. disorders in in Thomas, A., & Chess, S. (1977). Temperament and development and behavior disorders children. New York: New York University Press. children.

CHAPTER 16 16

PSYCHOPHYSIOLOGICAL ApPROACHES APPROACHES T O PERSONALITY TO RUSSELL G. GEEN RUSSELL G. GEEN UNIVERSITY OF OF MISSOURI MISSOURI UNIVERSIlY

I. I. INTRoDucnoN INTRODUCTION Theories based on biological processes are among the oldest approaches to explainexplain The Roman physician Galen of Pergamon (A.D. 130-2(0) 130-200) ing human personality. The expanded the the theory of the the four humors of the body, an idea originating in the writings of Hippocrates and Empedocles, to stipulate the existence of four temperatempera theo ments. In some ways the theory of Galen anticipated modem temperamental theories of personality (H. (H. J. J. Eysenck, 1981; 1981; Stelmack Stelmack & & Stalikas, Stalikas, 1991). 1991). The Thehumoral humoral theory dominated Western medicine and and provided the principal explanations for somatic processes and and dysfunctions, until the the 17th 17th century; century; itit was was eventually eventuallyreplaced replaced in the 18th century by theories that attributed these processes to the properties the 18th and functions of the nerves (Drinka, 1984). "Nervousness" thereafter became the preferred explanation for the etiology of such disorders as hysteria, hypochondria, and melancholia (e.g., (e.g., Jackson, Jackson, 1986; 1986; Veith, Veith, 1965). 1965). One observation made made during and One observation this period period seems particularly particularly relevant relevant to to the the hypothesized link between personality personality this and psychophysiology: psychophysiology: patients' reports of of irritability irritability or variations in in mood mood were and patients' reports or variations were often accompanied accompanied by by reports reports of of somatic somatic complaints complaints and symptoms (Fahrenberg, (Fahrenberg, often and symptoms 1992). Such aa correlation correlation was was attributed attributed to to underlying underlying individual differences in 1992). Such individual differences in nervousness; translated translated into into more more modem modem terminology, terminology, the observation might might be nervousness; the observation be that aa state state of of autonomic autonomic activation activation is is often often accompanied accompanied by by negative negative affectivity. affectivity. that 387 IIANDBOOK OFOF PERSONALITY PSYCHOLOGY HANDBOOK PERSONALITY PSYCHOLOGY

387

CormGIIT ComucHT C © 1997 1997 BY BY ACADEMIC ACADEMIC PlESS PRESS.. ALL IIGIITS OF REPRODUcnON IN ANY FORM RESERVED.

ALL RIGHTS OF REPRODUCHON IN ANY FORM RESERVED.

388 388

RUSSELL RUSSELL G. G.GEEN GEEN

The study of of biological factors factors in personality personality was facilitated facilitated by the emergence emergence developed electro of of psychophysiology psychophysiology in the 19th century. Clinical interest interest in newly newly developed electrotherapies led to early studies by Vigoroux and Fer6 Fere in which the electrodermal electrodermal response was measured measured in samples of of hysteric and nonhysteric patients (Neumann (Neumann & Blanton, 1970). These studies probably represent represent the first first attempts attempts at what we today call the psychophysiological psychophysiological approach to personality. In the writings on humoral and neurological of individual differences neurological theories of differences we see a prefiguration prefiguration of of the essential elements of of modern modern approaches to the of personality. The major major assumptions of of this approach approach have been biological bases of summarized by Gale (1987). A A few few of of these assumptions may be noted: summarized 1. Accounts of human behavior can be considered 1. considered adequate only if they include individual variations. transmitted genetically, account account for for much of of 2. Biological factors, which are transmitted this individual variation. These biological factors are manifested manifested in a number number of of complex complex ways in 3. These the neuroanatomy, biochemistry, and neurophysiology neurophysiology of of the individual. These represented in the person's systems interact with each other in many ways and are represented experience, behavior, and psychophysiology. psychophysiology. psychoevolutionary 4. Individual variations develop within a psycho evolutionary context. They have adaptive significance significance for for the species and are therefore therefore observed observed across not of human human cultures, but also in other other species that that have developed developed only the range of evolutionary pressures. under the same evolutionary 5. Living systems are characterized by transmission, storage, allocation, and of energy. The construct construct of of arousal arousal has heuristic heuristic value in describing discharge of biological links to personality, personahty, even though its explanatory explanatory status is debatable. 6. Living systems are not passive receptors receptors of of environmental environmental or biologically engendered engendered influences. They use feedback feedback from from the environment environment and from from the consequences of their own actions to adapt and maintain maintain a stable equilibrium. The relation of of individual difference difference variables to biological disposition disposition is therefore therefore not simple and reactive, but transactive and complex. We will return of their their implications for return to some of of these points, and to some of for evaluating research findings, findings, in subsequent subsequent sections of of this chapter. For For now, how howconsideration of of a major major ever, we will concentrate on item S, 5, because it leads to a consideration viewpoint viewpoint in the study of of the biological bases of of personality: the arousability arousability hy hypothesis. pothesis.

U. THEORIES OF AROUSABILITY II. Theory A. Eysenck's Deory

The modern era of of interest in the biological approach approach to personality began with the pioneering work 19408. Eysenck's Eysenck's original theory work of of H. J. Eysenck Eysenck in the late 1940s.

CHAPTER 16 16 CHAPTER

PSYCHOPHYSIOLOGICALApPROACHES APPROACHES PSYCHOPHYSIOLOGICAL

389 389

of personality structure, which was based on factor analysis of of a number of of existing of psychometric instruments, stipulated two factors: a general activity factor labeled Extraversion-Introversion Extraversion-Introversion (E-I) and an emotionality factor designated Neuroticism (N). From the beginning, H. J. Eysenck (1947) sought to locate the bases for these factors in underlying biological processes. At first the basis for E-I was thought to lie in the Pavlovian constructs of excitation and inhibition, with introverts showing a general tendency toward excitation and extraverts extraverts an overall tendency toward toward inhibition. This explanation relied on a metaphoric model of of brain physiology derived mainly from Pavlov which did not link Unk E-I to any specific anatomical or physiological physiological processes. The study of of the psychophysiological basis of of personality took a major step The Biological forward in 1967 with the publication of H. J. Eysenck's of Biological Basis of of Personality. In this book Eysenck described a revised version of the theory, along of studies designed to test the new approach both behaviorally behaviorally and with a review of physiologically. The revised theory located both E-I and N in specific cortical and of the brain. Central to the theory was the construct of of general subcortical centers of diffuse energization of of the body's major response systems correlated arousal, a diffuse physiological activity (e.g., electrodermal electrodermal conductance, cortical with observable physiological activation, cardiovascular activity). Following an approach taken by general arousal theorists in the 1950s (e.g., Lindsley, 1957), Eysenck considered general arousal to represent nonspecific nonspecific activity in the ascending reticular activating system (ARAS). represent Extraverts were described as people with relatively high thresholds for ARAS ARAS activation, with the result that they are relatively less aroused than introverts in of stimulation. The ARAS ARAS is activated by both external response to given levels of stimuli and ascending and descending pathways to the limbic system. Thus general of either external stimulation or emotional activity arising arousal can be the result of of individual within the person. The revised theory described Neuroticism in terms of differences in the activity of the limbic system, manifested in emotional arousal in differences stressful or threatening situations. Given the neural connections be beresponse to stressful tween the limbic system and the ARAS, a moderate positive correlation between often been reported. Finally, H. J. Eysenck (1967) related E-I to E-I and N has often of an optimal level of of arousal, another idea behavior by invoking the concept of popular in the 1950s and 1960s: extraverts were described as engaging in highly arousing activities and expressing expressing preference for complex and exciting stimuli in order to elevate their arousal levels to some desired intermediate range. The number of of empirical investigations stimulated by Eysenck's theory over the 45 years since its original publication is literally in the thousands. To a large extent the theory's popularity is due to several notable strengths. It is built on a solid theoretical theoretical base. It involves rigorously developed and validated psychometric assessment. It generates predictions of of not only psychophysiological psychophysiological differences differences differences that along the E-I and N dimensions, but also experiential and behavioral differences psychophysiological differences, and to each other, in theoretically are related to the psychophysiological relevant ways.

390 390

RUSSELL G. GEEN G.GEEN

B. Gray's Gray's Theory Theory B. 1981) proposed a modification Gray (1972, (1972,1981) modification of Eysenck's theory of E-I that attributes the individual differences differences to underlying activity in both the ARAS and the medial septal area, the hippocampus, and the orbital frontal frontal cortex. The revised theory follows from from the the observation small injections injections of of sodium sodium amobarbital amobarbital in in rats rats follows observation that that small leads to behavior has been characterized as as similar similar to to that that of of human human extraverts: extraverts: leads to behavior that that has been characterized impulsive, relatively uninhibited, uninhibited, and and generally generally not not susceptible to threats threats of of punish punishimpulsive, relatively susceptible to ment (Gray, 1972). 1972). Similar Similar "extravert-like" "extravert-like" behavior behavior has has been been observed observed in in animals animals ment (Gray, that have undergone undergone lesions lesions in in the the septal septal area area or or the the orbital orbital frontal frontal cortex. cortex. The The that have upshot of of such such findings is that that Gray Gray proposes proposes aa septal-hippocampal-frontal septal-hippocampal-frontal cortex cortex upshot findings is system in which which the the biological biological basis of introversion introversion and and extraversion extraversion is is to to be be found. found. system in basis of In aa later later version version of of the the theory theory (e.g., (e.g.. Gray, Gray, 1981), 1981), two two antagonistic antagonistic control control In centers in in the the central central nervous nervous system system are are postulated: postulated: aa behavioral behavioral activation activation system system centers (BAS) and and aa behavioral system (BIS), (BIS), the the latter latter of of which which was was described described (BAS) behavioral inhibition inhibition system 1972 paper paper cited cited above. Whereas the the BIS controls passive passive avoidance, avoidance, the the BAS BAS in the the 1972 in above. Whereas BIS controls is set set in in motion motion by by signals signals of of reward reward or or non nonpunishment. controls active active avoidance avoidance and and is controls punishment. is not not motivated motivated by by fear fear or or anxiety, anxiety, which which are are associated associated Active avoidance, avoidance, therefore, therefore, is Active with the BIS, but rather by incentive motivation. As has already been noted. Gray with the BIS, but rather by incentive motivation. As has already been noted, Gray is a high level of activity (1972) proposes that the physiological basis for introversion (1972) proposes that the physiological basis for introversion is a high level of activity in the the neural neural centers centers that that constitute constitute the the BIS. BIS. Introverts, Introverts, therefore, therefore, are are especially especially in susceptible to stimuh that signal punishment or frustrative nonreward. Extraverts, susceptible to stimuli that signal punishment or frustrative nonreward. Extraverts, on the the other other hand, are more more sensitive sensitive to to signals signals of of reward reward or or nonpunishment. nonpunishment. on hand, are

C. Related Related Constructs Constructs C. 1. Sensation-Seeking /. Sensation-Seeking

One conclusion that came out of of research on the effects effects of sensory deprivation that was conducted in the 1950s and 1960s is that people differ differ from from one another in the extent to which they seek out exciting events in their surroundings (Zuckerman, 1969). The The individual individual difference difference variable variable associated associated with with these these behaviors behaviors was was labeled labeled sensation-seeking sensation-seeking (SS) (Zuckerman, 1971). Zuckerman (1984, 1990) assumes a biological basis for sensation-seeking sensation-seeking and argues that the trait has a psychoevolu psychoevolutionary basis. The theory of SS SS therefore therefore generates predictions about both behavior and physiological activity. Zuckerman once proposed that sensation-seeking is rere lated to the quest for an optimal level of stimulation, a view that suggested a commonality between that trait and extraversion. That view is no longer held, as commonality will be noted in a subsequent section of this review.

2. Impulsiveness Impulsiveness and and Augmenting/Reducing 2. Augmenting/Reducing The traits of extraversion-introversion extraversion-introversion and sensation-seeking both involve to some degree individual differences differences in impulsive behavior. H. J. Eysenck Eysenck (1967) considered considered impulsiveness to be one of the two two main characteristics of the extraverted extraverted personality

CHAPTER CHAPTER 16 16

PSYCHOPHYSIOLOGICAL PSYCHOPHYSIOLOGICAL ApPROACHES APPROACHES

391

sensationand Zuckerman found tendencies toward disinhibition to be an element in sensation Some investigators have have concluded concluded that that the the individual individual differences differences in in arousa arousaseeking. Some attributed to extraversion-introversion extraversion-introversion are indicators of differences differences in impulbility attributed impul 1980; Schalling & A Asberg, sivity (Revelle, Humphreys, Simon, & Gilliland, 1980; sberg, 1985). The concept of augmenting/reducing was introduced by Petrie (1967) on the of studies of the degree to which subjects overestimate or underestimate basis of the intensity of a stimulus following prolonged stimulus bombardment. Using a methodology involving tactile stimulation to the hand, Petrie found that some people—designated "reducers"-tend "reducers"—tend to underestimate stimulus intensity under people-designated ("augmenters") tend to overestimate. Augmenting/ such conditions whereas other ("augmenters") differreducing, though assessed peripherally, was assumed to reflect underlying differ ences in the central nervous system. Subsequent studies (e.g., (e.g.. Sales, 1971) showed that augmenters tend to behave much as introverts and that reducers show many of the same characteristics as extraverts (cf. H. J. Eysenck & Eysenck, 1985). 1985). More recently, the augmenting/reducing dimension has been operationally defined defined in terms of individual differences in evoked responses at the level of the cerebral terms of individual differences in evoked responses at the level of the cerebral cortex (Buchsbaum & Silverman, 1968), and has been shown to be related to cortex (Buchsbaum Silverman, 1968), and has been shown to be related to both sensation-seeking and impulsivity. Research pertaining to this conclusion both sensation-seeking and impulsivity. Research pertaining to this conclusion is is reviewed in in aa later later section. section. reviewed 3. 3.

Strength oof the Nervous NervousSystem System Strength f the

Certain similarities have been observed between Eysenck's dimension of extraver extraverstrength of of the sion-introversion and the concept, first described by Pavlov, of strength nervous 1964; Strelau, 1987). 1987). This term refers to the strength of of nervous system (Gray, 1964; excitatory processes generated in the central nervous system by increasing levels of stimulation. Pavlov taught that excitation in the central nervous system is directly related to stimulus intensity over moderate to high stimulus intensity levels, but that as stimulus intensity is increased beyond these levels, the point is eventually reached at which the nervous system exceeds its working capacity and goes into a progresstate of inhibition. Beyond that level, increasing stimulus intensity evokes progres less excitation. The level of stimulus intensity at which excitation gives way sively less to inhibition is called the threshold transmarginalinhibition. inhibition. threshold of of transmarginal Psychologists working in the Pavlovian tradition have used several experimen experimental procedures to classify classify subjects according to the intensity of stimuli needed to bring about transmarginal inhibition (Strelau, 1983). 1983). Those Those in in whom whom transmarginal transmarginal inhibition is induced by a relatively less intense stimulus are classified as having "weak" nervous systems; those who require relatively more intense stimulation are designated "strong." Strelau (1987) has has observed that persons in in the former former category appear to respond to stimulation in the same way as introverts, augmenters, low sensation-seekers, and low impulsives, whereas whereas those those of the the latter category category resemble resemble extraverts, reducers, high sensation-seekers, and high impulsives. Strelau (1983) reactivity to this list, with high has added his own variable of high versus low reactivity reactives showing the properties of the weak nervous system.

392 392

RUSSELL G. G.GEEN GEEN

m. III. CRmCISM CRITICISM OF OF mE THE AROUSABILITY AROUSABILITY HYPOmESIS HYPOTHESIS Although the concept of general arousal continues to be used among psychologists and to to retain retain aa certain certain heuristic heuristic value value in in theory theory construction, it has been largely largely and rejected by psychophysiologists at least since 1967, when Lacey published a widely cited critique. The core of Lacey's (1967) argument was that indicators of activity in the various physiological systems of the body (e.g., cortical, motor, autonomic) are at best weakly intercorrelated, and that activity in these systems and behavioral activation are likewise not tightly coupled. In addition, certain indicators of physio of physiological activity show increases in the same situations that evoke decreases in other measures, aaphenomenon thatLacey Laceycalled calleddirectional directionalfractionation. fractionation.Finally, Finally,individ individphenomenon that uals manifest a high degree of autonomic thethe profile of of auto autonomic response responsespecificity: specificity: profile autonomic responding for a given person tends to be relatively stable across time and situations but but often often to to be be different different from from the the profiles profiles of of other other people undergoing situations the same same experiences. experiences. the Problems such as these suggest that the arousability hypothesis cannot be tested adequately in limited studies that involve only a few psychophysiological variables or situations. Larger investigations that involve behavioral and subjective measurement along with psychophysiological indicators, and which include a wide array of of situational situational manipulations, manipulations, are are necessary necessary (Fahrenberg, (Fahrenberg, 1992). The The results results array of studies of this type offer no support for the notion of general arousability. A growing number of psychophysiologists have proposed that if arousal is to be a useful construct, it will be in the context of a multiple-systems approach. For finding no evidence for a second-order construct of example, Myrtek (1984), whilefinding autonomic lability, found evidence of individual reactivity in a number of primary systems (e.g., heart rate, blood pressure, respiration volume). The key to the role played by energy transfer in the living system may lie in the complex ways in which these various systems interact with each other. In effect, such an emphasis would call for change in the basic underlying model that has been used to conceptualize arousal-from arousal—from a simple mechanistic one that Venables (1984) has called the "foot "footon-the-accelerator" view to one based more on the principles of cybernetics and control mechanisms. The several somatic systems that are characterized by individindivid ual differences differences in activation interact in such a way that some may activate, and some inhibit, other systems. Actually, such a control model has been implicit in discussions of human action for some time. Venables (1984) has pointed out, for example, that John Hughlings Jackson wrote on the hierarchical levels of brain organization more than a century ago, ago, and and that he emphasized that "what **what happened at any one level was determined by interaction between levels" (Venables, 1984, 1984, p. 138). More recently, Claridge (1967) has developed a sophisticated model of arousal organized around two hypothetical systems. One, the tonic arousal arousal system, system, maintains maintains the the person's person's gross level of activation in response to stimulus inputs. The other, the arousal arousal system, has two regulatory functions. One is to control the level of modulating system, modulating of

16 CHAPTER 16

PSYCHOPHYSIOLOGICAL PSYCHOPHYSIOLOGICALApPROACHES APPROACHES

393 393

activity in the tonic system; the other is to integrate the input input to both systems through either facilitation The systems work together facilitation or inhibition of of stimulation. The antecedent stimulus to maintain a balance or steady state of activation, even when when antecedent moduconditions vary. A study by Birchall and Claridge (1979) indicates a possible modu augmenting/reducing of lating mechanism involving augmenting/reducing of the cortical evoked evoked response and the hypothesized hypothesized tonic arousal arousal system. They found found that that subjects subjects who manifested manifested relatively high levels of skin conductance tended to be reducers as the intensity relatively of tended of incoming stimulation stimulation increased, whereas those low in skin conductance of conductance level tended to be augmenters. If, If, as Birchall and Claridge Claridge reason, skin conductance conductance level is a reflection reflection of of a hypothetical hypothetical arousal system, this finding suggests that the augmenting/reducing augmenting/reducing function function serves to dampen dampen or or sharpen amount of of sharpen the amount stimulation that the person receives from from the environment. environment. IV. PSYCHOPHYSIOLOGICAL STUDY OF PERSONALITY IV.

In spite of arousability hypothesis of the criticisms noted in the preceding section, the arousability generated a large amount amount of research research on personality. personaUty. Multimodal investigations has generated involving the role of interactions among systems are of personality in the complex complex interactions outset that the study of of activity in the relatively rare. It must be recognized at the outset subjects rests on indirect central and autonomic nervous systems among human human subjects inferred from evidence. In general, arousal and emotionality have been inferred from three types of data: verbal reports of of experience (e.g., Endler, Edwards, & Vitelli, 1989; Thayer, of 1970); observation observation of of behavior that is linked to underlying underlying brain brain processes on and the measurement measurement of of psy psytheoretical grounds (e.g.. (e.g., Gray, 1964; Strelau, 1983); and focused on of these three chophysiological indicators. This review will be focused on the last of (for earlier reviews, see Stelmack, 1981, 1981,1990; 1990; Stelmack & Geen, 1992). (for A. Extraversion-Introversion Extraversion-Introversion If introverts are more arousable than extraverts because of of differential differential thresholds thresholds of reactivity in the ARAS, we might expect that the clearest clearest evidence of of E-I of differences would be found found in electrocortical measures. The The majority of studies on differences majority of extraversion extraversion and arousal have involved the use of of such such measures. Overall, whereas some of of the findings findings support support Eysenck's theory, the results have been mixed and complex, reflecting reflecting the possible operation of of several situational situational moderator moderator variables. L Electroencephalograph (EEG) 1. Electroencephalograph (EEG) E E G depicts a complex waveform waveform consisting of frequencies. The most The EEG of several frequencies. commonly studied bandwidth bandwidth is the alpha wave (7.5 to frequency commonly to 13.5 Hz), with frequency dependent measures. In general, an alpha wave showing and amplitude as the main dependent frequency is associated with relatively low cortical arousal. large amplitude and low frequency high-frequency activity is taken as an indicator indicator of If Low-amplitude, high-frequency of high arousal. If differences are linked to cortical arousal, introverts should should show show higher basal E-I differences

394 394

RUSSELL G.GEEN GEEN RUSSELL G.

or resting levels of of low-amplitude, high-frequency high-frequency activity than extraverts. Extra Extravert-introvert differences differences in EEG activity have been the subject of several reviews (e.g., Gale, 1981, 1983, 1987). In each of these reviews, 1981,1983,1987). reviews. Gale has shown that much of the research on the problem has been flawed by weak methodologies and lack of theoretical sophistication and that replicability across studies is uncommon because critical situational variables have not been controlled. However, even studies based offer much support for the notion of of higher basal on sounder methods fail to offer of cortical arousal in introverts (e.g., Golding & Richards, 1985; Matthews & levels of Amelang, 1993). 2. Evoked Evoked Responses Responses (ER) (ER)

a. Cortical Cortical Evoked Evoked Response. Response. Somewhat stronger evidence of E-I differ differences is found in studies of of evoked responses (ER). The average evoked cortical potential is a momentary change in brain wave activity that occurs in response to complex brief presentations of of a stimulus or during cognitive processing. It presents a complex waveform consisting of of successive positive and negative peaks. ERs are superim superimwaveform posed on the EEG pattern and are extracted from EEG records by a technique of of such responses. In general, earlier peaks in the averaging over a large number of of the ER are most likely determined by physical characteristics of of the waveform of eliciting stimulus and later peaks (later than 300 ms after the onset of of the stimuls) reflect attentional and endogenous cognitive processes (Hillyard & Hansen, 1986). Several variables influence influence ERs, among them stimulus intensity, rate of of presentation, and overall arousal levels. It follows from the latter that extraverts and introverts should show different different levels of of amplitude over various segments of of the ER waveform. Stimulus frequency was found to interact with E-I to influence the amplitude of of the N N1P2 of the waveform waveform evoked approximately 90901P2 component (the section of 250 ms after the stimulus) in a study by Stelmack, Achorn, Achom, and Michaud (1977). When a tone of of low frequency (500 Hz) was used, the N1P N1P22 amplitude in introverts was greater than that of extraverts, but no E-I differences differences were found when a tone of of higher (8000 Hz) frequency was used. This finding could have reflected E-I differences differences in sensitivity to the low-frequency low-frequency tone, but it could also be explained in terms of of heightened attention on the part of introverts or more rapid habituation habituation to the tone on the part of extraverts. To test these possibilities, Stelmack and Achorn-Michaud (1985) presented subjects with 500-Hz tones under instructions Achom-Michaud N1P2 either to attend to the tones or to ignore them. Introverts displayed a larger N1P2 amplitude than extraverts in response to the first tone in a series of of four in both differential the attend and the ignore conditions, ruling out the possibility that differential attention caused the personality difference. In addition, Stelmack and Achorn AchornMichaud found that whether the tones were presented in a repetitious way that fostered habituation or in a way that retarded habituation, E-I differences differences were fostered N1P22 amplitude of introverts seems to be linked to not affected. Thus the greater N1P low-frequency stimuli. their greater sensitivity to low-frequency

CHAPTER 16 16 CHAPTER

PSYCHOPHYSIOLOGICAL PSYCHOPHYSIOLOGICALApPROACHES APPROACHES

395 395

The results of a study by Bartussek, Diedrich, Naumann, and Collet (1993) of differences between introverts and extraverts within the addresses another facet of of research on the ER. This study was based on Gray's (1981) revision of of context of of Eysenck's theory, according to which introverts are more reactive to signals of of punishment than extraverts, whereas extraverts are more susceptible to signals of rewards than introverts. Subjects were required to guess which of two tones varying in frequency (800 or 1600 Hz) would be delivered on each trial. The occurrence of the tone then indicated whether the guess had been correct or incorrect. Each correct guess was followed by a monetary reward and each incorrect guess by a loss of money. Analysis of of the P2 wave (relative to the baseline amplitude) showed that extraverts manifested manifested greater amplitude than introverts in response to tones that signaled a correct guess and subsequent reward, and that introverts showed a greater amplitude than than extraverts extraverts when when the the tone tone indicated indicated that that the the guess guess had had been been greater amplitude incorrect. Similar E-I differences in ER amplitude were found in the N2 wave and incorrect. Similar E-I differences in ER amplitude were found in the N2 wave and in a later segment designated P3 (details follow). However, in each of the latter, in a later segment designated P3 (details follow). However, in each of the latter, E-I differences differences were were moderated moderated by by other other variables variables in in the the experiment. experiment. Nevertheless, Nevertheless, E-I the data for the P2 component support Gray's extension of Eysenck's approach. the data for the P2 component support Gray's extension of Eysenck's approach. differences have also been found in later components of of the cortical ER E-I differences wave. The P3 is a late positive spike (approximately 300 ms after the stimulus) that is evident during tasks in which target events must be detected or in which the feedback signal to the subject. Amplitude of of the P3 is a function event serves a feedback of several several variables, variables, among among which which are are the the importance importance of of the the stimulus stimulus event, event, the the of information-processing demands or not not more more than than one one stimulus stimulus must must be be information-processing demands (whether (whether or processed simultaneously), and and the the novelty novelty or or unexpectedness unexpectedness of of the the event. processed simultaneously), event. P3 P3 amplitude therefore the allocation of limited attentional resources resources to to the the amplitude therefore reflects reflects the allocation of limited attentional task. Because introverts have been shown to to manifest manifest better better performance in monot monottask. Because introverts have been shown performance in onous signal-detection tasks (e.g.. Gauge, Geen, & Harkins, 1979), and because onous signal-detection tasks (e.g., Gange, Geen, & Harkins, 1979), and because sustained attention attention appears appears to to be be related related to to higher higher levels levels of of arousability arousability (e.g., (e.g., sustained Matthews, Davies, & HoUey, 1990), introverts should show greater P3 amplitudes Matthews, Davies, & Holley, 1990), introverts should show greater P3 amplitudes than extraverts. extraverts. than Subjects in a study by Daruna, Karrer, and Rosen (1985) took part in a lengthy of guessing which of two tones would occur on each trial. and monotonous task of As expected, introverts showed higher P3 amplitudes than extraverts, leading the authors to conclude that introverts allocated more attention to the task. Similar methodology findings come from a study by Ditraglia and Polich (1991). Using a methodology of Daruna et al, similar to that of aI., these investigators presented the tone pairs in two separate blocks separated by approximately 2 min. Although they found no overall E-I differences differences in P3 amplitude, they found an E-I difference difference by blocks interaction: extraverts showed a decline in P3 amplitude across blocks whereas introverts did not. This finding suggests that E-I differences differences may be due to a progressive decrement in attention to the stimuli among extraverts. in attention to the stimuli among extraverts.

b. Brainstem Brainstem Auditory Auditory Evoked Evoked Response. Another evoked response has of the brainstem. This brainstem auditory evoked been measured at the level of

RUSSELL G. G.GEEN RUSSELL GEEN

396

response (BAER) consists of of a series of of seven positive waves within the first 10 ms of of a brief brief auditory auditory stimulus (e.g., a click). dick). Like the cortical evoked response, through averaging over a large number number of of trials. The seven waves are are it is detected through evaluated in terms of of both their absolute absolute latency (i.e., from the evaluated (Le., the elapsed time from peak of of the wave) and their interpeak interpeak latencies (Le., (i.e., the time that stimulus to the peak from one wave peak to another). These are preferred preferred to measures of of elapses from of the relative unreliability of of the latter. It is widely believed believed amplitude because of that Waves I and II reflect reflect mainly activity in the auditory nerve; that Waves III, that and that Waves Waves VI and and IV, and V indicate activity in centers within the brainstem; and may reflect reflect the beginning of of activity within the cortex. VII may The evidence relating E-I differences differences to the BAER BAER is mixed. In an early study, Andress and Church (1981) found found that introverts showed shorter Wave I-V I-V interpeak latencies than than extraverts in response to relatively loud (80 dB) clicks chcks but but interpeak not to clicks of of lower intensity. In another another early study, Campbell, Baribeau-Brown, Baribeau-Brown, not and Braun (1981) found found no E-I differences differences in BAER BAER activity. Stelmack Stelmack and Wilson found that introverts introverts displayed shorter Wave V latencies than extraverts extraverts in (1982) found of 2000 2000-- and 4000 4000-Hz revealed response to tone pips of -Hz frequencies, but that extraverts revealed Of greater greater importance was Stelmack Stelmack shorter latencies when a 500-Hz pip was used. Of finding of of significant significant positive correlation between between extraversion extraversion and and Wilson's finding of either 75, 75,80, dB intensity were presented. This Wave I latencies when clicks of 80, or 85 dB finding showed showed that introverts introverts manifested manifested shorter shorter Wave I latencies than extraverts. finding reflects activity at the level of of the auditory nerve, this Inasmuch as Wave I activity reflects finding suggests greater peripheral sensitivity of of introverts without without having having a bearing bearing finding on the hypothesized hypothesized E-I difference difference in central central arousal. In a more recent investigation, investigation, Stelmack, Campbell, and and Bell (1993) tested for for E-I differences differences in BAER B A E R activity during presentation of of clicks in the 80- to 9O-dB 90-dB range, and, although some sugges suggesshorter latencies in introverts was found, found, none of of the differences differences reached reached tion of of shorter of statistical significance. significance. the .05 level of One study has shown relatively strong support for for a hypothesis relating E-I E-I to the to the BAER. BAER. Bullock Bullock and and Gilliland Gilliland (1993) (1993) delivered delivered clicks clicks at at an an intensity intensity 80 80 dB dB above above subject's subject's auditory auditory threshold threshold and and found found that that Wave Wave V V latency latency was was shorter shorter in in introverts in extraverts. I-III and introverts than than in extraverts. In In addition, addition, introverts introverts had had shorter shorter Wave Wave I-III and Wave extraverts. Bullock Wave I-V I-V interpeak interpeak latencies latencies than than extraverts. Bullock and and Gilliland Gilliland also also adminis administered tered caffeine, caffeine, in in either either aa high high or or aa moderate moderate dosage, dosage, to to two two groups groups of of their their subjects subjects while third in condition. They while running running the the third in aa placebo placebo condition. They found found that that the the subjects subjects who who had had received received caffeine caffeine in in either either the the moderate moderate or or the the large large dose dose group group manifested manifested faster faster Wave Wave I-III I-III and and Wave Wave I-V I - V latencies latencies than than those those given given the the placebo. placebo. The The effects effects of of caffeine, caffeine, aa known known source source of of arousal, arousal, therefore therefore showed showed the the same same pattern pattern of of interpeak E-I. This interpeak latencies latencies as as individual individual differences differences in in E-I. This convergence convergence of of effects effects from the argument that E-I from different different operations operations strengthens strengthens the argument that E-I differences differences are are related related arousability. to arousability. 3. 3.

Audiomotor Reflex Reflex Audlomotor

The audiomotor audiomotor reflex reflex (AMR) is a phasic startle reaction elicited by sudden visual, auditory, auditory, cutaneous, cutaneous, or or electrical electrical stimulation. stimulation. It It involves involves aa number number of of synaptic synaptic

CHAPTER CHAPTER 16 16

PSYCHOPHYSIOLOGICAL APPROACHES PSYCHOPHYSIOLOGICAL ApPROACHES

397 397

connections in the brainstem and subsequent effector effector processes, one of the latter orbicularisoculi oculimuscle. muscle. being activation of the facial nerve and, through it, of the orbicularis As the intensity of the eliciting stimulus is increased, the amplitude and probability of the AMR increase, and AMR latency decreases. Britt and Blumenthal (1991) have shown that individual differences differences in E-I moderate the stimulus intensity effect. effect. Introverts showed a clear discrimination in in the latency of eyeblink AMRs to auditory stimuli of 60 and 85 dB, with the latter evoking the shortest latency reaction. Extraverts showed no shorter latencies with the more intense stimulus than with the less intense one. Differences Differences between extraverts and introverts in the amplitude of the auditory AMR has been shown by Ljubin and Ljubin (1990), who found found manifested a greater greater response amplitude than extraverts to a stimulus that introverts manifested of 100 dB intensity. 4. Electrodermal Activity Activity

A large body of evidence indicates that extraversion interacts with levels of environ environ1983). The mental stimulation to influence electrodermal conductance (e.g., Smith, 1983). significance significance of skin conductance activity for for personality in general, or for Eysenck's theory of E-I in particular, has not been spelled out in detail, but there is some is initiated in situations involving aversive motiva motivaevidence that sweat gland activity is The relationship, relationship, moreover, moreover, appears appears to to be be linear: linear: as asthe theaversiveness aversivenessof ofstimula stimulation. The 1983). In addition, tion increases, electrodermal activity increases also (Fowles, 1983). orientincreased skin conductance has long been recognized as a component of the orient ing response to less intense stimulation (Sokolov, 1963). Thus, skin conductance activity is part of a broadly conceived "arousal" response in both defensive and differential arousability orienting behavior. We might expect, therefore, that the differential of extraverts and introverts should include an electrodermal component. The study of the role played by E-I involves systematic manipulation of of environmental conditions to elicit varying degrees of electrodermal activation and observation of E-I differences differences within that setting. Three methods of manipulating arousal have been reported. The first involves administration of stimuli stimuU of varying intensities (e.g., Fowles, Roberts, & Nagel, 1977; 1984; Smith, Wilson, & 1977; Geen, 1984; Davidson, 1984). 1984). The The effect effect of of such such treatments treatments isis to to increase increase arousal arousal monotonically. monotonically. The second consists of giving subjects a stimulant drug such as caffeine caffeine (e.g., (e.g.. Smith, Rypma, & Wilson, 1981; 1981; Smith, Wilson, & Jones, 1983). 1983). It is assumed that the arousing effect effect of caffeine caffeine summates with that of the eliciting stimulus. The third method differs differs from the other two in that instead of increasing stimulation above a baseline level it reduces stimulation below that level. level. This method consists of giving the subject a preliminary signal before the stimulus for the electrodemial electrodermal response (Smith et aI., al., 1984; 1984; Smith, Rockwell-Tischer, & Davidson, Davidson, 1986). 1986).By By remov removing uncertainty regarding the onset of the stimulus, the signal should reduce reduce arousal. arousal. Most of the studies discussed below involve the use of one or more of these three methods of arousal manipulation. Geen (1984) studied differences differences in electrodermal activity among extraverts and introverts who were either allowed to choose the intensity of stimuli they

398 398

RUSSELL G. GEEN RUSSELL G.GEEN

received or were received (ostensibly (ostensibly as as aa background background stimulus) stimulus) or were assigned assigned stimuli stimuli by by the the experimenter. Some persons of each personality classification were assigned a experimenter. Some persons of each personality classification were assigned a stimu stimulus equal equal in intensity to that chosen chosen by yoked person person of of the the other other type. type. The lus in intensity to that by aa yoked The dependent variable dependent variable was was the the number number of of specific specific skin skin conductance conductance responses responses occurring occurring during aa 4-min period just first trial trial of during 4-min period just before before and and during during the the first of aa learning learning task. task. Extraverts selected noise of a higher intensity than introverts. Extraverts Extraverts and intro introbackground stimulus levels were equally equally aroused response verts who had chosen background aroused in response stimuU, as were extraverts extraverts and introverts introverts who had been to their respective chosen stimuli, assigned stimuli at those levels. Introverts who were given noise at an intensity other groups, and extraverts extraverts chosen by extraverts were more aroused than all other intensity were the least aroused. Thus assigned noise at introverts' chosen level of intensity of stimulation that both introverts and extraverts choose, when possible, a level of of arousal is approximately approximately the evokes an optimal level of arousal, and that level of for both groups. same for 5. 5.

Cardiovascular Activity Activity Cardiovascular

A small number number of of studies have addressed the question of whether introverts and of whether A differ in cardiovascular activity. The results are mixed and inconclusive. extraverts differ found evidence of of E-I differences differences in skin conductance conductance and vasomotor vasomotor Although they found Pickard (1979) found found no comparable comparable activity, Stelmack, Bourgeois, Chien, and Pickard differences in heart rate. However, Gange Gauge et al (1979) observed observed higher heart rates differences of an experiment experiment on visual vigilance, in introverts than in extraverts in all conditions of including one in which no task was performed. performed. These These investigators found found no E-I including differences in heart rate during a baseline rest period. Their Their findings findings may therefore therefore differences differences reflect reflect to some extent a reaction on subjects' part to indicate that E-I differences any sort of experimental procedures. To To understand understand the the role role played played by by E-I E-I in in cardiovascular cardiovascular activity, activity, however, however, it it is is necessary necessary first first to to consider consider the the functional functional significance significance of of heart heart rate rate change. change. Two Two viewpoints the E-I and Feij viewpoints on on this this are are represented represented in in the E-I literature. literature. Orlebeke Orlebeke and Feij (1979) (1979) have have taken taken the the position position that that phasic phasic cardiac cardiac acceleration acceleration is is aa component component of of the the defensive defensive response response whereas whereas phasic phasic deceleration deceleration is is aa component component of of the the orienting orienting response response (Graham, (Graham, 1979). 1979). A A similar similar argument argument is is found found in in the the intake-rejection intake-rejection hypothesis Jennings, 1986): is prompted hypothesis of of Lacey Lacey (cf. (cf. Jennings, 1986): cardiac cardiac acceleration acceleration is prompted by by condi conditions overstimulation that shutting out tions of of threat threat or or overstimulation that generate generate aa reaction reaction of of shutting out environ environmental by conditions conditions that that call mental stimulation, stimulation, whereas whereas deceleration deceleration is is evoked evoked by call for for atten attention tion and and taking taking in in stimulation. stimulation. If If we we assume assume that that introverts, introverts, relative relative to to extraverts, extraverts, tend of stimulation, tend to to be be stimulus-reducers stimulus-reducers who who seek seek low low levels levels of stimulation, it it follows follows that that introverts to high-intensity high-intensity stimuli introverts are are more more likely likely than than extraverts extraverts to to respond respond to stimuli with with of defense. On On the other other hand, and for for the same the cardiac acceleration indicative of reasons, extraverts should be more likely to show cardiac deceleration deceleration (i.e., orienting and stimulation) than less intense intense stimuli stimuli are and taking taking in in stimulation) than introverts introverts when when less are presented. presented. of this hypothesis, Orlebeke and Feij (1979) found found that that introverts In partial support of responded to a 60-dB, 1000-Hz greater heart heart rate acceleration acceleration than responded 6O-dB, 1000 -Hz tone with greater support the theory is the finding finding by Harvey and Hirschextraverts. Also tending to support Harvey and

CHAPTER 16 CHAPTER

APPROACHES PSYCHOPHYSIOLOGICAL ApPROACHES

399 399

man (1980) that introverts responded to highly aversive photographs (of (of persons who had died violent deaths) with greater immediate heart rate increases than extraverts, who showed no increase. A study by Richards and Eves (1991) also supports the hypothesis of greater defensive cardiac increase in introverts. In their study, subjects were classified classified as defensive "accelerators" or "nonaccelerators" on the basis of whether they showed large of accelerative changes or tended to show deceleration in response to a tone of for 1 s. Accelerators, who showed the defensive acceleration most 110 dB intensity for found to have lower scores than nonaccelerators on a measure of clearly, were found of (i.e., accelerators tended to be be introverts) and to score higher in in neurot neurotextraversion (Le., icism. B. Sensation-Seeking Sensation-Seeking B. & Eysenck, 1978) operationally The Sensation-Seeking Scale (Zuckerman, Eysenck, & defines tendencies to seek and engage in activities that provide excitement and risk. In addition to giving a total score for sensation-seeking, the instrument has four subscales that yield scores for (1) thrill and adventure seeking (TAS), (2) experience-seeking, (3) disinhibition (Dis), and (4) susceptibility to boredom. The subscales that have been most commonly associated with physiological activity have been T TAS TAS AS and Dis. Dis. The T AS scale measures liking for risky and arousing physical activities and the Dis scale assesses the quest for excitement through such actions as social drinking, sexual activity, and a general lowering of social constraints on behavior. The Psychophysiology o of /. The 1. fSensation-Seeking Sensation-Seeking a. The Optimal Arousal Hypothesis. Hypothesis, The relation of sensation-seeking to SS would appear standard psychophysiological measures is not clear. At first glance, SS extraversion-introversion as defined defined by Eysenck, so to have much in common with extraversion-introversion SS could be thought of as as one who is characteristically that the person who is high in SS underaroused and who seeks stimulation in order to attain an optimal level. This person would be comparable to the typical extravert. At one time Zuckerman (1969) took this position, but he has subsequently abandoned it largely because of of drug studies that failed failed to support the theory. In a critical experiment, Carroll, Zuckerman, and Vogel (1982) administered a placebo, a stimulant, or a depressant to subjects who were either high or low in SS, SS, and then assessed both mood and efficiency efficiency of performance. It was expected that high sensation-seekers would feel happiest and perform perform best after after having received a stimulant because the drug level. For the same reason, low would move arousal in the direction of the optimal level. perform best after after having sensation-seekers were expected to feel happiest and perform received a depressant. However, the results of the experiment showed that both performed best after after receiving a stimulant. high- and low- sensation-seekers felt and performed In addition, evidence from from experiments using psychophysiological psychophysiological measures for people high in SS to be underaroused. Although the does not show a tendency for

400 400

RUSSELL GEEN RUSSELLG. G.GEEN

findings involving electrodermal, cardiac, and cortical measures are complex and inconsistent, they suggest that people who are high in SS reveal an overall pattern of of responsiveness responsiveness similar to the orienting orienting response, whereas those who are low in SS tend under the same conditions to manifest a pattern similar to the defensive defensive or startle response.

b. fe. Electrodermal Electrodermal Activity. Activity. In an early study, Neary and Zuckerman (1976) presented subjects with a simple visual stimulus for each of 10 trials, after which they ran 10 trials with a different different stimulus. Male subjects who were high in SS emitted a greater number of of skin conductance responses than low sensation-seekers sensation-seekers on the first presentation of of each stimulus but not on subsequent presentations of either. Neither basal skin conductance level nor rate of habituation was affected affected effect of of SS was therefore a short-lived reactivity to a novel by SS differences. The effect stimulus that habituated rapidly. Similar results have been reported by others. Feij, between Orlebeke, Gazendam, and Van Zuilen (1985) found a positive correlation between amplitude on the first of three presentations of of a tone, SS and skin conductance amplitude found a similar correlation for the T TAS but not on two succeeding trials; they found AS Differences in phasic conductance responses to novel stimuli as a function function subscale. Differences of scores on the Dis subscale have been reported by Robinson and Zahn (1983). of However, not all studies have reported clearcut effects of SS on electrodermal difference between high and low activity. Ridgeway and Hare (1981) reported no difference sensation-seekers in the number of of skin conductance responses elicited by a tone. Falkenberg (1983) found that high sensation-seekers sensation-seekers showed Stelmack, Plouffe, and Falkenberg meaningful words greater electrodermal reactivity than persons low in SS when meaningful meaningful pictures were used. were used as stimuli, but weaker reactivity when meaningful of the stimulus may have an influence influence on the This finding indicates that the nature of of SS to skin conductance, and that failure to consider this influence relationship of conflicting results. could lead to inconclusive or conflicting of the studies reviewed here the effects effects It should also be noted that in many of of of SS were limited to male subjects. In contrast, a study involving skin conductance conductance and heart rate measurement by Zahn, Schooler, and Murphy (1986) showed evi evidence of of a stronger relationship between these variables and SS among women than among men. Age may be another moderator variable of of some importance. of young women showed a negative Plouffe and Stelmack (1986) found that a sample of conductance level and SS whereas a sample of of elderly correlation between skin conductance women did not. Cardiovascular Activity. Activity. Studies of electrodermal activity therefore sug sugc. Cardiovascular gest a positive association between strength of the orienting response to novel or of the subscales interesting stimuli and either the total SS score or scores on one of of this is arguable because skin conductance is also the trait. However, conclusion is of part of of the defensive defensive and startle responses. To discover whether SS is related to of these other reactions we must analyze studies in which heart orienting or to one of

CHAPTER CHAPTER 16 16

PSYCHOPHYSIOLOGICAL ApPROACHES APPROACHES PSYCHOPHYSIOLOGICAL

401 401

rate is the dependent variable. As we have noted, heart rate decelerates during orienting and accelerates in the defensive response. As is the case with skin conductance, basal levels of heart rate are not related to SS (e.g., Ridgeway & Hare, 1981) whereas reactivity to stimuli reveals an SS effect. Ridgeway and Hare (1981) found that subjects who were high in SS showed heart rate deceleration on the first trial of a series of 60-dB tones, as did subjects subjects who scored high on the Dis subscale. Subjects who were low in either sensation sensationseeking or disinhibition manifested manifested a short latency acceleration on the first trial, followed by rapid habituation. Similar results were reported by Orlebeke and Feij al. (1985). (1979) and by Feij et a1.

d. Evoked of Potentials, The magnitude of change from the positive peak of Evoked Potentials. after the stimulus the cortical evoked potential occurring approximately 100 ms after Pi component) to the following negative peak occurring approximately 40 ms (the PI Ni) forms the basis for the variable of augmenting-reducing augmenting-reducing (Buchs (Buchslater (the NI) baum & Silverman, 1968). 1968). In general Pi~Ni PI-NI amplitude increases as the intensity of of the stimulus is increased. However, some people show this progressive increment only across low to moderate stimulus intensities. When stimulus intensities become progressive decline in PIN PiNi1 amplitude as as stimulus more intense, these subjects show a progressive intensity is increased. Other subjects show a continuing direct relationship between stimulus intensity and PiNi PINI amplitude even in response to highly intense stimuli. Persons in the first of these two groups are reducers and those in the second are augmenter-reducer (A-R) variable is based on the slope of augmenters. Thus, the augmenter-reducer of PiNi amplitude to stimulus intensity. the function relating PINI Beginning with a study by Zuckerman, Murtaugh, and Siegel (1974), several Zuckerman et al. studies have shown A-R to be related to SS and its components. Zuckerman a1. found that subjects scoring above the median on the Dis scale showed (1974) found augmenting in response to the brightest of a series of light flashes whereas those below the median showed reducing in response to the most intense flashes. flashes. These effects were replicated by Zuckerman, Simons, and Como (1988), but only on the effects first of two successive series of stimuli, suggesting that the relation between Dis and A-R may have dropped out because of habituation on the second set. Evidence of a positive correlation between augmenting for the PINt PiNi component and high scores on the Dis scale has also been reported by Blenner (1993). 2. fSensation-Seeking Sensation-Seeking 2. The The Biology o of

a. Monoamine Oxidase. Oxidase, Sensation-seeking has been studied in relation to a. levels of monoamine oxidase (MAO) in the blood. MAO is an enzyme that is present in the mitochondria of of neurons in the central nervous system as well as other parts of the body. Especially high concentrations are found in the limbic system, where MAO presumably plays a part in mediating various appetitive and function is to degrade the central monoaminergic transmit transmitemotional behaviors. Its function ters such as norepinephrine and dopamine. For that reason it is supposed that MAO regulates the action of these central transmitters. Direct assessment of brain

402 402

RUSSELLG. G.GEEN GEEN RUSSELL

MAO in living humans is not possible in the context of of normal research. However, the level of MAO in platelets of the level of MAO in the the platelets of the the blood blood correlates correlates positively positively with with the the level level therefore serve as indicators of of the enzyme that in the brain. Platelet MAO levels therefore can be studied in connection with other variables. A A few few studies have tested the relation between the various sensation-seeking sensation-seeking scales and platelet MAO, with mixed results. Neither Ballenger et al. (1983) nor Calhoon-La Grange, Jones, Reyes, and Ott (1993) found a relationship between MAO activity and sensation-seeking. However, Murphy et al. (1977) found signifi signifiof SS (Total SS, Disinhibition, cant correlations between MAO and three measures of of Murphy et al. held for male and Boredom Susceptibility). Whereas the findings of subjects only, Schooler, Zahn, Murphy, and Buchsbaum (1978) found a negative correlation between SS and platelet MAO levels among both men and women. of these correlations is somewhat somewhat obscured, however, by data reported The meaning of by Schalling, Asberg, Edman, and Oreland (1987) indicating that the relationship of scores on the overall SS scale to MAO levels is not linear. Subjects in this study of who were characterized moderate MAO MAO levels levels were were equal equal in in SS SS to to those those low low who were characterized by by moderate in MAO and higher in SS than those with high MAO levels. All of the subjects in in MAO and higher in SS than those with high MAO levels. All of the subjects in this study were males. Possibly the link between MAO and SS is moderated by sexthis study were males. Possibly the link between MAO and SS is moderated by sex related In this this connection, connection, it it should should be be noted noted that that aa positive positive correlation correlation related hormones. hormones. In between scores on on the the Dis Dis scale scale and and male male sex sex hormones hormones in in male male subjects subjects has has been been between scores reported (Daitzman (Daitzman & & Zuckerman, Zuckerman, 1980). 1980). reported fc. Central Central Catecholamines. Catecholamines, A A negative relationship between MAO levels b. and sensation-seeking would suggest that the biological substrate of of the trait may lie in the activity of biogenic amines in the brain, such as norepinephrine and of MAO should dopamine. If MAO depletes these neurotransmitters, then levels of be negatively correlated with central catecholamine activity, implying a positive relationship between sensation-seeking and the latter. As has been noted earlier, Zuckerman (1983) abandoned his earlier theory of sensation-seeking, which had been based on the assumption that high sensation sensationof seekers are motivated by a need to increase reticulocortical arousal to an optimal level. In subsequent formulations, Zuckerman has emphasized the importance of reward centers in the limbic system and the mediating role played by central cateneurotransmitters. Genetic inheritance determines the level of these central cate turn affect the activity of the limbic system. Zuckerman (1979) cholamines, which in tum articulated a viewpoint formulated by Stein (1983), according to which neural circuits in which dopamine is the transmitting agent are the pathways for incentive motivation and norepinephrine circuits govern expectancies of of positive reinforce reinforcement. High levels of of dopamine should therefore initiate activity and exploratory behavior high levels levels of of norepinephrine norepinephrine should should establish establish expectations expectations of of behavior whereas whereas high positive outcomes outcomes from from situations situations or or contacts contacts with other people. people. Both Both of of these these positive with other outcomes are are integral integral elements elements in in sensation-seeking. sensation-seeking. In In addition, addition, through through collaterals collaterals outcomes to the the cortex, cortex, the the active active limbic limbic centers centers initiate initiate alertness alertness and and sensitivity sensitivity to to novel novel to stimuli, which are are manifested manifested in in the the orienting orienting response response and and augmentation augmentation of of the the stimuli, which

CHAPTER CHAPTER 16 16

PSYCHOPHYSIOLOGICALApPROACHES APPROACHES PSYCHOPHYSIOLOGICAL

403 403

sensation-seeker is therefore an expression evoked potential. The behavior of the sensation-seeker of of biologically determined incentives and not a response to a deficit deficit in arousal as the original theory implied. As is the case with the study of MAO, direct assessment of of central neurotrans neurotransmitters is difficult. However, the action of these substances is accompanied by the formation formation of metabolites, the concentrations of of which vary positively positively with the of norepinephrine is 3-methoxy-4activity of the transmitters. One metabolite of hydroxyphenylglycol (MHPG). This substance, extracted from urine, is generally assumed to be derived from brain norepinephrine and therefore to reflect the action of the latter. Given the several assumptions made here, it should follow that sensation-seeking is positively correlated with levels of of urinary MHPG. Such a Goodwin (1981), although correlation was reported by Buchsbaum, Muscettola, and Goodwin the number of subjects in that study was small. Studies by Ballenger et al. (1983) and Thieme and Feij (1986) found no relation between sensation-seeking and MHPG activity. Relatively few findings have been reported on the relation of of sensation-seeking to biogenic amine activity. One study has, however, been influential in a revision of Zuckerman's theoretical model. In that study, Ballenger et al. (1983) analyzed of male and female subjects and levels of norepinephrine in the cerebrospinal fluid of A similar negative found that these levels varied negatively with sensation-seeking. A correlation was found between sensation-seeking and levels of of dopamine-�-hydrox dopamine-jS-hydroxof norepinephrine (cf. Umber Umberylase (DBH), an enzyme involved in the production of koman-Wiita, Vogel, & Wiita, 1981). These correlations are, of of course, opposite in direction to what the model had predicted. Zuckerman (1984) has therefore propro of activity and risk in posed that high sensation-seekers may require high levels of of norepinephrine and DBH or to stimulate order either to compensate for low levels of an optimal level of norepinephrine release. This is an optimal-level theory, but one grounded in central processes other than those mediated by the reticular activating system. Matters are somewhat complicated, however, by the finding in two more 1988,1991) recent studies (Calhoon, 1988, 1991) of a positive correlation between DBH activity and sensation-seeking in selected samples of subjects (e.g., those over 30 years of age). This This correlation correlation is is consistent consistent with with the the earlier above, but but not not age). earlier model model described described above, with Zuckerman's Zuckerman's (1983) (1983) revision. The negative correlation between between sensation sensationwith revision. The negative correlation seeking and and norepinephrine norepinephrine activity activity may may also also reflect reflect the the action action of of tyramine, tyramine, which which seeking is known known to to release release norepinephrine norepinephrine from from its its storage and thereby to bring bring is storage centers centers and thereby to about its its depletion depletion in in neuronal tissue. Thieme Thieme and and Feij Feij (1986) (1986) found evidence of of about neuronal tissue. found evidence higher of urinary urinary tyramine tyramine in in high high versus versus low low sensation-seekers. sensation-seekers. higher levels levels of Conclusions 3. Conclusions

Zuckerman's model of the biological basis of sensation-seeking sensation-seeking is a good example of an approach that brings together psychometric, behavioral, and biological findings of in the service of an evolving theory. For that reason it offers the best explanation for the physiological side of this important personality trait. Future findings may lead to further refinements of the model. For example, there is some evidence that

404 404

RUSSELL G. G. GEEN GEEN RUSSELL

of endogenous opioid peptides, or endorphins, are correlated sensationlevels of correlated with sensation seeking Johansson, Almay, Astrom, 1979). seeking ((Johansson, Almay, von von Knorring, Knorring, Terenius, Terenius, & Astrom, 1979). In In addition, addition, found that that the disinhibition disinhibition component component Pivik, Stelmack, and Bylsma (1988) have found of of the spinal spinal motoneurons; of sensation-seeking sensation-seeking is negatively negatively related related to excitability excitability of disinhibitors show evidence evidence of of less excitability excitability than that is, high disinhibitors than low disinhibitors. Extraverts in the Pivik et al. a l study also showed showed less excitability excitability than Extraverts than introverts. sensation-seeking is motivated motivated not not by a need for for These investigators argue that sensation-seeking stimulation, but by a low level of of motor motor excitability. This proposition must obviously proposition must be incorporated incorporated into future future explanations of of sensation-seeking. sensation-seeking. c. C. Impulsivity

1. /• Components Components o off Impulsivity Impulsivity

Assessment been hindered lack of consensus on what the Assessment of of impulsivity impulsivity has has been hindered by by aa lack of consensus on what the trait comprises (e.g., S. B. G. Eysenck Eysenck & Eysenck, 1977), but but some attempts attempts have trait defining its various components. One One is the Barratt Barratt Impulsiveness Impulsiveness been made at defining motor impulsiveness (e.g., Scale (Barratt, 1987), which measures three subtraits: (1) motor impulsiveness (e.g., rapid rapid decision decisionacting rashly and without thought), (2) cognitive impulsiveness making), and (3) nonplanning impulsiveness (Le., (i.e., concentration concentration on the present more than on the future). Using this scale, Barratt, Pritchard, Faulk, and Brandt conducted a study in which subjects were stimulated stimulated with bright and and dim (1987) conducted of light. Evoked responses were assessed at 13 sites on the scalp. Barratt Barratt flashes of found that each of of the three components components of of impulsivity correlated and his associates found significantly with augmenting of of the Nl Ni component component at most most of locasignificantly of the recording loca tions. A A subsequent study by Carillo-de-la-Pefla and Barratt reaffirmed Carillo-de-Ia-Peiia and Barratt (1993) reaffirmed the importance of of Ni discriminator of individual differences differences in Nl augmenting as a discriminator of individual impulsiveness. This measure revealed revealed a consistent consistent positive correlation correlation with scores on all three impulsiveness subscales and with the total score at two of of five recording recording sites on the scalp. No other component component of of the evoked response was related evoked response related to impulsiveness. Impulsiveness and and Monotony Monotony AAvoidance 2. Impulsiveness voidance of research spanning two decades, Schalling and In a program of and her her colleagues have found several psychophysiological psychophysiological correlates correlates of of impulsivity. Early found Early work associated of Solidity as measured measured by the Marke-Nyman Temperaimpulsivity with the trait of Marke-Nyman Tempera comparable to a ment Schedule (Schalling, 1977). A high level of of this trait trait is comparable of impulsivity. Research Research reviewed by Schalling (1976, (1976,1977,1978) found low level of 1977, 1978) found conducrelationships between this trait and cortical activity, heart heart rate, and skin conduc of this research dealt with impulsivity impulsivity as a component component of of psychopathy, psychopathy, tance. Much of therefore beyond the scope of this review. however, and is therefore More More recently, recently, Schalling Schalling and and her her associates associates have have developed developed two two scales scales related related Karolinska Scales of of Personality Personality (KSP): Impulsive Impulsiveto impulsiveness as part of the Karolinska Avoidance (MA). The first assesses tendencies to act without without ness (I) and Monotony Avoidance forethought, to make hasty decisions, and to live in a carefree carefree manner. The second forethought,

CHAPTER CHAPTER 16 16

APPROACHES PSYCHOPHYSIOLOGICAL ApPROACHES

40S 405

measures the desire for novelty and change and an intolerance for the usual and the predictable. The MA scale correlates positively with the total SS scale, as well as with experience-seeking, disinhibition, and boredom susceptibility. Some evidence links the I and MA scales to platelet MAO levels. In general, monotony avoidance and MAO levels are negatively related. Perris et al. (1980) found found such a negative correlation in a sample of male and female hospitalized depressives, but no correlation between MAO and impulsivity. Schalling et al. (1987) obtained obtained similar results with an all-male sample of university students. A study by af Klinteberg, Schalling, Edman, Oreland, and Asberg (1987) found found MAO negatively correlated with MA among high-school-aged boys but not among girls. correlated high-school-aged However, in both sexes those subjects who were high in MAO levels were less impulsive (Le., (i.e., low in I) than were those low in MAO. Thus, for for girls low impulsive impulsiveness was associated with high MAO levels whereas for boys both low I and low MA were associated with high MAO. Schalling, Edman, Asberg, and Oreland (1988) found that I correlated correlated negatively with MAO in a sample of male university students whereas MA was not related to MAO levels.

D. Neuroticism Neuroticism and and Anxiety Anxiety D. In general, studies designed to test a link between psychophysiological processes and H. J. Eysenck's (1967) construct of neuroticism have not yielded impressive effects (e.g., Stelmack, 1981). 1981). The same is true for the study of trait or consistent effects anxiety as assessed by other means. Navateur and Freixa i Baque (1987) concluded from their extensive review of the literature that previous studies did not reveal a from consistent relationship between anxiety and electrodermal activity. In their own study, Navateur and Frebca Freixa i Baque (1987) found that highly trait anxious subjects showed lower skin conductance levels, lower conductance amplitudes, and fewer spontaneous conductance responses than their less trait anxious counterparts in response to both neutral and stressful pictures. These results are inconsistent with what would be predicted from from a simple arousal theory of anxiety, and could be explained post hoc by the authors only tentatively. A recent study by Britt and Blumenthal (1992) suggests a new direction in which studies on the psychophysiol psychophysiology of anxiety may go go by showing that whereas low state anxious subjects manifested manifested a more rapid periorbital startle response to an 85-dB tone than to a less intense (60-dB) tone, subjects high in state anxiety responded with equal latency to both tones. This finding suggests that high anxiety may be associated with relative motomoto neural insensitivity. The study of the psychophysiology of anxiety requires some careful careful attention to certain details that have not always been observed in the past. One is the nature of the eliciting stimulus. Whereas weak or trivial laboratory stimuli cannot be expected to elicit the sort of somatic reaction associated with high levels of anxiety, ethical considerations set a limit on what can be done under such conditions. Another is is the state-trait distinction made originally by Spielberger (1972): individ individdifferences in trait anxiety will be related to psychophysiological differences differences ual differences

406 406

RUSSELL GUN RUSSELLG. CGEEN

only to the extent that they interact with situational conditions to produce state anxiety. The intensity of of the eliciting situation noted above is obviously important in effecting effecting this process. In addition, Spielberger has proposed that the level of trait anxiety influences the range of of conditions that evoke state anxiety, with highly trait anxious people showing anxious states in a greater number of of situations than their less trait anxious counterparts. This "range *'range hypothesis" of of trait anxiety has not been studied systematically. Another matter to keep in mind is the nature of of anxiety in normal life. of in vivo vivo studies of persons who Fahrenberg (1992) has called for increasing use of suffer from such anxiety-related disorders as panic attacks and agoraphobia agoraphobia in the generalizations that hope that a *'bottom-up" "bottom-up" inductive approach may yield some generalizations of better scientific scientific theories of of anxiety. Finally, will ultimately lead to the construction of we must keep in mind the fact that humans deal with anxiety with a number of coping and defensive strategies that may reflect personality influences. Some recent developments suggest that individual differences differences in coping styles and in personal beliefs in the ability to deal with problems may be correlated with specific patterns psychophysiological activity (e.g., Bandura, Cioffi, Cioffi, Taylor, & Brouillard, 1988; of psychophysiological Miller & Mangan, 1983). These intervening processes may interact with anxiety level to influence the overall somatic response pattern elicited by threatening or stressful situations. Further research along these lines would be in the spirit of the stressful multimodal approach approach discussed discussed earlier earUer in in this this review. review. multimodal

V. V. SUMMARY The modem modern period in the study of personality and psychophysiology psychophysiology began with differences in arousability. In his pioneering theory, the hypothesis of individual differences Eysenck proposed two systems governing these differences: one, centered in the reticular activating system, formed the basis for introversion-extraversion; the other, centered in the limbic system, formed formed the basis for neuroticism and anxiety. Later theorists offered offered revisions of of this model, extended it, and developed other constructs similar to it. The arousability arousability hypothesis, in one form or another, has informed informed most of the research on personality personality and psychophysiology psychophysiology published to date. Some of of that research has been reviewed in this chapter. Many of the studies designed to test the arousability hypothesis have shown that psychometrically assessed personality variables are correlated with activity in the physiological systems of the body. Nevertheless, the arousability hypothesis has been criticized on several grounds: the results of the research are mixed and some sometimes contradictory; the relationship of of personality to physiological physiological functioning functioning is model would would suggest; studies that have more complex than a simple general arousal model been carried out to test the model have often often been simplistic and badly designed of generalized ones that yield little more than simple correlations; and the idea of arousal itself itself is no longer widely accepted. These are all legitimate criticisms, but in themselves they do not vitiate the study of personality personality through its connections with

CHAPTER 16 16

PSYCHOPHYSIOLOGICALApPROACHES APPROACHES PSYCHOPHYSIOLOGICAL

407 407

psychophysiology. This approach to personality remains an exciting and challenging option that can yield important insights into the nature of of individual differences. examinaSeveral leading psychophysiologists are now calling for a thorough examina tion of of the problems that have beset the area in the past and for a rigorous appraisal of steps that must be taken for the study of of personality and psychophysiology to of move on to the next plateau. The suggestions made by these observers should be followed by any student of of personality personality wishing to enter this specialized field, and followed psychophysiology to their by senior researchers in personality who may wish to add psychophysiology of the points raised are those of of Gale and armamentarium of methods. Typical of Edwards (1986), whose main ideas are summarized next as a conclusion to this review. psychophysiology must be based on rigorous First, the study of personality and psychophysiology theories of behavior. Both the personality and the physiological variables must be well-defined theoretical constructs connected to each other by an grounded in well-defined definitions explicit syntax. The constructs must also be linked through operational definitions to observable and objective conditions that form the basis for testable hypotheses. psychophysiology should manifest manifest the In other words, research on personality and psychophysiology characteristics of good theory-driven scientific research. The best research that of investigators like Eysenck, Zuckerman, has been reported to date—the date-the work of Barratt—shows these characteristics. This work should be Stelmack, Schalling, and Barratt-shows the norm for all subsequent studies. Second, statements about personality must be based on good psychometric psychometric measures. This means that scales used to assess traits must have high levels of reliabiUty and construct validity. In addition, personality measures must, whenever reliability possible, be used for the purposes for which they have been devised; scales that have been developed for one purpose should be used for other purposes only with of their possible shortcomings. Attention must also be caution and with realization of paid to the extent to which the chosen personality variables converge on other such variables, variables, and and whether whether the the ones ones chosen chosen are are only only subsets subsets of of larger larger and and more more such inclusive ones. ones. inclusive Third, psychophysiological psychophysiological assessment assessment must be thorough enough to allow adequate tests of hypotheses. This requires measurement of several body systems (e.g., cardiac, muscular, electrodermal) as well as measurement of of several aspects of one system (e.g., skin conductance, skin potential, and specific conductance responses within the electrodermal system). Some aspects of a given system may personaUty measures more highly than others. Attempts should also correlate with personality be made to use similar measures across studies within a research program and across laboratories. The study of personality and psychophysiology to date has shown a relatively low level of replication of findings; this has been due in large part to the lack of comparability of methods used in various programs. Fourth, attention must be shifted from the quest for simple correlations be beof tween personality and psychophysiological variables that has characterized most of As the work to date to a search for the processes that underlie the correlations. As of has been noted in this chapter, some recent approaches call for the analysis of

408 408

RUSSEU GEEN RUSSELL G. G.G EEN

activation activation within within specific specific somatic somatic systems systems and and for for the the study study of of ways ways in in which which these these systems systems interact interact with with each each other. other. An An important important part part of of aa process process view view of of psychophysi psychophysiological ological functioning functioning is is recognition recognition that that the the person person is is not not aa passive passive recipient recipient of of stimuli stimuli from from the the environment. environment. It It has has already already been been noted noted that that among among the the determinants determinants of of reactions to to the the environment environment are are the the person's efforts to to adapt adapt to, to, and and to to cope cope with, with, reactions person's efforts stimuli (Gale, personality and stimuli (Gale, 1987). 1987). The The general general model model of of personality and psychophysiology psychophysiology that that from this new viewpoint is one involving the several physiological physiological is emerging from arranged in a larger larger control control system. This model points to a need need for for components arranged multiprocess and multimodal multimodal studies. large, multiprocess The The study study of of personality personality through through its its connections connections with with physiological physiological processes processes of previous previous research research and theory, and it still presents a unique rests on a solid basis of for those who wish to conduct conduct careful careful theory-driven theory-driven research on challenge today for efforts needed to produce produce the next generation of of findings findings complex processes. The efforts but so will the ultimate reward. will be great, but

REFERENCES Asberg, af Klinteberg, B., B., Schalling, D., Edman, G., Oreland, L., & A sberg, M. (1987). (1987). Personality of platelet monoamine oxidase (MAO) activity in female and male subjects. correlates of Neuropsychobiology, 18, 18, 89-96. Neuropsychobiology, W. (1981). (1981).Differences Differences in inbrainstem brainstemauditory auditoryevoked evokedresponses responses Andress, D. L., & Church, M. W. function of stimulus intensity. intensity. Psychophysiology, Psychophysiology, between introverts and extraverts as a function 18, 18, 156-157. Ballenger, J. c., C, Post, R. M., Jimerson, D. C., C, Lake, C. R., Murphy, D. L., Zuckerman, M., & Cronin, C. (1983). Biochemical correlates of personality traits in normals: An M., exploratory study. Personality ferences, 4,4,615-625. Personalityand andIndividual IndividualDif Differences, 615-625. Bandura, A., Cioffi, B., & Brouillard, M. E. (1988). Cioffi, D., Taylor, C. B., (1988). Perceived self-efficacy self-efficacy in coping with cognitive stressors and opioid activation. Journal of Personalityand andSocial Social of Personality Psychology, Psychology, 55, 55,479-488. 479-488. Barratt, E. S S.... (1987). Impulsiveness Impulsiveness and and anxiety: anxiety:Information Information processing processingand andelectroencepha electroencephalograph topography. Journal of ResearchininPersonality, Personality,21,21,453-463. 453-463. of Research Barratt, E. S., Pritchard, W. S., Faulk, D. D. M., & Brandt, M. E. (1987). (1987). The relationship between impubliveness impulsiveness subtraits, trait anxiety, and visual N100 NlOO augmenting/reducing: augmenting/reducing: . A topographic analysis. Personality Personalityand andIndividual IndividualDifferences, Differences,8,8,43-51 43-51. Bartussek, D D.,. , Diedrich, 0., O., Naumann, E., & Collet, W W.. (1993). Introversion-extraversion Introversion-extraversion and event-related potential (ERP): A test of J. A. Gray's theory. Personality Personality and and Individual Differences, Differences, 14, 14,565-574. 565-574. Birchall, P. M. A., & Claridge, G. S. (1979). (1979). Augmenting-reducing of the visual evoked potential as a function function of changes in skin conductance level. Psychophysiology, Psychophysiology, 16, 16, 482-490. Blenner, J. L. (1993). Visual evoked potential stimulus intensity modulation and sensation Personalityand andIndividual IndividualDifferences, Differences,14,14,455-463. 455-463. seeking in thrill seekers. Personality Motoneuronal insensitivity insensitivity in in extraverts extraverts as as revealed revealed Britt, T. W., & Blumenthal, T. D. (1991). Motoneuronal by the startle response paradigm. Personality Personalityand andIndividual IndividualDifferences, Differences,12,12,387-393. 387-393.

16 CHAPTER 16

PSYCHOPHYSIOLOGICALApPROACHES APPROACHES PSYCHOPHYSIOWGICAL

409 409

Britt, T. W., & Blumenthal, T. D. (1992). of (1992). The effects effects of anxiety on motoric expression of the startle response. Personality /3,13, 91-97. Personalityand andIndividual IndividualDifferences, Differences, 91-97. Buchsbaum, M. S., S.,Muscettola, Muscettola,G., G.,& &Goodwin, Goodwin,F. F.K. K.(1981). (1981).Urinary UrinaryMHPG, MHPG,stress stressresponse, response, personality factors and somatosensory evoked potentials in normal subjects and pa paNeuropsychobiology, 7, tients with affective affective disorders. Neuropsychobiology, 7, 212-224. Buchsbaum, M. S., & Silverman, J. (1968). (1968). Stimulus intensity control and the evoked cortical response. Psychosomatic , 12-22. PsychosomaticMedicine, Medicine,3030, Xl-ll, Bullock, W. A., & Gilliland, K. (1993). (1993). Eysenck's arousal theory of Introversion IntroversionExtraversion: A converging measures investigation. Journal of of Personality Personalityand andSocial Social Psychology, Psychology,64, 64,113-123. 113-123. Calhoon, L. L. (1988). Explorations into the biochemistry of sensation seeking. Personality Personality 941-949. and Individual Individual Differences, Differences,6, 6, 941-949. Calhoon, L. L. L. (1991). (1991). Sensation Sensation seeking, seeking,exercise, exercise,and anddopamine dopaminebeta beta hydroxylase. hydroxylase.Personal Personality and Individual ferences, 9, 9, 903-907. IndividualDif Differences, 903-907. Calhoon-La Grange, L. L. Jones, T. D., Reyes, E., & Ott, S. (1993). Monoamine oxidase levels in females: Relationships to sensation seeking, alcohol misuse, physical fitness, ferences, 14,14, 439-446. Personalityand andIndividual IndividualDif Differences, 439-446. and menstrual cycle. Personality B., Baribeau-Brown, Baribeau-Brown, J., & & Braun, C. C. (1981) (1981) Neuroanatornical Neuroanatomicai and physiological physiological Campbell, K. B., foundations of extraversion. Psychophysiology, Psychophysiology, 18, 18,263-267. 263-267. Carrillo-de-Ia-Pena, T., & &Barratt, Barratt,E. E.S.S.(1993). (1993).Impulsivity Impulsivityand andERP ERPaugmenting/reducing. augmenting/reducing. Carrillo-de-la-Pena, M. T., Personalityand andIndividual IndividualDifferences, Differences, 25-32. Personality 15, 15, 25-32. Carroll, E. N., Zuckerman, M., & Vogel, W. H. (1982). (1982). A test of the optimal level of arousal Journalof ofPersonality Personalityand and Social Psychology, 572-575. theory of sensation seeking. Journal Social Psychology, 42, 42, 572-575. Claridge, G. S. (1967). Personality Personalityand andarousal. arousal,London: London:Pergamon PergamonPress. Press. Daitzman, J., & Zuckerman, M. (1980). Disinhibitory sensation seeking personality personaUty and 1, 103-110. gonadal hormones. Personality Personalityand andIndividual IndividualDifferences, Differences, 1, 103-110. Daruna, J. H., Karrer, R, R., & Rosen, A. J. (1985). Introversion, attention and the late positive component of event-related potentials. Biological Biological Psychology, Psychology,20, 20,249-259. 249-259. introverted/extraverted personality types. Ditraglia, G. M., & Polich. J. (1991). P300 and introverted/extraverted Psychophysiology, Psychophysiology,28, 28,177-184. 177-184. Drinka, G. F., (1984). The birth birth of of neurosis. neurosis. New NewYork: York:Simon Simon &&Schuster. Schuster. Endler, N. S., Edwards, J. M., & Vitelli, R R. (1989). (1989). Endler multidimensional multidimensional anxiety anxietyscales: scales: Manual. Los Angeles: Western Psychological Services. Manual Eysenck, H. J., (1947). The dynamics dynamics of of anxiety anxiety and hysteria. hysteria. London: Routledge & Kegan Paul. biologicalbasis basisofofpersonality. personality.Springfield, Springfield,IL:IL: CharlesC.C. Thomas. Charles Thomas. Eysenck, H. J. (1967). The biological of Eysenck, H. J. (1981). General features of the model. In H. J. Eysenck (Ed.), A model of personality (pp. 1-37). New York: Springer-Verlag. personality Eysenck, H. H. J., & & Eysenck, Eysenck, M. M. W. W. (1985). (1985). Personality Personalityand andindividual individualdifferences: differences: A natural A natural Eysenck, science approach. approach.New NewYork: York:Plenum PlenumPress. Press. science (1977). The place of impulsiveness in a dimensional Eysenck, S. S. B. B. G., & Eysenck, H. J. (1977). system of personality description. British British Journal Journalof of Psychology, Psychology,16, 16,57-68. 57-68. Fahrenberg, J. (1992). Psychophysiology of neuroticism and anxiety. In A. Gale & M. W. Biological perspectives (pp.(pp. 179-179Eysenck (Eds.), Handbook of of individual individualdifferences: differences: Biological perspectives 226). Chichester, England: Wiley. R. W. (1985). Sensation-seeking: Feij, J. A., Orlebeke, J. F., Gazendam, A., & van Zuilen, R Measurement and psychophysiological correlates. In J. Strelau, F. H. Farley, & Measurement

410 410

RUSSELL RUSSELL G. G. GEEN GEEN

A. Gale (Eds.) The biological biologicalbases basesofofpersonality personalityand andbehavior behavior(Vol. (Vol.1, 1,pp. pp.195-210). 195-210). A

New York: McGraw-Hill. Fowles, D. C. (1983). Motivational effects effects on heart rate and electrodermal activity: Implica ImplicaResearchininPersonal Personalfor research on personality and psychopathology. Journal of of Research tions for ity, 117, 48-71. 7, 48-71. ity, Fowles, D. C., C, Roberts, R., & Nagel, K. (1977). (1977). The influence of of introversion/extraversion introversion/extraversion Research of Research on the skin conductance response to stress and stimulus intensity. Journal of Personality, 11, 11,129-146. 129-146. in Personality, Gale, A A. (1981). EEG studies of extraversion-introversion: What's the next step? In R. Lynn (Ed.), Dimensions of of personality: personality: Papers Papersininhonour honour of of H. H. J.J. Eysenck Eysenck (pp. (pp. 181-207). 181-207). Oxford, England: Pergamon Press. Gale, A A. (1983). Electroencephalographic studies of extraversion-introversion: A case study Personality and and Individual Individual Differ Differin the psychophysiology of individual differences. Personality ences, 371-380. ences, 4,4,371-380. values-An attempt at review and reap Gale, A A. (1987). Arousal, control, energetics and values—An reappraisal. In J. Strelau & H. J. Eysenck (Eds.), Personality Personalitydimensions dimensionsand andarousal arousal(pp. (pp. 287-316). New York: Plenum Press. A, & Edwards, J. A A. (1986). Individual differences. In M. G. H. Coles, E. Donchin, & Gale, A., S. W. Porges, (Eds.), Psychophysiology: and applications (pp.(pp. 431Psychophysiology:Systems, Systems,processes, processes, and applications 431Guilford Press. 507). New York: Guilford Gange, J. J., Geen, R. G., & Harkins, Harkins, S. S. G. (1979). (1979). Autonomic Autonomic differences differences between between extraverts extraverts and introverts during vigilance. Psychophysiology, Psychophysiology, 16, 16,392-397. 392-397. Geen, R. G. (1984). Preferred Preferred stimulation levels in introverts and extraverts: Effects Effects on arousal and performance. Journal of 46,46,1303-1312. 1303-1312. of Personality Personalityand andSocial SocialPsychology, Psychology, Golding, J. F., & Richards, M. (1985). EEG spectral analysis, visual evoked potential and Personalityand andIndividual IndividualDiffer Differphotic driving correlates of personality and memory. Personality ences, 67-76. ences, 6,6,67-76. Graham, F. K. (1979). Distinguishing among orienting, defense, and startle reflexes. In H. D. Kimmel, E. H. van Olst, orientingreflex reflexininhumans humans Gist, & J. F. Orlebeke (Eds.), The orienting (pp. 137-167). Hillsdale, NJ: Erlbaum. Gray, J. R. (1964). Pavlov's Pavlov*stypology. typology,London: London:Pergamon PergamonPress. Press. The psychophysiological psychophysiological nature nature of of introversion-extraversion: introversion-extraversion: A A modifica modificaGray, J. R. (1972). The tion of Eysenck's theory. In V. D. Nebylitsyn & J. A. Gray (Eds.), Biological Biological bases bases individual behavior (pp. 182-205). New York: Academic Press. of individual of A Gray, J. R. (1981). A critique of Eysenck's theory of personality. In H. J. Eysenck (Ed.), A model for for personality personality (pp. 246-276). Berlin: Springer. Haier, R. J., Robison, D. L., Braden, W., & Williams, Williams,D. D. (1984). (1984).Evoked Evokedpotential potentialaugmenting augmentingPersonality and Individual Differences, Differences, 5, reducing and personality differences. Personality 293-301. Harvey, F., & Hirschman, R. (1980). The influence of extraversion and neuroticism on Personalityand andIndividual IndividualDifferences, Differences, heart rate responses to aversive visual stimuli. Personality 1, 97-100. 1, A, & Hansen, J. C. (1986). Attention: Electrophysiological approaches. In Hillyard, S. A., Psychophysiology: Systems, Systems, M. G. H. Coles, E. Donchin, & S. W. Porges (Eds.), Psychophysiology: processes, and andapplicatiOns applications(pp. (pp.227-243). 227-243).New NewYork: York:Guilford GuilfordPress. Press. processes, Jackson, S. (1986). Melancholia Melancholiaand anddepression. depression.New NewHaven, Haven,Cf: CT:Yale YaleUniversity UniversityPress. Press. Jennings, J. R. (1986). Bodily changes during attending. In M. G. H. Coles, E. Donchin, & Psychophysiology:Systems, Systems,processes processes and applications 268S. W. Porges (Eds.), Psychophysiology: and applications (pp.(pp. 268Guilford Press. 289). New York: Guilford

16 CHAPTER 16

PSYCHOPHYSIOLOGICAL ApPROACHES APPROACHES PSYCHOPHYSIOWGICAL

411 411

Johansson, F., Almay, B. G. L., von Knorring, L., Terenius, L., & Astrom, M. (1979). Personality traits in chronic pain patients related to endorphin levels in cerebrospinal -239. fluid. Psychiatry Psychiatry Research, 1, 7, 231 231-239. Lacey, J. I. (1967). Somatic response patterning and stress: Some revisions of activation theory. In M. H. Appley & R R. Trumbull (Eds.), Psychological Psychological Stress: Issues in research (pp. 14-37). New York: Appleton-Century-Crofts. Appleton-Century-Crofts. Lindsley, D. B. (1957). Psychophysiology and motivation. In M. R R. Jones (Ed.), Nebraska Nebraska Symposium Symposium on Motivation Motivation (Vol. 5, pp. 44-105). Lincoln: University of Nebraska Press. Ljubin, T., & Ljubin, C. (199O). (1990). Extraversion and audiomotor reflex. Personality Personality and Individ Individual Dif ferences, 11, 977-984. Differences, Lukas, J. H. (1987). Visual evoked potential augmenting-reducing augmenting-reducing and personality: personaUty: The vertex vertex augmenter is a sensation seeker. Personality ferences, 8, 385-395. Personality and and Individual Individual Dif Differences, Lukas, J. H., & Mullins, L. F. (1983). Auditory augmenting-reducing and sensation seeking. Psychophysiology, Psychophysiology, 20, 457. Lukas, J. H., & Mullins, L. F. (1985). Auditory Auditory augmenters are sensation seekers and perform perform Psychophysiology, 22, 580-581. better under high work loads. Psychophysiology, Matthews, G., & Amelang, M. (1993). Extraversion, arousal theory and performance: A A ferences, study of of individual individual differences differences in the EEG. Personality Personality and Individual Individual Dif Differences, 14, 347-363. Matthews, G., Davies, D. R, R., & Holley, HoUey, P. J. (199O). (1990). Extraversion, arousal, and visual sustained of resource availability. Personality Personality and Individual Individual Dif Differences, attention: The role of ferences, 11, 11159-1173. 159-1 173. Miller, S. M., & Mangan, C. E. (1983). The interacting effects of information and coping style of Personality Personality and and in adapting to gynecologic stress: Should the doctor tell it all? Journal of Social Psychology, Psychology, 45, 223-236. Social Mullins, L. F., & Lukas, J. H. (1984). Auditory augmenters are sensation seekers if they Psychophysiology, 21, 589. attend the stimuli. Psychophysiology, R. H., Buchsbaum, M. S., Martin, N. F., Ciaranello, K., & Wyatt, Murphy, D. L., Belmaker, R R. J. (1977). Biogenic amine related enzymes and personality personaUty variations in normals. R Phychological Medicine, 7, 149-157. Phychological Constitutional psychophysiology: psychophysiology: Research in review. Orlando, FL: Aca AcaMyrtek, M. (1984). Constitutional demic Press. Navateur, J., & Freixa i Baque, E. (1987). Individual differences differences in electrodermal activity of subjects' subjects* anxiety. Personality Personality and and Individual Individual Dif Differences, ferences, 8, 615-626. as a function of Neary, R R. S., & Zuckerman, M. (1976). Sensation seeking, trait and state anxiety, and the Psychophysiology, 13, 205-211. electrodermal orienting reflex. Psychophysiology, Neumann, E., & Blanton, R R. (1970). The early history of of electrodermal research. Psychophysi Psychophysiology, 6, 453-475. ology, Orlebeke, J. F., & Feij, J. A. (1979). The orienting orienting reflex as a personality correlate. In humans H. D. Kimmel, E. H. van Olst, & J. F. Orlebeke (Eds.), The orienting reflex in humans (pp. 567-585). Hillsdale, NJ: Erlbaum. Perris, C C.,, Jacobsson, L., von Knorring, L., Oreland, L., Perris, H., & Ross, S. I. (1980). deEnzymes related to biogenic amine metabolism and personality characteristics in de Acta Psychiatrica Scandinavica, 61, 477-484. Psychiatrica Scandinavica, pressed patients. Acta Individuality in pain pain and and suffering. Chicago: University of of Chicago Press. Petrie, A. (1967). Individuality R. T., Stelmack, R R. M., & Bylsma, F. W. (1988). Personality and individual differences differences Pivik, R Psychophysiology, 25, 16-24. in spinal motoneuronal excitability. Psychophysiology,

4U 412

RUSSELL G. G. GEEN GEEN RUSSELL

Plouffe, L., & Stelmack, R. M. (1986). Sensation-seeking and the electroidermal orienting Individual Differences, response in young and elderly elderiy females. Personality Personality and Individual Differences, 7,7, 119-120. Revelle, W. Humphreys, M. S., Simon, L., & Gilliland, K. (1980). The interactive effect of effect of of ExperiExperi personality, time of day, and caffeine: A test of the arousal model. Journal of 1-31. mental Psychology: Psychology:General, General,109, 109, 1-31. Richards, M., & Eves, F. F. (1991). Personality, temperament re temperament and the cardiac defense re12,12, 999-1007. sponse. Personality sponse. Personalityand andIndividual IndividualDifferences, Differences, 999-1007. Ridgeway, D., & Hare, R. D. (1981). (1981). Sensation seeking and psychophysiological responses to auditory stimulation. Psychophysiology, Psychophysiology, 18, 18,613-618. 613-618. Robinson, T. N., Jr., & Zahn, T. P. (1983). Sensation seeking, state anxiety, and cardiac and EDR orienting reactions. Psychophysiology, Psychophysiology,20, 20,465. 465. Sales, S. M. (1971). (1971). Need for for stimulation as a factor factor in social behavior. Journal of of Personality Social Psychology, Psychology,19, 19,124-134. 124-134. and Social Schalling, D. (1976). (1976). Anxiety, pain, and coping. In I. G. Sarason & C. D. Spielberger (Eds.), Stress and and anxiety anxiety(Vol. (Vol. 3,3,pp. pp.49-71). 49-71). Washington, Washington, DC: DC: Hemisphere. Hemisphere. Stress The trait-situation trait-situation interaction interaction and and the the physiological physiological correlates correlates of of behav behavSchalling, D. (1977). The Personality at atthe thecrossroads crossroads(pp. (pp.129-141). 129-141). ior. In D. Magnusson & N. Endler (Eds.), Personality Hillsdale, NJ: Erlbaum. Schalling, D. (1978). (1978). Psychopathy-related Psychopathy-related personality variables and the psychophysiology of of D. Schalling Schalling (Eds.), (Eds.), Psychopathic Psychopathicbehavior: behavior:Approaches Approaches socialization. IN R. D. Hare & D. research(pp. (pp.85-106). 85-106).Chichester, Chichester,England: England:Wiley. Wiley. to research Asberg, (1985). Biological and psychological correlates of impulsiveness sberg, M. (1985). Schalling, D., & A biological and monotony avoidance. In J. Strelau, F. H. Farley, & A. Gale (Eds.), The biological ofpersonality personalityand andbehavior: behavior:Vol. Vol1. 1.Theories, Theories, measurement techniques^ and develbasesof bases measurement techniques, and devel opment ((pp. pp. 181-194). Washington, DC: Hemisphere. sberg, M., Edman, G., & 8L Oreland, L. (1987). Markers for for vulnerability to Asberg, Schalling, D., A psychopathology: Temperament traits associated with platelet MAO activity. Acta 76,76, 172-182. Psychiatrica PsychiatricaScandinavica, Scandinavica, 172-182. Schalling, D., sberg, M., D., Edman, G., G., A Asberg, M., & &Oreland. Oreland. L. L. (1988). (1988).Platelet Platelet MAO MAOactivity activityassociated associated erences, 9,9,597-605. with impulsivity and aggressivity. Personality Personalityand andIndividual IndividualDiff Differences, 597-605. Schooler, C., P.,Murphy, Murphy, D. D. L., L.,& &Buchsbaum, Buchsbaum,M. M.S.S.(1978). (1978).Psychological Psychologicalcorrelates correlates C , Zahn, T. P., disease, 166,166,177-186. 177-186. normals.Journal Journalof ofNervous Nervousand andMental Mental disease, of monoamine oxidase in normals. Smith, B. D. (1983). (1983). Extraversion and electrodermal activity: Arousabilty and the inverted inverted4, 41 U. Personality Personalityand andIndividual IndividualDifferences, Differences, 4, 1-419. 411-419. B. D., Rockwell-Tischer, S., & Davidson, R. (1986) Extraversion and arousal: Effects Effects Smith, B. of attentional conditions on electrodermal activity. Personality Personality and andIndividual IndividualDiffer Differences, 7, 7, 293-303. 293-303. ences, Smith, B. D., Rypma, C. B., B., & &Wilson, Wilson,R. R.J.J.(1981). (1981).Dishabituation Dishabituationand andspontaneous spontaneousrecovery recovery Effects of of extraversion, extraversion, impulsivity, impulsivity, sociability, sociability, of the electrodermal orienting response: Effects and caffeine. caffeine. Journal Journalof ofResearch ResearchininPersonality, Personality, 233-240. and 15,15, 233-240. B. D., Wilson, R. J., & Davidson, R. (1984). Electrodermal Electrodermal activity and extraversion: Smith, B. Caffeine, preparatory signal and stimulus intensity effects. Personality Personalityand andIndividual Individual Differences,S,5,59-65. 59-65. Differences, B. D., D.,Wilson, Wilson, R. R. J., J., & &Jones, Jones,B. B.E. E.(1983). (1983).Extraversion Extraversionand andmultiple multiplelevels levelsof ofcaffeine caffeineSmith, B. induced arousal: Effects Effects on overhabituation and dishabituation. Psychophysiology, Psychophysiology,20, 20, 29-34. N. (1963). Perception Perceptionand andthe theconditioned conditionedreflex. reflex.New Ntv/York: York:Pergamon PergamonPress. Press. Sokolov, E. N.

CHAPTER CHAPTER 16 16

PSYCHOPHYSIOLOGICAL PSYCHOPHYSIOLOGICAL ApPROACHES APPROACHES

413 413

Spielberger, C. D. (1972). Anxiety as an emotional state. In C. D. Spielberger (Ed.), Anxiety: Anxiety: (Vol. 1 pp. 23-49). New York: Academic Press. Current Currenttrends trendsinintheory theoryand andresearch research (Vol. 1 pp. 23-49). New York: Academic Press. Stein, L. (1983). (1983). The The chemistry chemistry of of positive positive reinforcement. reinforcement. In In M. M. Zuckerman Zuckerman (Ed.), (Ed.), Biological Biological bases bases of of sensation, sensation, seeking, seeking, implusivity, implusivity, and anxiety (pp. 151-175). Hillsdale, NJ: Erlbaum. Stelmack, R. M. (1981). The psychophysiology of extraversion and neuroticism. In H. J. Eysenck (Ed.), A model for for personality personality (pp. 38-64). New York: Springer-Verlag. Stelmack, R. M. (1990). Biological bases of extraversion: Psychophysiological evidence. Journal of of Personality, Personality,58, 58,293-311. 293-311. Stelmack, R. M., & Michaud, Achorn, A. (1985) Extraversion, attention, and the habituation of the auditory evoked response. Journal of of Research Research in in Personality, Personality,19,19,416-428. 416-428. Extraversion and individual differences differences Stelmack, R. M., Achorn, E., & Michaud, A. (1977). Extraversion in auditory evoked response. Psychophysiology, Psychophysiology, 14, 14, 368-374. C, & Pickard, C. (1979). Extraversion and the Stelmack, R. M., Bourgeois, R., Chien, J. Y. C., OR habituation rate to visual stimuli. Journal of of Research Research ininPersonality, Personality,13,13,49-58. 49-58. Stelmack, R. M., Campbell, K. B., Extraversion and brainstem auditory B., & Bell, I. (1993) Extraversion Personalityand andIndividual IndividualDifferences, Differences, evoked potentials during sleep and wakefulness. Personality 14, 447-453. 447-453. 14, Stelmack, R. M., & Geen, R. G. (1992). The psychophysiology of extraversion. In A. Gale & of individual individualdiff differences: Biological perspectives (pp. erences: Biological perspectives (pp. M. W. Eysenck (Eds.), Handbook of 227-254). Chichester, England: Wiley. Plouffe, L., & Falkenberg, W. (1983). Extraversion sensation seeking and Stelmack, R. M., Plouffe, Personality and Individual Differences, Differences, electrodermal response: Probing a paradox. Personality 4, 607-614. 4, Stelmack, R. M., & Stalikas, A. (1991). Galen and the humor theory of temperament. Personalityand andIndividual IndividualDifferences, Differences,12, 12,255-263. 255-263. Personality effects of frequency frequency and Stelmack, R. M., & Wilson, K. G. (1982). (1982). Extraversion and the effects brainstem evoked response. Personality Personalityand andIndividual IndividualDiffer Differintensity on the auditory brainstem ences, 3,3,373-380. 373-380. ences, Stenberg, G., Rosen, I., & Risberg, J. (1988). Personality and augmenting/reducing in visual and auditory evoked potentials. Personality Personalityand andIndividual IndividualDifferences, Differences,9,9,571-579. 571-579. Strelau, J. (1983). Temperament, New York: Academic Press. Temperament,personality, personality,activity. activity. New York: Academic Press. Strelau, J. (1987). Personality Personahty dimensions based on arousal theories: Search for integration. In J. Strelau & H. J. Eysenck (Eds.), Personality Personalitydimensions dimensionsand andarousal arousal(pp. (pp.269-286). 269-286). New York: Plenum Press. Thayer, R. R. E. (1970). Activation Activation states states as as assessed assessed by by verbal verbal reports reports and and four four psychophysiolog psychophysiological variables. Psychophysiology, Psychophysiology, 7,7,86-94. 86-94. Thieme, R. E., & Feij, J. A. (1986). Tyramine, a new clue to disinhibition and sensation seeking? Personality Personalityand andIndividual IndividualDifferences, Differences,7,7,349-354. 349-354. Umberkoman-Wiita, B., Vogel, W. H., H., & Wiita, Wiita, P. P. J.J. (1981). (1981).Some Somebiochemical biochemicaland andbehavioral behavioral (sensation seeking) correlates in healthy adults. Research ResearchCommunications CommunicationsininPsychol Psychology, ogy, Psychiatry, Psychiatry,and andBehavior, Behavior,6,6,303-316. 303-316. Veith, I. (1965). Hysteria: Hysteria: The Thehistory historyof ofaadisease. disease.Chicago: Chicago:University UniversityofofChicago ChicagoPress. Press. Venables, P. H. (1984). Arousal: An examination of its status as a concept. In M. G. H. Coles, J. R. Jennings, & J. A. Stem (Eds.), Psychophysiological Festschrift Psychophysiologicalperspectives: perspectives: Festschrift for Beatrice Beatrice and andJohn John Lacey Lacey (pp. (pp. 134-142). 134-142).New New York: York:Van Van Nostrand-Reinhold. Nostrand-Reinhold. for Visual evoked evoked responses responses and and platelet platelet monoamine monoamine oxidase oxidase in in patients patients von Knorring, L. (1981). Visual suffering from alcoholism. In H. Begleiter (Ed.), The biological biologicaleffects effectsof ofalcohol alcohol(pp. (pp. suffering 270-291). New York: Plenum Press.

414 414

RUSSELL GEEN RUSSELLG. G.GEEN

Zahn, T. T. P., P., Schooler, Schooler, C., C, & &Murphy, Murphy,D. D.L. L.(1986). (1986).Autonomic Autonomiccorrelates correlatesof ofsensation sensationseeking seeking and monoamine oxidase activity: psychophysio activity: Using confirmatory factor analysis on psychophysiological data. data. Psychophysiology, Psychophysiology,23, 23,521-531. 521-531. Zuckerman, M. (1969). Theoretical formulations. In J. P. Zubek (Ed.), Sensory Sensory deprivation: deprivation: Fifteen Fifteenyears yearsof ofresearch research(pp. (pp.407-432). 407-432).New NewYork: York:Appleton-Century-Crofts. Appleton-Century-Crofts. Zuckerman, M. M. (1971). Dimensions of sensation seeking. Journal Journalof ofConsulting Consultingand andClinical Clinical Psychology, Psychology, 36, 36,45-52. 45-52. Zuckerman, M. (1979). Sensation Hillsdale, Sensationseeking: seeking:Beyond Beyondthetheoptimal optimallevel levelofofarousal. arousal. Hillsdale, NJ: NJ: Erlbaum. Zuckerman, M. (1983). A biological theory of sensation seeking. In M. Zuckerman (Ed.), Biological andand anxiety (pp.(pp. 37-76). Hillsdale, Biological bases bases of of sensation sensationseeking, seeking,impulsivity, impulsivity, anxiety 37-76). Hillsdale, NJ: NJ: Erlbaum. Behavior Zuckerman, M. M. (1984). (1984).Sensation Sensationseeking: seeking:AAcomparative comparativeapproach approachtotoa ahuman humantrait. trait. Behavior Brain Sciences, Sciences,7, 7, 413-471. and Brain 413-471. Zuckerman, M. M. (1990). (1990). The The psychophysiology psychophysiology of of sensation sensation seeking. seeking. Journal Journalof ofPersonality, Personality, Zuckerman, 58, 58, 313-345. Zuckerman, M., Eysenck, S. B. G., & & Eysenck, H. J. (1978). Sensation seeking in England and America: Cross-cultural, age, and sex comparisons. Journal Journal of of Consulting Consultingand and Clinical Clinical Psychology, Psychology,46, 46,139-149. 139-149. T. T., T.,& &Siegel, Siegel,J.J.(1974). (1974).Sensation Sensationseeking seekingand andcortical corticalaugment augmentZuckerman, M. Murtaugh, T. ing-reducing. Psychophysiology, Psychophysiology,11, 11,535-542. 535-542. Zuckerman, M., M., Simons, Simons,R. R.F., F.,&&Como, Como,P.P.G.G.(1988). (1988).Sensation Sensationseeking seekingand andstimulusintensity stimulusintensity as modulators modulators of of cortical, cortical,cardiovascular, cardiovascular,and andelectrodermal electrodermalresponse: response:AAcross-modality cross-modality study. Personality 9, 361-372. Personalityand andIndividual IndividualDifferences, Differences, 9, 361-372.

PART V PARTV

SOCIAL DETERMINANTS OF PERSONALITY

This Page Intentionally Left Blank

CHAPTER 17

PERSONALITY AND SOCIAL STRUCTURE SOCIAL PSYCHOLOGICAL CONTRIBUTIONS

THOMAS F. F. PETIIGREW PEHIGREW THOMAS THE UNIVERSITY UNIVERSITY OF CALIFORNIA, SANTA SANTA CRUZ THE OF CALIFORNIA, CRUZ

L INTRODUCTION INTRODUCTION I.

AUport (1968, (1968, p. p. 9), 9), "be "be both both aacause cause "How can the individual," asked Gordon Allport and a consequence of society?" This is a fundamental fundamental question for social psychol psychology-indeed, ogy—indeed, for all all social sciences and and personality personality psychology psychology as aswell. well. ItIt opens opens questions of the complex causal linkages between personality and various levels of of social structure (Elder, 1973). 1973). Once Once social social structure structure isis defined defined as asrelatively relatively persistent persistent considerasocial patterning, these questions become essential for placing personality considera in their needed social context. tions in A. A Basic Basic Area Area of of Social Social Science Science A. A Virtually all major major social theorists have fashioned fashioned concepts and developed hypothe hypotheses for explaining the linkages of Unkages between personality and social structure. Many of these broad efforts efforts have centered on the presumed negative effects effects on individuals the decline of traditionalism traditionaUsm and and the the rise of modem institutions. Hence, the of the the decline of religious institutions and political stability stabiUty for Emile Durkheim (1951) (1951) was the cause of personal isolation and depression, while for Gustave Le 1897) Le Bon ((1897) HANDBOOK HANDBOOK OF OF PERSONALITY PERSONALITY PSYCHOLOGY PSYCHOLOGY

417 417

COPYIIGIIT COPYRIGHT C © 1997 1997 BY BY ACADEMIC ACADEMIC PlESS PRESS.. A L L RlGIm RIGHTS OF OF 1E'lODUCll REPRODUCTION IN ANY ANY FORM FORM RESElYEO. RESERVED. ALL ON IN

418 418

THOMASF. F.PEITIGREW THoMAS PETnGREW

of humanity to burst forth forth in the crowd. For For Karl it allowed the ugly, irrational side of aUenation Marx (1964), the new technology and property relations led to worker alienation from from their work, their colleagues, and even themselves. For Max Weber Weber (1968), the the growth growth of of bureaucracy bureaucracy threatened threatened dehumanizing dehumanizing and and depersonalizing depersonalizing outcomes. outcomes. Even the triumph of equality, for for Alexis de Tocqueville (1945), could lead to envy Even And for for Sigmund Freud Freud (1930), modern modern life life deepened deepened the conflict conflict and resentment. And between an individual's impulses and their repression. repression. It It remained remained for for the the social social psychologist psychologist and and philosopher philosopher Georg Georg Simmel Simmel (1955) (1955) to to see see any any good good in in the the momentous momentous structural structural changes changes taking taking place place around around these these 19th and early 20th century century theorists. He noted that the social differentiation differentiation that characterized modern modern institutions made it possible for for individuals to develop "a characterized of affiliations" affiliations" in nonascriptive groups. This *'web," web of "web," argued Simmel, ran the of role conflict, even social marginality, in its newly differentiated differentiated and cross crossrisk of of person personcutting social relationships. But it also made possible the individualization of recognition of of uniqueness. Observe the mediated mediated form form of of Simmel's ality and the recognition altered macro-structure macro-structure leads first to cross-cutting group memberships memberships and theory: altered face-to-face interactions that in turn cause individuation. This mediated mediated argument, face-to-face form of of personality and social structure structure we shall note later, remains the principal form theories in social psychology.

B. Twentieth Century Development of of the Area Personality Personality and and social social structure structure constitute constitute an an interdisciplinary interdisciplinary field field par par excellence; excellence; political scientists, anthropologists, sociologists, and psychiatrists as well as psychol psycholformally emerged as a ogists have all contributed. As outlined by House (1981), it formally specialty specialty between between the the world world wars, wars, using using gross gross molar molar conceptions conceptions of of both both personality personaUty and prevailed, and and social social structure. structure. Broad Broad Freudian Freudian notions notions prevailed, and whole whole societies societies were were considered as influential and considered as homogeneous homogeneous entities. entities. Margaret Margaret Mead's Mead's (1935) (1935) influential and contro controversial work, work. Sex Sex and and Temperament Temperament in Three Three Primitive Primitive Societies, typifies this period. During World War II, II, American American specialists attempted to contribute to the war effort advancing sweepingly Russians, and effort by by advancing sweepingly expansive expansive analyses analyses of of the the Japanese, Japanese, Russians, and of these analyses is illustrated illustrated Germans. The extremely holistic and crude nature of by Gorer's (1943) contentions about Japanese national character. Their character structure was capsuled capsuled as anal-compulsive and attributed to early and severe toilet of these national character character studies training in childhood. The most lasting thesis of of authoritarianism authoritarianism as applied to Nazi Germany Germany (Adorno, (Adomo, Frenkel Frenkelwas the analysis of Sanford, 1950). Brunswik, Levinson, & Sanford, Psychoanalytic this work. Psychoanalytic theory theory inspired inspired the the most most prevalent prevalent model model of of this work. Societal Societal maintenance systems were held to shape particular child-rearing practices that in maintenance of the society's members. This mediated model is best turn shaped the personalities of of Kardiner Kardiner (1945) and later Whiting and Child (1953). But tests seen in the work of of the model typically typically failed failed to measure the key components directly. In addition, of homogeneity of of both both a society's practices and its personalities was an unreaUstic unrealistic homogeneity assumed. So cross-national probability surveys were introduced by Cantril (1965),

CHAPTER CHAPTER 17 17 SOCIAL SOOAL PSYCHOLOGICAL CONTRlBunONS CONTRIBUTIONS

419

Almond and Verba (1965), and others to establish national differences differences more firmly. Yet survey data alone could not explain the differences differences uncovered. The past generation has witnessed a shift to more viable, if more modest, approaches (House, 1981). 1981). Both theory and research in the field have become less holistic in their conceptions of both personality and society. Freudian concepts of of personality are less utilized, and increased attention is now given to intrasocietal as well as intersocietal differences. Explanations also center now more on structural than cultural factors. Moreover, while interest in the area waned during the 1970s, witnessed aa revival-as revival—as evinced evinced by by such such seminal seminal papers papers as as those those by by House House the 1980s witnessed Ryff (1987), and Smelser and Smelser (1981). (1981), Ryff

C. The The Potential Potential Social Social Psychological Psychological Contribution Contribution C. is better able to partici particiWithin this new, more specified specified approach, social psychology is pate in the field. But, as Ryff Ryff (1987) makes clear, experimental social psychologists have not as yet exploited this enhanced potential. Figure 1 outlines in broad strokes contribution to this where the discipline is best equipped to make its distinctive contribution interdisciplinary enterprise. Four causal routes are possible between the necessarily interdisciplinary

Rnttnmim Bottom-Up B oCausal ttom-Up I Causal Paths Paths

FIGURE 11 Six Six causal paths. FrGURE

/

/

Clf..otl«nal

\

\

I

Top-Down Top-Down Causal Causal CaUiial Paths Paths

420 420

F. PETnGREW PETHGREW THOMAS �

personality personality and larger structural structural levels of analysis. Paths A, A , B, and C are "bottom *'bottomup" routes routes where influence influence is assumed assumed to flow flow upward upward from from personality personality changes changes to social structural changes. Path A is a direct path; personality personality factors in this route route shape structural alterations without situational factors. without the mediation of meso-level, situational and C provide the the same causal route except that the the personality influence influence Paths B and situational factors before before indirectly shaping structural structural changes. mediated by situational is first mediated D , E, and and F are are "top-down" routes where influence influence is assumed to flow flow Paths D, downward from from initial social structural changes to personality changes. Paths D downward flow of influence influence is first mediated by situational situational and E assume that this downward flow allows direct structure-to-personality structure-to-personality effects. This chapter will pro profactors; Path F allows between specific empirical empirical illustrations of each of these four causal paths between vide specific and social structure. personality and personality Like other social has the the potential potential to contribute contribute social sciences, social psychology psychology has found understanding of all four potential paths. But its distinctive expertise is found to the understanding and E, the the bottom-up and and top-down links between between personality personality and in Paths B and and face-to-face situations. Note, for example, the the overwhelming overwhelming proportion of studies studies face-to-face the Path B and and E levels in the the December December 1987 1987 issue of working exclusively exclusively at the The Journal of Personality and Social Psychology devoted entirely to "integrating The Journal of Personality Social Psychology and social psychology." The The field, field, then, is best at uncovering uncovering how how the the personality and two-way influences influences between between personality personality and and social structure structure are are mediated at the two-way meso-situational meso-situational level of analysis. But not all, or even most of, social psychology psychology is engaged in this process. Only Only the discipline, the the survey-oriented survey-oriented contextualists contextualists (Pettigrew, (Pettigrew, 1980), one branch of the concerned themselves themselves with the field field of of personality personality and social structure. Indeed, have concerned only the contextualists, found the United United States, States, found largely largely in sociology sociology departments in the have continued continued to concern themselves themselves with personality personality questions questions of of any type. Herein Herein the limits of social lies an apparent paradox, one we must consider to understand the the study of personality personality in recent years. psychological contributions to the n. AN ApPARENT APPARENT PARADOX D.

The apparent paradox involves the way way the the two two social psychologies, sociological sociological of personality. Many social psychologists in sociol socioland psychological, treat the study of ogy, the macro social science that studies social structure, have over the past genera generaand concept concept of personality personality far more seriously than social tion taken the study and in psychology, psychology, the the micro social science that studies individuals. psychologists in There are many indications of of this paradoxical situation. Consider Consider the differen differenThere personality in the two handbooks of the social psychological psychological tial attention provided personality discipline. The The Rosenberg Rosenberg and Turner (1981) volume, volume. Social Social Psychology: Psychology: Sociologi Sociologi21 chapters, 10 of which involve personality personality concepts, cal Perspectives, contains 21 cal Perspectives, identity data, and theory. Three chapters dwell on socialization processes; 1 concerns identity and another another the self-concept; and and no fewer fewer than 5 deal directly with formation and formation and society. By contrast, the the third edition of Lindzey Lindzey and and Aronson's Aronson's personality and

17 SocIAL SOOAL PSYCHOLOGICAL CONTRIBUTIONS CHAPTER 17

421 421

(1985) Handbook of of Social Psychology devotes only 3 of its 30 chapters to largely personality personality concerns-1 concerns—1 on altruism and aggression, another on adult socialization, and a 3rd on personality and social behavior. An updated version of the classic chapter chapter in the first and second Handbook Handbook editions on national character by Inkeles and Levinson (1954, 1968) was omitted from from the third edition in 1985. Further evidence for the neglect of personality in psychological social psychol psychology is provided provided by the protests some years ago in Division 8 of the American Psychological Association (APA) among personality psychologists that their inter interests were not fairly represented. One response to these objections took the form of altering the basic format Personality and and Social SocialPsychology, Psychology, format of The Journal of of Personality with a separate section for "personality processes and individual differences." differences." The fact that personality and social psychology have long been joined together major AP APA A division and journals heightens the paradox. Personality in the same major historically played a large role in the development of psycholog psychologconsiderations have historically far less strong tradition in sociological ical social psychology. By contrast, there is a far often suspect for social psychology. Indeed, social psychologists in sociology are often their micro-level interests (DiRenzo, 1977). Thus, in the Rosenberg and Turner defend their focus on personality (1981) volume, several writers felt it necessary to defend selfas relevant to the larger discipline of sociology. In his excellent chapter on the self for example, Rosenberg (1981, p. 593) begins by allowing that the concept concept, for "may appear appear to to be be peculiarly peculiarly ill-suited ill-suited as as aa subject subject for for sociological sociological concern." concern." "may

A. So So Why Why the the Paradox? Paradox? A. Three interrelated factors explain this apparent paradox, and these factors are important for understanding the study of personality and social structure within contextual social psychology. First, social psychologists in the sociological mold different—though not conflicting-aims conflicting—aims from from their psychological colleagues. have different-though psycholSecond, there has been over the past generation a general trend throughout psychol ogy toward situationalism. And, finally, sociological social psychology has been from these psychological influences. Let us consider each of these fac facshielded from further. tors further. /. Diff Different Theoretical Aims Aims 1. erent Theoretical Social psychology serves as as both the most macro-focused subdiscipline of psychology micro-focused subdiscipline of sociology. As such, social psychologists and the most micro-focused on the two sides of the disciplinary divide often often have somewhat different, different, though aims. Social psychologists in psychology are concerned complementary, theoretical aims. heavily with the links between the individual and the situational levels of analysis (Paths B and E in Fig. 1). Consistent with their interest in broader social structure, sociological social psychologists are sometimes also concerned with the links to such larger social structures as institutions (Paths A, C, D, and F in Fig. 1). from Kurt Lewin's fascination fascination with Experimental social psychology broke from human motivation and has increasingly been swept up with the cognitive revolution

411 422

THOMASR F.PETnGREW PETTIGREW TnOMAS

throughout psychology. psychology. Many advances in in the the disciphne discipline in throughout Many of of the the most most notable notable advances in recent cognition. This This motivation-to recent decades decades have have come come in in the the subfield subfield of of social social cognition. motivation-tocognition paradox under cognition shift shift is is important important for for understanding understanding the the paradox under discussion. discussion. A A focus focus on on motivation motivation lends lends itself itself to to personality personality considerations; considerations; the the theory theory and and work work of teacher, Gordon Allport, offer evidence of of this close connection. connection. As of my my teacher, Gordon AUport, offer evidence this close As the the longtime editor of of the old Journal Journal of of Abnormal Abnormal and and Social Psychology, he was both Social Psychology, AUport experienced no role conflict conflict in this a personality and a social psychologist. Allport motivation and dual role, given his emphasis on both motivation and individual uniqueness. cognition in recent recent decades has weakened weakened this But the increasing focus on cognition connection connection between between the the two two fields fields that that compose compose Division Division 8. 8. Cognitive Cognitive variables variables have substituted substituted for for personality personality variables as micro indicators for for much experimental experimental further that social psychological work over the past generation. Jones (1985) notes further fundamental attribution attribution error (Ross, 1977) acted acted to erode further further attention to the fundamental internal dispositions. This marked interest in approaches that assigned weight to internal shift from from motivation motivation and personality concerns concerns to cognition cognition was made even more shift pronounced by the lack of of interest interest in individual differences differences in cognition. Cognitive pronounced variables and processes have generally been assumed (but rarely empirically demonempirically demon operate universally. strated) to operate Given their theoretical aims, many sociologists are less prone to make such easy the personality personality and and social easy assumptions assumptions of of universality. universality. Social Social psychologists psychologists in in the social structure types of people are are attracted structure tradition tradition often often wish wish to to see see how how different different types of people attracted to to particular institutions (a Paths A A and B-C self-selection self-selection problem), and how different different particular institutional-shaping problem). institutions alter individuals (a Paths D-E and F F institutional-shaping Some cognitive variables Some cognitive cognitive variables variables are are of of interest interest for for these these issues. issues. But But cognitive variables cannot such personality-structure cannot fully fully substitute substitute for for personality personality variables variables in in such personality-structure work. work. 2. General General Psychological Psychological Trend Trend toward toward Situationalism Situationalism

The move from from motivational to cognitive concerns in experimental experimental social psychology was was furthered furthered by by other other trends trends within within psychology psychology as as aa whole, whole, especially especially personality personality psychology. psychology. McGuire McGuire (1985) (1985) holds holds that that there there have have been been 2525- to to 30-year 30-year swings swings of of the of theoretical theoretical empha the pendulum pendulum between between the the personal personal and and the the situational situational poles poles of emphasis. sis. And And Snyder Snyder and and Ickes Ickes (1985) (1985) detect detect aa returning returning interest interest in in personal personal consistency. consistency. But the sharp decline in experimental social psychology's interest interest in personality personality in the situationalism. Mischel's the 1960s 1960s coincided coincided with with the the rise rise of of an an extreme extreme situationalism. Mischel's (1968) (1968) the high high point of influential influential book, book. Personality Personality and and Assessment, Assessment, marks the of this latest return to to the the Hartshorne Hartshorne and and May May (1928) (1928) emphasis emphasis on on the the situation. situation. The The most most memorable and cited part of his contention on the the limits limits of of of Mischel's critique was his contention on personality predictions of correlation coeffi of behavior across situations: the magic correlation coefficient of of .30, just 9% of of the variance, became part of of psychological-but psychological—but not socio sociological—lore. logical-lore. The Mischel's arguments The differential differential influence influence of of Mischel's arguments can can be be traced traced in in the the citations citations to his 1968 book and related articles. In the Rosenberg and Turner Turner (1981) volume references to this work are made; in on sociological social psychology, only two references Aronson (1985) Handbook Handbook published 4 years later, a dozen such the Lindzey and Aronson

CHAPTER CHAPTER 17 SOCIAL SOCIAL PSYCHOLOGICAL PSYCHOLOGICAL CONTRIBUTIONS CONTRIBUTIONS

423 423

references are made. There is an irony here, perhaps. Jones (1985) asserts that it is the Lewinians who have most focused on situations, while it is the non-Lewinians who have taken Lewin's famous B = = f (P,E) equation more seriously. In any event, the rise of situationalism in personality personaUty psychology itself itself in the late 1960s proved influential influential because the overwhelming power of situations to shape unexpected behavior behavior had already become an empirical preoccupation of psychological social psychology. From Asch's (1956) famous famous conformity conformity study and Milgram's (1974) obedience studies to Zimbardo's (1970) prison simulation and Tajfel's Tajfel's (1970) "mini "minipowerful mum situation," much of the exciting empirical work in the field provided powerful framework that did not require, beyond cognitive variables, support for a situational framework attention to the Person component to Lewin's formulation. Even the scant attention differences in psychological social psychology has that has been paid to individual differences generally been within this situational perspective. Hence, a leading personality measure in experimental studies of the 1970s was Rotter's (1966) external-internal variable—a measure that fits easily within a largely situational locus of control variable-a framework.

3. The The Shielding Shielding oof ContextualSocial SocialPsychology Psychologyffrom 3. f Contextual rom These Psychological Psychologicallnftuences Influences These If these influences converged to lessen social psychology's interest in personality within psychology, why did they not also cause a similar decline in social psychology within sociology? There are both theoretical and empirical answers to this question. But, first, several distinctions must be made among sociology's various branches distrusted stable of social psychology. Symbolic interactionists have traditionally distrusted in their emphasis on specifically specifically situated and emerging definitions definitions personal attributes in and meanings (Blumer, 1969). Likewise, the ethnomethodologists have little use for dispositional concepts. They focus on how people make sense of their social for environment; in doing so, they are inspired in part by modern modem linguistic theory and often often employ conversational analysis in their empirical work (Handel, 1982). Unfortunately, symbolic interactionists and ethnomethodologists rarely read the psychological literature, and vice versa. For these two branches, then, the paradox under discussion does not exist. They have never systematically used personality indepenvariables, and they arrived at their own versions of situationalism largely indepen dent of psychological influence. The sociological group of -so of interest interest is the "contextual social psychologists" psychologists"—so for the social context of social psychological named because of their special concern for familiar with the psychological literature, edit phenomena. They are thoroughly familiar Psychology Quarterly, Quarterly, and compose most of the membership of the The Social Psychology of the American Sociological Association (Pettigrew, Social Psychology Section of often use survey and other field methods in an attempt to balance 1980). They often external and internal validity concerns. Less mobilized than symbolic interactionists and ethnomethodologists, this largest group of sociological social psychologists have traditionally employed employed and and continue continue to to employ employ personality personality variables. variables. And And both both the the traditionally

424 424

THOMAS F. F. PEmGREW PETTIGREW THoMAS

theory theory and the methods of of contextual contextual social psychologists acted to shield them them from from the recent situationalist situationalist influences influences in psychology. psychology.

a. Theoretical Social psychologists Theoretical Shielding. Shielding, psychologists in psychology psychology use personality personality measures measures as both independent independent and dependent dependent variables. But the theoretical theoretical interests interests of contextual contextual social psychologists psychologists in sociology sociology lead to a virtually virtually exclusive exclusive use use of of personality personality variables as dependent dependent measures of of the effects effects of of social structure structure of (DiRenzo, 1977; Elder, 1973; but see exceptions in Part IV of of Smelser & Smelser, Smelser, (DiRenzo, For this purpose, the attacks attacks on personality personality variables as predictors predictors were were 1963). For power of irrelevant. More generally, sociological sociological theory, with its emphasis on the power social structure, had long long served to immunize immunize sociologists against the fundamental fundamental social attribution error error that operated operated among among extreme extreme dispositionalists dispositionalists in psychology. psychology. Thus, attribution Milgram, Zimbardo, Zimbardo, and Tajfel Tajfel experimental experimental demonstrations demonstrations did not not the Asch, Milgram, seem nearly so surprising, dramatic, and "nonobvious" to social psychologists psychologists in seem sociology as they did to those in psychology. Contextual Contextual social psychologists psychologists had sociology long before before gravitated to a person-situation-structure person-situation-structure interactionist interactionist position position closer closer long Lewinian formulation formulation left left behind by many Lewinians. This position shielded shielded to the Lewinian from the rampant rampant situationalism situationalism these publicized publicized studies triggered triggered in psy psythem from chology. difference in familiarity familiarity with social structure between between This point highlights the difference psychologists in the two fields. fields. Accustomed Accustomed to dealing with social structure, social psychologists contextual social psychologists psychologists in sociology sociology have seen no reason to deemphasize deemphasize contextual personality just because strong effects effects of of situations are demonstrated. By By contrast, contrast, personality experimental social psychologists psychologists have little training in social structure, and, after after experimental of a century of famous situational experiments, the branch has yet to to a third of formulate taxonomy of formulate a taxonomy of situations or to specify specify a consistent consistent means of of measuring measuring and comparing comparing situations. 6. Empirical Empirical Shielding. Shielding, While some some sociological sociological social psychologists psychologists conduct conduct b. experiments, most researchers in the contextual contextual tradition use a variety of of field field methods-particularly methods—particularly surveys. This work work often often requires requires regression regression and other tech techniques quite different different from from the standard analyses analyses of of variance variance designs designs routinely routinely used used in experimental experimental social social psychology. psychology. This fact fact has meant meant that contextualists contextualists have have benefited benefited from from the advances advances in empirical empirical techniques techniques and statistical statistical treatments that have characterized characterized sociology sociology generally generally in recent years. This statistical upgrading upgrading was made made more more rapid by sociology's sociology's adaptations of of econometric econometric techniques-for techniques—for sophisticated in social science. While the the all their problems, still by far the most sophisticated sociologists sociologists were were upgrading upgrading along econometric econometric lines, psychological psychological social social psycholo psycholowere not keeping keeping pace. gists were differential statistical statistical sophistication sophistication is part of of the answer to the paradox paradox This differential under under discussion. Mischel's Mischel's magic coefficient coefficient of of .30 as the presumed presumed predictive limit of personality personality variables, which impressed impressed many psychologists, had no impact on on of sociologists more skilled with regression analyses and accustomed to similar coeffi coeffisociologists cients from from such crudely measured measured variables as social class. The role of of error in

CHAPTER 17 SOCIAL SOOAL PSYCHOLOGICAL CONTRIBUTIONS CONTRIBUHONS CHAmR

425

both independent and dependent variables, the need for aggregated measures, and similar considerations were more widely appreciated in sociology. To be sure, Epstein (1979) and others countered the Mischel arguments with these points in psychology, but it took a decade for these correctives to appear. B. B. The The Foci Foci of of Contextual Contextual Social Social Psychology Psychology While psychological social psychologists have focused less in recent decades on personality, personaUty, they have focused focused more on the specification of interactional processes. It is is precisely these processes that are needed to explain the causal linkages between the various levels of analyses. Indeed, numerous commentators agree that the general failure to specify specify linkage processes is the glaring weakness of the personality and social structure realm (DiRenzo, 1977; 1987). To make our 1977; Elder, 1973; 1973; Ryff, 1987). discussion concrete, let us review illustrations of the various paths and some of the linkage processes that have been advanced to explain the shaping of social structure by personality.

DI. III. PERSONALITY SHAPES SOCIAL STRUCTURE Sweepingly global "great man" theories of history have put in disrepute the conten contention that an individual can make a major major difference difference in society. Had Napoleon not Zeitgeistwould wouldnot nothave haveproduced producedaa lived, goes the counter, who is to say that the zeitgeist similar charismatic French leader with little change on history? At this gross level Uttle value. Yet it has been demon demonof abstraction, of course, such arguments are of little strated that personality personaUty does shape social structure once the measures at both levels strated effect have been more narrowly specified. Indeed, laboratory and the hypothesized effect research even supports a narrowed version of "great man" theory when placed in the specific context of small task-oriented groups (Borgatta, Bales, & Couch, 1954). from Figure 1, "bottomWe have already noted that, following from 1, two principal "bottom up" causal paths are possible. One involves the direct effects effects of Path A A— up" unmediated alterations of social structure by personality factors. The other involves effects. Paths B and C of Figure 1, where personality personality first first alters alters situations situations mediated effects, 1, where that in tum turn shape larger social structures. Consider examples of each of these bottom-up causal paths between personality and social structure.

A. A. Direct Direct Effects Effects (Causal (Causal Path Path A) A) Max Weber's (1930) theory of the rise of capitalism constitutes the classic Path A between personality personality and social structure. In his Protestant Protestant Ethic and the linkage between of Capitalism, Capitalism,Weber Webercountered counteredMarx Marxby byholding holdingthat thatsuch suchmicro microphenomena phenomena Spirit of independently shape macro structure. With as beliefs, motives, and values could independently 17th century Hollanders particularly in mind, he explained why ascetic Protestant often met with economic success even though they viewed wealth as dangerous sects often

426 426

THoMAS GREW THOMASF. F.PE111 PEITIGREW

for soul. He argued Protestant ethic, epitomized in the anxiety-arousing for the soul. argued that the Protestant anxiety-arousing Calvinist doctrine predestination, explained Calvinist doctrine of of predestination, explained the the riddle. riddle. A A self-denying, self-denying, frugal, frugal, and work-oriented routine in final judgment and work-oriented routine in busy busy preparation preparation for for final judgment led led to to accumulated accumulated wealth. wealth. Weber Weber also also held held that that social social structure structure shapes shapes personality personaUty in in aa reciprocal reciprocal cycle, and Protestant ethic's capitalism was cycle, and he he understood understood that that the the Protestant ethic's influence influence on on capitalism was mediated by family, work, and other face-to-face, face-to-face, meso-level situations. Yet his mediated famous prototype of of a Path A A causal theory involving personality famous theory remains the prototype and social structure. and Building on thesis, David 1961) posited posited that Building on Weber's Weber's thesis, David McClelland McClelland (1955, (1955,1961) that the the association capitalism was mediated by association between between the the Protestant Protestant ethic ethic and and capitalism was mediated by the the achievement by child-training achievement motive. motive. Fostered Fostered by child-training practices practices consistent consistent with with the the Protes Protestant tant ethic, the need for for achievement, he maintained, is both challenged challenged and satisfied satisfied entrepreneurial activity activity that characterizes characterizes capitalism capitaUsm in its early stages. Both by entrepreneurial high-need achievers and entrepreneurs entrepreneurs are independent-minded, independent-minded, relatively relatively autono autonohigh-need mous, and strivers for for success. The only situational mediation McClelland McClelland specifies specifies responsible is that such people seek out situations in which they can feel personally responsible for clearly measurable results of of their efforts efforts (Atkinson & Hoselitz, 1958). Carrying Carrying for attempted to show that the his analysis to the macro level, McClelland (1961) attempted achievement motive played a significant significant part in the rapid economic growth of of the achievement emphasized cross-national cross-national and longitudinal relationships between levels West. He emphasized of achievement achievement and a variety of of measures of of technological technological growth. of Psychohistorical studies, from from biographies of of such famous famous figures figures as Luther Luther Psychohistorical 1958,1969) of such broad social phenomena as 1969) to analyses of and Gandhi (Erikson, 1958, racism racism (Kovel, (Kovel, 1970), 1970), offer offer additional additional examples examples of of such such sweepingly sweepingly broad, broad. Path Path A A theories. theories. But But increasingly increasingly research research and and theory theory in in this this vein vein are are more more delimited delimited in in scope example, Gluckstern scope and and more more differentiated differentiated in in prediction. prediction. For For example, Gluckstern and and Packard Packard (1977) (1977) studied studied change change in in aa prison. prison. In In viewing viewing the the personalities personalities of of change change agents, agents, they they noted noted that that different different styles styles acquired acquired greater greater salience salience at at different different times times in in the the process process of of organizational organizational change. change. Such Such results results suggest suggest the the interactive interactive nature nature of of personality personality and and social social structural structural relationships, relationships, aa major major point point to to which which we we shall shall return. return. Mediated Effects (Causal Paths B and C) B. Mediated

We have noted noted that that social psychology makes its distinctive contribution contribution by explain explaining the links between the individual and the situational situational levels of of analysis-both analysis—both Path Path B and Path E (Pettigrew, 1988). 1988). In doing so, the discipline provides distinctive variables and explanations together with a general situational perspective largely from the analyses of of other other social sciences. These distinctive variables and missing from subjective interpretations of of the social environment. environment. explanations usually involve subjective for the Path B approach, the field conceptualizes personality as individual Hence, for respond differentially differentially to particular particular situations. These individual susceptibilities to respond difference variables, then, lend themselves to links with the situational level. Thus, difference Rotter's (1966) internal-external internal-external locus of of control control is one of of the most used individual individual Rotter's measures in social psychological research.

CHAPTER 17 17 CHAPTER

SoaAL PSYCHOLOGICAL PSYCHOLOGICAL CONTRIBUTIONS CONTRIBUTIONS SOCIAL

427 427

Applied social psychology's remedies for problems using this bottom-up bottom-up direc direction usually take the form of special training-for training—for the critically ill, the lonely, the shy, the elderly, burnouts, rape victims, and those facing major surgery and other serious life events-to situation. Often Often self-imposed, self-imposed, maladap maladapevents—to reconceptualize their situation. tive dispositional attributions are converted to situational attributions when approappro priate. These remedies resemble superficially typical therapeutic attempts to change These remedies resemble superficially typical to change internal qualities of individuals. But actually these remedies are directed primarily at changing the behavior of individuals in problematic situations. Ultimately, most . social psychological remedies try to alter the nature of human interaction in situasitua tions. Both personality of personaUty and social psychologists have produced literally hundreds of Path B studies over the years. This most popular of bottom-up paths demonstrates how personality makes a significant difference at the situational level. Much of the work of this genre following World War II was conducted in the small group research tradition, the most important of which came from Freed Bales and his colleagues. In time, this work led to the most developed, data-based model that exists of the complex interaction of personality and task group structure (Bales, This interactive interactive model model encompasses encompasses linkages linkages of both Paths B B and and E E of of Figure Figure 1970). This by combining group group roles with with personality traits in aa three-dimensional value 11 by alone, for for example, example, the the model model generates generates the basis basis of personality measures measures alone, space. On the predictions concerning likely coalitions and behavior in the task-oriented group. Other early work showed how a single disruptive person (a schoolroom devidevi ant, an alcoholic, a mental patient) could radically alter the operation of a family 1960; Jackson, 1956; 1956; Yarrow, Clausen, & Robbins, 1955). or classroom (Gnagey, 1960; More recently, following the cognitive revolution, such individual factors as causal attributions are shown to affect situational perception and interaction. Such studies The Journal Journalof ofPersonality Personalityand and Social Social predominate in the December 1987 issue of The Psychology Psychology dedicated to "integrating" the two psychologies, even though as we have noted this substitution of cognitive for personality factors is a principal way experimental social psychology has neglected personality considerations in recent decades. However, in one of the collection's papers, an individual's gender role orientation and level of moral reasoning are are shown shown to to relate relate to tothe the influence influence exerted exerted by the subject on a group decision-making task (McGraw & Bloomfield, 1987). But Path C studies from the meso situation to the macro structural level are understandably rare in psychology. This missing link results in an incomplete understanding of the mediated, bottom-up causal path from personality via situasitua tions to institutions. The necessity of interdisciplinary cooperation is underscored, for Path C examples are commonplace in sociology and organizational research. Often these studies involve negotiation over the nature of social roles and the informal operations of the social structure. Hence, informal communication and exchange processes at the situational level can create continual modifications modifications in the structure of highly bureaucratized governmental agencies (Blau, 1955). Such negotiation can even occur in "total institutions." Sykes (1958) demonstrated how prisoners, working working collectively collectively at atthe thesituational situationallevel, level,managed managedto toalter alterthe thestructure structure

428 428

THoMAs R PETnGREW THOMAS F.PETHGREW

of unpopular formal of a maximum security prison. Especially unpopular formal rules went unenforced, unenforced, and their own and prisoners prisoners gained gained considerable considerable authority authority over over their own lives lives within within the the con constraints personality with with roles straints of of neither neither escaping escaping nor nor rebelling. rebelling. The The mismatching mismatching of of personality roles and and social structure is not always correctible, of of course, and the individual alienation and stress that results from "top-down" effects. from such mismatching constitute *'top-down" effects.

IV. IV. SOCIAL STRUCTURE SHAPES PERSONALITY The most hypothesized personality and and social social structure hypothesized and studied links between between personality structure involve the top-down top-down causal links. Instead of of being the initial causal agent, personal personality factors factors are now the dependent dependent variables. We have noted noted that that two general general paths possible—the direct, unmediated unmediated route route (Path (Path F F of and the mediated are possible-the of Figure 1) and D and E) through which structural changes shape personality personality by first route (Paths 0 face-to-face human interaction interaction in situations. altering the face-to-face Direct Effeds Effects (Causal Path F) A. Dired Mediated causal theories are to be preferred, preferred, because their linkage processes be beMediated generally far far better specified. Direct Direct effect effect theories, whether whether Path tween levels are generally A or F, generally generally signal a failure to demonstrate demonstrate how the causal sequence sequence carries carries A face-to-face interaction stage-though stage—though this is not to deny the possibility through the face-to-face of true direct effects effects that go unmediated unmediated at the situational situational level. In recognition recognition of of of proportion of of top-down top-down theories in social science gener generthis fact, the overwhelming proportion psychology in particular, are mediated, Paths D D and E E contentions. ally, and social psychology Why should this trend trend toward toward more situational situational specification specification be more evident than speculations suggest than for for bottom-up bottom-up theories? theories? Two Two interrelated interrelated speculations suggest themselves. themselves. First, the fundamental fundamental attribution attribution bias-the bias—the tendency tendency to magnify magnify the causal power as others. power of of individuals-operates individuals—operates for for social social scientists scientists as as well well as others. Indeed, Indeed, psychologists bias than than most psychologists may may be be even even more more prone prone to to the the bias most human human beings. beings. Second, Second, there there has has been been considerably considerably less less awareness awareness of of and and attention attention to to "the "the compositional "ecological fallacy." fallacy." The compositional fallacy" fallacy" than than to to the the corresponding corresponding "ecological The compo compositional sitional fallacy fallacy occurs occurs when when the the dynamics dynamics of of groups groups and and collectivities collectivities are are simply simply extrapolated of individual extrapolated from from the the aggregate aggregate of of tendencies tendencies of individual persons persons composing composing the worst when when social the group. group. Such Such extrapolation, extrapolation, often often seen seen at at its its worst social problems problems are are viewed tendencies, leads viewed as as mere mere cumulated cumulated extensions extensions of of individual individual tendencies, leads to to error error because because it it denies denies the the operation operation of of distinctive distinctive group-level group-level processes. processes. In In reverse, reverse, the the ecological ecological fallacy fallacy occurs occurs when when individual individual characteristics characteristics are are simply simply extrapolated extrapolated from from aggregated aggregated data. data. Psychologists Psychologists and and bottom-up bottom-up theorists theorists seldom seldom discuss discuss the the compositional literature on compositional error. error. But But there there is is in in sociology sociology an an extended extended literature on the the ecological (e.g., Robinson, theorists are ecological fallacy fallacy (e.g., Robinson, 1950), 1950), and and most most top-down top-down theorists are keenly keenly aware of of it. A major major means of of avoiding the ecological error error is to specify specify the linkage processes through a mediated mediated model.

CHAmR CHAPTER 17 17

SOCIAL SOQALPSYCHOLOGICAL PSYCHOLOGICALCONTRIBUTIONS CONTRIBUTIONS

429 429

At At any rate, direct Path F examples are hard to find find in the social science literature. Uterature. Perhaps, for illustrative purposes, two famous famous studies in political science come closest. Almond and Verba (1965) studied five democratic democratic nations with varying "political cultures." They used survey methods to ascertain each nation's political attitudes; then they compared these data with each nation's political institutions. The most most relevant of of their contentions concerns their their structural explanations for the wide attitudinal differences differences across the five populations that remain even after such variables as education are controlled. For example, British respondents evinced considerable confidence confidence in both their administrative and their political-legislative officials; Germans particular confidence confidence in administrative administrative officials; Americans in legislative officials; and Italians and Mexicans in neither. Almond Almond and Verba ac accounted for these aggregated individual differences differences in terms of of the contrasting structural histories across the countries. Thus, Germany Germany had witnessed the establish establishstructural ment of of an early and stable bureaucracy, but late and unstable political development. The United United States had much the reverse history, with a late developing developing civil civil service but an early foundation foundation of of its political structure. The United Kingdom was seen as United Kingdom having experienced early development development of of both institutions, while while Italy and Mexico having suffered late development development of of both. had suffered of Gurr (1970) on the important important role of The work of of relative deprivation in affords a second example of of Path F revolution across 1119 19 nations affords F theorizing. Using a popular social psychological concept ((Pettigrew, Pettigrew, 1978; Walker & Pettigrew, 1984), measurement of of relative deprivation deprivation an unmediated unmediated develop developGurr assumes in his measurement of this motivational motivational condition condition of of individuals from condiment of from particular structural condi of individuals from from such tions. Hence, he directly infers the relative deprivation of proportion of of the population population excluded excluded from from valued aggregated measures as the proportion Then Gurr shows shows a positive, positive. positions and the increase in migrants to the cities. Then A relation between "relative deprivation" and political political unrest. In truth, then, Path A pseudo-demonstration of of a personality and structure relationship. By com comthis is a pseudo-demonstration ecological fallacy, Gurr's research research operations operations actually actually check relationmitting the ecological check on relation ships between structural variables and simply assumes that relative deprivation is connection underlying these relationships. the psychological connection feature of of these two For present purposes, the striking feature two studies is their unmediated character. Just how did the differentially differentially rooted rooted administrative and political political structures across the five countries shape individual individual attitudes? Were they reflected in different different socialization patterns concerning politics? In different different face-to-face level? Or, in Gurr's work, just how mobilization at the face-to-face how does exclusion from valued positions and migration to the city produce relative deprivation? from of situations—the of In other words, how did the meso level of situations-the principal level of psychology—transform the structural differences differences across nations into attitudi attitudisocial psychology-transform differences? Without such mediation, the linkage processes nal and motivational differences? left open involved remain unspecified, and the theory left open to numerous plausible alternative explanations. explanations. alternative

430 430

THOMAS F. THOMAS F. PEmGREW PETHGREW B. B.

Mediated Mediated Effects Effects (Causal (Causal Paths Paths D D and and E)

Social psychology's distinctive situational focus causes it to prefer prefer manipulable situational variables (Pettigrew, 1988). This This focus focus isis aa great great advantage advantage when when applica applications are made to practical problems, because it is far far easier and more ethical to top-down approach, the disci discialter situations than to alter people. For the Path E, top-down pline searches for characteristics of situations that shape and elicit eUcit specific behavior from individuals. Aronson's jigsaw cooperative learning design for for classroom offers offers from 1978). When When pressed pressed a superb example (Aronson, Blaney, Stephan, Sikes, & Snapp, 1978). for advice in solving practical problems, the field differs differs from bottomfor from psychology's bottom up remedies by typically advancing top-down, situation-to-individual situation-to-individual recommenda recommendations: change change the situation situation (increase (increase choice, choice, cease cease labeling, labeling,allow allowmore more participation, participation, etc.) and individual improvements are predicted predicted (improved (improved morale, less sense of of being stigmatized, greater involvement, etc.). Often Often the linkage Unkage processes invoked to explain these Path Path E relationships family expectations. Hence, in his famous famous study of industrial involve role playing or family foremen and stewards, Lieberman (1950) demonstrated demonstrated how attitude changes plant foremen followed role changes. Newly appointed foremen foremen soon became more procompany followed while newly newly appointed appointed stewards stewards soon soon became more prounion. Within in their views, while groups had established estabhshed almost "diametrically opposed sets of of 3 years, the two groups foremen and stewards who returned to the attitudinal positions." And for those foremen revert to earlier held production line as regular workers, there was a tendency to revert provided by the "double "doubleattitudes. The classic example of family process linkage is provided bind" theory of the family origins of schizophrenia, in which childrenfind find themselves conflicting parental expectations expectations (Bateson, Jackson, Haley, & caught in the midst of conflicting Weakland, 1956). mediated. Paths D and But, unlike the case with bottom-up causal theories, mediated, E, top-down work is plentiful literature. Many plentiful in the contextual social psychology Uterature. of the best exemplars focus on the individual effects effects of social stratification. The for instance, demonstrated longitudinal research of Sewell and Hauser (1975), for the powerful powerful influence influence of socioeconomic status on educational aspirations and attainments. Mediated through such situational contexts as neighborhoods and effects of social class remain even after after measured ability and schools, the strong effects gender are controlled. major contextual social psychological projects over the past generation Three major of structural structural and cultural have played a dominant role in defining top-down analyses of shaping of personality. The first of these involve Melvin Kohn and his colleagues. In a notable series of interrelated studies, he has attempted to specify specify class and He emphasizes emphasizes the importance importance of occupational demands medi medipersonality linkages. He family socialization patterns for for shaping self-direction conformity. ated by family self-direction versus conformity. First, Kohn (1969; Pearlin & Kohn, 1966) showed that in both the United States and Italy middle-class parents more often often stressed intent and self-direction self-direction while often stressed obedience to external aulower- and working-class parents more often au thority.

CHAPTER CHAPTER 17 17

SOCIAL SOOAL PSYCHOLOGICAL PSYCHOLOGICAL CONTRIBUTIONS CONTRIBUTIONS

431 431

selfNext, in studies with Carmi Schooler (Kohn & Schooler, 1969, 1973), self direction values were shown to be positively associated with occupational position. self-direction the most; this trend reflected, reflected, Those in higher-status positions valued self-direction Kohn and Schooler (1969) argued, their greater opportunities to exercise occupa occupational self-direction-to self-direction—to use initiative, thought, and independent judgment in their work. Likewise, those in lower-status positions, who largely had to obey others, valued obedience to authority. Hence, the demands of the workplace generalized to the home and the preferred preferred parental means of socializing children. Kohn thus stratification inherent provides an answer to the broad question, How does the social stratification differentiation of the labor market translate into differential differential values? in the structural differentiation class-associated— The class and value relationship derives, he maintains, from class-associated particularly occupational-conditions occupational—conditions of life: "It is chiefly by shaping the everyday realities people must face that social structure exerts its psychological impact" (Kohn, 1977, 1977, p. xlviii). The second famous project utilizing causal Paths D and E extends this analysis effects both conceptually and cross-nationally. Alex Inkeles (1960, of occupational effects 1969,1978; Inkeles & & Smith, Smith, 1974) 1974) argues arguesthat that industrialization industrialization leads leadsto tosimilar similarforms forms 1969, 1978; Inkeles "modern" of social organization across six culturally divergent societies. societies. These "modern" face-to-face situations that in time produce similar patterns of of organizations shape face-to-face "modern" beliefs, perceptions, values, values, and even thought. structural theory-not theory—not to to be be confused confused with with Note that this is strictly a social structural 1961). That is, both Kohn such cultural ideas as "the culture of poverty" (Lewis, 1961). patterning rather rather then then the the content contentof of and Inkeles emphasize the importance of the patterning modern society. Kerckhoff Kerckhoff (1976) contrasts the two types of explanation in terms allocation (indirect (indirect transmission transmission through throughstructural structural constraints constraintsand and imperatives imperatives of allocation socialization (direct transmission of learned skills, plus selection criteria) versus socialization topmotives, attitudes, etc.). In our terms, allocation theses lend themselves to top down analyses, while socialization theses assume bottom-up and top-down forms, fully separable-indeed, separable—indeed, they are though in practice the two approaches are not fully complementary. But the distinction is important important for for considering the full full social context of personality. Inkeles (1960, 1969, 1978; 1978; Inkeles & Smith, 1974) surveyed 6,000 industrial and agricultural male workers in six developing countries-Israel, countries—Israel, Argentina, Chile, India, Nigeria, and what is now Bangladesh. Of special interest is his careful of careful specification specification of the independent variables (the structural positioning of the workers) and the dependent dependent variable (the "individual modernity" personality syndrome). Proving considerably more important than national cultures were the structural components considered together: education, factory experience, mass media exposure, urban residence, possession of consumer goods, and age. A syndrome was predicted across these six nations that deserves attention from personality psychologists. Its seven components are: (1) openness to new experience; (2) assertion of independence from from traditional figures to an allegi allegiance to modern modem leaders; (3) belief belief in scientific efficacy efficacy rather than fatalism; fatalism; (4) ambitious occupational and educational goals for for one's self and children;

432 432

THOMAS PEmGREW THOMAS F. F.PETTIGREW

(5) concern for punctuality punctuality and planning; (6) interest interest and participation participation in local politics; and (7) interest in national and international news. international The third of of the defining defining research projects in this area is Morris Rosenberg's Rosenberg's (1965, 1979, 1981) ground-breaking Not only has this work ground-breaking work work on self-esteem. self-esteem. Not work produced the most most useful useful measure measure of of this elusive elusive phenomenon, phenomenon, but it neatly neatly illus illusproduced of directly testable social psychological psychological linkage processes. processes. trates the application of Rosenberg and Pearlin (1978) considered in detail a finding finding that that both both the Kohn Kohn Rosenberg repeatedly uncovered-a imcovered—a strong positive positive relationship relationship and the Inkeles projects had repeatedly among adults between between social social class and self-esteem. self-esteem. But But this relationship relationship is weak weak among among adolescents adolescents and disappears among younger younger children. Why? These These investiga investigaamong finding as an instructive instructive means of of formulating formulating and testing four different different tors used this finding testing four structure-to-personality linkage linkage (see (see also D emo & process possibilities for this structure-to-personality Demo Savin-Williams, 1983). (1) Social comparison comparison process. Adults Adults meet meet a far far wider wider class range of of persons persons perceive class differences differences more more than children. Moreover, Moreover, adult than children, and perceive perception of of class differences differences accounts accounts for a significant significant part of of the class and self selfperception esteem relationship. relationship. (2) Reflected ourselves in part as Reflected appraisal appraisal process. Likewise, since since we see see ourselves others see us, the wider class contacts of of adults may also also shape shape self-esteem through others self-esteem through how others react to us. how (3) Self-perception by Bem Bem (1972), we Self-perception process. In the manner manner advanced advanced by we partly behavior as others observe observe it. Rosenberg form our self-image self-image by observing observing our own behavior Rosenberg form and Perlin (1978) note that the class standing of children is ascribed ascribed while that of of children of adults is more to make make inferences more likely to be achieved. Hence, adults are more more likely likely to inferences concerning their self-worth self-worth from from their social social class standing. concerning Psychological centrality centrality processes. The The more more centrally centrally important social important social (4) Psychological class dimensions self-esteem. Indeed, for dimensions are, the more they should influence influence self-esteem. for adults who who most value money, the income income and self-esteem self-esteem association association is strongest. In addition, children often often either deny or are confused confused about social class position; about their social thus, it is not not likely Ukely to be central for for them.

Another Another facet of of top-down top-down analyses involves involves historical historical events. Social Social change change of course, just as inevitable inevitable and relentless as personal personal change, so dynamic dynamic treat treatis, of ment historical events ment of of social structure structure becomes becomes essential. But But when when historical events are involved, involved, macro-level causes are often often difficult difficult to untangle. Three overlapping possibilities possibilities macro-level Three overlapping arise: generation, life-cycle, and cohort Smelsers (1981) cohort effects. For For example, the Smelsers point out for the out how how all three interpretations interpretations can can account account for the finding finding that a young young cohort changed 8-year period changed its political poUtical attitudes extensively extensively over over an 8-year period toward toward those those of because their youthful of the parental generation generation ( Jennings & Niemi, 1975). Was Was this because youthful rebelliousness because of passage from rebelliousness receded (a generation effect), effect), because of their passage from adoles adolescence because of of their special cence to young young adulthood adulthood (a life-cycle life-cycle effect), and/or and/or because special experi expericohort effect)? ence with a particular historical event event or situation situation (a cohort effect)? Longitudinal Longitudinal ence obviously necessary, but even then the decomposition decomposition can be difficult. difficult. data are obviously

17 CHAPTER 17

SOCIALPSYCHOLOGICAL PSYCHOLOGICALCONTRIBUTIONS CONTRIBUTIONS SOOAL

433 433

effects of the Great Depression Depression presents a model model Elder's (1974) study of the effects for a Paths D D and E, cohort study study using historical events events as the macro causes. From From for longitudinal data on 167 families followed followed from from 1932 1964, Elder looked at 1932 to 1964, effects of unemployment unemployment and economic deprivation on family structure and the effects meso-situational level and personality personaUty at the micro level. The interaction at the meso-situational economic loss that generated severe severe social strains and altered Depression led to economic family life through changes in marital relations, parent-child relations, and the family Traditional attitudes about gender roles were reinforced, for division of labor. Traditional disproportionately more responsibility for example, by the need for girls to take disproportionately home duties and for boys to earn outside income. Taking early jobs for the boys boys home movement toward the adult world." At A t the personality level, meant "an accelerated movement lasting enhanced need for stability and security among among this these changes led to a lasting enhanced need cohort of of children. These These effects effects were were greater for those families families whose whose income income loss cohort Depression was especially especially profound. profound. during the Depression

V. V. PERSONALITY AND SOCIAL STRUcruRE STRUCTURE SHAPE TOGETHER For heuristic purposes, we have discussed the causal paths of multilevel analyses analyses as if only simple main effects effects applied. In reality, of course, social causation isJar is,far multipHcative, and interactive effects effects more complex. Indeed, multilevel multilevel additive, multiplicative, of the most intriguing and theoretically theoretically suggestive findings findings in the provide many of personality and social structure domain. personality Consider Parker and Kleiner's (1964) penetrating study of the mental health of blacks in Philadelphia. They They found mobiUty, either upward or down downof found that social mobility, combined with high levels of personal goal striving was associated with ward, when combined ambivalent or weak weak racial identity. Similarly, Cohn both mental illness and either ambivalent shown that unemployment, unemployment, when combined combined with a strong sense of internal (1978) has shown locus of control, can lead to severe dissatisfaction dissatisfaction with oneself. oneself. Such mutual shaping underscores once once again the fundamental fundamental tenet of social psychology: **. ". . . if men define define situations as real, they are real in their conse conse567). Culture and social structure impinge impinge quences" (Thomas & Thomas, 1928, 1928, p. 567). on individuals through their subjective subjective interpretation. A A major social psychological psychological contribution to this area is its insistence that the linkages between personality and social structure must involve processes that include the individual's perceptions of environment. the social environment.

VI. VI. ADVANTAGES OF THE PERSONALITY AND SOCIAL SOCUL STRUCTURE PERSPECTIVE Ryff Ryff (1987) provides a succinct succinct discussion of five interrelated advantages of the personality and social structure perspective for personality and social psychology. personaUty psychology. First, neither personality nor social psychology psychology has probed deeply deeply into the general-

434 434

THOMAS PETnGREW THOMAS F. F.PETnGREw

ization of assumptions. Is it it really part of of its theories, findings, or assumptions. of the human human condition, peoples everywhere condition, for for example, example, for for peoples everywhere to to operate operate with with aa "just-world "just-world hy hypothesis"? pothesis"? Or Or is is this this just just an an assumption assumption of of Westerners? Westerners? The The personality personality and and social social structure structure perspective perspective provides provides a a theoretical theoretical and and empirical empirical entry entry into into such such questions. questions. Likewise, Likewise, the the influence influence of of normative normative culture culture can can be be approached approached through through this this tradition. What tradition. What do do subjects subjects bring bring to to the the personality personality inventory inventory or or the the laboratory? laboratory? . Ryff Ryff (1987) (1987) points points out, out, for for instance, instance, that that the the fundamental fundamental attribution attribution bias bias may may not not be think in be as as "fundamental" "fundamental" as as we we think in cultures, cultures, such such as as India's, India's, where where individuals individuals are are viewed more more holistically as part of of their social context. context. A A third advantage concerns broad description as opposed to process explana explanation. The The study of of personality personality and social structure has focused focused on description-with description—with such outstanding exceptions as the Rosenberg Rosenberg and Pearlin (1978) work work discussed discussed previously. By B y contrast, both personality personality and social psychology psychology have stressed pro process. Here, argues cess. Here, argues Ryff Ryff (1987) (1987) convincingly, convincingly, the the fields fields can can mutually mutually enrich enrich each each other. example, both other. Attribution Attribution and and self-discrepancy self-discrepancy theories, theories, for for example, both focus focus on on process. process. what is the prevalence prevalence and distribution distribution in the general population population of of particular particular But what attributional styles and discrepancy discrepancy types? types? attributional A A fourth fourth advantage advantage to to be be gained gained by by exploiting exploiting the the personality personality and and social social structure structure perspective perspective has has been b e e n emphasized emphasized throughout throughout this this discussion. discussion. Only Only aa multilevel multilevel approach approach can provide an explicit explicit specification specification of of the social context context for personality. And A n d such a context context is essential for utilizing the fifth fifth advantage. for changing continuously, we we need need Since persons, situations, and societies are all changing broad, multilevel models models to attempt to capture the dynamics of of change. Rather Rather than regard such efforts efforts as virtually hopeless (e.g., Gergen, 1973), the personality personality and social social structure perspective perspective provides a promising approach for coming coming to to grips with this challenge.

REFERENCES T., Frenkel-Brunswik, E., Levinson, D. J., & Sanford, R. N. (1950). The authoritarian authoritarian Adorno, T., personality. New York: Harper. personality. of modern social psychology. In G. Lind LindAllport, G. W. (1968). The historical background of handbook of of social social psychology psychology (Vol. {VoX, 1,1,pp. pp. 1-80). 1-80). New New zey & E. Aronson (Eds.), The handbook York: Random House. Almond, G. A., & Verba, S. (1965). The civic civic culture. culture, Boston: Little, Brown. Aronson, E., Blaney, N., Stephan, C c.,, Sikes, J., & Snapp, M. (1978). The jigsaw jigsaw classroom. classroom, Beverly Hills, CA: Sage. Asch, S. (1956). Studies of of independence independence and conformity: A A minority minority of of one against a unanimous majority. Psychological Psychological monographs, monographs, 70(9, 70(9, Whole No. 416). Atkinson, J. W., & Hoselitz, B. F. (1958). Entrepreneurship and personality. Explorations Explorations 12. in Entrepreneurial Entrepreneurial History, History, 10, 107-1 107-112. Bales, R. F. (1970). Personality Personality and and interpersonal interpersonal behavior. behavior. New York: Holt, Hoh, Rinehart & Winston.

CHAPTER CHAPTER 17 17

SOQAL PSYCHOLOGICAL CONTRIBUTIONS SOCIAL CONfRlBUll0NS

435 435

(1956). Toward Toward aa theory theory of of schizophre schizophreBateson, G., Jackson, D. D., Haley, J., & Weakland, J. (1956). nia. Behavioral Science, Science, 1,1,251-264. 251-264. Bern, D. J. (1972). Self-perception Self-perception theory. In L. Berkowitz (Ed.), Advances in experimental experimental social psychology (Vol. (Vol. 6, 6, pp. pp. 1-62). 1-62). New New York: York: Academic Academic Press. Press. social psychology Blau, P. (1955). The dynamics of of bureaucracy. bureaucracy,Chicago: Chicago:University UniversityofofChicago ChicagoPress. Press. Blumer, H. (1969). Symbolic interaction ism: Perspective Englewood interactionism: Perspectiveand andmethod. method, EnglewoodCliffs, Cliffs,NJ: NJ: Prentice-Hall. Borgatta, E. F., Bales, R. F., & Couch, A. S. (1954). (1954). Some findings relevant to the great man theory of leadership. American Sociological SociologicalReview, Review,19, 19,755-759. 755-759. pattern of of human concerns. concerns. New NewBrunswick, Brunswick, NJ: NJ:Rutgers Rutgers Univer UniverCantril, H. (1965). The pattern sity Press. Cohn, R. R. M. M. (1978). The The effect effect of of employment employment status status change change on on self-attitudes. self-attitudes. Social SocialPsychol PsycholCohn, ogy, 41, 41, 81-93. 81-93. ogy, Demo, D. D. H., & Sayin-Williams, Savin-Williams, R. R. C. C. (1983). (1983). Early Early adolescent adolescent self-esteem self-esteem as as aa function function of Demo, H., & of social class: class: Rosenberg and Pearlin revisited. American journal of of Sociology, Sociology, 88, 763-774. DiRenzo, G. J. (1977). Socialization, personality, and social systems. Annual Review of of DiRenzo, Sociology, 3, 3, 261-95. Sociology, Suicide((J. A.Spaulding Spaulding&&G. G.Simpson, Simpson,Trans.). Trans.).New NewYork: York:Free FreePress. Press. Durkheim, E. (1951). Suicide J. A. BehavioralScience, Science, Elder, G. (1973). On linking social structure and personality. American Behavioral 16, 785-800. 785-800. 16, Children of of the the great greatdepression. depression.Chicago: Chicago: UniversityofofChicago ChicagoPress. Press. University Elder, G. (1974). Children Epstein, S. (1979). The stability of behavior: I. On predicting most of the people much of of of Personality Personality and andSocial SocialPsychology, Psychology,37, 37,1097-1126. 1097-1126. the time. Journal of Luther. New York: Norton. Erikson, E. H. (1958). Young man Luther. GandhVs truth. truth.New New York: York: Norton. Norton. Erikson, E. H. (1969). Gandhi's Civilization and andits itsdiscontents. discontents.New NewYork: York:Macmillan. Macmillan. Freud, S. (1930). Civilization Social psychology psychology as as history. history.Journal Journalof ofPersonality Personalityand andSocial Social Psychology, Gergen, K. (1973). Social Psychology, 26, 309-320. 26, Gluckstern, N. B., & Packard, R. W. (1977). The internal-external change agent team: of Applied Behavioral Behavioral Science, Science, Bringing change to a "closed institution." Journal of 13, 41-52. Gnagey, W. J. (1960). Effects Effects on classmates of a deviant student's power and response to a teacher-exerted control technique. Journal of of Educational Educational Psychology, Psychology,51, 51,1-8. 1-8. Themes in in Japanese Japanese culture. culture. Annals Annals of of the the New New York York Academy Academy of of Sciences, Sciences, Gorer, G. (1943). Themes 5, 5, 106-124. Gurr, T. R. (1970). Why men rebel. rebel, Princeton, Princeton,NJ: NJ:Princeton PrincetonUniversity UniversityPress. Press. Handel, W. (1982). (1982). Ethnomethodology: Ethnomethodology: How people make sense. sense. Englewood Cliffs, NJ: Prentice-Hall. New York: Macmillan. Hartshorne, Hartshome, H., & May, May, M. A. (1928). Studies Studiesininthe thenature natureof ofcharacter. character. New York: Macmillan. M. Rosenberg & R. Turner (Eds.), House, J. S. (1982). Social structure and personality. In M. Social (pp. 525-561). New York: Basic Books. Social psychology: psychology:Sociological Sociologicalperspectives perspectives (pp. 525-561). New York: Basic Books. Inkeles, A. (1960). Industrial man: The relations of status to experience, perception, and of Sociology, Sociology, 66, 66,1-31. 1-31. value. American Journal of of individual Inkeles, A. (1969). Making men modern. On the causes and consequences of of Sociology, Sociology, 75, 208-225. change in six developing countries. American Journal of Inkeles, A. (1978). National differences differences in individual modernity. Comparative Comparative Studies Studies in Sociology, 1, 1, 47-72. Sociology,

436 436

THOMAS� F.PETnGREW PETHGREW THOMAS

Inkeles, A, A., & Levinson, D. (1954). National character: The study of modal personality and social systems. In G. Lindzey (Ed.), Handbook Handbook of of social social psychology (Vol. 2, pp. 975-1020). Cambridge, MA: Addison-Wesley. A, & Levinson, D. (1968). National character: The study of modal personality Inkeles, A., and sociocultural systems. In G. Lindzey & E. Aronson (Eds.), Handbook Handbook of of social psychology (Vol. 4, pp. 418-506). Reading, MA: Addison-Wesley. A, & Smith, D. (1974). Becoming modern: Individual change Inkeles, A., change in six developing countries. countries. Cambridge, Cambridge,MA: MA:Harvard Harvard University UniversityPress. Press. Jackson, J. K. (1956). The adjustment adjustment of the family to alcoholism. Marriage Marriage and and Family Family 18, 361-369. Living, 18, 361 -369. Niemi, R. R. (1975). (1975).Continuity Continuityand andchange changeininpolitical politicalorientations: orientations:AAlongitudi longitudiJennings, M., & Niemi, nal study of two generations. American Political Political Science ScienceReview, Review,60, 60,1316-1335. 1316-1335. Major developments in social psychology during the past five decades. Jones, E. E. (1985). Major Aronson (Eds.), Handbook Handbook of of social psychology psychology (Vol. (Vol. 1,1, pp. pp. In G. Lindzey & E. Aronson 47-108). New York: Random House. Kardiner, A A. (1945). The psychological psychologicalfrontiers frontiersof of society. society.New New York: York: Columbia Columbia Univer University Press. A. C. (1976). The status attainment attainment process: Socialization or allocation? Social Social Kerckhoff, A Forces, 55, 55,368-381. 368-381. Forces, (1969). Class Class and andconformity: conformity:AAstudy studyininvalues. values.Homewood, Homewood,IL:IL:Dorsey. Dorsey. Kohn, M. L. (1969). 1977. In Class Class and andconformity conformity (2nd (2nd ed.). ed.).Chicago: Chicago: Kohn, M. L. (Ed.) (1977). Reassessment, 1977. University of Chicago Press. L., & &Schooler, Schooler, C. C. (1969). (1969). Class, Class,occupation occupation and andorientation. orientation.American AmericanSociological Sociological Kohn, M. L., Review, 34, 659-678. Review, 34, Kohn, M. L., & Schooler, C. (1973). Occupational experience and psychological funtioning: funtioning: SociologicalReview, Review,38, 38,97-118. 97-118. An assessment of reciprocal effects. American Sociological Kovel, J. (1970). White racism: racism:AA psychohistory. psychohistory.New NewYork: York:Pantheon. Pantheon. Le Bon, G. (1897). The crowd. crowd. London: London:Unwin. Unwin. Lewis, O. (1961). The children children of ofSanchez. Sanchez.New New York: York: Random Random House. House. Lieberman, S. (1950). (1950). The effects effects of changes in roles on the attitudes of role occupants. Human Relations, Relations, 9,9,385-403. 385-403. Lindzey, G., & Aronson, E. (Eds.). (1985). Handbook of of social psychology psychology (Vols. (Vols. 11 and and 2). 2). New York: Random House. B. Bottomore (Ed. 2nd Marx, K. (1964). Economic and philosophical manuscripts. In T. B. Trans.), Early writings writings (pp. (pp.63-219). 63-219).New NewYork: York:McGraw-Hill. McGraw-Hill. Studiesininmotivation. motivation.New NewYork: York:Appleton-Century-Crofts. Appleton-Century-Crofts. McClelland, D. C. (Ed.). (1955). Studies McClelland, D. C. (1961). The achieving achievingsociety. society.Princeton, Princeton,NJ: NJ:Van VanNostrand. Nostrand. McGraw, K. M., & & Bloomfield, J. (1987). Social influence on group moral decisions: The effects of moral reasoning and sex role orientation. Journal of of Personality Personality interactive effects and Social Psychology, Psychology,53, 53,1080-1087. 1080-1087. Mcguire, W. J. (1985). Attitudes and attitude change. In G. Lindzey & E. Aronson (Eds.), social psychology psychology (Vol. (Vol. 2, 2, pp. pp. 233-346). 233-346). New New York: York: Random Random House. House. Handbook of of social Handbook Mead, M. (1935). Sex and temperament temperamentin in three threeprimitive primitivesocieties. societies.New New York: Morrow. York: Morrow. Milgram, S. Obediencetotoauthority: authority:An Anexperimental experimental view. New York: Harper & Row. view. New York: Harper & Row. S. (1974). Obedience Personalityand andassessment. assessment.New NewYork: York:Wiley. Wiley. Mischel, W. (1968). Personality Parker, S., & & Kleiner, Kleiner, R. R. J. J. (1964). Mental illness illnessininthe theurban urbanNegro Negrocommunity. community.New NewYork: York: Parker, S., Free Press.

CHAPTER CHAPTER 17 17

SOCIAL PSYCHOLOGICAL PSYCHOLOGICALCONfRlBUTIONS CONTRIBUTIONS SOCIAL

437 437

Pearlin, L. L, I., & Kohn, M. L. (1966). Social class, occupation and parental values. American Sociological SociologicalReview, Review,31,31,466-479. 466-479. Pettigrew, T. F. (1978). Three issues in ethnicity: Boundaries, deprivations, and perceptions. view In J. M. Yinger & & S. J. Cutler (Eds.), Major Major social social issues: issues:AAmulti-disciplinary multi-disciplinary view (pp. 25-49). New York: Free Press. of Pettigrew, T. F. (1980). (1980). Social psychology's potential contributions to an understanding of poverty. In V. T. Covello (Ed.), Poverty evaluation of ofsocial social Poverty and public policy: An evaluation science science research research(pp. (pp.189-233). 189-233).Boston: Boston:Hall. Hall. Pettigrew, T. F. (1988). Influencing policy with social psychology. Journal of of Social Social Issues, Issues, 44(2), 205-219. Robinson, W. S. (1950). (1950). Ecological correlations and the behavior behavior of individuals. American Sociological SociologicalReview, Review,15,15,351-357. 351-357. Rosenberg, M. (1965). (1965). Society Princeton, NJ:NJ: Princeton Univer Societyand andthe theadolescent adolescentself-image. self-image. Princeton, Princeton University Press. Press. sity Rosenberg, M. (1979). (1979). Conceiving Conceivingthe theself. self New NewYork: York: Basic Basic Books. Books. Rosenberg, M. (1981). Self-concept: Social product and social force. In M. Rosenberg & R. Rosenberg, perspectives (pp. 591-624). New York: Turner (Eds.), Social Socialpsychology: psychology:Sociological Sociological perspectives (pp. 591-624). New York: Basic Books. Rosenberg, M., & Pearlin, L. I. (1978). Social class and self-esteem self-esteem among children and AmericanJournal JournalofofSociology, Sociology, 53-77. adults. American 84,84, 53-77. Social psychology: psychology:Sociological Sociologicalperspectives. perspectives. Rosenberg, M., & Turner, R. (Eds.) (1981). Social New York: Basic Books. Ross, L. (1977). The intuitive psychologist and his short-comings. In L. Berkowitz (Ed.), Ross, Advances in experimental experimentalsocial socialpsychology psychology(Vol. (Vol.10, 10,pp. pp.173-220). 173-220).New NewYork: York:Aca AcaAdvances demic Press. demic (1966). Generalized expectancies for internal vs. external control of reinforcement. reinforcement. Rotter, J. (1966). Psychological , Whole PsychologicalMonographs, Monographs,80(1 80(1, WholeNo. No.609). 609). The place place of of personality personality and and social social structure structure research research in social social psychology. psychology. (1987). The Ryff, C. D. (1987). Journalof ofPersonality Personalityand andSocial Social Psychology, 1192-1202. Journal Psychology, 53,53, 1192-1202. Education,Occupation, Occupation,and andEarnings: Earnings: Achievement & Hauser, R. M. (1975). Sewell, W. H., & (1975). Education, Achievement in the Early Career. Career.New NewYork: York:Academic AcademicPress. Press. Bendix, Simmel, G. (1955). Conflict Conflict and the web of of group affiliations affiliations (K. H. Wolff Wolff & R. Bendbc, Trans.). New York: Free Press. Smelser, N. J., & Smelser, W. T. (Eds.). (1963). Personality Personality and andsocial socialsystems. systems.New NewYork: York: Wiley. Smelser, N. J., & Smelser, W. T. (1981). Group movements, sociocultural change, and personality. In M. Rosenberg & R. Turner (Eds.), Social psychology: psychology: Sociological Sociological perspectives perspectives (pp. (pp.625-652). 625-652).New NewYork: York:Basic BasicBooks. Books. Snyder, M., & Ickes, W. (1985). Personality Personality and and social social behavior. behavior. In G. G. Lindzey & & E. E. Aronson Aronson Handbook of of social psychology (Vol. 2, pp. 883-948). New York: Random (Eds.), Handbook House. Sykes, G. (1958). (1958). The society society of of captives. captives. Princeton, Princeton,NJ: NJ:Princeton Princeton University UniversityPress. Press. intergroiip discrimination. Scientific Scientific American, 225(5), 223(5) Tajfel, H. (1970). Experiments in intergroup 96-102. S. (1928). The child in America. New York: Knopf. Thomas, W. I., & Thomas, D. S. DemocracyininAmerica America(Vols. (Vols.l1and and2). 2).New NewYork: York:Vintage. Vintage. Tocqueville, A. de (1945). [1840] Democracy Walker, I., & Pettigrew, T. F. (1984). Relative deprivation theory: An overview and concepcritique. British BritishJournal JournalofofSocial SocialPsychology, Psychology, 301-310. tual critique. 23,23, 301-310. ,

438 438

THOMAS PEmGREW THOMAS F. F.PETHGREW

(T.(T. Parsons, Trans.). New Weber, M. (1930). (1930). The Protestant Protestantethic ethicand andthe thespirit spiritof ofcapitalism capitalism Parsons, Trans.). New York: Scribner. Roth && Weber, M. (1968). (1968). Economy and society: society: An An outline outlineof of interpretive interpretivesociology. sociology,(0.(G. Roth C. Wittich, Trans.). Berkeley: University of California California Press. cross-cultural training and personality: Whiting, J. W. M., & Child, T. L. (1953). Child training personality: AA cross-cultural study. New Haven, CT: CT: Yale University Press. Yarrow, M. M. R., Clausen, J. A., & Robbins, P. R. (1955). The social social meaning of of mental illness. illness. Journal of of Social Social Issues, Issues,11(4), 11(4),33-48. 33-48. The human human choice: choice: Individuation, Individuation,reason, reason,and andorder orderversus versusdeindividua deindividuaZimbardo, P. (1970). The tion, impulse, and chaos. In W. J. Arnold & <&D. D.Levine Levine(Eds.), (Eds.),Nebraska NebraskaSymposium Symposium 17, pp. 237-307). Lincoln: on Motivation (Vol. 17, Lincoln: University of Nebraska Press.

CHAPTER CHAPTER 18 18

CROSS-CULTURAL CROSS·CULTURAL PERSPECTIVES ON PERSONALITY HARRY C. C, TRIANDIS TRIANDIS UNIVERSITY OF ILLINOIS UNIVERSITY OF ILLINOIS AT AT CHAMPAIGN-URBANA CHAMPAIGN-URBANA

The subfield subfield of culture and personality is the oldest of cross-cultural psychology. "magnificent failure," based on the observation that Bruner (1974) assessed it as a "magnificent personality attributes change meaning depending on the cultural antecedents of personality context in which they occur. Klineberg (1980) chastised those who wrote about "national character" without taking seriously Linton's (1945) distinction between universals,which whichapply applytotoall alladult adultmembers membersofofaasociety; society;specialties, specialties, found certain found in in certain universals, alternatives, behaviors behaviors among amongwhich which one one can can roles (e.g., priest, warrior, male); alternatives, variants, found found only onlyinincertain certainrelatively relativelyrare rareindividuals. individuals. choose; and variants, Many writers (e.g., Klineberg, 1954; 1954; Mead, 1953) prescribed how culture and followed these prescrip prescrippersonality studies should be done, but almost no one has followed for instance, suggested that it tions, because they are too demanding. Klineberg, for was necessary to first study the general pattern of a culture and determine the distribution of key charactersitics of the population ap population based on a multimethod multimethod approach (experiments, biographies, opinion surveys, tests, etc). We may have an approximation of this approach if we synthesize studies done in Japan (Hayashi, 1992; Stoetzel, 1955), but there are no other other examples examples of of this this very very expensive expensive strategy. strategy. 1992; Jahoda (1980) criticized many attempts in this area (e.g., (e.g.. Berry, 1976; 1976; Cole, 1975; LeVine, 1973) because they did not provide clear definitions and operationali operationaU1975; zations of culture. He saw little chance to develop this field without a proper definition of culture. He urged the development of a theory of culture that is tailor tailordefinition made for psychologically oriented cross-cultural work. Triandis (1994) has provided such a theory, but it has yet to be evaluated in the literature. HANDBOOK PERSONALfll' PSYCHOLOGY PSYCHOLOGY HANDBOOK OF OF PERSONALITY

439

439

CoPYII GIIT C COPYRIGHT © 1997 1997 IY BY ACADEMIC ACADEMIC PlESS PRESS.. ALL lIGHlS OF IERODUcnON IN ANY FOaM JESElVED.

ALL RIGHTS OF REraooucnoN w

Km FORM RESERVED.

440 440

HARRY HARRY C. C. TRiANDIS TRIANDIS

Inkeles and Levinson (1969) examined examined national national character character by using the con construct struct of of aa "modal "modal personality." personality." The The operationalization operationalization of of this this construct construct is is also also very very demanding, since determination of demanding, since any any society society is is likely likely to to be be multimodal multimodal and and the the determination of the the distributions distributions of of traits traits in in aa culture culture is is likely likely to to be be very very expensive. expensive. Tapp Tapp (1981) (1981) has has provided provided a a review review of of the the major major reviews reviews of of the the field, field, such such as as the 1969), Hsu the work work of of Kluckhohn Kluckhohn and and Murray Murray (1948), (1948), Inkeles Inkeles and and Levinson Levinson (1954, (1954,1969), Hsu (1972), (1972), Kaplan Kaplan (1961), (1961), Child Child (1968), (1968), and and DeVos DeVos and and Hippler Hippler (1969), (1969), but but found found convergence in methodology or findings. findings. little convergence in methodology Draguns Draguns (1979, (1979, 1990) 1990) has has suggested suggested that that the the strategy strategy should should consist consist of of an an examination of dimensions of personality across cultures. This is the strategy examination of dimensions of personality across cultures. This is the strategy that that will be described described in will be in this this chapter. chapter. The (a) dimensions The chapter chapter will will discuss discuss (a) dimensions of of cultural cultural variation, variation, (b) (b) dimensions dimensions of social behavior, and (c) dimensions of personality variation, and of social behavior, and (c) dimensions of personality variation, and suggest suggest how how the the three three sets sets of of dimensions dimensions can can be be placed placed into into an an integrated integrated theoretical theoretical framework. framework. But But first, first, some some preliminary preliminary major major differences differences of of opinion opinion among among specialists specialists in in this this field must be considered, and some methodological issues must be mentioned. field must be considered, and some methodological issues must be mentioned.

I. PRELIMINARY CONSIDERATIONS CONSIDERATIONS A. Do Personality Dimensions Maintain More or Less the tiie Same Meaning across Cultures or Do They Change Their Meaning in in Every Cultural Context? This opinion that ignored. Anthropologists This is is an an important important difference difference of of opinion that cannot cannot be be ignored. Anthropologists (e.g., (e.g., Shweder Shweder & LeVine, LeVine, 1984; 1984; Stigler, Stigler, Shweder, Shweder, & Herdt, Herdt, 1990), 1990), with with their their interest interest way "the in in seeing seeing the the world world the the way "the natives" natives" see see it it (Malinowski, (Malinowski, 1944), 1944), take take the the view view that that the the construct construct of of personality personality changes changes meaning meaning in in each each cultural cultural context. context. Behavior Behavior is is aa consequence consequence of of the the specific specific meaning meaning that that the the situation situation has has for for each each individual, individual, this this meaning meaning is is culturally culturally determined, determined, and and thus thus aa "cultural "cultural psychology" psychology" (Shweder (Shweder & Sullivan, Sullivan, 1990) 1990) requires requires the the examination examination of of the the "semiotic "semiotic subject" subject" that that does does not not have have traits, traits, mental mental states, states, or or psychological psychological processes processes that that are are independent independent of of culture. culture. This This is is the the "relativist" "relativist" position position (Berry, (Berry, Poortinga, Poortinga, Segall, Segall, & & Dasen, Dasen, 1992). 1992). It It contrasts contrasts with with the the "absolutist" "absolutist" position position that that assumes assumes that that traits traits and and psychological psychological processes are processes are universal universal and and unaffected unaffected by by culture. culture. An An example example of of that that position position is is the the work work of of Eysenck Eysenck and and Eysenck Eysenck (1983), (1983), who who measure measure traits traits across across the the world world with with relatively little relatively little concern concern about about the the equivalence equivalence of of what what they they measure. measure. Critics Critics (Bijnen, (Bijnen, Van Van der der Net, Net, & Poortinga, Poortinga, 1986) 1986) have have argued argued that that this this work work is is indefensible. indefensible. Cross-cultural Cross-cultural psychologists psychologists (e.g., (e.g., Berry Berry et et aaI., l , 1992; 1992; Pepitone Pepitone & & Triandis, Triandis, 1987) 1987) take take an an intermediate intermediate position, position, rejecting rejecting both both relativism relativism and and absolutism. absolutism. First, First, they distinguish distinguish emie they emic and and etic etic elements elements of of any any construct. construct. The The emic emic are are culture culture specific; example, the specific; the the etic etic are are universal. universal. For For example, the trait trait "self-reliance" "self-reliance" has has some some common cultures. What these meanings common meaning meaning across across cultures. What is is common common in in these meanings across across cultures cultures

CHAPTER CHAPTER 18 18

CROss-CULTIlRAL CROSS-CULTURAL PERSPEcnvES PERSPECTIVES

441 441

is etic. But self-reliance has somewhat different different goals in collectivist coUectivist cultures (e.g., traditional) than in individualistic cultures (e.g., the West). In coUectivist collectivist cultures often takes the form form "I am self-reliant self-reliant so I will not burden my group." In it often individualist cultures it takes the form "I am self-reliant self-reliant so I can have fun fun and do my own thing." Clearly, the meaning is different different across cultures; thus our task is to measure both the etic and the emic aspects of of the construct. Culture-free Culture-free measurement is very difficult, difficult, but cross-cultural psychologists have developed methodological strategies that they believe can deal with the influ influence of culture. Specifically, they emphasize separate construct construct validations in each culture, emic measurement of etic constructs, multimethod measurement, and the control of rival hypotheses (see Triandis, 1992, 1994, for 1992,1994, for details). If the building blocks of personality are trait descriptions, we need to recognize recognize that such descriptions depend on categorization, that is, treating discriminably different entities as the same. While all cultures use categories (Triandis, 1964), different grouped and considered "the same" for for the purpose of categori categorithe entities that are grouped zation are not the same. For example, "yellow" does not include the identical chips from the Munzel charts in every culture. We need methods that will take that fact from for some time (e.g., Triandis, into account. Such methods have been available for Davis, & Takezawa, 1965), but are time-consuming, and hence unpopular. B. Is Is the the Variance Variance of of Traits Traits Similar Similar across across Cultures? Cultures? B. If we measure an important trait in culture A that has essentially no variance in difficult to compare the two cultures, beyond beyond saying culture B, it would be very difficult that culture B does not seem to have the trait. Ethnocentrism is an important aspect of the human condition (Triandis, 1994), because we all start life by knowing only we learn about other cultures we we tend to incorporate incorporate our own culture, and even when we information into the framework framework that our own culture has created. Other the new information cultures "make sense" only in relation to our own. Thus, in this situation we are likely to use the following logic: "The trait is important in my culture. I do not detect it in that other culture. It must be that my measurements are inadequate. measured in culture B." But is it Therefore, it is a universal trait, that is poorly measured really a universal trait? This issue becomes salient in discussions of cultures where behavior follows a script so that one cannot see many individual differences. Geertz (1983), for for instance, paints such a picture of social behavior in Bali. While this position is follow a script still show some probably extreme, since actors in our theaters who follow variations in behavior, and even insects are genetically extremely variable (Jeanne, 1988) it is a view that cannot be ignored. A related issue is that the within-culture (Mintum & Lam Lamvariance may be much larger than the between-cultures variance (Minturn bert, 1964), in which case the association of culture with personality becomes extremely hazardous.

442 442

HARRY C. TRIANDIS

C. C. Are Are the the Correlations Correlations among among Traits Traits the tiie Same Same or or DitTerent Different across Cultures? Cultures? across This issue links with the viewpoint that we we ought to develop "indigenous psycholo psychologies" (Heelas & Lock, 1981; 1981; Kim & Berry, 1993). One can visualize a table in which each psychological phenomenon has a row, and each culture a column (Kim & Berry, 1993, 1993, p. 279). As we synthesize the information information on each row, the common elements are the etics of the phenomenon. What is different different in each cell of that row from from the etics of the phenomenon provides the emics of the phenomenon. As we synthesize each column we identify identify the indigenous theory of personality that summarizes how the various phenomena occur in one culture. For example, particu particular correlations among variables may have unique values in one culture. Finally, as we integrate across both rows and columns we arrive at a universal psychology, of which the indigenous psychologies are special cases. The important thing to psychologies—the remember here is that current psychology is one of the indigenous psychologies-the one from from the West.

D. D. Rival Rival Hypotheses Hypotheses It is easy to assume that a difference difference that has been observed between two or more fact due to "culture." Before making such an assumption one must cultures is in fact measurecheck rival hypotheses. A mundane one is that people have reacted to the measure differently. This is why multimethod measurement is so important. Only ifif ments differently. different kinds of measurements detect the same difference, and we have controlled different 1989; Triandis, 1972), can we be reasonably sure for response sets (Hui & Triandis, 1989; for difference. that we have detected a difference. It is also possible that the difference difference is due to some demographic attribute attribute that happens to be correlated correlated with culture, such as differences differences in social class, age, sex, sex, religion, occupation, education, and so so forth. Social class is especially important. important. Researchers have often often assumed that differences differences in prejudice were due to race or that differences differences in child-rearing patterns were due to nationality, when in in fact they differences due to social class (W. (W. E. E. Lambert, 1992; 1992; W. W. E. Lambert, Lambert, Hamers, Hamers, & & were differences 1979; Triandis & Triandis, 1960, 1960,1962). after checks have been Frasure-Smith, 1979; 1962). Only after differences can we made that Linton's "specialties" do not explain the obtained differences difference. be relatively sure that we have observed a cultural difference. turn now to an examination of different different kinds of cultural, behavioral, and We tum personality variations. n. DIMENSIONS OF CULTURAL VARIATION n

A psychological, metaphorical definition definition of "culture" is that it is to society what memory is to individuals (Kluckhohn, 1954). It includes what "has worked" in the identified by examining the extent to which psychological processes, past and can be identified

CHAPTER 18 18 CROss-CULTURAL CROSSCULTURALPERSPECTIVES CHAFfER PERSPECI1VES

443 443

such as beliefs, attitudes, and values, are shared and transmitted from one generation generation to the next. Historical analysis is also likely to show that the shared patterns were adaptive at some point in the history of the society. This definition, however, is too abstract. One way to make it more useful useful is to consider "cultural syndromes"-shared syndromes"—shared sets of beliefs, attitudes, norms, values, and behavior organized around a central theme and found found among speakers of of one language, in one language, in one time time period, period, and and in in one one geographic geographic region. region. In In specific specific geographicllanguage geographic/language regions regions one one can can observe observe shared shared human-made human-made attributes, attributes, such such as as tools, tools, laws, laws, norms, norms, values, values, perspectives, perspectives, information information processing processing strategies, strategies, and and other other elements elements of of subjective subjective culture. culture. Subjective Subjective culture culture (Triandis, (Triandis, 1972) is a cultural group's characteristic way of perceiving the human-made part of of the environment. It consists of of the categories used to cut the pie of of experience, the associations among these categories, the norms of social behavior, the roles, and the values of of the cultural group. Language, of of course, is intimately linked to subjective culture. Each of of the elements of of subjective culture can be measured and tested to determine if the within-culture variance is smaller than the between-cultures vari varidetermine useful to use that element in the description description of of the specific ance, in which case it is useful widely dispersed cultures to culture. Such studies, ideally, should use half a dozen wid�ly estimate the between cultures variance. The elements of subjective culture are not randomly distributed. On the importance to the contrary, they are organized around central themes of great importance of elements of of subjective culture that we call culture. It is these organizations of "cultural syndromes." In any culture there are diverse, often contradictory elements of of subjective culture and cultural syndromes that can be sampled by any individual. Sampling themcan reflect both genetic predispositions and experiential factors. People place them consisselves in situations where they can obtain the most rewards (e.g., resources), consis tent with their biologically determined needs and learned situation-behavior se sequences. Behaviors with rewarding consequences increase in frequency. frequency. Culture then can be conceptualized as providing a "tool kit" of of habits, skills, styles, perspectives, norms, roles, and values out of which each individual can construct construct a potentially unique strategy for action. This strategy will depend on the way the situation is perceived by the individual. More More generally, generally, culture culture can can be be conceived conceived in in numerous numerous ways: ways: (a) (a) as as aa set set of of schedules of reinforcement reinforcement (Skinner, 1981), (b) as knowing the "rules of the game" (Goodenough, 1981), (c) as a system of of symbols through which people experience and express meaning (Keesing, 1981), and (d) as a strategy for adapting to an environment. People can potentially construct an infinitely large set of idiosyncratic patterns of of action. However, since people share features of the ecology in which of these strategies of action are similar. For example, one can they live, some of identify cultural scripts, that is, patterns of of social interaction that are characteristic identify of a particular cultural group (Triandis, Marin, Lisansky, & Betancourt, 1984). of

444 444

HARRY HARRY C. C. TRIANDIS TRIANDIS

Thus the way to proceed in this area is to identify particular features of of the ecology that that result in particular cultural cultural scripts scripts and and syndromes, and show how how these features features are are likely likely to to increase increase the the frequency frequency of of specific specific behaviors. behaviors. The The implications implications of of this this perspective perspective are are clear: clear: There There is is no no one-to-one one-to-one correspondence correspondence between between culture culture and where the and personality. personality. There There are are only only probabilistic probabilistic links, links, where the cultural cultural syndrome syndrome increases sampled, and increases the the probability probability that that it it will will be be sampled, and behaviors behaviors consistent consistent with with it it might population of of aa society. might be be observed observed among among significant significant segments segments of of the the population society. We We will will discuss discuss three three interrelated interrelated cultural cultural syndromes: syndromes: complexity, complexity, tightness, and individualism-collectivism. the modem, and individualism-collectivism. Complexity Complexity contrasts contrasts the modern, industrial, industrial, afflu affluent cultures with the simpler cultures, such as the hunters and gatherers, or the residents of a monastery. Tightness (Pelto, 1968) contrasts cultures where norms are imposed very tightly, allowing very little deviation from "proper" behavior, with cultures that considerable deviation from norms. Triandis (1994) has suggested that tight allow considerable of the culture cultures are more homogeneous (in order to impose a norm, members of must agree on what is proper behavior) and are likely to have a high population lowdensity (it is more functional to have norms when there is high density; in low frontier, the desert, or the Arctic, loose density environments such as in the open frontier, imposition of norms is widely accepted). It is also likely that people who have interdependent jobs will insist that others behave as expected, and thus be "tight." interdependent Individualism is a cultural syndrome whose elements are organized around Individualism environments that are complex and autonomous individuals, and is maximal in environments Collectivism is a syndrome in which the elements are organized around one loose. Collectivism environments that are simple and tight. or more collectives and is maximal in environments coUectivist cultures people define themselves as members of of collectives; In collectivist conflict between the goals of of the collective collective and the individual the when there is a conflict goals of of the collective are considered "obviously" as the ones that have priority. In individualism, on the other hand, the goals of of the individual have priority in such 1994). such cases cases (Triandis, (Triandis, 1990, 1990,1994). Research aI., 1986; Research (Hofstede, (Hofstede, 1980; 1980; Triandis, Triandis, 1990; 1990; Triandis Triandis et et al, 1986; Triandis, Triandis, Bon Bontempo, Villareal, Asai, & Lucca, 1988; Triandis, McCusker, & Hui, tempo, Villareal, Asai, Lucca, 1988; Triandis, McCusker, Hui, 1990) 1990) suggests suggests and in East Asia, that that collectivism collectivism is is especially especially high high in in traditional traditional societies societies and in East Asia, and and is is high America, and and the high in in most most of of the the cultures cultures of of Africa, Africa, Asia, Asia, Latin Latin America, the Pacific Paciflc Islands. Islands. Individualism is especially high in North America north of of the Rio Grande, in in North and West Europe, and Australia-New Zealand. Moderate levels of of individualism individualism can collectivism can be found found in in parts be found in the remaining parts of Europe, but collectivism of Eastern Europe and in Souther Italy and rural Greece. In addition to ecological ecologicalof collectivismenvironmental factors, a number of other factors are likely to affect affect collectivism individualism: gender, age, education, affluence, and so forth. These syndromes are related to personality. Cultural complexity corresponds conscientiousness and to cognitive complexity; cultural tightness corresponds to conscientiousness intolerance for ambiguity, as well as to the use of of narrow categories in in thinking about events; individualism corresponds to idiocentrism.

18 CHAPTER 18

CROSS-CULTURAL PERSPEcnVES PERSPECTIVES CROss-CULTURAL

445 445

cultural and cognitive complexity complexity correspondence. Cognitive Consider the cultural complexity can be measured in a number of (Witkin & Berry, 1975). of different different ways (Witkin Specifically, field-independent who are field-independent people, placed on a tilted chair, who are asked asked to adjust lines so that they are parallel to the the sides sides of of the the room, are not not influenced influenced by the tilted chair as much as people people who are field-dependent; field-dependent; that is, their judgments judgments are relatively independent independent of of the frame (the tilted tilted chair). Thus they they are more more field-independent people can pick figure in a complex complex complex. Similarly, field-independent pick out a figure field-dependent have trouble pattern quickly and reliably, while people who are field-dependent doing this task ((Witkin Field-independent people people are good good when when doing Witkin & Berry, 1975). Field-independent forth. making spatial judgments, so they can be good mechanics, architects, and so forth. Field-dependent people are interpersonally interpersonally sensitive, so they they are good good at occupa occupaField-dependent tions requiring good human relationships. likely to emerge emerge in agricultural The field-dependent field-dependent cognitive style is more more likely cooperate and to obey societies (Berry, 1976), where people have to learn to cooperate authorities that organize the work (e.g., digging ditches for for irrigation). For that field-dependent style characterizes characterizes reason child rearing emphasizes obedience. The field-dependent of attention to the context context of of conversations, that is, the frame, people who pay a lot of for that reason they can intuitively figure figure out what what other people are feeling. feeling. and for other people Similarly Kohn (1969, 1987) found that the upper classes in modern modem societies (1969,1987) obedivalue self-reliance, creativity, and independence, while the lower lower classes value obedi ence, reliability, and conformity. It is clear that it is functional functional for for a person to conforming in lower-class lower-class occupations. be creative in upper-class occupations and conforming functional to be obedient obedient when a society society needs perform many Similarly, it is functional needs to perform functional to be creative when working alone or doing coordinated tasks, but it is functional when working of the kind rewarded in an information information society. intellectual tasks of be sure, it is likely that In sum, while there is too little research on this point to be corresponding to cultural complexity complexity there is a personality personality dimension dimension of of cognitive corresponding Cognitively complex individuals tend to use many complexity. Cognitively to use many dimensions dimensions for for the of events, they make finer finer discriminations discriminations along these judging of these dimensions, and they meaningful conceptual conceptual wholes wholes (Harvey, Hunt, & integrate the dimensions into meaningful be much much variation variation on Schroeder, 1961). However, within cultures there is likely to be personality type, so we do not expect particular cultures cultures as a whole to be be high this personality or low on this dimension. of no reliable research linking cultural tightness to conscientiousness, conscientiousness, I know of though observations of of tight cultures suggest that there may be be such a link. Corre Correlow tolerance tolerance for for ambiguity ambiguity (Budner, 1962; Draguns, sponding to tightness is also low 1990). High tolerance for ambiguity ambiguity corresponds corresponds to looseness. A related personality personality variable is narrow narrow versus wide categorization. Detweiler Detweiler A found that narrow narrow categorizers (Pettigrew, (Pettigrew, 1959) adjust adjust to other cultures (1980) has found difficulty than do wide categorizers. A A test item such as that used in with greater difficulty measure this attribute. The more check check marks marks on the 24 figures figures of of Figure 1 can measure the individual. People from tight this item the wider is the category width used by the People from difficulty adjusting to other cultures have more difficulty other cultures, cultures, since since they they insist insist that that the behavior of of other other people be just as they they expect expect it. High tolerance tolerance for for ambiguambigubehavior people should should be just as it. High

446 446

HARRY C. TRlANDlS TRIANDIS

Suppose Suppose this figure is is a ZUPF.

Place a check check in in all all the the figures figures below below that that you you consider consider to to be be ZUPFS. ZUPFS. Place a

o cD C5 c=0 G O J> D G G q Q G g "^GS OB Q (5 .g 8 � \S 0 £) � GJ C] FIGURE 11 Example of the kind of item used in in a test of category width. FIGURE

ity is a great advantage in cross-cultural adaptation, and the avoidance of culture shock. Corresponding to individualism is the personality pattern called idiocentrism. Corresponding to to collectivism is the the personality pattern called allocentrism allocentrism (Triandis (Triandis et al., ai., 1985). Idiocentrics see themselves as independent of groups. Allocentrics AUocentrics see themselves as parts of groups. When personal goals and group goals are in conflict, idiocentrics are likely to direct their energies toward personal personal goals, goals, and and allocentrics allocentrics toward group goals. Allocentrics feel interdependent with others (Markus & KitayKitay ama, 1991), emphasize relationships, and behave according to their duty. Idiocentrics ama, and behave according to their duty. Idiocentrics feel independent of others, emphasize personal attributes, and behave so as to maximize pleasure. Numerous hypotheses on these patterns have been proposed al., 1986; Triandis, (Triandis, 1990,1994) 1990, 1994) and some have been tested (Triandis et ai., 1986; Triandis, 1988; Triandis, McCusker, & & Hui, 1990). Bontempo et al, ai., 1988; Attitudes, norms, and values among allocentrics reflect interdependence, inin cluding ingroup security, obedience, duty, ingroup harmony, concern for virtuous and persistence persistence to to achieve achieve ingroup ingroup goals goals (Triandis, action as defined by the ingroup, and & Hui, 1990). Attitudes, norms, and values among idiocentrics Bontempo, Leung, & pleasure, achievement, achievement. reflect independence, emotional detachment from ingroups, pleasure,

CHAPTER 18 18 CHAPTER

447 447

CROSS-CULTURAL PERSPEcnVES PERSPECTIVES CROSS.CULTURAL

competition, freedom to do own thing, autonomy, fair exchange, concern for '*the "the truth," and action consistent with principles. The greatest calamity for allocentrics is to be excluded or rejected from the group; the greatest calamity for idiocentrics is to become dependent on the group. Allocentrics have few ingroups, but very close relationships with them. IdioIdio centrics have many ingroups and loose, emotionally noninvolving relationships relationships with them. Ingroups are perceived as more homogeneous than outgroups by allocentrics and more heterogeneous than outgroups by idiocentrics. Conflict Conflict with outgroups is expected by allocentrics, and accepted but not desired by idiocentrics. Allocentrics behavprefer social behaviors that are submerged in large groups (where individual behav ior is unnoticed), while idiocentrics prefer social behaviors on a one-on-one basis (the cocktail party is a Western invention). For example, Korean skiers are more likely to ski in groups, and American skiers to ski alone (Brandt, 1974). motherThe most important relationships for allocentrics are vertical (e.g., mother relationson), and for idiocentrics horizontal (e.g., spouse-spouse). Even horizontal relation ships are "converted" into vertical relationships in collectivist cultures (e.g., in respect from India, the friend who is one day older is supposed to receive more respect the "younger" friend than the other way around), and vertical relationships are converted into horizontal relationships in individualistic cultures (e.g., the student converted calling the the professor professor by by a a first name). Allocentrics Allocentrics are are comfortable comfortable with with asymmetric asymmetric calling first name). status relationships, while idiocentrics idiocentrics are are uncomfortable. uncomfortable. status relationships, while of this syndrome for cognition, emo emoThe evidence concerning the relevance of tion, and motivation was very well reviewed by Markus and Kitayama (1991). The behavior in situations where collectivists coUectivists relevance of this syndrome for social behavior interact with individualists has been presented by Triandis, Brislin, and Hui (1988). differing along the two impor imporChild-rearing patterns can be conceptualized as differing tant dimensions (Adampoulos & Bontempo, 1984) shown in Figure 2: Acceptance versus Rejection and Dependence versus Independence. High and low values on each dimension result in four kinds of of child-rearing patterns. Collectivists CoUectivists conformity to ingroup norms, so they generally are especially concerned with conformity Acceptance

Conformity Conformity

Dependence

Self-confidence Self-confidence

+^ • ---1--+ Independence Independence

Concreteness

Information vigilance

Rejection

FIGURE 22 Socialization Socialization patterns patterns and and behavioral behavioral outcomes outcomes (Adamopoulos (Adamopoulos & & Bontempo, Bontempo, 1984). 1984). FIGURE

448 448

C TRIANDIS HARRY C. TRiANOIS

use acceptance and foster dependence. Individualists are especially concerned concerned with self-reliance and self-confidence, self-confidence, so they generally use acceptance and foster foster independence. Collectivism Collectivism and and individualism individualism may may be be distinguished distinguished into into vertical vertical and and horizon horizontal types, thus obtaining a 2 by 2 typology. This typology can be linked to Fiske's (1990,1992) four etic patterns of social behavior that acquire an emic (1990, 1992) definition of four character in each culture. Fiske's four patterns are: 1. Communal sharing (CS), where if one belongs to a group one is entitled to a share of most resources of of that group. It is the pattern of behavior most frequently found in families around the world. 2. Authority ranking (AR), where status determines how resources are to be distributed. Those with most status get a disproportional share of the group's resources. Inequality is expected. Rank has its privileges. 3. Equality matching (EM), where each person is entitled to an equal share of of whatever is to be distributed. If one person goes through the door first, then the other must go through the door first next time, and so on. 4. Market pricing (MP), where each person obtains resources in proportion to own contributions. The more you give the more you get. If you invest a lot your interest is high.

Vertical collectivism (CV) corresponds to CS and AR, and horizontal collectiv collectivism (He) (HC) to CS and EM. Vertical individualism (VI) corresponds to MP and AR, and horizontal individualism (HI) to MP and EM. Triandis (1995) argues that all four types can be found in all four types can be found in all cultures, cultures, but but will will have have distinct distinct distributions. distributions. For For example, one 10% HC, example, one culture culture may may be be 80% 80% VC, VC, 10% HC, 3% 3% VI, VI, and and 7% 7% HI, HI, while while another another might be 5% 5% VC, 15% HC, HC, 45% 45% VI, 35% HI. might be VC, 15% VI, and and 35% HI. If If we we take take the the modal modal pattern pattern into into account, account, we we may may call call aa culture culture by by that that pattern, pattern, but but this this will will be be an an obvious obvious oversimplication. case, he oversimplication. In In any any case, he has has reviewed reviewed ethnographic ethnographic and and other other work work suggest suggesting ing that that vertical vertical collectivism collectivism is is widely widely found found in in India India and and many many traditional traditional societies. societies. Horizontal Horizontal collectivism collectivism is is frequently frequently found found in in the the Israeli Israeli kibbutz. kibbutz. Vertical Vertical individual individualism is widely found in the United States, France, and many of the Central ism is widely found in the United States, France, and many of the Central European European countries. countries. Horizontal Horizontal individualism individualism is is modal modal in in Scandinavia Scandinavia and and Australia. Australia. There There is is much much within-culture within-culture variation variation on on idiocentrism-allocentrism. idiocentrism-allocentrism. There There are idiocentrics in collectivist coUectivist cultures (e.g., many graduate students who return from the United States to their collectivist coUectivist cultures have trouble adjusting to their home cultures because they have become idiocentrics in the United States). Simi Similarly, there are allocentrics in individualist cultures (they may join communes, gangs, social movements, religious sects). Triandis et aI. al. (1985) found that U.S. allocentrics report receiving more social feel less loneliness than U.S. idiocentrics. This relationship was repli replisupport and feel cated in Puerto Rico (Triandis, Bontempo et al., aI., 1988), suggesting that there may allocentrics to receive more social support than idiocen idiocenbe a general tendency for allocentries trics. tries.

18 CHAPTER 18

CROSS.CULTURAL PERSPEcnVES PERSPECTIVES CROSS-CULTURAL

449

. I

A. Are Are There There Additional Additional Cultural Cultural Syndromes? Syndromes? A. of as yet, and There are probably more cultural syndromes that we do not know of a lot more research is needed in this area. Hofstede (1991) describes what seems to be a cultural syndrome characteristic of the East (Orient) as opposed opposed to the West (Europe). Using data about Chinese values (Chinese Culture Connection, 1987), he argues that there is a unique Chinese viewpoint that emphasizes the importance of virtue, virtue, in incontrast contrastto toaaWestern Westernviewpoint viewpointthat thatemphasizes emphasizesthe theimpor imporThis can be seen most clearly in the philosophic positions of Confucius tance of truth. truth. This can be seen most clearly in the philosophic positions of Confucius and the Eastern religions of Hinduism, Buddhism, Shintoism, and Taoism versus the views of the three monotheistic religions-Judaism, religions—Judaism, Christianity, and Islam. The Eastern ideologies emphasize the importance of virtuous action and the need for carrying out the prescribed rituals. The monotheistic religions emphasize the importance of "belief." "beUef." For the East what you do is more important than what you believe; for the West what you believe is more important than what you do. I am not sure that this is an additional syndrome. It is based on empirical measurements, but such measurements often often produce "echoes" in factor factor analyses. It is is as if you are looking at an elephant from from different different directions, and each direction produces something that looks different different when in fact it is the same elephant. As research is done, we may be able to sort out whether virtue-truth is really more research different from from collectivism-individualism. In the mean time, the contrast between different for viewing the differences virtue and truth is intriguing and useful for differences between East and West. B. B. The Self and Cultural Syndromes The self consists of all statements made by a person, overtly or covertly, that include the words "I," "me," "mine," and "myself." That is a large universe of statements. People in any culture sample such statements in ways that maximize their well wellbeing. How the self self is sampled can influence the social behaviors that are most probable. For example, sampling the "I am a member of my group" statement will increase the probability probabiUty of paying attention to group norms. It is useful useful to distinguish three kinds of "self' "self statements: private (e.g., "I am kind"), public (e.g., "Most people think I am kind"), and collective collective (e.g., (e.g., "My "My co coworkers think I am kind"). We can identify identify separate domains, within our memory structure, that contain the private, public, and collective elements (Trafimow, Trian Triandis, dis, & Goto, 1991). People People in in different different kinds kinds of of cultures cultures have have more more or or less less differenti differentiand collective collective selves. selves.When Whenone oneof ofthese theseselves selvesisis highly highlydifferen differenated private, public, and tiated, that is, has many elements, it is more likely to be sampled by people in that culture. Thus, Triandis (1989) has argued that people in complex cultures tend to sample more more frequently frequently the the public public and and the the private private self self and and less less frequently frequently the the sample self. The The more more individualistic individualistic the the culture culture the the more more frequent frequent is is the the sampling sampHng collective self. collective self. CollectivCoUectivof the the private private self self and and the the less less frequent frequent the the sampling sampling of of the the collective collective self. of

450 450

HARRY C. HARRY C. TRIANDIS TRIANDIS

ism, external threat, threat, competition with outgroups, and common fate increase the the sampling sampling of sampHng of of the the collective collective self. self. Tightness Tightness also also results results in in sampling of the the collective collective self. self. Cheek (1989) and his associates have developed a scale that measures personal, five-point scale (from social, and collective identity. Subjects are asked to rate on a five-point "Not important to my sense of who I am" am" to "Extremely important to my sense **Not reflected by emphasis on am") a set of statements. Personal identity is reflected of who I am") importance of "My dreams and imagination," "My thoughts and ideas," the importance ideas," and so forth. Social identity requires emphasis on items such as "The things I own, my possessions" and "My reflected "My sex, sex, being male or female." Collective identity is reflected "My race or ethnic background," and so forth. by endorsement of "My religion," "My Very good alphas have been obtained for personal and social identity and further work on collective identity selfidentity is in progress. progress. Susceptibility to shame and public public self work consciousness were found found related related to identity, and and susceptibility susceptibility to to guilt guilt and consciousness were to social social identity, and private self-consciousness were related related to to personal personal identity. identity. High High personal identity private self-consciousness were personal identity seems to to buffer buffer subjects subjects against against threats threats to to social social esteem. esteem. seems The sampling of the collective self, which is more probable in simple and tight cultures, has important implications. Such sampling increases the salience of the collective elements of subjective culture, such as norms, roles, and values. The patterns of social behavior associated with collectivism; for example, the great differentiation of ingroups and outgroups, can be seen to be the consequences of differentiation of such sampling. A similar argument made about the other cultural syndromes; for example, in complex and loose cultures there will be greater sampling of the private affect toward the behavior self, resulting in more sampling by individuals of their affect (is this going to be fun?) and their perceptions of the utility of the behavior for them (what will I get out of this?).

m. VIOR m . DIMENSIONS OF SOCIAL DEnA BEHAVIOR Triandis (1978) has argued that every social behavior can be described as a point in a four-dimensional four-dimensional mathematical space defined defined by the dimensions association association (positive, supportive behavior) versus dissociation dissociation (avoidance or aggression, nega negasuperordination (domination, taking the initiative) initiative) versus subordina subordinative behavior); superordination oihevY,formality formality (following (following aascript script tion (conformity, reacting to initiative taken by other); intimacy (spontaneous, disclosing provided by the norms of the culture) versus intimacy overt (visible muscle movements) versus covert covert (fantasies, nonobserv nonobservbehavior); and overt able behavior). Lonner's (1980) review review of cross-cultural universals strongly supported the universality of the first two dimensions. These also correspond to dimensions disdis cussed by Leary, Hogan, and others. The third and fourth fourth dimensions have not yet been established as universals, though they are plausible etics. If the Geertz (1973) description of Bali is correct, most behavior in in that culture is formal; hence there would not be a "dimension" of intimacy-formality intimacy-formality in that culture, but it surely can be seen in other cultures. of

CHAPTER 18 CHAPTER

CROSS-CULTURAL PERSPECDVES PERSPECTIVES CROSS-CULTURAL

451 451

Furthermore, I find it difficult difficult to believe that Balinese behavior is equally formal formal after careful careful study, this dimension will be in all situations. Thus, I suspect that, after found in all cultures. found understandThe lack of studies of overt versus covert (fantasies) behavior is understand able, since covert behavior is difficult difficult to study.

A. Relation Relation between Cultural Syndromes Syndromes and and A. between Cultural Social Behavior In our previous discussion we presented many specific points about the way the cultural syndromes are related to social behavior. While such details are. very association-dissociation, useful, we also need to link the cultural syndromes to the association-dissociation, superordination-subordination, intimacy-formality, intimacy-formality, and overt-covert overt-covert description superordination-subordination, of social behavior in order to link the cultural syndromes to the dimensions of of personaUty we discuss later. personality Collectivism is characterized by higher rates of association with the ingroup and of dissociation with the outgroup than is individualism; individuahsm; also, collectivism in ingreater emphasis on the subordination-superordination subordination-superordination dimension of social volves greater found in individualism. Finally, in collectivist coUectivist cultures more behav behavbehavior than is found ior is formal and covert than in individualist cultures. This is because norms are conformity to the collective can be oppressive more important than attitudes, and conformity fantasy life. and thus individuals escape in fantasy

B. B. Dimensions Dimensions of of Personality Personality Variation Variation

Hogan (1983) has shown that most dimensions of personality personaUty are various combina combinations of the two social behavior dimensions mentioned above: association associationdissociation and superordination-subordination. superordination-subordination. We can broaden this perspective by arguing that the quality quaUty of social behavior can be determined by examining the frequencies frequencies of associative-dissociative, super superordinate-subordinate, and intimate-formal intimate-formal behaviors for for different different kinds of actors (e.g., gender, status, culture, age) responding to different different kinds of targets (e.g., gender, status, culture, age) in settings that involve particular configurations configurations of goal interdependence (e.g., competition, cooperation). Hogan's (1983) socioanalytic theory of personality personaUty is based on the first two dimensions of social behavior discussed above. He has reviewed many studies that framework of high-low sociability (associative behavior) and can be placed in a framework conformity (subordinate behavior). The classic dimensions of of personality personaUty high-low conformity isolated via inventories can be seen as mixtures of these basic dimensions. For adjustment, extroversion, and other dimensions are related to associative associativeexample, adjustment, superordinate-dissociative subordinate behaviors, while neuroticism is linked to superordinate-dissociative subordinate-dissociative behaviors, and originality behaviors, anxiety is linked to subordinate-dissociative associative-superordinate behaviors. to associative-superordinate

45Z 452

HARRY C. HARRY C. TRlAJIiDIS TRIANDIS

The mistake of some psychologists has been to try to account for one behavior of of aa particular particular individual individual in in aa specific specific setting. setting. The The use use of of traits traits to to predict predict behavior behavior has not been fruitful fruitful (Nisbett, 1980) when used in such a simple form. But indexes based on many behaviors have been much more successful (Epstein, 1979, 1980). Most promising is the use of of dimensions, the expectation being that the higher an individual is on a dimension, the higher the probability that the individual's behavior summarized by that dimension. will be summarized What kinds of of indexes and dimensions should we look for? One direction that seems promising was suggested by Goldberg (1981), who reviewed all typologies typologies of of personality personality and asked the question, what typology is most likely to be universal? may be universal. The He suggests that the so-called "Big Five" (Norman, 1969) may five dimensions are: 1. 2. 3. 4. 5.

Surgency: talkative, sociable, adventurous, open Agreeableness: good natured, cooperative, mild Conscientiousness: responsible, scrupulous, persevering Emotionally stable: calm, composed, poised Cultured: intellectual, artistic, imaginative, polished

He believes that there are some questions all humans are likely to ask: (1) Is X active active and anddominant dominantor orpassive passiveand andsubmissive submissive(Can (CanI Ibully bullyXXor orwill will agreeable(warm (warmand andpleasant) pleasant)or ordisagreeable disagreeable(cold (cold X try try to to bully bully me)? (2) Is Is X X agreeable X and distant)? Can II count count on on X X (Is (Is X X responsible responsible and and conscientious conscientiousoror and distant)? (3) Can undependable and and negligent)? negligent)? (4) (4) Is Is X X crazy (unpredictable) (unpredictable) or or sane sane (stable)? (stable)? undependable (5) Is Is X X smart smart or or dumb dumb (How (How easy easy will will it it be be for for me me to to teach teach X)? Are these these (5) X)? Are universal questions? questions? universal (Goldberg,1981, 1981,p.p.161) 161) (Goldberg,

We cannot be sure, but the argument seems plausible. There is some empirical support for Goldberg's speculation from a study of Yang and Bond (1990). These authors started from a pool of of Chinese personality descriptors and then asked Chinese Chinese subjects subjects to to rate rate several several targets targets on on these these descriptors. descriptors. They They extracted extracted five five dimensions personality and reasonable correlations dimensions of of personality and found found reasonable correlations between between the the Chinese Chinese indigenous dimensions and the Big Five. Table I presents the Chinese indigenous factors. Table II shows the correla correlations between between the Chinese factors and the Big Five. This table shows that the Chinese Chinese dimension of of Social Orientation is related to the Big Five dimension of Agreeableness; Competence to both Emotional Stability and Culture; Expressive Expressiveness to Extroversion; Self-Control Self-Control to Agreeableness also; and Optimism to Emo Emotional Stability. If one takes into account the unreliability of of measurements, the correlations are not bad, but they are certainly not good. The one Big Five factor that did not emerge among the Chinese is Conscien Conscientiousness. It is only mildly related to three of the five Chinese factors. My speculation is so because Chinese culture is very high in emphasizing "virtuous" is that this is (Hofstede, 1991), "reliable" behavior; people are socialized to be conscientious to

CHAPTER 18 18 CHAPTER

453 453

CROSS-CULTURAL PERSPECDVES PERSPECTIVES CROSS-CULTURAL

TABLE II TABLE Most Salient Variables and Their Average Varimax Varimax Loadings on the Five Factors

Salient variable

Average loading

Salient variable

Average loading

Orientation--Self-Centeredness Social Orientation-Self-Centeredness Honest Good and gentle

Loyal Cordial Kind Friendly Frank Morally clean Responsible Gracious

.61 .61 .57 .55 .55 .54 .48 .48 .47 .45 .43

Untruthful Selfish Opportunistic Sly

Greedy

Naughty Ruthless

Merciless Hostile Harsh

-.53 -.53 - . .50 50 -.49 -.49 - . .49 49 -.47 -.47 -.47 -.45 -.45 -.44 -.44 -.44 -.44 -.44 -.44

Competence -Impotence Competence-Impotence Determined Resolute and firm Capable Tactful Tactful Brave Smart Rational Independent Wise Quick and sharp

.46 .46

.46 .46 .46 .44 .43 .43 .42 .42 .41 .41

Dependent Fearful Timid Childish Foolish Dull

Shallow Vulgar Shy

Self-disdainful Self-disdainful

-.49 -.48 -.48 - . .48 48 -.43 -.43 -.41 -.40 -.40 -.39 -.39 -.36 -.36 - . .35 35 -.34 -.34

Expressiveness-Conservatism Expressiveness -Conservatism Vivacious Passionate Straightforward Humorous Talkative Mischievous Optimistic Broad-minded Gracious

Generous

.56 .47 .43 .43 .43 .41 .41 .39 .38 .37 .36

Old-fashioned Conservative Rigid Solemn Awkward Introverted Stubborn Indifferent

-.46 -.46 -.46 -.46 -.45 -.43 -.43 -.41 -.41 -.41 - . .35 35 -.35 -.35

Self-Control-Impulsiveness Self-Control-:Impulsiveness

refined Quiet and refined Cultured Modest Upright and correct Self-possessed Steady Objective

.42 .41 .41 .40 .38 .37 .36 .35

Impulsive Irritable Frivolous Bad-tempered Headstrong Stubborn Opinionated Extreme

-.55 -.55 -.53 -.53 -.42 -.42 -.42 -.42 -.39 -.39 -.38 -.38 -.37 -.37 -.37 -.37 {continues)) (continues

4S4 454

HARRY C. TRIANDIS TRiANDlS

continued TABLE II continued TABLE Average loading

Salient variable

Salient variable

Average loading

Optimism-•Neuroticism Optimism-Neuroticism .47 .38 .34

Optimistic Pleasant Self-confident

Moody Worrying Pessimistic Anxious Sensitive Self-pitying

-.67 -.64 -.55 -.50 -.42 -.38

Note. Note. Data are derived from the 150 adjective scales for the the six six target persons (i.e., father, mother, teacher, teacher, neighbor, friend, and self). After Yan and Bond (1990).

such an extent that individual differences on this variable do not stand out. If everyone in a culture is high on a particular attribute, there is no point for people in that culture to notice that attribute. That observation is important, because it tells us that some etic dimensions may not be important in some cultures, and some not be beknown known to to us. us. IIargued argued emic dimensions that indigenous cultures generate may not Association-Dissociation and Superordination-Subordination are etic earlier that Association-Dissociation frequency of of interaction (Chapple, (Chappie, dimensions. Another important etic dimension is frequency 1970). For example, Native Americans and people who live in the Arctic do not consider it necessary to talk unless they have something to say. In the United States we tend to think that silence implies hostility, so we make small talk even when we have nothing important to communicate. The Big Five are related to these three dimensions of social behavior. First, frequency of interaction is clearly related to SurgencylExtroversion. discus Surgency/Extroversion. In our discussion of cultural syndromes we suggested greater Surgency in individualistic cultures TABLE D TABLEn Average Correlations between the Emic and the Imposed Etic Factors

Average Correlations between the Emic and the Imposed Etic Factors Emic factors

Emic factors

Imposed etic factor

S-S

C-I

E-C

S-I

O-N

Extraversion Extraversion Agreeableness Conscientiousness Agreeableness Emotional Stability Conscientiousness Culture Emotional Stability

.21 (.12-.46) .21 (.12-.46) 66 (.41-.77) .28 (- .08-.48) 66 (.41-.77) .35 .28 (.27-.40) (-.08-.48) .29 .35 (.16-.47) (.27-.40)

.09 (.00-.14) .09 (.13-.37) (.00-.14) .29 .31 .29 (.06-.53) (.13-.37) .55 .31 (.36-.68) (.06-.53) .50 .55 (.40-.62) (.36-.68)

.51 (.46-.59) .51 (.12-.44) (.46-.59) .30 - .09 (- .32-.07) .30 (.12-.44) .36 (.32-.41) -.09 (-.32-.07) .37 .36 (.22-.53) (.32-.41)

.01 (-.07-.15) .01 (.40-.66) (~.07-.15) .56 .29 .56 (.13-.40) (.40~.66) .43 .29 (.34-.52) (.13-.40) .28 .43 (.07-.39) (.34-.52)

.16 (.09-.25) .16 (.09-.21) (.09-.25) .14 .01 .14 (-.13-.11) (.09-.21) .44 .01 (.40-.48) (-.13-.11) .11 (.05-.20) .44 (.40-.48)

Imposed etic factor

Culture

S-S

.29 (.16-.47)

C-I

.50 (.40-.62)

E-C

.37 (.22-.53)

S-I

.28 (.07-.39)

O-N

.11 (.05-.20)

Note. (i.e., father, mother, Note. Data derived from the six target persons (i.e., mother, teacher, neighbor, friend, and self). S-S, Social Orientation-Self-Centeredness; C-I, Competence-Impotence; Competence-Impotence; E-C, Expressiveness-Conservatism; Expressiveness-Conservatism; S-I, Self-Control-Impulsiveness; O-N, Optimism-Neuroticism. Optimism-Neuroticism. After Yang and Bond (1990).

CHAPTER CHAPTER 18 18 CROss-CULTURAL CROSS-CULTURALPERSPECDVES PERSPECTIVES

455 455

than collectivist cultures, especially when new groups are present. Associative be behaviors are clearly related to Agreeableness. We can expect expect more Agreeableness among collectivists coUectivists interacting with ingroup members than among individualists, especially since they have important long-term relationships with ingroup members. Second, if Hogan is correct in linking Associative-Subordination Associative-Subordination to Adjust AdjustDissociative-Subordination to Anxiety, we might argue argue that Associative Associativement and Dissociative-Subordination Subordination (Adjustment) corresponds to Conscientiousness and Dissociative DissociativeSubordination (Anxiety) corresponds to Emotional Instability. This would suggest Associative-Superordination (calm, composed) is Emotional Stability, StabiHty, and that Associative-Superordination Dissociative-Superordination reflects the opposite of of Adjustment—aggression, atDissociative-Superordination Adjustment-aggression, at tack. Again, we might expect more Emotional Stability and less Aggression within coUectivists than individualists, and the opposite pattern for the ingroup among collectivists relationships with outgroups. That leaves only the Culture dimension unaccounted for, but that may well be a fourth universal dimension, since it reflects intelligence. Granted, intelligence is differently in each culture. Nevertheless, it has common elements defined somewhat differently across cultures. Even those who define an intelligent person as one "who is slow and correct and knows what the elders expect" would agree that one needs to be able to learn what the elders expect, and would agree with Goldberg's question, question. Is X X smart? The Big Five personality dimension labeled Culture, then, can be etic if we allow it to have drastically different different content content in each culture; for example, familiarity with Beethoven in the West and Indian classical music in India. There is no reason to expect that the cultural syndromes will be related to it. Now if the Big Five are the way to study personality, we need to ask, how framework do we can we study personality across cultures? And, what theoretical framework use to link culture and personality? of the ways we can study the Big Five across First, let us consider some of cultures. In specific geographic/language regions we can 1. Ask individuals to observe themselves and rate themselves on scales such as those that reflect the Big Five. 2. Ask them to observe their own group's behavior and rate it on the same autostereotypes (see Triandis & Vassiliou, 1967). scales, which would provide their autostereotypes hetero3. Ask them to observe and rate other groups, which will provide their hetero stereotypes (see Triandis, 1971). stereotypes Convergence among these three ratings, that is, extraction of of common elements of the group among these data sets, would indicate a good deal about the attributes of being rated. We could, of course, do the same study for each of of the nations we wish to understand. Specifically, if Americans (i) rate themselves ("I tend to be") on the scales of the Big Five and (ii) rate the stimulus "Americans tend to be" on the same scales, and if (iii) we collect data from 10 diverse countries who supply "Americans tend to be" ratings on the same scales, it is likely that some common

456 456

HARRY HARRY C. C. TRiANDlS TRIANDIS

elements would emerge. It seems reasonable to assume that the common common elements are indicative of the attributes of Americans. of of This view is supported by Fischer and Trier's study (1962) on the attributes of of French and German Swiss. There was considerable considerable overlap of of auto- and hetero heterostereotypes. Similarly, a study of of Hispanic and non-Hispanic auto- and heteroste heterostereotypes (Triandis et al., 1984) showed convergence. The addition of of self-ratings self-ratings of the three data sets would make it even more likely that the common elements of are reliable and valid. While stereotypes have limited validity, they are not entirely invalid. For example, the stereotypes Bangladeshis had about the piousness piousness of of members of of different Bangladeshi tribes were found to be valid (Schuman, 1966) when compared different with observations of the frequency of of praying. However, note that in the case of of Muslims, prayers can be readily observed. Most stereotypes reflect traits that are not reliably observed. Nevertheless, when there is agreement across several sources of data, the chances are that something reliable and valid has been identified. identified. of experiments to check In addition, we can observe, ask questions, and set up experiments for the consistency between stereotypes and actual behavior. Feldman (1968), for example, checked the stereotype of of how "honest" taxi drivers were in three foreign cities by asking 50 taxis in each city to take him between randomly determined points and checking checking whether the taxis took him the long or the direct way. Another source of data is ethnographic. We define a valid stereotype as a Another sociotype. If the ethnographers ethnographers have done a good job, they have provided sociotypes. sociotype. W. W. Lambert (1984) suggested that it would be fruitful fruitful to triangulate across different methods of study. In short, one can examine the convergence among different of sociotypes (obtained from ethnographies), autostereotypes, heterostereotypes. sociotypes autostereotypes, and heterostereotypes. When there is convergence there is likely to be validity. Hampson (1982) has argued that there are similarities in the structures of self-reports, reports by others, and reports by personality scientists. It may well be the case that such similarities also exist across sociotypes, autostereotypes, and heterostereotypes. heterosterSuppose that we did find evidence that ethnographies and auto- and heteroster eotypes converge. How would we assemble the information into a meaningful meaningful framework that is presented next suggests how this may be done. whole? The framework

IV. FRAMEWORK FOR STUDIES OF CULTURE AND PERSONALITY rv. A A FRAMEWORK Figure 3 presents a framework framework for studies of culture and personality which expands frameworks proposed by John and Beatrice Whiting (Whiting & and modifies the frameworks of the framework framework Whiting, 1975) and Robert LeVine (1973). The basic elements of are ecology (e.g., there is fish that can be eaten), environments (i.e., schedules of of reinforcement reinforcement associated associated with specific specific behavior settings), history (e.g., wars), maintenance system (e.g., hunting, fishing, agriculture, industry), the interindividual system (e.g., socialization practices), innate behaviors (e.g., tropisms, need arousal), learned behaviors (e.g., conformity), and the projective system (e.g., myths, reli-

CHAPTER CHAPTER 18 18

CROSS.CULTURAL PERSPECDvES CROSSCULTURALPERSPECTIVES

" � " " ' / " 1i — ^

" E
History HI ..

�

BiOlOgiCal Biological Inheritance

Environments

Maintenance system

/

�

Genetic Genetic system

!...t� Innate behavior

II(

Inter-Individual Inter-individual System

__ �

�

457 457

—

—

T

!

Learned behavior

1-

Projective system

FIGURE FIGURE 33 A A framework framework for for the the study study of of culture culture and and personality. personality. Some Some definitions definitions follow: follow: ecology, ecology, climate, climate,

fauna, terrain, terrain, resources; environment, environment, factory, school, school, church, and other behavior behavior settings; history, history, flora, fauna,

migrations, wars, revolutions, revolutions, inventions, inventions, transfer of of ideas; migrations, ideas;

maintenance system, system, subsistence subsistence patterns, patterns, maintenance interinter individual system, system, child-rearing child-rearing patterns, patterns, interaction interaction rates, attributions attributions about about the the other person's person's behavior; behavior; individual behaviors, tropisms, need arousal, levels of of activity; activity; learned learned behaviors, behaviors, skills, skills, performances, performances, conflict conflict innate behaviors, resolution styles, styles, value value priorities; priorities; projective projective system, religion, religion, magic, magic, rituals, rituals, ceremonies, ceremonies, art, art, games, games, play, play, resolution crime, suicide. suicide. crime, means of of production, production, settlement settlement patterns, patterns, social social structure, structure, division division of of labor, labor, subjective subjective culture; culture; means

gion). Personality refers to all the individual differences reflected in innate and learned behavior, as well as behaviors within the projective system. personality is is the thesum sumof ofinnate innateand andlearned learnedbehaviors, behaviors,plus plusthe thebehaviors behaviors Thus, personality that are part of the projective system. It reflects the million years of biological and & Richerson, Richerson, 1985), 1985),the thesocialization socialization experiences, experiences, and and the the cultural evolution (Boyd & influence of recent and historical events (the mass media, wars, etc.). Culture is the influence sum of the maintenance system, the environments, and the interindividual system. This framework shows that culture and personality are interrelated, but in a very loose way. frameNumerous examples can be provided of the links summarized in this frame work. Euler, Gumerman, Karlstrom, Dean, and Hevly (1979) showed that cultural and demographic changes on the Colorado plateaus coincided with environmental fluctuationsthat thatcan canbebedefined definedprecisely preciselybybygeoclimatic geoclimaticindicators. indicators.Triandis Triandis(1972) (1972) fluctuations discussed the emergence of Greek subjective culture as a result, in part, of the low communication among geographic regions that was created by an ecology of mountains and islands. Environments have been shown to modify the maintenance system and to be good predictors of social behavior (Whiting & & Edwards, 1988). Extreme poverty (an attribute of the ecology) can modify interindividual behavior, such as the degree of a mother's attachment to an infant; extremely poor women feel emotionally distant from their babies and do not take care of them (Scheper-Hughes, 1985).

4S8 458

C TRIANDIS HARRY C. TRiANDIS

Maintenance systems are linked to interindividual systems. For example, when many females are available to take care of children, children are indulged more. Simple societies assign more chores to children, mothers receive more help, and children interact more with infants (Whiting & Whiting, 1975). As discussed previously, aspects of of the ecology, such as the isolation of of a group of of humans from from other humans, can result in high agreement about how to raise children, an aspect of system, and this leads leads to tightness raise of the interindividual system, of culture) and to conformity, conformity, a learned behavior. (an aspect of Some measures of of conformity conformity are especially interesting. In most cultures the right hand is considered "the correct" one, and people hold the norm that people should use their right hand to carry out important behaviors. If this norm did not exist, because of of variations in genetic make-up, about 12% of the population would 12% of be left-handed. Tightness can be measured by the percentage of of the population leftthat is not left-handed. The lower the percentage of the population that is left handed, the greater the tightness. of handedness-for handedness—for example, which Dawson (1974) has used several tests of hand was used to deal cards, unscrew aa jar, manage chopsticks, sweep, thread a needle, write, strike a match, throw a ball, use a hammer, toothbrush, racket, and found rates of of left-handedness left-handedness in hunting and fishing societies of of around so forth. He found 10% (Australian Aboriginal males, 10.7%, females, 10.3%; Chinese boat people males, 5.4%, females, 12.9%; Eskimo Eskimo males 12.5%, 12.5%, females, females, 10.3%). 10.3%).Western Western samples samples had rates around 6%. In agricultural societies, where conformity conformity is high, the rates .8%, females, 0%; 0%; Temne of Africa were around 1% 1% (e.g., African Congo males .8%, of Hong Kong Chinese men, 2.7%, women, males, 3.8%, females, 0%, University of Extremely low levels of of left-handedness left-handedness have also been reported from other 0%). Extremely societies that are very strict in their child rearing, e.g., Nigeria, 0.28%, reported by Bakare (1974). of humans have a tendency to In sum, it appears that, if left alone, samples of use the left 12% of left hand about 12% of the time. Western samples are between this level and zero, while the more conforming agricultural samples, that typically use severe methods of of socialization, are close to 0% in left-handedness. left-handedness. The interindividual system results in differences differences in behavior. For example, commonly in cultures that limit the indulgence indulgence of of drunkenness is found more commonly dependence in infancy, emphasize demands for achievement in childhood, and limit dependent behavior in in adulthood (Bacon, 1973). Many individualistic cultures limit dependence, because they believe that children should become independent of of their parents as soon as feasible. They pay the price in higher rates of alcoholism than coUectivist cultures. collectivist Rohner (1986) has reported studies linking parental acceptance-rejection acceptance-rejection to Rohner the child's behavior. Rejected children are more likely to be dependent, hostile, and aggressive, or passive aggressive, anxious, and emotionally unstable, and to have low self-esteem self-esteem and a negative world view (distrust others). Numerous findings of this link can be found also in in Munroe, Munroe, and Whiting (1981). supportive of

18 CHAPTER 18

CROSS-CULTURAL PERSPECI1VES PERSPECOVES CROss.CULTURAL

459 459

The projective system has also been studied in terms of child-rearing patterns. For example, cultures where supernatural beings are perceived to be malevolent have been found found to use socialization patterns characterized by rejection rejection (Rohner, 1986, p. 158). 158). Similarly, in cultures where the gods were malevolent, child-rearing 1986, patterns were found found to be very strict and severe punishment was common (W. W. Lambert, Triandis, & Wolf, 1959). This study found found also a relationship between children being beaten by nurtur nurturing agents (usually the parents) and properties of the supernaturals supernatural in several nonliterate societies. The relationship indicates that in cultures where children are frequently frequently beaten, supernaturals supernatural are perceived to be aggressive. Figure 3 shows that there is considerable distance between the interindividual system, which includes child-rearing patterns, and the projective system, which includes the kinds of of supernaturals found in a culture. The theory behind Figure 3 has a logical structure: structure: the more distant the constructs, constructs, the weaker will be the relationships. In short, finding a relationship between the interindividual system and the projective system is very supportive of the theory. Finally, the theory assumes that there are feedback feedback loops, such as behavior changing ecology. Examples include the greenhouse effect effect and changes in the ozone layer. Obviously, behavior also creates history and new genetic forms, and changes the kinds of environments in which people behave, but those relations are not for simplicity's sake. shown in Figure 3 for Of Of course, the few studies that interrelate the theory's elements do not as yet validate the theory. It is only a beginning. Much more research will be needed before we can discuss it with confidence.

V. V. SUMMARY personality is not supported by empirical The intuition that culture is related to personality research because there is is too much variation in in personality within cultures. However, if we we know the cultural syndromes of the people whose behavior we wish to predict, we can make accurate probabilistic behavioral predictions. Though we cannot say whether this person will act in this or that way, we can say that in that culture, on way. When that information information the average, people are more likely to act in this or that way. information about personality (How is this person Ukely is combined with information likely to see this social situation? What are the person's habitual ways of behaving?) there is a good chance that behavior can be predicted.

REFERENCES Adamopoulos, J., J., & & Bontempo, Bontempo, R. R. (1984). (1984). AAnote noteon onthe therelationship relationshipbetween betweensocialization socialization practice Psychology Bulletin, 18, 4-7. practice and and artistic artistic preference. preference. Cross-Cultural Cross-Cultural Psychology Bulletin, 18, 4-7.

460 460

HARRY HARRY C. C. TRIANDIS TRIANDIS

Adorno, T. T. W., W., Frenkel-Bruswik, Frenkel-Bruswik, E., E., Levinson, Levinson, D. D. J., J.,& &Sanford, Sanford,R. R.N. N.(1950). (1950).The Theauthoritar authoritarian personality. personality. New NewYork: York:Harper Harper &&Row. Row. Bacon, M. K. (1973). Cross-cultural studies in drinking. In P. G. Bourne & R. Fox (Eds.), Alcoholism: progress progress in research researchand andtreatment treatment(pp. (pp.171-194). 171-194).New NewYork: York:Academic Academic Press. Bakare, C. (1974). The development of laterality and right-left right-left discrimination in Nigerian children. In J. L. M., Dawson & W. J. Lonner (Eds.), Readings Readings in cross-cultural cross-cultural psychology (pp. 150-166). Hong Kong: University of Hong Kong Press. Berry, J. W. (1976). Human ecology ecology and cognitive cognitive style. style,Beverly Beverly Hills, Hills, CA: CA: Sage. Sage. Berry, J. W., Poortinga, Y., Segall, M., & Dasen, P. (1992). Cross-cultural Cross-culturalpsychology. psychology.New New York: Cambridge University Press. Bijnen, E. J., Van der Net, Z. J., & Poortinga, Y. H. (1986). On cross-cultural comparative studies with the Eysenck Personality Questionnaire. Journal Journalof of Cross-Cultural Cross-CulturalPsychol Psychology, ogy, 17, 17, 3-16. 3-16. Boyd, R., & Richerson, P. J. (1985). University (1985). Culture Cultureand andthe theevolutionary evolutionaryprocess. process.Chicago: Chicago: University of Chicago Press. Brandt, V. S. S. (1974). Skiing cross-culturally. Current Current Anthropology, Anthropology, 15, 15,64-66. 64-66. Bruner. J. (1974). Concluding comments and summary of conference. In J. L. M. Dawson & Bruner, W. J. Lonner (Eds.). (Eds.), Readings Readings in incross-cultural cross-culturalpsychology psychology(pp. (pp.381-391). 381-391).Hong HongKong: Kong: University of Hong Kong Press. Budner. Personality, Budner, S. (1962). Intolerance of ambiguity as a personality variable. Journal Journal of of Personality, 30, 30, 29-50. Chapple. Chappie, E. D. (1970). Culture Culture and and biological biologicalman. man.New NewYork: York:Holt. Holt,Rinehart Rinehart&&Winston. Winston. Cheek. Cheek, J. M. (1989). Identity orientations and self-interpretation. In D. M. Buss & N. Cantor (Eds.) Personality Personalitypsychology: psychology:Recent Recenttrends trendsand andemerging emergingdirections directions(pp. (pp.275-285). 275-285). New York: Springer. Child. I. (1968). Personality in culture. In E. F. Borgatta & W. W. Child, W. Lambert Lambert (Eds.). (Eds.), Handbook Handbook of of personality personality theory and research research(pp. (pp.82-148). 82-148).Chicago. Chicago,Rand RandMcNally. McNally. Chinese Culture Connection. (1987). Chinese values and the search for culture-free culture-free dimen dunensions of culture. Journal of of Cross Cross Cultural CulturalPsychology, Psychology,18,18,143-164. 143-164. Cole. Cole, M. M. (1975). (1975). An An ethnographic ethnographic psychology psychology of of cognition. cognition.In In R. R. D. D.Brislin. Brislin,S. S.Bochner. Bochner,&&W. W. (pp. 157-176). New York: Wiley. Lonner (Eds.). (Eds.), Cross-cultural Cross-culturalperspectives perspectivesononlearning learning (pp. 157-176). New York: Wiley. Dawson. J. (1974). Dawson, (1974). Ecology. Ecology, cultural pressures toward conformity and left-handedness: A bio-social psychological approach. In J. L L. M. Dawson Dawson & &W. W.J.J.Lonner Lonner(Eds.). (Eds.),Readings Readings in cross-cultural cross-culturalpsychology psychology(pp. (pp.124-149). 124-149).Hong HongKong: Kong:University UniversityofofHong HongKong KongPress. Press. Detweiler. culture: A Detweiler, R. (1980). The categorization of the actions of people from from another culture: conceptual analysis and behavioral outcome. International Intercultural International Journal Journal of of Intercultural Relations, Relations, 4.4,275-293. llS-l^Tt. DeVos. DeVos, G G., & Hippler. Hippler, A. A. (1969). Cultural psychology: Comparative studies of human behavior. In G. Lindzey & E. Aronson (Eds.). (Eds.), The handbook of of social social psychology. psychology. Reading. Reading, MA: Addison-Wesley. Draguns. Draguns, J. G. (1979). (1979). Culture and personality. In A. J. Marsella. Marsella, R. G. Tharp. Tharp, & T. J. Ciborowski (Eds.). psychology (pp. 179-208). New York: (Eds.), Perspectives Perspectiveson oncross-cultural cross-cultural psychology (pp. 179-208). New York: Academic Press. Draguns. J. (1990). Normal and abnormal behavior in cross-cultural perspective: Specifying Draguns, Specifying the nature of their relationship. In J. Berman (Ed.). (Ed.), Nebraska Symposium Symposium on on Motiva Motivation, 1989 (pp. tion, 1989 (pp. 235-278). 235-278). Lincoln: Lincoln: University University of of Nebraska Nebraska Press. Press. .•

CHAPTER CHAPTER 18 18

CROSS-CULTURAL CROSS-CULTURAL PERSPECTIVES PERSPECTIVES

461 461

Epstein, S. (1979). The stability of behavior. I. On predicting most of the people much of of Journalof ofPersonality Personalityand andSocial Social Psychology, 1097-1126. Psychology, 37, 37, 1097-1 126. the time. Journal Epstein, S. (1980). The stability of behavior. II. Implications for for psychological research. American 790-806. American Psychologist, Psychologist,35,35, 790-806. Euler, R. C., C, Gumerman, G. J., Karlstrom, T. N., Dean, J. S., & Hevly, R. H. (1979). (1979). The Colorado plateaus: Cultural dynamics and paleoenvironment. Science, 1089-1 101. Science,205, 205,1089-1101. Eysenck, H. 1., of J., & & Eysenck, S. B. G. (1983). Recent advances in the cross cultural study of & c. C. D. Spielberger (Eds.), Advances in personality personality. In J. N. Butcher & assessment(Vol. (Vol.2,2,pp. pp.41-69). 41-69).Hillsdale, Hillsdale,NJ: NJ:Erlbaum. Erlbaum. assessment Feldman, R. E. (1968). Response to compatriot and foreigner foreigner who seek assistance. Journal Journal 202-214. of 10,10, of Personality Personalityand andSocial SocialPsychology, Psychology, 202-214. Fischer, H., H., & & Trier, Trier, U. U. P. P. (1962). (1962). Das Das Verhaeltnis Verhaeltniszwischen zwischen Deutschschweizer WestschFischer, Deutschschweizer undund Westsch weizer: Untersuchung. Bern:Bern: HansHans Huber. weizer:Eine Einesozialpsychologische sozialpsychologische Untersuchung. Huber. Structuresof ofsocial sociallife: life:The The four elementary forms of human relations. Fiske, A. A. P. P. (1990). (1990). Structures Fiske, four elementary forms of human relations. New York: Free Press. A. P. (1992). The four four elementary forms of sociality: Fiske, A sociality: Framework for a unified theory PsychologicalReview, Review,99, 99,689-723. 689-723. of social relations. Psychological Geertz, C. (1973). The interpretation interpretationof ofcultures. cultures.New NewYork: York:Basic BasicBooks. Books. knowledge:Further Further essays interpretive anthropology. York: Geertz, C. C. (1983). (1983). Local knowledge: Geertz, essays in in interpretive anthropology. NewNew York: Basic Books. Language and and individual individual differences. differences. The The search search for for universal universal in in person personGoldberg, L. (1981). Language ahty lexicons. lexicons. In In L. L. Wheeler Wheeler (Ed.), (Ed.), Review Review of of personality personality and andsocial socialpsychology psychology(pp. (pp. ality 141-161). Beverly Hills, CA: Sage. Culture, language, language, and andsociety society(2nd (2nded.). ed.).Menlo MenloPark, Park, CA: CA: Goodenough, W. H. (1981). Culture, Benjamin/Cummings. Hampson, S. S. (1982). (1982). The The construction constructionofof personality: introduction. London: Routledge Hampson, personality: AnAn introduction. London: Routledge & & Kegan Paul. & Schroeder, H. M. (1961). (1961). Conceptual Conceptualsystems systemsand andpersonality personality Harvey, O. J., Hunt, D. E., & organization. New NewYork: York:Wiley. Wiley. organization. Belief systems and the Japanese way of thinking. In H. Motoaki, J. Hayashi, C. (1992). Belief Misumi, & «feB. B.Wilpert Wilpert(Eds.), (Eds.),Proceedings Proceedingsofof the 22nd International Congress ofApplied the 22nd International Congress ofApplied Misumi, Psychology Psychology (pp. 3-34). Hove, England: Erlbaum. The anthropology ofthe London: Heelas, P., & & Lock, Lock, A A. (1981). (1981).Indigenous Indigenouspsychologies: psychologies: The anthropology of self. the self London: Academic Press. Hofstede, G. (1980). Culture's Culture's consequences. consequences.Beverly BeverlyHills, Hills,CA: CA:Sage. Sage. Hofstede, G. (1980). Cultures Cultures and andorganizations. organizations.London: London:McGraw-Hill. McGraw-Hill. Hogan, R. (1983). A socioanalytic theory of personality. In Nebraska Nebraska Symposium Symposium on on Motiva Motivation, tion, 1982 1982 (pp. (pp. 55-89). 55-89). Lincoln: Lincoln: University University of of Nebraska Nebraska Press. Press. Hsu, F. F. L. L.K. K.(Ed.). (Ed.).(1972). (1972).Psychological Psychological anthropology: Approaches to culture and personality. anthropology: Approaches to culture andpersonality. Cambridge, MA: Schenkman. & Triandis, H. C. (1989). Effects Effects of culture and response format format on extreme Hui, C. C. H., & Journalof of Cross CrossCultural Cultural Psychology, 296-309. Psychology, 20,20, 296-309. response style. Journal & Levinson, D. J. (1954). National character: The study of model personality Inkeles, A., A, & Handbook of of social social psychology psychology (pp. (pp. and sociocultural systems. In G. Lindzey (Ed.), Handbook 975-1020). Reading, MA: Addison-Wesley. Inkeles, A., & Levinson, D. J. (1969). National character: The study of model personality personaUty A, & & E. Aronson (Eds.), Handbook of of social and sociocultural systems. In G. Lindzey & social psychology (pp. 418-506). Reading, MA: Addison-Wesley.

462 462

HARRY HARRY C. C. 1iuANDIS TRIANDIS

Jahoda, G. (1980). Theoretical and systematic approaches in cross-cultural psychology. In H. C. Triandis & W. W. Lambert (Eds.), Handbook of of cross-cultural cross-culturalpsychology psychology(pp. (pp. 69-141). Boston: Allyn AUyn & Bacon. Jeanne, R. L. (Ed.). (1988). Interindividual Interindividual behavioral behavioral variability variabilityininsocial socialinsects. insects.Boulder, Boulder, CO: Westview Press. IL:IL: Row, Peterson. B. (Ed.). (1961). Studying personality personalitycross-culturally. cross-culturally,Evanston, Evanston, Row, Peterson. Kaplan, B. Keesing, R. M. M. (1981). Theories of culture. In R. W. Casson (Ed.), Language, Language, culture, culture, and and cognition: cognition: Anthropological Anthropologicalprespectives prespectives(pp. (pp.42-66). 42-66).New NewYork: York:Macmillan. Macmillan. & Berry, J. W. (1993). Indigenous Indigenouspsychologies: psychologies:Research Researchand andexperience experience cultural in in cultural Kim, U., & context. context. Newbury Park, CA: Sage. psychology. New New York: York: Holt. Holt. Klineberg, O. (1954). Social psychology. (1980). Historical perspective: Cross-cultural psychology before 1960. In Klineberg, O. (1980). & W. W. Lambert (Eds.), Handbook Handbook of of cross-cultural cross-culturalpsychology psychology(Vol. (Vol. H. C. Triandis & 1, AUyn & Bacon. 1, pp. 31-68). Boston: Allyn of social (1954). Culture and behavior. In G. Lindzey (Ed.), Handbook Handbook of social Kluckhohn, C. (1954). psychology (Vol. 2, pp. 921-976). Cambridge, MA: Addison-Wesley. Personalityand andnature naturesociety societyand andculture. culture. New C, & & Murray, H. (Eds.). (1948). (1948). Personality New Kluckhohn, c., York: Knopf. Kohn, M. M. L. (1969). (1969). Class and conformity. Homewood, IL: Dorsey. Kohn, M. M. L. (1987, August). Cross-national research as an analytic strategy. Presidential Address to the Annual Meeting of the American Sociological Association, in Chicago. Lambert, W. E. (1992). Challenging established views on social issues: The power and Psychologist,47, 47,533-542. 533-542. limitations of research. American Psychologist, & Frasure-Smith, N. (1979). Child rearing rearing values. values.New New York: York: Lambert, W. E., Hamers, J., & Praeger. brief reflections of the construction of of culture. Society Fall). Some brief Society Lambert, W. W. (1984, Fall). for Cross-Cultural Cross-CulturalResearch ResearchNewsletter, Newsletter, p. p. 9. 9. for Lambert, W. W., Triandis, L. M., & Wolf, M. (1959). Some correlates of beliefs in the malevolence and benevolence of supernatural beings: A cross-cultural study. Journal of of Abnormal Abnormal and Social Psychology, Psychology, 58, 58,162-169. 162-169. LeVine, R. A. (1973). IL:IL: Aldine. Culture, behavior, behavior,and andpersonality. personality.Chicago, Chicago, Aldine. (1973). Culture, Linton, R. (1945). The cultural cultural background of of personality. personality. New York: Appleton-Century Appleton-CenturyCrofts. (1980). The search for Lonner, W. (1980). for psychological universals. In H. C. Triandis & W. W. Lambert (Eds.), Handbook Handbook of of cross-cultural cross-culturalpsychology psychology(Vol. (Vol.1,1,pp. pp.143-204). 143-204).Boston: Boston: Allyn AUyn & Bacon. Malinowski, MaUnowski, G. (1944). A scientific scientifictheory theoryof ofculture. culture.Chapel ChapelHill: HiU:University UniversityofofNorth NorthCaro Carolina Press. Markus, H., & & Kitayama, S. (1991). Culture and self: Implications for for cognition, emotion and motivation. Psychological PsychologicalReview, Review,98, 98,224-253. 224-253. Kroeber (Ed.), Anthropology today (pp. Mead, M. (1953). National character. In A. L. Kroeber 642-667). Chicago: University of Chicago Press. Minturn, L., & & Lambert, W. W. (1964). Mothers Mothers of of six cultures. NewYork: York:Wiley. Wiley. cultures. New & Whiting, B. B. B. (1981). Handbook Handbook of of cross-cultural cross-culturalhuman human Munroe, R. H., Munroe, R. L., & development. development. New York: Garland. professional psychology. In L. Festinger Nisbett, R. E. (1980). The trait construct in lay and professional (Ed.), Retrospections Retrospectionson onsocial socialpsychology psychology(pp. (pp.109-130). 109-130).New NewYork: York:Oxford OxfordUniver University Press.

CHAPTER 18

CROSS.CULTURAL PERSPECDVES PERSPECTIVES CROSS·CULTURAL

463 463

Norman, W. T. T. (1969). (1969)."To "Tosee seeourselves ourselvesas asothers otherssee seeus!" us!"Relations Relationsamong amongself-perceptions, self-perceptions, peer perceptions, and expected peer-perceptions of personality attributes. Multivariate Multivariate Behavioral Research, 417-443. Behavioral Research,4, 4, 417-443. difference between "tight" and "loose" societies. Transaction, Transaction, Pelto, P. J. (1968, April). The difference pp. 37-40. Pepitone, A., A, & Triandis, H. C. (1987). On the universality of social psychological theories. Journal 18,18, 471 471-498. -498. Journalof of Cross-Cultural Cross-CulturalPsychology, Psychology, Pettigrew, T. (1959). The The measurement measurement of of correlates correlates of of category category width width as as aacognitive cognitive variable. variable. Journalof ofPersonality, Personality,26,26,532-544. 532-544. Journal The warmth warmthdimensions: dimensions: Foundations of parental acceptance-rejection Rohner, R. P. (1986). The Foundations ofparental acceptance-rejection theory. theory. Newbury Park, CA: Sage. Rokeach, M. (1960). The open and closed closed mind. mind.New New York: York: Basic Basic Books. Books. Scheper-Hughes, N. (1985). Culture, scarcity, and maternal thinking: Maternal detachment Ethos, 13, 13, 291-317. and infant survival in a Brazilian shantytown. Ethos, Schuman, H. H. (1966). (1966). Social Social change change and and the the validity validity of of regional regional stereotypes stereotypes in in East East Pakistan. Pakistan. Schuman, Sociometry,29, 29,426-440. 426-440. Sociometry, Shweder, R. A., A, & Bourne, E. J. (1982). Does the concept of person vary cross-culturally? In A A. J. J. Marsella Marsella & & G. G. M. M. White White (Eds.), (Eds.), Cultural Culturalconceptions conceptionsofofmental mental health health andand In therapy (pp. 97-137). London: Reidel. therapy A., & & LeVine, R. (1984). Culture Culturetheory. theory.Chicago: Chicago:University UniversityofofChicago ChicagoPress. Press. Shweder, R. A, Shweder, R. R. A, A., & M. A A. (1990). (1990). The The semiotic semiotic subject subject of of cultural cultural psychology. psychology. In In Shweder, & Sullivan, M. L. A A. Pervin Pervin (Ed.) Handbook of of personality personality theory theoryand andresearch research (pp. 399-416). New (pp. 399-416). New L. Guilford Press. York: Guilford Skinner, B. Science,213, 213,501-504. 501-504. B. F. (1981). Selection by consequences. Science, A., & & Herdt, G. (Eds.). (1990). (1990). Cultural Culturalpsychology: psychology:Essays Essayson on Stigler, J. W., Shweder, R. A, comparativehuman human development. development.New NewYork: York:Cambridge CambridgeUniversity UniversityPress. Press. comparative Withoutthe thecrysanthemum crysanthemumand and sword: A study of attitudes of youth Stoetzel, J. J. (1955). (1955). Without Stoetzel, thethe sword: A study of attitudes of youth Japan. London: London: Heinemann. Heinemann. in post-war Japan. personality development. In H. C. Triandis & &A A. Heron (Eds.), Tapp, J. L. (1981). Studying personality Handbook of of cross-cultural cross-culturalpsychology psychology(Vol. (Vol.4,4,pp. pp.343-424). 343-424).Boston: Boston:Allyn AUyn&&Bacon. Bacon. Handbook C, & & Goto, S. S. (1991). Some tests of the distinction between the Trafimow, D., Triandis, H. C., private and and the the collective collective self. self. Journal Journal of of Personality Personality and and Social SocialPsychology, Psychology,60,60, private 649-655. Triandis, H. C. (1964). Cultural influences upon cognitive processes. In L. Berkowitz (Ed.), Advances Advances in experimental experimentalsocial socialpsychology psychology(Vol. (Vol. 1,1,pp. pp.1-48). 1-48).New NewYork: York:Aca Academic Press. Triandis, H. C. (1971). Attitudes and attitude attitude change. change.New NewYork: York:Wiley. Wiley. Triandis, H. C. (1972). The analysis York: Wiley. analysisof of subjective subjectiveculture. culture.New New York: Wiley. Triandis, H. C. (1978). Some universals of social behavior. Personality Personalityand andSocial SocialPsychology Psychology Bulletin, Bulletin,4,4,1-16. 1-16. Triandis, H. C. (1989). Self Self and social behavior in differing differing cultural contexts. Psychological Psychological Review, 96, 96, 269-289. 269-289. Review, Triandis, H. C. (1990). Cross-cultural studies of individualism and collectivism. In J. Berman Motivation, 1989 1989(pp. (pp.41-133). 41-133).Lincoln: Lincoln:University Universityofof (Ed.), Nebraska Symposium on Motivation, Nebraska Press. Triandis, H. C. (1992). Cross-cultural research in social psychology. In D. Granberg & G. Social judgment judgmentand andintergroup intergrouprelations: relations: Essays in honor of Muzafer Sarup (Eds.), Social Essays in honor of Muzafer Sherif(pp. Sherif (pp. 229-244). New York: Springer-Verlag.

464 464

HARRY HARRY C. C. TRIANDIS TRIANDIS

Triandis, H. C. McGraw-Hill. C. (1994). Culture Cultureand andsocial socialbehavior. behavior.New NewYork: York: McGraw-Hill. Westview Triandis, H. C. C. (1995). Individualism Individualismand andcollectivism. collectivism.Boulder, Boulder,CO: CO: WestviewPress. Press. Triandis, H. c., C, Bontempo, R, R., Betancourt, H., Bond, M., Leung, K., Brenes, A., Georgas, 1., Hui, Spangenberg, J.,J., Touzard, J., Hui, C. C. H., H.,Marin, Marin,G., G.,Setiadi, Setiadi,B., B.,Sinha, Sinha,J.J.B.B.P.,P.,Verma, Verma,J.,J., Spangenberg, Touzard, & de de Montmollin, Montmollin, G. G. (1986). (1986). The The measurement measurement of of etic etic aspects aspects of of individualism individualism and and H., & 38,38, 257-267. collectivism across cultures. Australian AustralianJournal JournalofofPsychology, Psychology, 257-267. R., Leung, K., & & Hui, C. H. (1990). A method for determining Triandis, H. C, C., Bontempo, R, Psychology, cultural, demographic, and personal constructs. Journal Journalof ofCross-Cultural Cross-Cultural Psychology, 21, 302-318. 21, Triandis, H. c., C, Bontempo, R, R., Villareal, M. J., Asai, M., & & Lucca, N. (1988). Individualism and collectivism: Cross-cultural perspectives on self-ingroup relationships. Journal Journal of of Personality 323-338. Personalityand andSocial SocialPsychology, Psychology,54,54, 323-338. R., & &. Hui, C. C. H. (1988). Cross-cultural training across the Triandis, H. C, c., Brislin, R, InternationalJournal JournalofofIntercultural Intercultural Relations, Relations, 12, 12, individualism-collectivism divide. International 269-289. C, Davis, E. E., & & Takezawa, S. S. (1965). Some determinants of social distance Triandis, H. c., Journalof ofPersonality Personalityand andSocial Social among Americans, German, and Japanese students. Journal Psychology, 2,2,540-551. 540-551. Psychology, C, Leung, K., Villareal, M., & & Clack, Clack, F. F.L. L. (1985). (1985). Allocentric AUocentric versus versusidiocentric idiocentric Triandis, H. c., tendencies: Convergent Convergent and anddiscriminant discriminantvalidation. validation.Journal JournalofofResearch Research Personality, tendencies: in in Personality, 19, 395-415. 395-415. 19, C, Lisansky, J., Setiadi, B., Chang, B., Marin, G., & & Betancourt, H. (1982). Triandis, H. c., Anglos: The uniformity, intensity, direction, and Stereotyping among Hispanics and Anglos: Journal of Psychology,13, 13, quality of auto- and heterosterotypes. Journal of Cross-Cultural Cross-Cultural Psychology, 409-426. Triandis, H. C., C, Marin, G., Lisansky, J., & & Betancourt, H. (1984). Simpatia Simpatia as a cultural script of Hispanics. Journal 47,47, 1363-1375. Journalof of Personality Personalityand andSocial SocialPsychology, Psychology, 1363-1375. Triandis, H. c., C, McCusker, c., C, & & Hui, C. H. (1990). Multimethod probes of individualism Psychology, 59,59, 1006-1020. and collectivism. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 1006-1020. Triandis, H. c., C, & & Triandis, L. M. (1960). Race, social class, religion, and nationality as 61, determinants of social distance. Journal Journal of of Abnormal and Social Psychology, Psychology, 61, 1110-118. 10-118. Psychologi Triandis, H. H. c., C, & &Triandis, Triandis,L.L.M. M.(1962). (1962).AAcross-cultural cross-culturalstudy studyofofsocial socialdistance. distance. Psychological Monographs, Monographs,76. 76.(No. (No.21, 21,Whole WholeNo. No.540). 540). Triandis, H. C., C, & & Vassiliou, V. (1967). Frequency of contact and stereotyping. Journal Journal of of Personality 316-328. Personalityand andSocial SocialPsychology, Psychology,7, 7, 316-328. Triandis, H. C, C., Vassiliou, V., & & Nassiakou, Nassiakou, M. M. (1968). (1968).Three Threecross-cultural cross-culturalstudies studiesof ofsubjec subjecJournalof of Personality Personalityand andSocial SocialPsychology, Psychology, Monograph Supplement, tive culture. Journal Monograph Supplement, 8(4) 8(4),, 1-42. Whiting, B. B., & & Edwards, C. P. (1988). Children Children of ofdifferent differentworlds. worlds.Cambridge, Cambridge,MA: MA: Harvard University Press. & Whiting, J. W. W. M. (1975). Children Children of ofsix sixcultures-a cultures—apsycho-cultural psycho-cultural Whiting, B. B., & analysis. analysis. Cambridge, Cambridge,MA: MA:Harvard HarvardUniversity UniversityPress. Press. H., & &Berry, Berry,J.J.W. W.(1975). (1975).Psychological Psychologicaldifferentiation differentiationononcross-cultural cross-culturalperspective. perspective. Witking, H., Journalof of Cross-Cultural Cross-CulturalPsychology, Psychology, 6, 4-87. Journal 6, 4-87. & Bond, M. H. (1990). Exploring implicit personality theories with indigenous Yang, K. S., & imported constructs: constructs:The TheChinese Chinesecase. case.Journal JournalofofPersonality Personality and Social Psychology, and Social Psychology, or imported 558, 8, 1087-1095.

CHAPTER CHAPTER 19 19

TRUST AND BETRAYAL THE PSYCHOLOGY PSYCHOLOGY OF OF GETIING GEHING ALONG THE ALONG AND AND GETIING GEHING AHEAD AHEAD

H . JONES WARREN H. UNIVERSITY TENNESSEE UNIVERSITY OF OF TENNESSEE

LAURIE COUCH FORT HAYS HAYS STATE STATE UNIVERSITY UNIVERSITY FORT SUSAN SCali SUSAN SCOTT UNIVERSITY OF TULSA UNIVERSITY OF TULSA

An interesting problem emerges in in considering the role of personality in in the context of social life. On the one hand, various personality factors clearly cleariy contribute to one's acceptance by peers, which in turn is associated with happiness, health, and adjustment. Furthermore, it is reasonable to suppose that the ability to "get along" with others othershas has been beenaa key key feature feature in in the the evolution evolution of of human humanpersonality; personality;specifically, specifically, the propensity for "banding together" (i.e., group formation and group living) was the key to human evolution and the survival of human beings as a species, as well as in the development of culture. Thus, individual differences differences in such themes as trust, friendliness, kindness, and sincerity were likely selected in human evolution because such propensities better ensured individual survival to reproduce through the mechanisms of group acceptance and popularity. On the other hand, individual HANDBOOK OFOF PERSONALITY PSYCHOLOGY 465 HANDBOOK PERSONALITY PSYCHOLOGY

465

CoPYll GHT C 1997 IY COPYRIGHT © 1997 BY ACADEMIC ACADEMIC PRESS. ALL UGHTS OF lEPlODUcnON IN ANY FORM RESEIYED.

ALL RIGITTS OF REnoDucnoN w ANY FORM RESERVED.

466 466

SCOTT JONES, COUCH, AND ScOlT

human beings inevitably compete with one another for scarce natural and social resources, not only between groups, but between individuals within groups. groups. All human groups are stratified according to status and power, and having status and power are also closely linked to health, survival, and the quality of mate selection. These factors not only contribute to individual happiness and well-being, but also influence the likelihood of individual survival for reproduction and must have been selected by evolutionary forces. Herein lies the problem. There is a fundamental fundamental and unavoidable tension between behaviors which promote "getting along" and those associated with "getting ahead." faciliahead." Whereas being accepted by others is facili and power often depend on ambition, tated by trust, friendliness, and sharing, status and competitiveness, and cunning. Thus, they involve different behaviors and disposi dispositions, and furthermore, success in one domain is likely to result in failure in the other. For example, popularity may be acquired at the price of foregoing status, whereas status often creates jealousy and envy in others. Our chapter emphasizes the tension between trust-what trust—what many consider the qua non non of of the the personality personality factors factors necessary necessary for for mature mature and and mutually mutually satisfying satisfying sine qua relationships-and relationships—and violations of trust or what we will call betrayals. We begin with our approach to personality including our assumptions about the nature a review of our of human nature and individual differences. Next, we present definitions of trust found in the literature and present conceptual distinctions we believe will facilitate area. We then examine research on trust in subsequent research and theory in this area. some detail, including generalized trust (e.g., global trust in human nature) and particularly relational trust, that is, trust of specific relational partners. Also, we present extant theoretical perspectives on relational trust including developmental turn our attenapproaches, a component model, and the appraisal model. We then tum atten trust—what we call betrayal experiences-in experiences—in which tion to research on violations of trust-what our own own recent program of research. We We conclude conclude with withaabrief briefoverview overview we highlight our and discussion of the implications of the trust-betrayal dialectic for understanding the nexus between personality and social life.

I. PSYCHOLOGY OF PERSONALITY Traditional theories of personality (e.g., psychoanalysis) were designed to explain the origins of neurosis. In general, these theories are not very helpful in trying to reasons: (a) they have little to say about understand trust and betrayal for three reasons: positive striving or moral character (one could even argue that traditional theories tend to stigmatize successful people); (b) they emphasize overt and individualized psychopathology without regard for the subtle and nonneurotic ways in which people undermine themselves and others; others; and (c) they are hostile to the influence of society, suggesting that people who are normally socialized into the values and norms of of aa culture culture are, are, at at the the same sametime, time,deeply deeplyneurotic. neurotic.By Bycontrast, contrast,our ourperspective perspective 1976,1983; & Cheek, on personality, socioanalytic theory (Hogan, 1976, 1983; Hogan, Jones, & 1985), assumes that the purpose of personality theory is to explain social action

CHAPTER CHAPTER 19 19

TRUST TRUST AND AND BETRAYAL BETRAYAL

467 467

and that the important important differences differences among people concern differences differences in social overarching goals of of status and acceptance. acumen and the initiative to pursue the overarching From the socioanalytic perspective, neurosis and other psychopathologies are not From of oversocialization; rather, they reflect reflect an inability inability to get along with the products of life and are important important primarily primarily for for these reasons. others and to get ahead in social life

A. Personality Dynamics On the basis of of our evolutionary evolutionary history (Le., (i.e., as group-living, culture-dependent culture-dependent ani anifitness), it is reasonable to as asmals concerned with maximizing individual and group fitness), acceptance in our sume that, at a deep, organic level, we all need status and social acceptance acceptance enhance the chances of of individual reproductive social groups. Status and acceptance success or fitness; fitness, defined defined in these terms, is the ultimate agenda in biology and evolutionary theory. Status and acceptance (both of of which promote fitness) fitness) cannot be evolutionary for these negotiations is demanded; they must be negotiated, and the principal vehicle for order to take part in the larger social process, we must have individual identities. In order defined by our social identities (i.e., individroles to play in society, and those roles are defined (Le., individ differences in personality). Much of of what people do in everyday life, therefore, therefore, is ual differences defend, reinforce, or modify modify their identity claims in response to the reactions to assert, defend, of others. Identity Identity negotiation negotiation is not a lighthearted lighthearted exercise; it is always serious and of matter of of life and death. sometimes it is a matter Identity negotiations are carried carried out through strategic self-presentation. self-presentation. Iden IdenIdentity self-images are like templates that constrain, tities are idealized self-images; these self-images behavior in an often often unconscious attempt attempt to control the direct, and guide social behavior manner in which others perceive perceive us. Obviously there are extensive individual individual differ differmanner ences in every aspect of of this process. Some people are tactless bumpkins who rarely Other people are concern themselves with how others perceive and evaluate them. Other neurotically obsessed obsessed with social evaluation (cf. Hogan et aI., al., 1985). Most of of us are neurotically of strategic self selfsomewhere in between between these two poles. Moreover, this process of presentation is largely unconscious, and necessarily so. Self-conscious Self-conscious self-presenta self-presentapresentation performances. tion leads to awkward, stilted, and unconvincing social performances.

B. Social Reputation from the perspective of of both the actor and the observer. observer. Personality must be defined defined from performances Status and social acceptance depend on how observers evaluate the performances of actors. These evaluations are summarized and communicated in a social system of of individual reputations. Reputations are consequential, they are highly in terms of correlated with each person's status and social acceptance, and to a large extent correlated predict quality of of life and reproductive success. they predict The deep structure structure of of reputations reputations turns out to be relatively simple and stable. supported the notion that reputation reputation Research evidence since the 1960s has steadily supported Takemoto-Chock, is encoded in three to seven broad categories (cf. Digman & Takemoto-Chock, descriptions 1981). In what is sometimes called the "Big Five" theory, observers' descriptions

468 468

JONES, JONES, COUCH, COUCH, AND AND Scorr SCOTT

of of actors actors are are found found to to reflect reflect judgments judgments about about the the following following characteristics characteristics of of actors: actors: (1) (1) intellectual intellectual capacities, capacities, (2) (2) emotional emotional stability, stability, (3) (3) dependability, dependabihty, (4) (5) likability. (4) assertiveness, assertiveness, and and (5) likability. We We can can now now speak speak with with some some confidence confidence about about the the structure structure of of personality personality from from an an external external or or observer's observer's perspective, perspective, and and this this structure structure is is the the basis basis for for individual individual reputations. reputations. Our Our perspective perspective on on personality, personality, then, then, may may be be briefly briefly summarized summarized as as follows: follows: People People need need status status and and social social acceptance acceptance from from the the members members of of their their social social groups. groups. They negotiate negotiate for for these these commodities commodities using using self-presentations self-presentations which which are are constrained constrained They individual differences differences in social identity. The results of of their negotia negotiaand guided by individual recorded in terms of of their reputations, and these reputations reputations can be profiled profiled tions are recorded of five relatively relatively independent independent categories of social evaluation. in terms of

II. THE PSYCHOLOGY OF TRUST n. Definition and Conceptualization Conceptualization A. Definition

Several observers have suggested that trust is critical to human existence and vital to effective functioning functioning of of all levels of human systems: individuals, dyads, groups, the effective For example, example, organizations, and nations (Rotenberg, 1991; Rotter, 1967; Stack, 1978). For suggested that trust is an essential ingredient in the process of of organiza organizaScott (1980) suggested effectiveness. Several researchers have found found that high-trust groups perform perform tional effectiveness. better on certain tasks than low-trust groups, and, in fact, it has been argued that trust better of group processes (Golembiewski & McConkie, 1975). Dyads are lies at the heart of of the other. Research on dyadic trust also clearly influenced influenced by each member's trust of suggests that a person's trust in another strongly influences the other's trust in that that person. Several authors (e.g., Erikson, 1950; Rotter, 1967; Wrightsman, 1974) have suggested suggested that that trust trust is is the the essential essential basis basis for for the the development development of of any any personal personal relation relationship. Also, ship. Also, it it has has been been suggested suggested that that the the development development of of trust trust is is an an important important compo component "healthy personality." nent in in adjustment adjustment and and aa "healthy personality." For For example, example, Schill, Schill, Toves, Toves, and and Ra Ramanaiah (1980) manaiah (1980) reported reported that that individuals individuals low low in in trust trust tend tend to to have have higher higher stress stress scores scores and emotional and and report report more more emotional and physical physical distress distress than than persons persons scoring scoring high high in in trust. trust. In short, short, trust permeating the people and In trust is is aa ubiquitous ubiquitous concept concept permeating the interface interface between between people and their their social social environments. environments. Several Several definitions definitions oftrust of trust have have been been proposed. proposed. Erikson Erikson (1963) (1963) offered offered aa general general theory theory of of trust trust as as aa stage stage in in the the development development of of personality. personality. In In this this view, view, the the first first crisis crisis of development development concerns whether of whether or not to trust the primary caregiver. According to the theory, theory, every every person person must must learn learn to to trust trust at at least least one one other other person person to to care care for for them, them, the fail to survive. The "initial trust" described by Erikson is presumably presumably imperative imperative or fail for all other other aspects of of personality development development because it initiates subsequent subsequent social for of trust is is vital vital to to the the understanding understanding growth. Deutsch (1958) believed that the concept of of social life and personality development development as well. Specifically, Specifically, Deutsch explained the of importance of of risk-taking behaviors associated associated with trust and argued that trust and importance object of trust will likely trustworthiness are very strongly related; specifically, the object

CHAPTER CHAPTER 19 19 TRUST TRUST AND AND BETRAYAL BETRAYAL

469 469

have some bearing on the amount of trust placed in him or her by an individual. He also proposed that trust would be dependent on the perceived power of the object (i.e., the other person person)) to cause desired events, the the relationship relationship between the the individual individual (Le., and the object, the influence ofcommunication, third party influences, and the individindivid self-esteem. ual's level of self-esteem. Similarly, Rotter (1967) developed the concept of interpersonal trust as an exex pectancy of a person or group regarding the likelihood that a promise will be kept. developed for its measurement) This particular conceptualization (and the scale he developed has enjoyed investigations enjoyed widespread usage in the literature, including, for example, investigations of the dynamics of interpersonal trust, client -therapist trust, trust in cUent-therapist in work/team situa situadevelopment of trust in children, and political or tions (including game playing), the development Many of these these studies used used Rotter's Rotter's definition definition in in aa slightly slightly different different sense sense public trust. Many to mean expectancies about the trustworthiness of people in general (Stack, 1978; Wrightsman, 1974). different types Some authors have suggested the utility of distinguishing among different example, Driscoll DriscoU (1978) (1978)suggested suggesteddividing dividingtrust trustinto intotwo two further further catego categoof trust. For example, attitudinal/affective component and a specific situational/cognitive ries: a general attitudinal/affective further proposedfive comcomponent. Stack (1978) further five determinants of situational trust: com Each munication, risk, credibility of promises, social evaluation, and generalized trust. Each future events so as to of these determinants is viewed in terms of past, present, and future ascertain the likelihood that trust will emerge. Risk is perhaps the most prevalent of of certhese determinants in the literature on trust. Trusting another person involves a cer tain amount of vulnerability vulnerability for for the trusting individual. The individual must be willing in order to experience intimacy (Rempel, Holmes, & Zanna, 1985). 1985). to accept this risk in Information about the degree of vulnerability an individual will be placed in can be Information obtained through communication and experience. When one accepts a vulnerable position and experiences a negative outcome, one will be less likely to risk trusting the same partner in the future, and perhaps other people as well. We concur regarding the utility of drawing conceptual distinctions in this area and would elaborate on the distinctions among these concepts. concepts. First, it is important to distinguish between, on the one hand, the trust of an observer in a social actor (or all social actors) and, on the other hand, the trustworthiness of the actor. This distinction is similar to the two definitions of personality as internal structures and processes versus social reputation. Second, distinctions may be drawn on the basis of the scope or target of trust, for often described (cf. Wrights Wrightsfor which we propose two levels. Most broadly, trust is often man, 1974) as a generalized expectancy that other people in general are reliable and is what Erikson, Deutsch, Wrightsman, and most of the earlier scholars honest. This is have meant by the concept of trust. We We refer refer to to this this as as generalized generalizedtrust-its trust—itsopposite opposite is misanthropy or paranoia—and is also what Rotter meant by by his is paranoia-and would argue that it is concept of interpersonal trust. Within generalized trust, one may imagine trust of of trustworthiothers or people in general, without regard to their specific behaviors (i.e., (Le., trustworthi such aa concept is is explicit in the the work work ness) or their personal characteristics. Although such we suspect that this isis primarily primarily relevant in in some some hypothetical hypothetical and! and/ of earlier scholars, we

470 470

JONES, COUCH, AND AND ScOIT JONES, COUCH, SCOTT

If not, dark alleys would or philosophical sense, but less so in everyday experience. If be populated than be more more densely densely populated than they they appear appear to to be, be, and and there there would would be be no no need need for for social social anxiety. anxiety. We We suspect suspect that that experientially, experientially, generalized generalized trust trust is is circumscribed circumscribed by by general so they are some way social identity; social identity; that that is, is, one one trusts trusts people people in in general so long long as as they are in in some way similar to belong to the same same group, group, nationality, similar to oneself, oneself, for for example, example, like-minded like-minded or or belong to the nationality, or or race. race. At At the the very very least least we we require require that that others others abide abide by by the the same same expectations expectations for for social the insane, social interchange-creatures interchange—creatures from from outer outer space, space, the insane, foreigners, foreigners, deviants, deviants, and and intentions, but but cannot other other "outsiders" "outsiders" are are not not necessarily necessarily malevolent malevolent in in their their intentions, cannot always always be acknowledged in in the be relied relied upon upon to to play play by by the the rules. rules. Although Although rarely rarely acknowledged the literature, literature, we we suspect suspect that that this this is is really really what what has has been been meant meant by by generalized generalized trust. trust. In In any any case, case, generalized trust trust is is important important because because it it accounts accounts for for the the expectations expectations people people carry carry with them into with them into their their social social experiences. experiences. relational trust refers refers to the trust one has for for a By contrast, what we call relational specific person or groups of of people. In a sense relational relational trust trust derives from from one's specific participation in specific relationships and interpersonal interpersonal exchanges (e.g., marriage, participation family relationships, friendships, boss-subordinate relationships, work colleagues, family trustworthiness of of one's relational relational neighbors), and yet it does not reduce to the trustworthiness refers to the motivation motivation and and ability to permit partners. Instead, relational trust refers oneself to become vulnerable to others through the development development of of the relationship oneself & Burdette, in press) that although itself. We have argued elsewhere (cf. W. H. Jones & for psychological well-being, involvement involvement in personal connecessary for personal relationships con risks—rejection and and betrayal. Thus, relational tains two principal psychological risks-rejection betrayed by one's relational partners. trust requires risking being betrayed B. B.

Generalized Trust

of generalized generalized trust According to the literature, one's level of trust is determined determined by his or her general general expectations of of the social motives niotives of of others or the nature of of the world (i.e., person expects others to behave honorably honorably and (i.e., one's one's philosophy philosophy of of life). life). If If aa person expects others to behave and without without malice malice (i.e., (i.e., is is trusting trusting of of others others across across many many situations), situations), he he or or she she is is said said to to be be high high in in generalized generalized trust. trust. This This trust trust implies implies a a confidence confidence and and assurance assurance in in one's fellow very trusting one's fellow human human beings. beings. Conversely, Conversely, if if aa person person is is not not very trusting or or is is typically typically distrustful in generalized generalized trust. distrustful of of others others he he or or she she would would be be considered considered low low in trust. Some Some level level of of generalized generalized trust trust appears appears to to be be necessary necessary for for aa normal normal and and satisfying satisfying life. Although especially high levels of of trust trust could possibly be seen as gullibility, it has been demonstrated demonstrated in many studies that that people with high trust are no more gullible than people who score low on trust trust (Rotter, 1967, 1980). trust participants interpret interpret cues from from their interpersonal interpersonal Research suggests that high trust environment as well as low trust participants do in order order to determine if trust is environment appropriate in a given situation. But when there are no cues on which to base that of their general expectations of of the actions of of decision, the high trusters, because of fooled by trusting individuals who are not worthy of of people in the world, may be fooled fooled also by not trusting trusting the that trust. However, the person low in trust may be fooled honest individual.

CHAPTER CHAPTER 19 19

TRUST TRUST AND AND BETRAYAL BETRAYAL

471 471

Research indicates that most people trust individuals whom they perceive as trustworthy and that the level of generalized trust is correlated with one's own trustworthiness (Rotter, 1971). 1971). The more trusting an individual is, the less likely it is that he or she will lie or cheat (Rotter, 1971, 1971,1980). 1980). Studies also have shown that generalized trust is inversely related to alienation (Rotter, 1971), socioeconomic status (Rotter, 1971), and communication of information information (MacDonald, Kessel, & Fuller, 1972). 1972). Almost without exception, the literature appears to favor favor the person higher in trust. Studies have found found that people high in trust are seen by others and see themselves themselves as as happier, happier, more more ethical, ethical, more more attractive attractive to to the the other sex, and and more more see other sex, desirable as a close friend (Rotter, 1980). Other studies seem to support the idea desirable as a close friend (Rotter, 1980). Other studies seem to support the idea that people high in trust are better adjusted individuals (Schill et al., 1980). that people high in trust are better adjusted individuals (Schill et aI., 1980).

C. C. Relational Trust The literature on trust in close relationships is a fairly fairly new one. But, trust as a Much of of the the research research concerning self-disclosure self-disclosure relational variable is not a new idea. Much indicates aa strong strong relationship between between that that construct construct and relational relational trust. trust. For For indicates example, trust is is considered central component component of of self-disclosure (Broder, 1987), example, trust considered aa central self-disclosure (Broder, 1987), and the the two two constructs constructs are are positively positively correlated correlated (e.g.. Steel, 1991). 1991). Trust Trust has has also also and (e.g., Steel, been strongly tied tied to to relationship satisfaction (Canary (Canary & Spitzberg, 1989; D. C. been strongly relationship satisfaction Spitzberg, 1989; D. C. Jones, 1991; 1991; Larzelere Larzelere & Huston, Huston, 1980), 1980), conflict conflict and and its resolution in in relationships relationships its resolution Jones, (Canary & Cupach, Cupach, 1988; 1988; Canary Canary & Spitzberg, Spitzberg, 11989; Collins & & Read, Read, 1990; Pon(Canary 989; Collins 1990; Pon zetti & Cate, Cate, 1986; 1986; Simpson, Simpson, 1990), 1990), attachment attachment (Simpson, (Simpson, 1990), 1990), love love (Larzelere (Larzelere & & zetti Huston, 1980; 1980; Rempel Rempel et et al., 1985), and and commitment commitment (Becker, Fichman & Huston, aI., 1985), (Becker, 1987; 1987; Fichman & Levinthal, 1991; 1991; Larzelere Larzelere & &Huston, Huston, 1980). 1980). Levinthal, Three main approaches to the study of relational trust have emerged. Each supposes that trust is a critical element in close relationships, but each addresses different aspects of this process. The contributions to the literature discussing very different these approaches have been proposed primarily by John Holmes and John Rempel. The first approach includes a number of theories about the development of trust of trust. The last to through stages of a relationship. Another is a component model of be discussed is a model of the effects effects of trust on previously established relationships. 1. Trust /. Developmental Developmental Approaches to Relational Relational Trust

conjunction with intiinti The developmental approach to trust has been described in conjunction overshadmate relationships. During the early stages of romantic love, trust may be overshad of rejection, one denies owed by love itself. Instead of facing the potential pain of found that trust does tend to that trust is an issue. Larzelere and Huston (1980) found relationship development, be strongly related to feelings of love at this early stage of of relationship of trust emerged together and Dion and Dion (1976) reported that love and a sense of infatuation period in romantic relationships. Trust will during an initial unstable infatuation eelings only become become salient salient if if feelings feelings of of asymmetric asymmetric attachment attachment emerge; that that is, is, ffeelings only of uncertainty uncertainty may may arise arise that that are are not not satisfied. satisfied. For For example, example, when when an an individual individual of

JONES, COUCH, JONES, COUCH, AND AND Scon SCOTT

472 472

begins feel that the partner begins to to feel that the partner in in aa relationship relationship is is losing losing interest, interest, mistrust mistrust regarding regarding the the partner's partner's actions actions or or verbalizations verbalizations ensues. ensues. Subsequently, Subsequently, as as relationships relationships develop, develop, the the risk risk of of becoming becoming increasingly increasingly de dependent on the partner partner becomes more salient. This is characteristic characteristic of of the evaluative evaluative pendent of trust at this point. The first stage. Three theoretical perspectives explain the role of claims that love and trust trust are likely to promote one another another in a circular circular fashion fashion If trust trust is not established, a satisfying satisfying relationship cannot result. (Holmes, 1991). If An alternative theory theory holds that trust trust can be described as a reduction reduction of of uncertainty uncertainty (Kelley & Thibaut, 1978). Once the basic issues of of the relationship have been been behavresolved, questions arise as to the future. According to this view, a partner's behav iors indicate a pattern pattern which will likely continue. If the partner partner behaves predictably predictably and dependably dependably during the initial stage, the more likely it is that an individual will trust that that partner. partner. The The third third theoretical theoretical perspective perspective emphasizes emphasizes reciprocal reciprocal attach attachtrust ment. Each partner must provide the other with a confidence that their feelings ment. Each partner must provide the other with a confidence that their feelings will be reciprocated at at every every stage stage in in the the relationship relationship in in order order for for trust trust to to develop. develop. will be reciprocated In the development of romantic relationships, the risk of rejection is obvious. But, In the development of romantic relationships, the risk of rejection is obvious. But, if the partner reciprocates the individual's feelings and his or her actions seem to if the partner reciprocates the individual's feelings and his or her actions seem to demonstrate an adequate level of commitment, this exchange reduces uncertainty demonstrate an adequate level of commitment, this exchange reduces uncertainty and risk risk by by ensuring ensuring equality equality in in risk risk and and in in involvement involvement in in the the relationship relationship (Holmes, (Holmes, and 1991; Holmes & Rempel, 1989). 1991; Holmes & Rempel, 1989). Once Once the the relationship relationship is is more-or-Iess more-or-less established, established, new new issues issues arise. arise. Conflicts Conflicts may occur as each person begins to see his or her partner partner in a new and different different light. Attempts to negotiate differences differences and/or compromises are carried out during accommodation stage. This is seen as a time when a couple engages in diagnostic diagnostic this accommodation of trust (Holmes, 1991). Because relationships at processes which **test" "test" the level of of development often characterized characterized by frequent frequent and heated conflict conflict— this point of development are often of the relationship—actions of the partner which show relationship-actions on the part of risking the end of empathy or sensitivity to one's personal needs are seen as more significant significant and act empathy to pull the couple closer. Security in this stage comes from from the feelings feelings of of control **fate" which are brought about through intimacy. over one's "fate" particular (Larzelere (Larzelere & Huston, 1980) produced produced results support supportOne study in particular of the developmental developmental model of of trust. Trust and love were strongly correlated, correlated, ive of intimacy of of self-disclosure, particularly for for longer longer married married couples. as were trust and intimacy trust was reciprocated reciprocated more than either love or level of of intimacy, and Furthermore, trust of commitment commitment to the relationship. trust varied with the level of A Component Component Model Model o of Relational Trust 2. A f Relational proposed a component component theory of relational trust. This model Rempel et al. (1985) proposed four assumptions: (1) trust trust is seen to derive from from past experience experience and is based on four order to be trusted the partner partner must be develop along with the relationship; (2) in order seen as trustworthy trustworthy (e.g., reliable, dependable, and concerned concerned with providing providing ex expected rewards); (3) trust involves accepting the risk of of relying on a partner's word through intimate disclosure and gains; and through intimate disclosure and by by sacrificing sacrificing present present rewards rewards for for future future gains; and confidence in the partner's caring. (4) trust is interpreted as confidence

19 TRUST TRUSTAND ANDBETRAYAL BETRAYAL CHAPTER 19

473 473

deThree components of trust, predictability, dependability, and faith, are de scribed. Predictability-the Predictability—the most concrete and specific component-simply component—simply refers to an expectation of of the occurrence of specific behavioral patterns of the partner. If If the the partner partner conforms conforms to to this this expectancy, expectancy, the the individual individual is is able able to to move move from from concerns about the the stability his or her partner more pressing concerns about stability of of his or her partner to to more pressing matters matters of of trust. trust. Predictability Predictability is is influenced influenced by by many many factors, factors, including including the the amount amount of of experience experience in in the the relationship, relationship, consistency consistency of of recurrent recurrent behaviors, behaviors, and and stability stability of of the the social social environment (Rempel et al., aI., 1985). This stage is the most basic and probably accounts accounts for for most most of of the the trust trust early early in in aa relationship. relationship. As relationships develop, trust shifts from specific actions to a focus on the other person in general. This component is known as the partner's dependability, the second component of of relational trust. Dependability is closely related to predict predictability because a partner's predictability predictabiUty is a meaningful meaningful basis of of information from which to draw dispositional attributions. For example, if a partner behaves in a stable manner and in doing so is responsive responsive to the individual's needs, one may attribute dependability to that partner. These dispositional attributions can be transformed into what an individual views as his or her partner's trustworthiness. This component component also marks the necessary presence of risk as a factor in an individu individual's decision to trust another. The first two components of of trust, predictability and dependability, involve attention to the past. But, the issue of trust also concerns the future. After After all, the true "test" of of trust comes when new situations arise, for which relevant experience and/or knowledge may be unavailable. This requires faith, the third component of trust. Faith is considered the most important component of of trust. It may be the inherent quality in an individual that allows him or her to accept the unknown future and deal with it gracefully. Faith goes beyond an emphasis on dispositional future attributions, but is possibly related to both predictability and dependability dependability in that the past is an important basis for generalizing to future situations. However, predictability and dependability reflect the partner whereas faith is a quality of of the individual who does the trusting, and may also be related to personality factors, such as self-esteem and personal security. Rempel their component Rempel et et al. al. (1985) (1985) presented presented empirical empirical support support for for their component model model of couples, women of trust. trust. Among Among a a sample sample of of romantically romantically involved involved couples, women followed followed the the predicted predicted hierarchical hierarchical pattern pattern discussed discussed in in the the model, model, but but among among men men each each compo component function independently independently of others. The nent appeared appeared to to function of the the others. The results results also also indicated indicated that project their their own their partners. that individuals individuals tend tend to to project own motives motives onto onto their partners. Love Love was was strongly moderately related strongly related related to to the the faith faith component component of of trust, trust, moderately related to to the the depend dependability weakly correlated ability component, component, and and weakly correlated with with predictability. predictability. The The correlations correlations were were stronger for women than men. Love was also strongly related to reports of of partners having intrinsic motivations and instrumental motivations, but much less so for those who saw their partner's motivations as extrinsic. Women were more likely to feel love strongly when their partner was viewed as having instrumental motivations. This, along with gender differences, was interpreted as evidence evidence that women are more concerned than men about the basics of interpersonal behavior and appear

474 474

JONES, COUCH, AND AND SCOlT SCOTT JONES, COUCH,

integrated view of of trust. Faith and intrinsic motivations were strongly to have a more integrated related, but instrumental and extrinsic motivations generated generated no significant significant correla correlations. Predictability related to tions. Predictability was was related to instrumental instrumental and and extrinsic extrinsic motivations, motivations, and and dependability was related related to instrumental motives. dependability 3. The The Appraisal Appraisal Model Model

A more recent recent addition addition to to the the literature literature is is the the appraisal appraisal model model formulated formulated by A more by Rempel (1989) and Holmes (1991) which focuses focuses on the consequences consequences Holmes and Rempel of trust in established established relationships. High, medium, and low trusters are seen to of their relational relational interactions differently differently and thus "color" their interpretation interpretation assess their of events. Trusting persons tend to evaluate the course of of events in a relationship of over a more extended extended period of of time (Holmes, 1991). High trusting individuals do over weight in determining determining the outcome outcome or not judge single events as having much weight of a relationship. Positive events are viewed as confirming confirming the trust that has quality of threat to the relationship. been given, and single negative events are not seen as a threat trusting (those with more moderate levels of of trust) Individuals uncertain about trusting different pattern. Because they are motivated to reduce uncertainty, display a very different of responsiveness responsiveness these individuals actively assess their partner's motives and levels of hopeful that that their assessments assessments will yield in most interactions. Moderate trusters are hopeful and allow them to trust their partner, but their expectations are positive results and limited by their fear fear of of vulnerability (Holmes, 1991). Positive behaviors are readily viewed as relevant to greater greater matters of of loving or caring. Negative behaviors are also perceived perceived as very important, relating to the overall appraisal of of the relationship. greatest at this level of of trust trust because because a single negative event has heightened heightened Risk is greatest for the continuation continuation of of the relationship. implications for Low-trust couples may have had negative experiences which caused a lack of of Low-trust trust in the first place. This deficiency deficiency in trust usually implies that a breach of of security has transpired. It is important important to note, however, that little research has been done to factors that that contribute to the deterioration deterioration of of trust trust (Holmes & Rempel, establish the factors 1989), so researchers researchers have been left left only to speculate as to any connections between between Low-trust partners are likely to approach their relationship relationship betrayal and low trust. Low-trust with a relatively closed mind, which is very similar to the strategy of high-trust couples. concluded that their partner partner is not concerned concerned about about them They react as if they have concluded of positive behaviors behaviors and or the relationship (Holmes, 1991). They are suspicious of proceed cautiously. On the other other hand, negative behaviors are likely to reinforce reinforce the of trust that previously previously exists (Holmes & Rempel, 1989). lack of D. Summary of of Trust Research phenomenon of of trust suggestfour suggest four general conclusions. Research and theorizing on the phenomenon of generalized trust trust appears appears to be necessary necessary for for "normal" ^'normal" First, some minimal level of possibly for for the development development of of specific specific relationships as well. Second, Second, social life and possibly research evidence overwhelmingly supports the contention that higher levels of of both both research generalized and relational relational trust trust are associated with enhanced enhanced functioning functioning of of individugeneralized

CHAPTER CHAPTER 19 19 TRUST TRUSTAND ANDBETRAYAL BETRAYAL

475 475

als and groups at least under optimal conditions (e.g., when others behave honorably); relahigher trust is related to better communication, greater satisfaction with various rela tionships, higher levels of performance, and so forth. Third, relational relational trust appears related to and is possibly necessary for many of of the basic mechanisms of of relationship development (e.g., communication and self-disclosure), self-disclosure), although there is some evi evidence that relational relational trust may be limited by dispositional dispositional characteristics of of the ob observer, as well as by both the similarity and the trustworthiness of of the actor. Fourth, although their relative merits are difficult difficult to assess at present, recent conceptualiza conceptualizations of relational trust afford numerous testable hypotheses regarding regarding the role of of trust in the development and dissolution of relationships. One particular issue unresolved in the understanding of of trust is what happens to the relationship when trust is violated, in other words, when relational circum circumstances are less than optimal. In our view, most of of the literature on trust clearly generalized trust of of implies that violations of relational trust reduce the level of of generalized the victim and place the relationship in question in jeopardy. Although this pattern evolutionary-social perspective suggests that such may be common, we believe an evolutionary-social overly simplistic simplistic for for several reasons including the conflict conflict between status a model is overly of betrayal more-or-less inevitaand acceptance which makes at least some degree of more-or-Iess inevita importance of of status and acceptance and ble. Also, our assumptions regarding the importance interpersonal options are constrained negotiated identities suggest that a victim's interpersonal complex social and psychological processes which may inhibit inhibit leaving every by complex relationship in which one's partner behaves dishonorably. In any case, the literature of trust per se, and thus this is on trust has not focused extensively on violations of the topic to which we will now turn.

ni. THE PSYCHOLOGY OF BETRAYAL BETRAYAL HI. As noted at the outset, although apparently necessary for health and well-being, it is also evident that the pursuit of close personal relationships relationships engenders certain interpersonal risks. We believe there are two basic risks, rejection and betrayal. of what makes social Not all interpersonal approaches are welcomed and this is part of encounters with strangers and certain interpersonal interpersonal overtures (e.g., asking for a date) anxiety arousing (e.g., Russell, Cutrona, & Jones, 1985). We find it useful to draw a distinction between rejection-referring rebuffs of rejection—referring to rebuffs of various social overtures early in or prior to the development of interdependence interdependence and the negotiated identity of of being a pair (e.g., a couple or two friends) which characterize characterize relationships-and relationships—and of trust and allegiance allegiance as well betrayal. We conceptualize betrayal as any violation of as other forms of intrigue, treachery, and harm-doing in the context of of established and ongoing relationships. As painful as rejection may be, we suspect that betrayal is potentially much of its likely psychological consequences. For example, similar to worse because of rejection, betrayal is likely to have direct emotional (e.g., anger and depression) and other psychological implications. However, to the extent that one has invested

476 476

JONES, JONES, COUCH, COUCH, AND AND Scorr SCOTT

one's identity and sense of one's of self self in one's relationship with the instigator instigator of of the betrayal, betrayal, there there may may also also be be profound profound indirect indirect consequences. consequences. In In one one sense, sense, the the literature literature on on betrayal betrayal is is extensive. extensive. However, However, previous previous research research tends tends to to be be scattered scattered across across distinct distinct literatures literatures on on such such topics topics as as deception, deception, adultery, adultery, abuse, abuse, and and other other forms forms of of malevolence malevolence (cf. (cf. Lawson, Lawson, 1988; 1988; Metts, Metts, 1989; 1989; Miller, Miller, Mongeau, Mongeau, & Sleight, Sleight, 1986). 1986). By By contrast, contrast, in in our our approach approach to to studying studying betrayal betrayal we we have have sought sought to to subsume subsume various various types types of of betrayal betrayal as as well well as as betrayals betrayals among among varying varying popu populations lations and and types types of of relationships relationships in in order order to to discern discern communalities communalities in in the the determi determinants, nants, concomitants, concomitants, and and consequences consequences of of these these experiences. experiences. Finally, Finally, in in research research de described scribed later (Carver & Jones, 1992; Hansson, Jones, & Fletcher, 1990; W. H. Jones, 1988; W. H. Jones & Burdette, in press; W. H. Jones, Cohn, & Miller, 1991) we have operationalized betrayal betrayal in three different different ways: (a) descriptions of of specific specific betrayal betrayal operationalized betrayal narratives; (b) individual differences differences in the tendency tendency incidents which we call betrayal of the social network. to betray; and (c) betrayal as a dimension of

A. Betrayal Narratives

One approach to understanding the betrayal of trust has been to ask people to describe and answer general, but direct, questions about their experiences with betrayal. Accordingly, we have asked samples of college students and others to describe their most significant significant experience of having betrayed a relational partner significant experience of having been betrayed by a significant significant other. and their most significant For both instances, participants were then asked to indicate their relationship to the other person (e.g., friend, spouse, parent, sibling), the presumed motives under underlying the betrayal incident, when the event took place, and the consequences of of of having the incident, if any, for the relationship in question. For descriptions of betrayed another, respondents were also asked whether or not the other person (i.e., the victim) was aware of what the respondent had done to betray them. (Le., Subsequently, these descriptions and answers were thematically categorized and compared on the basis of gender, perspective (Le., (i.e., whether whether the respondent was the compared victim or the perpetrator of the betrayal), and other relevant variables. / . Types ooff Relationships 1. function of Results suggested that the types of relationships described varied as a function of the age of respondents. For example, among adults, spouses were most frequently frequently cited as both victims and instigators of betrayals, betrayals, but same-sex friends, friends, parents, parents, one's one's bosses,subordinates) subordinates) were were children, and work-related relationships (e.g., colleagues, bosses, cited. College College students students more more frequently frequently cited cited dating partners, partners,friends, friends, frequently cited. also frequently children and adolescents frequently frequently described betrayals involv involvand parents, whereas children excluing siblings, friends, and parents. Interestingly, psychiatric patients almost exclu victuns and instigators of betrayal episodes. In any case, it sively cited parents as victims is clear from data involving multiple samples that the betrayal incidents described

CHAPTER 19 19 TRUST TRUSTAND ANDBETRAYAL BETRAYAL CHAmR

477 477

almost always involve a significant relationship in the life of of the individual describing the betrayal. 2. Type Type o off Betrayal Betrayal 2.

But what is it that people do or have done to them that constitute significant of trust? Again, the specifics of of betrayal incidents vary somewhat with betrayals of often cited extramari extramariage and life circumstances. For example, married adults most often often described tal affairs as instances of significant betrayals, whereas students more often instances of jilting or being jilted by one's boy- or girlfriend. Other common types of of betrayal which appeared in various samples included telling lies, betraying confi confidences, inadequate emotional support and attention, acts of of disloyalty, excessive obviously more dramatic criticism, and ignoring and avoiding. Less common but obviously instances of betrayal cited by respondents included abandonment, giving a child up for adoption, physical and sexual abuse, and incest. Motives for for Betrayal Betrayal 3. Motives

Not surprisingly, the motives that respondents gave for their betrayals and those of others depend, almost exclusively, on the perspective perspective of of of the respondent describing the betrayal. By perspective we mean whether one is the victim or the instigator of the betrayal. When describing their own betrayals of significant others, a majority of significant of respondents attributed their own motives to intentional and internal, but unstable, of betrayal causes, thereby reducing their own culpability. For example, instigators of often cited temporary emotional and cognitive cognitive stages (e.g., anger, depres depresincidents often physiological/psychological conditions (e.g., being sion, a desire for excitement) and physiological/psychological of intoxicated) as extenuating reasons for their actions. By contrast, descriptions of intentionaUty, internality, intemality, having been betrayed most often involved explanations of of intentionality, and stability as illustrated by references to the instigator's "mean streak," "inherent therefore seen as both responsible weaknesses," or "lack of principles." Others are therefore for their actions of betrayal and likely to betray again. 4. Relationship Relationship Change Change

Narrative responses to questions regarding how betrayal affects affects relationships were classified as one of three categories of change: change: termination/worse termination/worse (i.e., (i.e., the therelationship relationship ended, or if it continued it was characterized as less less satisfying satisfying and and intimate), intimate),no nochange change (Le., (i.e., some initial distrust, followed by a return to what the relationship had been previously, or no effects of of betrayal); and improvement (i.e., the relationship was described as better following the betrayal than before). Again, analyses clearly indi indiof perspective. cated that perceived change in the relationship varied as a function function of of the respondents adWhen describing their betrayals of others, approximately half half of ad remainmitted that the relationship had ended or was worse than before, whereas the remain ing half claimed that it had remained the same or even improved. By contrast, when of betrayals, respondents respondents over overdescribing instances in which they were the victims of whelmingly (>90% of the cases) indicated that the relationship had been terminated

478 478

JONES, COUCH, AND JONES, COUCH, AND Scon SCOTT

or was worse as a result, and these contrasting patterns emerged in studies involving differing samples of participants of differing samples of participants of varying varying ages ages and and life life circumstances. circumstances. 5. 5.

Other Other Issues Issues

Several were examined Several additional additional issues issues were examined in in these these studies. studies. For For example, example, we we found found that both men men and women reported reported that that men likely to that both and women men were were more more likely to be be instigators instigators of of betrayal these narrative betrayal in in these narrative accounts, accounts, and and that that the the accounts accounts of of betrayals betrayals by by others others against respondent involved "serious" violations against the the respondent involved more more "serious" violations of of trust trust than than did did those those of significant significant others in their lives. We also detailing the respondents' betrayals of found that betrayals of of others were more recent events, whereas the narratives of of found betrayed were, on average, alleged to have taken place significantly significantly having been betrayed earlier in the lives lives of these events were described earlier of respondents. Similarly, most of of these as having taken place during adolescence or early adulthood, and this held even among elderly participants. findings are of of note. First, instances for for which a respondent respondent de deTwo additional findings scribed two betrayals involving the same type of of event (e.g., the respondent respondent and his scribed her spouse both both having an extramarital extramarital affair) affair) or the same person (e.g., as a victim or her instigator in the other) were not associated with greater greater insight in one account and the instigator respondents tended tended to diminish their own culpability by citing or forgiveness. Instead, respondents extenuating circumstances while also holding relational partners morally responsible extenuating for for their their transgressions. Second, there was a gender gender difference difference regarding the role of of of others. For men, relationships in which the partner partner victim awareness in betrayals of of the respondent's betrayal were described as having ended ended or changed was aware of for the worse, whereas when the partner partner was not aware the relationship was seen as for women there was no relation relationimproving or remaining the same. By contrast, among women ship between ship between partner partner awareness awareness and and relationship relationship change. change. Differences in Betrayal B. Individual DitTerences In a second series of of studies we developed a brief, self-report self-report measure of of the tendency to tendency to betray betray called called the the Interpersonal Interpersonal Betrayal Betrayal Scale Scale (IBS). (IBS). Items Items referred referred to betraying others others as to commonplace commonplace instances instances of of betraying as illustrated illustrated by by the the following: following: "lying "lying to to aa family family member," member," "making "making aa promise promise to to aa friend friend or or family family member member with with no no intention of keeping it." Respondents were instructed to respond to each item by intention of Respondents instructed respond indicating the frequency frequency with which they had engaged in the behavior behavior described. described. L Biographical Betrayal 1. Biographical Correlates Correlates of of Betrayal

Analyses of responses to the IBS IBS suggested important Analyses of responses important biographical correlates of of self-reported self-reported betrayal. Specifically, Specifically, IBS scores were found found to be inversely correlated correlated of education, and length of of marriage. Married persons scored signifi signifiwith age, level of cantly persons, and cantly lower lower on on the the IBS IBS than than did did divorced divorced persons, and IBS IBS scores scores were were found found to to be be related related to to indications indications of of personal personal and and interpersonal interpersonal problems problems (e.g., (e.g., scores scores were were higher higher for for psychiatric psychiatric patients, patients, delinquents, delinquents, and and alcoholics). alcoholics). On On the the other other hand, hand, varied as a function function of of gender in our research thus far. far. IBS scores have not varied

CHAPTER 19

TRUST AND BETRAYAL

479 479

2. Psychological and Interpersonal Correlates Correlates 2.

As would be expected, scores on the IBS were found found to be significantly significantly related to personality measures of guilt, resentment, suspicion, and personality disorder as borderline, paranoid, antisocial, and passive-aggressive personal personaldimensions such as ity dimensions. Conversely, IBS scores were inversely related to such personality dimensions as self-control, well-being, responsibility, tolerance, intellectual effi effiself-descriptions ciency, psychological mindedness, and communality. Finally, both self-descriptions significant others suggested that betrayers are best described as and ratings by significant solitary, aloof, exploitive, gossipy, jealous, suspicious, envious, and cynical. Not surprisingly, IBS scores were also found to be related to variables assessing relational and interpersonal functioning. For example, scores on the IBS IBS were found found to be directly related to the number of marital problems reported as well as the for problems (Monroe, 1990). IBS scores were tendency to blame one's spouse for significantly and directly correlated with loneliness and inversely related to also significantly satisfaction with the family family with whom one was raised. Finally, individuals scoring satisfaction found to report less social support (Le., (i.e., comfort comfort and high on the IBS have been found from members of one's social network) and to have fewer voluntary companionship from relational partners (e.g., friends). Betrayal and and the the Social Social Network Network C. Betrayal operationalizing In our final approach to ope rationalizing betrayal, we have modified the Social Network List (cf. W. H. Jones & Moore, 1987) to include identification identification of persons betrayed by the respondent as well as persons who have betrayed the respondent. Our version of the Social Network List instructs participants to identify identify persons who are important to the respondent and with whom the respondent has at least occasional face-to-face face-to-face contact. In addition, participants are asked to indicate the age, gender, type of of relationship relationship (e.g., (e.g., spouse, spouse, friend, friend, brother, brother, or or work work colleague) colleague) age, gender, type and length of acquaintance with each person listed, as well as to rate various and length of acquaintance with each person listed, as well as to rate various interpersonal dimensions for each (e.g., satisfaction with the relationship, reciprocinterpersonal dimensions for each (e.g., satisfaction with the relationship, reciproc ity, jealousy, jealousy, love, love, disagreements). disagreements). ity, half the participants indicated that they have Results suggested that almost half betrayed at least one current member of their social network, and a comparable number of respondents indicated that they have been betrayed by a member of of their social network. On average, roughly 20% of network members are identified identified as being either victims or instigators of betrayal. In addition, these results suggest that betrayals are most likely to involve one's closest relationships: spouses, family, and close friends. identification of of a network member as having betrayed the Furthermore, identification significantly lower scores on positive dimensions respondent was associated with significantly describing network members (e.g., satisfaction, reciprocity, love, dependability, can turn to for help) and higher scores on negative descriptive dimensions (e.g., regret, jealousy). Also, IBS scores were were significantly significantly related related to to the the propor propordisagreements, jealousy). disagreements, tion of of the the network network identified identified as as victims victims and and instigators instigators of of betrayals. betrayals. tion

480

JONES, COUCH, AND ScOTT SCOTT JONES.

D. Summary Summary of of Betrayal Betrayal Research Research D. Our research on violations of trust may be summarized as suggesting the following four basic conclusions: First, betrayal is fairly commonplace and most likely to occur in one's most important relationships. Second, betrayal is associated with stable biographic and personality based individual differences, with the exception for which which our our data data are are inconsistent inconsistent across across differing differing measures measures of of betrayal. betrayal of gender, for function of Third, the consequences and meaning of betrayal seem to vary as a function of one's perspective as either the victim or the instigator. Fourth, although clearly associated with negative perceptions and attributions, violations of trust do not always result in the termination of a relationship. It is important to note also that the sample under although a few findings appear to depend on the characteristics of the investigation, most of these results have been replicated across samples varying in age and other life circumstances.

rv. CONCLUSION IV.

affords a unique opportunity to The integration of research on trust and betrayal affords the complex and dialectical dialectical nature of of getting getting along along and and getting getting ahead ahead in in human human view the groups and hence the complexity of understanding the role of personality in social life. Research and theorizing about trust suggest that (1) trust is a necessary precur precursor to group living and the development of specific relationships, and (2) higher levels of trust are associated not only with greater like ability-as would be exex likeability—as pected-but pected—but greater adjustment and happiness as well. well. On the other hand, the very process of trusting specific people increases the likelihood of being betrayed and apparently does so dramatically, and the persons most likely to do so are one's closest friends and relations. That most people continue to trust and take other interpersonal risks in view of both the likelihood and the potential damage of being betrayed by by one's one's partners partners is the surprising phenomenon here and suggests a fundamental fundamental and primitive need for human involvements and and companionship. companionship. Recognizing Recognizing that that much muchof ofwhat whatpeople people do in their interpersonal lives represents efforts efforts to balance interpersonal risks and rewards or to balance getting along and getting ahead raises an interesting issue about moral development, and this is the point with which we would like to close. Far from from the detached, intellectualized, and philosophical abstractions (e.g., moral moral judgments) portrayed by some researchers, viewed from from the perspective of trust and betrayal, issues of morality are immediate, messy, messy, practical, and decidedly interpersonal. Indeed, the central questions regarding moral development here might be stated as follows: (1) how do people proceed to trust specific others given the realistic possibility that they could be betrayed? and (2) how do some people after they have been continue to trust and engage others in relationships even after betrayed by them?

CHAPTER 19 19 CHAPTER

TRUST AND AND BETRAYAL BETRAYAL TRUST

481 481

REFERENCES Becker, C. C. S. S. (1987). (1987). Friendship Friendship between between women: women: A A phenomenological phenomenological study study of of best best friends. friends. Becker, Journal 18,18, 59-72. Journalof ofPhenomenological PhenomenologicalPsychology, Psychology, 59-72. Broder, S. N. (1987). Helping students with self-disclosure. School School Counselor, Counselor,34, 34,182-187. 182-187. Canary, D. D. J., J., & & Cupach, Cupach, W. W. R. R. (1988). (1988). Relational Relational and and episodic episodic characteristics characteristics associated associated Canary, 5, 305-325. with with conflict conflict tactics. tactics. Journal Journalof ofSocial Socialand andPersonal PersonalRelationships, Relationships, 5, 305-325. Canary, Canary, D. D. J., J., & & Spitzberg, Spitzberg, B. B. H. H. (1989). (1989). A A model model of of the the perceived perceived competence competence of of conflict conflict strategies. 15, 15, 630-649. Human Communication CommunicationResearch, Research, 630-649. strategies. Human Carver, M. M. D., D., & & Jones, Jones, W. W. H. H. (1992). (1992). The The Family Family Satisfaction Satisfaction Scale. Scale. Social Social Behavior Behaviorand and Carver, Personality, Personality,20, 20,71-84. 71-84. Collins, N. N. L., & & Read, Read, S. S. J. J. (1990). (1990). Adult Adult attachment, attachment, working working models, models, and and relationship relationship Collins, quality in in dating dating couples. couples. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 644-663. Psychology, 58, 58, 644-663. quality Deutsch, Deutsch, M. M. (1958). (1958). Trust Trust and and suspicion. suspicion. Conflict ConflictResolution, Resolution,2,2,265-279. 265-279. Digman, J. J. M., M., & & Takemoto-Chock, Takemoto-Chock, N. N. R. R. (1981). (1981).Factors Factorsininthe thenatural naturallanguage languageof ofpersonal personalDigman, ity: Re-analysis, Re-analysis, comparison, comparison, and and interpretation interpretation of of six six major major studies. studies. Multivariate MultivariateBe Beity: havioral Research, Research,16,16,149-170. 149-170. havioral Dion, K. K. L., L., & & Dion, Dion, K. K. K. K. (1976). (1976). Correlates Correlates of of romantic romantic love. love. Journal Journalof ofConsulting Consultingand and Dion, Clinical ClinicalPsychology, Psychology,41,41,51-56. 51-56. Driscoll, J. W. (1978). Trust Trust and and participation participation in in organizational organizational decision decision making making as as predictors predictors 21,21, 44-56. of satisfaction. Academy of of Management ManagementJournal, Journal, 44-56. Erikson, E. E. H. H. (1950). (1950). Childhood Childhoodand andsociety. society.New NewYork: York:Norton. Norton. Erikson, Erikson, E. E. H. H. (1963). (1963). Childhood Childhood and andsociety society(2nd (2nd ed.). ed.).New New York: York:Norton. Norton. Erikson, Fichman, M., M., & & Levinthal, Levinthal, D. D. A. A. (1991). (1991). Honeymoons Honeymoons and and the the liability liability of of adolescence: adolescence: A A Fichman, new perspective on duration dependence in social and organizational relationships. Academy of 442-468. of Management ManagementReview, Review,16,16, 442-468. Golembiewski, R. R. T., & & McConkie, M. M. (1975). (1975). The The centrality centrality of of interpersonal interpersonal trust trust in in group group Golembiewski, processes. -185). London: processes. In In C. C. L. L. Cooper Cooper (Ed.), (Ed.), Theories Theoriesof ofgroup groupprocess process(pp. (pp.131 131-185). London: Wiley. 0., Jones, W. H., & Hansson, R. O., & Fletcher, W. L. (1990). Troubled relationships in later life: Implications for support. Journal 7, 451-463. Journalof ofSocial Socialand andPersonal PersonalRelationships, Relationships, 7, 451-463. Hogan, R. (1976). Egoism, altruism, and culture. American American Psychologist, Psychologist,31,31,363-366. 363-366. Hogan, R. (1983). A socioanalytic theory of personality. In M. M. Page (Ed.), Nebraska Nebraska Symposium theory andand research (pp. 55-90). Lin LinSymposiumon onMotivation: Motivation:Personality-Current Personality—Current theory research (pp. 55-90). coln: coln: University of Nebraska Press. Hogan, R., Jones, W. H., & & Cheek, J. M. (1985). Socioanalytic theory: theory: An alternative to armadillo psychology. In B. Schlenker (Ed.), Self and identity: ofof self in in identity:Presentation Presentation self social life. life. New NewYork: York:McGraw-Hill. McGraw-Hill. Holmes, J. G. G. (1991). (1991). Trust Trust and and th� the appraisal appraisalprocess processin in close close relationships. relationships. In In W. W. H. H.Jones Jones& & D. D. Perlman (Eds.), Advances Advances in in personal personalrelationships relationships(pp. (pp.57-106). 57-106).London: London:Jessica Jessica Kingsley Publishers. Holmes, J. G., & & Rempel, J. K. (1989). Trust in close relationships. In C. Hendrick (Ed.), Close Close relationships relationships(pp. (pp.187-220). 187-220).Newbury NewburyPark, Park,CA: CA:Sage. Sage. Jones, D. D. C. C. (1991). (1991). Friendship Friendship satisfaction satisfaction and and gender: gender: An An examination examination of of sex sex differences differences Jones, in Journal of ofSocial Socialand andPersonal PersonalRelationships, Relationships, in contributors contributors to to friendship friendship satisfaction. satisfaction. Journal 8, 167-185. 167-185. 8, Jones, W. W. H. H. (1988, (1988, July). July). Psychological Psychologicaland andinterpersonal interpersonal issues betrayal treachery. issues in in betrayal andand treachery. Jones, Paper presented presented at at the the Fourth International International Conference Conference on on Personal Personal Relationships, Relationships, Vancouver, British British Columbia, Columbia, Canada. Vancouver,

482 482

JONES, COUCH, COUCH, AND AND SCOTf SCOTT JONES,

Jones, W. H., & Burdette, M. P. (in press). Betrayal in close personal relationships. In A. Weber & J. H. Harvey (Eds.), Perspectives Perspectiveson onclose closerelationships relationships(pp. (pp.243-262). 243-262).New New York: Allyn & Bacon. Jones, W. H., Cohn, M. G., & Miller, C. E. (1991). Betrayal among children and adults. In K. J. Rotenberg (Ed.), Children 's interpersonal trust: Sensitivity to lying, deception, and and Children's interpersonal trust: Sensitivity to lying, deception, promise violations violations (pp. 118-134). New York: Springer-Verlag. Behavior Jones, W. H., H., & & Moore, Moore,T. T.L.L.(1987). (1987).Loneliness Lonelinessand andsocial socialsupport. support.Journal JournalofofSocial Social Behavior and Personality, Personality,2,2,145-156. 145-156. Kelley, H. H., & AA theory of of interdependence. & Thibaut, J. W. (1978). Interpersonal Interpersonalrelations: relations: theory interdependence. New York: Wiley. Toward understanding interper Larzelere, R. R. E., E.,& &Huston, Huston,T. T.L.L.(1980). (1980).The Thedyadic dyadictrust trustscale: scale: Toward understanding interpersonal trust in close relationships. Journal 595-604. Journalof of Marriage Marriageand andthe theFamily, Family,42,42, 595-604. Lawson, A. (1988). Adultery. Adultery, New York: Basic Books. (1972). Self-disclosure and two kinds of of MacDonald, A. P., Kessel, V. S., & Fuller, J. B. B. (1972). trust. Psychological PsychologicalReports, Reports,30, 30,143-148. 143-148. Metts, S. (1989). An exploratory investigation of deception in close relationships. Journal of 6, 6, 159-179. of Social Social and andPersonal PersonalRelationships, Relationships, 159-179. Miller, G. R., Mongeau, P. A., & Sleight, C. (1986). (1986). Fudging with friends and lying to lovers: of Social and Personal Personal Deceptive communication in personal relationships. Journal of Relationships,3,3,495-512. 495-512. Relationships, Monroe, P. (1990). A study satisfaction, andand commitment. Unpub studyof ofmarital maritalproblems, problems,marital marital satisfaction, commitment. Unpublished doctoral dissertation, University of Tulsa, Tulsa, OK. Gate, R. M. (1986). The developmental course of conflict in the marital Ponzetti, J. J., & Cate, dissolution process. Journal of of Divorce, Divorce, 10, 10,1-15. 1-15. Rempel, J. K., Holmes, J. G., & Zanna, M. P. (1985). Trust in close relationships. Journal of 95-112. Personalityand andSocial SocialPsychology, Psychology,49,49, 95-112. of Personality Rotenberg, K. J. (Ed.). (1991). Children's York: Springer-Verlag. Children's interpersonal interpersonaltrust. trust.New New York: Springer-Verlag. of Rotter, J. B. (1967). A new scale for for the measurement of interpersonal trust. Journal of Personality, Personality,35, 35,651-665. 651-665. Rotter, J. B. (1971). Interpersonal trust, trustworthiness, and gUllibility. gullibility. American Psycholo Psychologist, gist, 35, 35,1-7. 1-7. Rotter, J. B. (1980). Interpersonal trust, trustworthiness, and gUllibility. gullibility. American Psycholo Psychologist, gist, 35, 35,651-665. 651-665. Russell, D., Cutrona, C. E., & Jones, W. H. (1985). A trait-situational analysis of shyness. In W. H. Jones, J. M. Cheek, & S. R. Briggs (Eds.), Shyness: Shyness: Perspectives Perspectivesononresearch research and treatment treatment (pp. (pp. 239-249). 239-249). New New York: York: Plenum Plenum Press. Press. Schill, T., Toves, C., & Ramanaiah, N. (1980). Interpersonal trust and coping with stress. Psychological PsychologicalReports, Reports,47,47,1192. 1192. Scott, D. (1980). The The causal causal relationship between between trust trust and and the the assessed assessed value value of of management management 6, 6, 157-175. by objectives. Journal Journalof ofManagement, Management, 157-175. Simpson, J. A. (1990). Influence of attachment styles on romantic relationships. Journal of of 59,59, 971-980. Personality Personalityand andSocial SocialPsychology, Psychology, 971-980. ), Dimensions of Stack, L. (1978). Trust. In H. London & 1. J. E. Exner (Eds. (Eds.), of personality personality (pp. 561-599). New York: Wiley. PsychologicalReports, Reports, Steel, J. L. (1991). Interpersonal correlates of trust and self-disclosure. Psychological 68, 1319-1320. 68, L. S. S.(1974). (1974).Assumptions Assumptionsabout about human nature: A social-psychological approach, Wrightsman, L. human nature: A social-psychological approach. Monterey, CA: Brooks/Cole.

PART VI PART VI

DYNAMIC PERSONALITY PROCESSES

This Page Intentionally Left Blank

CHAPTER 20 CHAPTER 20

MOTIVES AND LIFE GOALS A. EMMONS ROBERT A. EMMONS

UNIVERSITY DAVIS UNIVERSITY OF OF CALIFORNIA CALIFORNIA,, DAVIS

Motivation has always been seen as central to personality psychology, although interest in motivational concepts has waxed and waned over the past 50 years. McAdams (this volume, Chap. 1) includes concern with motivation and dynamics as one of the the three distinguishing features of the field. Hogan (1986) argues arguesforcefully forcefully field. Hogan as are the "explanatory concepts par par excellence excellence in in personal personalthat motivational concepts are ity psychology "" (p. 50). Historically, concepts such as instinct, need, and drive carried the burden of motivational theorizing. As Pervin (1983) and Cofer (1981) noted though, the demise of drive theory in in the late 1950s tended to result in a diminution of interest in motivational concepts in general. In contrast, today theotheo rists and researchers are more willing to invoke motivational concepts into their descriptive and explanatory models. Terms such as tasks, goals, concerns, projects, the everyday motivational parlance. I have argued strivings, and motives are part of the elsewhere (Emmons, 1993) that the recent revitalization of the field field of personality chap. 1, (see also McAdams, this volume, chap. 1, and Runyan, 1990) has been due in large appUcation of the part to a resurgence of interest in motivational concepts. The application motive concept in predicting important life outcomes in domains outside of the laboratory and the heightened interest in studying behavior as it unfolds over time in natural contexts and in life transitions are two of the factors responsible for the long awaited return of motivation to the field of personality. This chapter begins with a brief review of the the literature on on motive dispositions, as it represents how the concept of motivation has been traditionally handled within personality psychology. The review will of necessity be selective-new selective—new and goalinnovative directions will be emphasized. Metatheoretical perspectives on goal directed behavior will be reviewed next. An analysis of the construct of goals in be considered, considered, and and both bothnomothetic nomotheticand and idiographic idiographic personality psychology will then be the most important trends has been the recent perspectives will be covered. One of the movement toward the "personalization" of motivation. Thus, of particular focus ILtNDBOOK HANDBOOK OF OF PERSONALITY PERSONALITY PSYCHOLOGY PSYCHOLOGY

48S 485

CoPYRIGHT COPYRIGHT C © 1997 1997 BY BY ACADEMIC ACADEMIC PlEss PRESS.. Au IIGtmI OF REPlODUcnON IN ANY FOIM IESEIYED. ALL RIGHTS OF REPRODUCHON IN ANY FORM RESERVED.

486 486

ROBERT ROBERT A. A. EMMONS EMMONS

will be idiographic approaches to personal goals. Since it would appear that this is where the action will be in the future, the time is right for a critical review of of this literature. Psychometric and methodological considerations considerations will be highlighted. After After briefly considering the relationship between goals and personality traits, the recommendations for future directions. chapter concludes with recommendations

I. L THE MOTIVE DISPOSmON DISPOSITION ApPROACH APPROACH Motivation in personality psychology psychology has been traditionally traditionally conceived of of in terms of stable individual of stable individual differences differences in in the the strength strength of of motive motive dispositions. dispositions. A A motive motive disposition refers to a class or cluster of affectively affectively tinged goals (McClelland, 1985). develMotives energize, direct, and select behavior. Considerable literatures have devel oped around what may be called the Big Three motives: Achievement, Affiliationl Affiliation/ of brief Intimacy, and Power. These motives are assessed via content analysis of imaginative stories produced in response to pictures similar to those used in the psychometric propprop Thematic Apperception Test (TAT). Although battles over the psychometric of this measure have been waged, there exists now a greater awareness of erties of the factors factors affecting affecting the the reliability reliability and and validity validity of in this this manner manner the of scores scores derived derived in (Spangler, 1992). (Spangler, 1992). Achievement motivation can be defined as a recurrent preference or readiness of attaining excellence-of excellence—of competing with in thought and behavior for experiences of a standard of excellence excellence (McClelland, 1985). Affiliation Affiliation motivation can be defined defined of establishing, as a recurrent preference in thought and behavior for experiences of maintaining, and restoring a positive affective relationship (Atkinson, Heyns, & Veroff, 1954). Intimacy motivation can be defined defined as a recurrent preference in . thought thought and behavior for and behavior for experiences experiences of of warm, warm, close, close, and and communicative communicative interac interactions others. The as an tions with with others. The preference preference is is for for interpersonal interpersonal exchange exchange as an end end in in itself itself means to rather rather than than aa means to another another end end (McAdams, (McAdams, 1980). 1980). Intimacy-oriented Intimacy-oriented individuals individuals are close interpersonal interpersonal relationships are concerned concerned with with establishing establishing and and maintaining maintaining close relationships and (McAdams & Constantian, and report report positive positive emotions emotions in in the the presence presence of of others others (McAdams Constantian, 1983). preference in 1983). Power Power motivation motivation can can be be defined defined as as a a recurrent recurrent preference in thought thought and and impact on behavior for behavior for experiences experiences of of feeling feeling strong strong and and having having impact on others. others. Among Among other other correlates, correlates, power-oriented power-oriented individuals individuals are are concerned concerned with with attaining attaining status status motivation, and and and prestige, prestige, choose choose as as friends friends persons persons low low in in power power motivation, and are are highly highly promiscuous (Winter & Stewart, 1978). As promiscuous in in heterosexual heterosexual relationships relationships (Winter Stewart, 1978). As has has been been motives, but argued argued (Winter, (Winter, 1991), 1991), these these are are not not the the only only important important human human motives, but they they do include many of the most important human goals and concerns. of the motive literature have recently appeared Several scholarly reviews of ((Jemmott, Jemmott, 1987; 1987; McAdams, McAdams, 1994; 1994;McClelland, McClelland, 1985; 1985;Smith, Smith,1992; 1992;Stewart Stewart&&Chester, Chester, 1982; Winter, 1996). The edited volume by Smith (1992) is an especially impressive collection and includes the scoring systems for the major motive systems as well as psychosocial orientations and cognitive styles. Therefore, a comprehensive review of this voluminous literature will not be attempted here. Instead, three recent trends of

CHAPTER 20 20 CHAFfER

MOTIVES AND GOALS

487 487

will be discussed: (1) the role of social motives in physical health and disease; (2) contextual factors and motive dispositions; and (3) the measurement of motives is placed on these three topics as they are typically in personal documents. Emphasis is directions ignored in textbooks on personality (Emmons, 1989a), yet represent three directions being pursued in this literature with great applicational promise.

A. Motive Motive Dispositions Dispositions and and Physical Physical Health Health A. The effect effect of the needs for affiliation/intimacy affiliation/intimacy and power on physical health has been studied extensively by David McClelland and his associates (see Jemmott, 1987, for for a review). The available evidence points to power motivation as having 1987, (especially when combined with power stressors) a deleterious effect on health (especially whereas affiliation/intimacy affiliation/intimacy motivation has a buffering buffering effect. However, it is not the individual effect effect of these motives that is critical but rather their joint influence influence as expressed in various motive patterns. Two of these patterns have been identified: identified: the Inhibited Power Motive Syndrome (IPMS) and the Relaxed Affiliative Affiliative SynSyn drome (RAS). drome (RAS). for power than need for for affilia affiliaThe IPMS is characterized by a stronger need for tion/intimacy, and a high degree of self-restraint, or activity inhibition. Activity frequency of negations (use of the word "not" or its inhibition is measured by the frequency contracted form) in TAT protocols. This syndrome, then, characterizes individuals it. This syndrome who are high in the need for power but are inhibited in expressing it. self-reported physical has been linked with high blood pressure, high degree of self-reported as measured measured by by secretory secretory immuno immunoillnesses, and and lower lower levels levels of of immunocompetence immunocompetence as illnesses, globulin A A (S-IgA; (S-IgA; Jemmott, Jemmott, 1987). 1987). The The release release of of stress stress hormones, hormones, cortisol Cortisol and and globulin epinephrine, due due to to chronic chronic sympathetic sympathetic activity activity is is the the likely likely culprit culprit responsible responsible for for epinephrine, these effects. It It is is important important to to note note that for power power is is not not in in itself related to to these effects. that need need for itself related sympathetic activation activation and and subsequent subsequent disease disease but but only it is is blocked blocked by sympathetic only when when it by internal inhibition inhibition or or external external stressors stressors (see (see also also Fodor, Fodor, 1984). 1984). The The relationship relationship internal functioning has not been limited to S-IgA; this between IPMS and immune system functioning motive pattern has also been related to natural killer cell activity. This is important in that S-IgA as a measure of immunocompetence is controversial (Stone, Cox, Valdimarsdottir, & Neale, 1987). 1987). In addition, manipulations designed to increase power motivation in the laboratory (viewing a war film) resulted in diminshed & Kirschnit, immunocompetence in individuals characterized characterized by IPMS (McClelland & 1988). 1988). Independent investigations (Fontana, Rosenberg, Marcus, & Kerns, 1987) is associated associated with with systolic systolic blood blood pressure pressure reactivity. reactivity. In In contrast contrast also report that IPMS is effects of power motivation, affiliation/intimacy to the deleterious effects affiliation/intimacy motivation seems effect on health. RAS is defined defined as having a higher need for to have a salubrious effect affiliation/intimacy than need for power, combined with low activity inhibition. affiliation/intimacy Several studies report that RAS is associated with better health, across a variety of outcomes. RAS RAS has has been been linked linked with with lower lower blood blood pressure, pressure, reporting reporting fewer fewer of illnesses, and better better immunologic functioning ((Jemmott, 1987). Manipulations Manipulations deillnesses, and immunologic functioning Jemmott, 1987). de of Mother Mother Teresa Teresa of of Calcutta) Calcutta) signed to to increase increase affiliation/intimacy affiUation/intimacy (viewing (viewing aafilm signed film of

488

ROBERT ROBERT A. A. EMMONS EMMONS

resulted in an increase of S-IgA (McClelland & & Kirschnit, 1988). More recent research, however, suggests a more complicated relationship between these social motives and health. McClelland (1989) suggested the need for for a more fine-grained analysis of power and affiliation/intimacy affiliation/intimacy motivation. Certain types of of affiliation affiliation and power motivation, labeled affiliative affiliative trust and agentic power, are associated with better health, whereas cynical trust and stressed power motivation, respectively, effects. have disruptive health effects.

B. B. Measuring Motives in Personal Documents

Research on motive dispositions is sometimes limited due to its reliance on the picture-story exercise for assessing motives. There are many purposes for which it would be desirable to have an estimate of motive strength in individuals for whom it is impossible to administer the TAT, for instance, deceased historical figures. Winter (1991) has developed a method of assessing motives in "running text" which allows researchers to code speeches, interviews, and other personal documents and for the motive dispositions. In short, it can be applied to biographical materials for any written text or written transcripts of spoken material that are at least in part imaginative. Measuring motives "at a distance" has been most widely applied in the realm 1970; Winter, 1987, particular. Winter (Donley & & Winter, 1970; of political psychology. In particular, 1988; Winter & & Stewart, 1978) has applied the scoring system to the inaugural 1988; U.S. Presidents with the purpose of predicting outcomes in office, such addresses of U.S. as greatness ratings, assassination attempts, and entry into war. Motive scores obtained in this way can also be used in psychobiographical studies, as Winter has demonstrated in his analyses of John Kennedy (Winter, 1991) and Richard Nixon & Carlson, 1988). 1988). In In another another interesting interesting application, application. Winter Winter (1987) (1987) showed showed (Winter & presithat congruence between presidential and societal motive profiles predicted presi dential appeal, defined in terms of electrol success. running text scoring system, its psycho psychoA complete overview of the integrated running found in metric properties, and a review of studies that have employed it can be found (1991). Winter marshals a good deal of evidence for for the validity of the Winter (1991). integrated running text scoring system. C. Studying Studying Motivation Motivation in in Context Context C. A third emerging theme in the study of motivation is what might be termed as "contextual" perspective. A contextual perspective emphasizing social-historical and social structural factors factors has been emerging in developmental, personality, and Ryff, 1987). 1987).AAcontextual contextualperspective perspective Caspi, 1987; 1987;Ryff, social psychology (Bolger, 1988; 1988; Caspi, applied to motivation implies that motivational processes cannot be understood without taking into account a variety of contexts within which a person is embedded. recognized the importance of contextual Motivational theorists have long recognized of personality emphasized not only or situational factors. Murray's (1938) theory of

CHAPTER CHAPTER 20 20

MOTIVES MOTIVES AND AND GOALS GOALS

489 489

personality needs but the environmental press relevant to those needs. A leading spokesperson spokesperson for for the contextual approach to motivation has been Joseph Veroff. Veroff (1983, 1986; 1986; Veroff Veroff & Smith, 1985) has advocated the necessity of taking Veroff into account the influence of historical, cultural, developmental, organizational, and interpersonal contexts in understanding patterns of constancy and change in social motives. Veroff Veroff (1983) provided a contextual interpretation of achievement and affiliation motives and persuasively demonstrated how the meaning of a motive affiliation Veroff and Feld (1970) examined the may vary depending upon the five contexts. Veroff ways in which the three social motives interacted with three important social roles (marriage, parenting, and work) within a dynamic, contextual framework. In such affect the perceptions of these roles, but the a framework, not only do motives affect Veroff's work roles themselves shape the nature and expression of the motives. Veroff's points to the importance of addressing how motivational content and processes are (1987) provides provides aa pointed pointed example example influenced by by aa variety variety of of contextual contextual factors. factors. Ryff Ryff (1987) influenced of how how cultural cultural values values in in Japan Japan and and the the United United States States influence influence child-rearing child-rearing prac pracof tices, which which in in turn turn lead lead to to personality personality differences differences in in such such characteristics characteristics as as auton autontices, omy versus versus interdependence. interdependence. omy different contextual approach to motivation is represented in McA quite different Mc Adams' (1985b, 1990, 1993) theory of identity as a life story. Identity, that which future, lends coherence, unity, and purpose of binds together past, present, and future, of personality and allows adaptation to changing contexts. One's identity is organized self. Imagoes Imagoes are the the . around around imagoes-idealized imagoes—idealized and personified personified images images of of the self. two fundamen fundamencentral elements of a person's identity and represent instantiations of two (power/mastery/separation) and commu commutal thematic lines in people's lives: agency (power/mastery/separation) nion (intimacy/surrender/union). Imagoes are broad, superordinate constructs which encompass interpersonal styles, values and beliefs, and personal needs and McAdams isis most most concerned concerned with with the the historical historical context; context; motives (McAdams, 1985a). McAdams framework for conceptualizing the develthe life story model of identity provides a framework devel from birth to death. opment and history of the person from

_

D. n. METATHEORETICAL PERSPECTIVES ON GOALS

metatheoretical offer a framework framework within which to interpret Several meta theoretical perspectives offer goal-directed action. This review will focus on three of them: control theory, living systems framework, framework, and action theory. The basic ideas of each of these will be pre presented. A. Control Control Theory Theory A. One of the most simple yet potentially powerful powerful models for representing goals and goal-directed behavior is control theory (Carver & Scheier, 1981, 1981, 1982; 1982; Hyland, 1987, 1988; Powers, 1973, 1978). According to control theory, behavior is seen as 1987,1988; 1973,1978). feedback loop, a discrepancy reduction process operating in terms of a negative feedb_ack

490 490

ROBERT EMMONS ROBERT A. A. EMMONS

where individuals act to minimize the discrepancy their present present condition where discrepancy between between their and components in in standard and aa desired desired standard standard or or goal. goal. There There are are five five components standard control controltheory models. Goals or standards are checked against feedback theory models. Goals or standards are checked against feedback received received by by aa sensor; standard, outputting sensor; aa comparator comparator compares compares the the feedback feedback with with the the standard, outputting the the result mechanism which result to to aa decision decision mechanism which activates activates an an effector effector system system to to operate operate on on environment to further further reduce the discrepancy. From standpoint of of this the environment From the standpoint feedback are the crucial components: feedback feedback is information information chapter, goals and feedback that provides individuals with a basis for for decisions about about changing either the course of their behavior or the goals themselves. of Control One charge Control theory theory has has been been criticized criticized on on two two main main grounds. grounds. One charge is is that that it it ignores ignores the the role role of of emotion. emotion. Carver Carver and and Scheier Scheier (1990) (1990) have have attempted attempted to to eradicate eradicate of emotion within control control theory. A A second this charge by addressing the role of that this criticism is that control theories are mechanistic. Powers (1978) argues that beUef is mistaken mistaken as control theory can account account for for dynamic, flexible flexible aspects of of belief for the development development of behavior. This second criticism has served as an impetus for of framework, which is discussed next. the living systems framework,

B. Living Systems Framework formulations have proposed proposed system-theoretic system-theoretic conceptions of of person personSeveral recent formulations ality in which goals play a major instance, Powell, Royce, Royce, and Voorhees major role. For instance, complex hierarchical hierarchical information information processing system (1982) depict personality as a complex major component. Perhaps the most impressive (CHIPS) with goal directedness as a major of these efforts efforts is the living systems framework framework (LSF; D. H. Ford, 1987; of 1987; D. H. formulation is the the idea that Ford & Ford, 1987). Central to this formulation that people are active organisms who intentionally set and strive for for goals. There There is a strong emphasis on of organization. Goals organize and direct the activity goal directedness at all levels of of the system. Much of of their theory can be viewed as a formalization formalization and elaboration elaboration of of of control-theory ideas. A classical control system (mechanistic) model is trans transformed into a living system by adding self-organizing self-organizing and self-constructing self-constructing functions. functions. formed The functioning functioning of of a control system involves five interrelated interrelated and interdependent interdependent information collection, regulatory, control, and action. The processes: directive, information self-regulating in that the activity of of the system is part part of of the input system is self-regulating controlling the functioning functioning of of the system. Of what is labeled the directive Of greatest relevance for for a chapter on goals is what function function: the personal goals and aspirations of the individual. The directive function includes the formulation, formulation, revision, and elaboration of of goals. Goals in the LSF are defined of desired and undesired undesired conditions. Subordi Subordidefined as cognitive representations of function are the control (planning and problem problem solving) and nate to the directive function Ford and Ford regulatory (progress monitoring) functions. The volume by D. H. Ford and in professional professional (1987) illustrates the utility of the LSF in designing research and of the framework framework is broad-problems broad—problems in social application. The potential scope of competence, emotion development, prosocial behavior, and clinical and counseling

CHAPTER CHAPTER 20 20

MOTIVES MOTIVES AND GOALS GOALS

491 491

psychology are just a few of the domains in which the LSF can lead to greater understanding. c. C. Action Theory

Action theory, an information-processing information-processing approach to motivation motivation and goals, has 1985; Halisch & Kuhl, 1987; 1987; Semmer & been gathering momentum (Frese & Sabini, 1985; 1991; Kuhl Kuhl & & Beckman, Beckman, 1985). 1985).Action Action theory theory was was devel develHeckhausen, 1991; Frese, 1985; 1985; Heckhausen, oped largely in Germany and treats goal-directed action as the unit of analysis. Action is contrasted with behavior, the latter not being goal directed. The major major task of action theory is to fill the gap between the cognitive representation of of desired states and the execution of acts designed to achieve the desired states. In four questions need to be answered to the agenda order to close this gap, at least four (1985): (1) How do cognitive structures arouse set forth by Kuhl & Beckman (1985): tempomotivational states? (2) How does one understand the processes underlying tempo ral changes in these motivational states? (3) What are the mechanisms that mediate the formation and enactment of intentions? (4) What are the mechanisms that mediate the final execution of a sequence of behaviors? Kuhl (1985a) proposed that action versus state orientation is an important determinant of action control. A person is action oriented when his or her attention future state, the is focused on some aspect of the present state, some aspect of a future discrepancy between the two, and at least one action alternative that can reduce four elements is missing, the discrepancy. State orientation exists when one of the four for example, when a person dwells on the current state without reference to the for future or fails to develop action alternatives that could reduce the discrepancy. This future for the lack of of enactment enactment of distinction is important in that it is a primary reason for of an intention. A high degree of action orientation directs and controls cognitive processes the facilitate the enactment of the intention. State orientation impedes performance since it involves repetitive and dysfunctional dysfunctional focusing on fixed fixed aspects performance of the situation. Frese and Sabini (1985) provide multiple applications of of action theory, with counterpoint commentaries provided provided by researchers outside the paradigm. Indus Industrial/organizational, educational, and clinical cUnical and counseling counseling applications of action theory are presented. Critics of action theory (Klinger, 1985; 1985; Warren, 1987) contend that it is lost in thought: it is overly rational, and the role of dynamic and affective affective processes are minimized. Thus, it is unlikely to ever address motivational processes such as goal conflict, since in an expectancy-value framework, framework, it will be a rare case that compet competing goals will have equal arithmetic valences; thus there will be no conflict. Another from the paradigm has focused nearly exclusively criticism is that research emanating from has focused on trivial laboratory behavior. Lastly, action theory is concerned nearly exclusively with process, and has been silent with regard to motivational content. Kuhl (1985b) has attempted to answer these criticisms.

492 492

ROBERT ROBERT A. A. EMMONS EMMONS

m. in. GOAL ApPROACHES APPROACHES OF PERSONALITY A. Nomothetic Approaches to Goals Early efforts efforts into the inquiry of a phenomena are often often taxonomic in nature. Therefore, it is not surprising that one objective of of the research research using normatively constructed goal lists has been in constructing a taxonomy of of human goals. Several empirically derived taxonomies reflecting reflecting content categories into which goals are organized have recently been uncovered. Much of this work is summarized in Novacek and Lazarus (1990) and M. E. Ford (1992). The general procedure in this of goals and have subjects research is to present subjects with a standardized list of rate these goals on a preselected dimension, typically importance or relevance. subjected to factor factor or cluster analysis to Intercorrelations of these ratings are then subjected Often these goals are based on prior lists generated by a derive basic groupings. Often sample possessing similar characteristics to the sample doing the rating, but in at author without explicit least one case "goals are generated intuitively by the senior author recourse to prior classifications" (Wicker, 1984, p. 288). Table I shows the goal 288). categories that have emerged in a number of different different investigations. There seems to be evidence of five general factors, which have been labeled (1) Enjoyment, Enjoyment, (2) Self-Assertion, (3) Esteem, (4) Interpersonal, and (5) Avoidance of Negative fact that these clusters reemerge in several studies points to possibly Affect. The fact an uncovering of a basic motivational structure. These categories will need to be replicated in cross-cultural samples before such a claim can be made, however. It **basic motives" postu postuis encouraging, though, that these factors are similar to the "basic lated by several theories of motivation, including those of of Maslow (1970) and Hogan (1983).

B. Idiographic Goal Approaches

Nomothetic approaches such as those just noted are appealing because of their After all, the logic amenability to comparison, both across individuals and studies. After of taxonomy building requires normalization across individuals. However, a number number of other investigators have argued quite forcefully forcefully that any attempt to understand an individual's goal system in terms of a consensually defined defined a priori list of of goals is likely to be highly misleading and restrictive, since individuals strive for for goals in Thus, there there has has been been an an increasing trend trend recently toward toward highly individualized ways. Thus, the adoption of more chrcumscribed, circumscribed, idiographic units to account for for human motiva motivation. Without a doubt, the most important development in motivational approaches to personality has been the recent trend toward the personalization of motivation. The constructs of "current concern" (Klinger, 1975), "personal project" (Little, (Em1983), "life task" (Cantor, Brower, & Korn, 1985), and "personal striving" (Em for personality personaHty mons, 1986) have been proposed as idiographic goal-directed units for An overview of each of the four major major constructs is considered next. psychology. An

TABLE I TABLE

Summary Summary of of Studies Reporting Common Goal Clusters Clusters

Factors Study

Enjoyment Enjoyment --

Pervin

(1983) (1983) Emmons and Diener (1986) (1986) Novacek and Lazarus Lazarus (1990) (1990) Wicker (1984) (1984) Buhler (1964 (1%4)) M. E. E. Ford Ford and and Nichols Nichols (1987) (1987)

* Second-order Second-order factor. factor.

•

-

Self-Assertion Self-Assertion - - -

-

-

Esteem Esteem

Interpersonal Interpersonal

Avoidance of of Negative Negative Affect Affect Avoidance

-

RelaxationlFun RelaxationIFun

Aggression/Power AggressionIPower

Self-Esteem Self-Esteem

Affection/Support AffectionISupport

Anxiety/Threat AnxietyIThreat Reduction

Sensation-Seeking

Achievement/Power Achievement/Power

Personal Growth Growth Personal

Affiliation Affiliation

Stress Stress Avoidance Avoidance

Exploration/Play ExplorationIPlay

Competitive Competitive Ambition

Individual Individual Striving· Striving*

Interpersonal Concern Concern Interpersonal

Harmony Seeking· Seeking*

Social Social Relationships

Subjective Subjective Organization

Hedonistic Hedonistic

Self-Achievement Self-Achievement

Need Satisfactions Satisfactions

Creative Creative Expansion Expansion

Affective Affective

Task Task

Upholding of of Internal Internal Order Cognitive Cognitive

Interpersonal Interpersonal

Need Satisfaction Satisfaction Need

Self-Limiting Self-Limiting Adaptation Adaptation

494 494

ROBERT ROBERT A. A. EMMONS EMMONS

C. Four Four Major Major Constructs Constructs C. i. Current Current Concerns Concerns 1. Klinger (1975, 1977, 1987) has argued that experience is organized around the pursuit of incentives, and that these pursuits are represented by a "current concern." A current concern is a hypothetical motivational state in between two points in time: the commitment to a goal and either the consumation of the goal or the from it. This hypothetical state guides a person's ongoing thoughts, disengagement from emotional reactions, and behavior during the time it is active. Klinger developed dissatisfaction with the failure of the the notion of a "current concern" out of dissatisfaction motive dispositions to predict spontaneous thought content. However, there is no assumption about the representation of the concern in consciousness, and it is assumed that for the majority majority of the time, the concern is not reflected reflected in on-line cognitive processing. People simultaneously possess a number of current concerns, different concern for for each goal a person is committed to. The range as there is a different of potential concerns is diverse, as each individual possesses an idiographic set frequently changes. Examples of current concerns are going on a trip, keeping which frequently a dentist appointment, losing weight, and maintaining a love relationship. Other examples of current concerns along with examples of the other personal goal con conII. Concerns may be defined narrowly or broadly, and structs are shown in Table II. from a few seconds to a life time. The primary purpose of the may last anywhere from current-concern construct construct has has been been as as aa carrier carrier for for motivational motivational influences influences on on current-concern thought processes. processes. Klinger, Klinger, Barta, Barta, and and Maxeiner Maxeiner (1981) (1981) found found that that the the degree degree to to thought which are valued, valued, committed to, and and threatened threatened predicted predicted the the which current current concerns concerns are committed to, frequency with with which concerns are are thought thought about. According to to Klinger, Klinger, frequency which these these concerns about. According triggered by environmental cues that are related to current concerns, thoughts are triggered is the emotional properties of concerns that affect affect the processing of these cues. and it is Other research has shown that words related to current concerns surreptitiously left side of a computer screen interfered interfered with a lexical decision decisionpresented on the left concernmaking task (Young, 1988), indicating the automatic quality with which concern affect cognition. Klinger (1989a) reviews research showing that concern related cues affect findings relatedness and emotional arousingness of a word predict its recall. These findings are taken as evidence that current concerns influence cognitive processing because the concern state predisposes the individual to react with emotional arousal to cues associated with the concern. currentWhile initially serving as a link between motivation and cognition, the current contexts. The The concept concept has has been been framework has also proven useful in applied contexts. concern framework extended into the realms of depression (Ruehlman & Wolchik, 1988), alcoholism (Klinger & Cox, 1986), and work satisfaction satisfaction (Roberson, 1989).

2. Personal Pro Projects 2. jects A similar though independently developed concept is the personal project (Little, 1983, 1989; Palys & Little, 1983). 1983, 1987, 1987, 1989; 1983). Rooted in Murray's (1951) concept of a serial program, personal projects are "an interrelated sequence of actions intended

CHAPTER 20 20

495 495

MOTIVES AND GOALS MOTIVES

TABLE II II

Examples of of Personal Goals Goals Current concerns

Personal Personal projects projects

Putting out out the the cat Putting

Eating lunch lunch Going camping next weekend Finishing Finishing the the writing writing of of aa book book Getting Getting the the garden garden planted planted Gaining Gaining aa promotion promotion Eating dinner tonight Taking Taking aa skiing skiing vacation vacation Maintaining Maintaining aa marital marital relationship relationship Buying Buying aa fly swatter swatter Getting Getting closer closer to to God God Life Life tasks tasks

Coping Coping with with cancer

Sailing the the Atlantic Atlantic Sailing Making aa dress Making Learn how how to to ski Learn Graduate Graduate from from University University Get Get Tom Tom to to stop stop biting biting his his nails nails Not Not losing losing my my job job Becoming Becoming more more responsive responsive to to Pierre Pierre Revenging Revenging my my father's father's death death Personal Personal strivings strivings

Maturing Maturing beyond beyond my my high high school school mentality mentality

Make attractive women women notice notice me me

Finding Finding aa girlfriend girlfriend

Do Do as as many many nice nice things things for for people people as as

Establishing future future goals goals Getting good good grades grades Planning for the future future Developing an an identity identity Being productive productive at at work work Being Making Making friends friends Being on on one's own own away away from from family family Managing Managing time time

I can can Get to to know know new new people people Maintain Maintain an an above above average average beauty beauty Force to be intimate in Force men men to be intimate in relationships relationships Have Have as as much much fun fun as as possible possible Avoid Avoid being being dependent dependent on on my boyfriend boyfriend Make it it appear appear that I am am intelligent intelligent Avoid Avoid arguments arguments when when possible possible Make Make life life easier easier for for my my parents parents

to achieve a personal goal" (Palys & 1223) .. They have also been & Little, 1983, p. 1223)..They defined as "extended "extended sets sets of ofpersonally personallyrelevant relevantaction" action"(Little, (Little,1989, 1989,p.p.15). 15).Personal Personal projects are things that people think about, plan for, carry out, and sometimes, but not always, complete (Little, 1983). Everyday activity is organized around these prom with withBrad," Brad," personal projects. Examples of personal projects are "going to the prom job," and "shopping for the holidays" (Little, 1983). "finding a part-time job," 1983). The concept was developed and promoted by Little as an interactional unit linking the individual to his or her sociocultural context. Little (1987) stresses three types of of contexts that are necessary for understanding action: the intentional context, in which the purposes underlying the projects are discerned; the systemic context, the each other within within aaproject project system; system;and andthe theecological ecologicalcontext, context, relation of projects to each the environmental and historical milieu miUeu in which the action takes place. Interestingly, the term "personal project" was first first used by Nuttin (1957) to describe "fundamental dynamic orientations" of personality personaUty and were defined as "personal tasks to be accomplished" (p. 194). These projects, according to Nuttin, represent the personalized version of needs and account for the unity of personality functioning. The personal project concept has been applied in studies of subjective well-being. For instance, Palys and Little (1983) found that individuals who were

496 496

ROBERT EMMONS ROBERT A. A. EMMONS

involved in short-term important projects that were highly enjoyable and moderately difficult difficult were more satisfied with their lives than individuals who possessed projects that were longer range in scope and from which they derived little immediate interpersonal support enjoyment. Ruehlman and Wolchik (1988) reported that interpersonal well-being and distress, and hindrance in personal project pursuit were related to well-being respectively, and that hindrance was also related to low well-being. 3. Lif Lifee Tasks Tasks

Cantor and her colleagues (Cantor, 1990; Cantor et aI., al., 1985; Cantor & Kihlstr6m, Kihlstr5m, 1987, 1989; Cantor & Langston, 1989; 1989; Cantor, Norem, Niedenthal, Langston, & developed the concept concept of of life life tasks, defined as "problems Brower, 1987) recently developed
20 CHAPTER 20

MOTIVES AND GOAUl GOALS

497

the idiographic nature of the personal striving. striving. A personal striving is a unifying unifying construct—it unites what may be phenotypically different different goals or actions around construct-it a common quality or theme. Thus, a striving striving can be achieved in a variety of ways and satisfied via any one of a number of concrete goals. The personal striving has been primarily employed to account for for individual differences differences in in psychoconcept has psycho logical and physical well-being. For example, Emmons and King (1988) found found that conflict within and between strivings was associated with psychological distress and physical illness, both concurrently as well as prospectively. prospectively. Another line of research research (Emmons & King, 1989) has investigated the relationship between personal striving differentiation or complexity and affective affective reactivity. Reactive individuals (those differentiation experiencing intense and variable moods) were found found to possess a highly differenti differentipresents aa thorough thorough ated (more unrelated unrelated strivings) striving system. Emmons (1996) presents review of the literature on personal strivings and subjective well-being.

C. Distinguishing Distinguishing between between the the Personal Personal Goal Goal Units Units C. Before turning to the assessment of these personal goal units, it is useful useful to discuss differences among them. In the framework framework of control theory, all similarities and differences from which initiate and organize can be viewed as reference values, discrepancies from different representation of goals: goals action. Each construct involves a somewhat different as concerns, goals as projects, goals as tasks, and goals as strivings. Each construct Ust of these is both idiographic and nomothetic. Individuals possess a personalized list goals, yet the goals can be appraised on common dimensions (discussed later) that formlessness permit nomothetic comparisons and generalizations, thus avoiding the formlessness of strictly idiographic approaches. KUnger (1989b) has taken the initial step in highlighting the differences differences be beKlinger & Zirkel, 1990, 1990, for a comprehensive analysis tween these constructs (see also Cantor & refer to hypothetical underlying of cognitive units of personality). Current concerns refer states and thus connote a continuing dispositional state. Personal projects refer refer not to hypothetical states but to a set of related acts over time-the time—the observable behavior that presumably corresponds to a concern. They are not what the person has, but rather what the person does. Life tasks focus on nontrivial problems that the individual wishes to solve. These tasks are rooted in developmental stages and could be viewed as a subset of concerns or projects that are made salient by life transitions. Read Read and that life transitions. and Miller Miller (1989) (1989) point point out out that life tasks tasks capture capture the the organizations organizations of in of goals goals and and strategies strategies that that are are organized organized around around specific specific periods periods and and contexts contexts in people's less suited individual differences people's lives, lives, and and are are less suited for for describing describing individual differences more more gener generally. ally. A A personal personal striving, striving, defined defined as as aa class class of of goals goals that that is is characteristic characteristic for for aa particular particular person, person, describes describes enduring enduring and and recurring recurring personality personaUty characteristics. characteristics. Un Unlike current concerns and life tasks, the personal striving approach was not devel developed oped as as aa reaction reaction against against the the motive motive disposition disposition approach, approach, and and in in fact fact is is quite quite compatible with it (Emmons, 1989b). Hyland's (1988) analysis of reference criteria within control theory theory is useful useful in helping further further distinguish between these constructs. Life tasks represent end-

498

ROBERT A. EMMONS

state reference criteria. An end state is a completed piece of work with a definite definite termination point. Once one has graduated from from college, it need not be negotiated again. Concerns and projects may also serve as end-state reference criteria, though they need not be limited to these. Personal strivings are more likely to represent doing or being reference criteria: a recurring and enduring concern with a class of of goals rather than the attainment of a particular end state. They are not normally successful or unsuccessful unsuccessful experiences. For example, a person who terminated by successful difficult situations" situations*' is not likely to be content with a single is trying to "see the best in difficult from then on adopt a pessimistic orientation of accomplishment of this striving and from of situations. Personal projects projects or life tasks can be initiated by noting a discrepancy between affairs and a desired outcome. For example, a person's ideal self, a current state of affairs **lose 20 20 pounds" pounds" to be slim and athletic, can be a reference value, and a project of "lose may be initiated in the service of it. While such distinctions between the units are often obscured. possible in principle, in reality the boundaries between them are often Many if not most of the examples shown in Table II could be switched around without doing much damage to the construct. Part of the difficulty difficulty lies in the ambiguously defined defined category breadth and time frame frame of the respective constructs. Breadth of the construct (narrow and concrete versus broad and abstract) is important in that it defines the range of outcomes that are acceptable as goal satisfied attainments. For example, "trying to get to know others better" may be satisfied from "getting Sue to go to the dance." The majority majority by a wider range of outcomes from of the goal units are not explicit with reference to category width. Life tasks may vary in scope, from from "becoming a good person" to "getting good grades" (Cantor & Langston, 1989). micro1989). According to Little (1987), projects can be defined at a micro behavioral and a macro-intentional level (e.g., "go to church on Sunday" versus "explore my religious convictions"). Projects, tasks, and concerns can all vary between these extremes. Strivings, on the other hand, possess greater category breadth, as they are postulated to occupy a higher level in the hierarchy of motiva motivational control (Emmons, 1989c). D. Assessing Assessing Personal Personal Goals Goals D. 1. Generating Generating Goals

The assessment of personal goals begins with having respondents freely freely generate strivings. In this initial step, the definition definition a list of their concerns, tasks, projects, or strivings. of the construct is given, usually with examples, and subjects write down as many specified time period, ranging from from 10 min (projects) to goals as they can within a specified (strivings). Klinger's (1987) interview interview questionnaire (IntO) (IntQ) requires requires a few days (strivings). subjects to list their current concerns in 14 major major life areas (friends, employment, family, and so so forth). Some ofthe of the categories are further further subdivided and are accompa accompanied by illustrative concerns.

20 CHAPTER 20

MOTIVES MOTIVESAND ANDGOAI.'i GOALS

499 499

In personal projects analysis, subjects are told their personal projects are activities and concerns that people have and are provided with examples, such as "complete "complete my English essay" and "getting more outdoor exercise" (Little, 1983). of their personal projects that they are engaged in or They are instructed to list all of are thinking about at the time, and are told that these projects should represent necessarily major life projects. everyday activities and concerns and not necessarily The life life task methodology requires subjects to list their current life tasks, defined defined as things they felt felt they were working on in their lives. More specifically, specifically, respondents in one study were asked to focus on "the areas to which you have been and expect to be directing your energies" (Cantor et al., aI., 1985, p. 326). They are then told to categorize the tasks into one of six normative life task categories: three academic and three interpersonal. In the case of of personal strivings, individuals are given the definition definition of of a personal striving as "the things that you typically or characteristically are trying to do in your everyday behavior." They are then provided with several examples, such of as "trying to persuade others one is right" and "trying to help others in need of help." It is stressed that these strivings are phrased in terms of of what the person is of whether the person is actually successful. They are "trying" to do, regardless of also instructed that the strivings may be either positive or negative, and that the striving must refer to a repeating, recurring goal, not to a one-time concern. 2. Personal Personal Goal Goal Dimensions Dimensions 2.

of the personal goals, respondents are asked to rate each goal Following elicitation of on several dimensions. The dimensions used in any one study are derived from a of sources. An analysis of the motivational literature typically results in number of instrumenthe inclusion of such key goal attributes as value, expectancy for success, instrumen tality, and commitment. Other dimensions are included because of their presumed relevance to the particular study. While current concerns, personal projects, life tasks, and personal strivings are idiographic, they are also nomothetic. That is, the specific list of concerns, projects, and tasks is unique to each individual. Yet these concerns, be compared concerns, projects, projects, and and tasks tasks can can be compared along along such such nomothetic nomothetic dimensions dimensions as value, as value, expectancy expectancy for for success, success, complexity, complexity, and and difficulty, difficulty, and and in in this this sense sense compari comparisons can be made. In sons across across individuals individuals can be made. In addition, addition, it it is is these these nomothetic nomothetic properties properties which which tie tie these these concepts concepts to to cognition, cognition, emotion, emotion, and and action. action. With With these these common common properties, properties, general general statements statements tying tying goals goals to to affective, affective, cognitive, cognitive, and and behavioral behavioral outcomes content of outcomes can can be be made made which which are are likely likely to to transcend transcend the the idiographic idiographic content of these units. The dimensions that have been used are shown in Table III. There are these units. The dimensions that have been used are shown in Table III. There are none none listed listed for for life life tasks tasks since since these these duplicate duplicate those those used used in in personal personal projects projects analysis. analysis. Several investigators have examined the factorial structure of these Several investigators have examined the factorial structure of these dimensions. dimensions. A A summary of these findings is is presented in Table IV. In general, three to five dimen of these dimensions emerge. Although the labels and loading variables vary somewhat somewhat from study to study, the most robust dimensions appear to embody (1) the degree of of commit-

soo 500

ROBERT ROBERT A. A. EMMONS EMMONS

TABLE m TABLEm

Goal Dimensions Current Current concern concern

Personal Personal project project

Personal Personal goal goal

Personal Personal striving striving

Commitment Commitment Positivity Positivity Loss Loss Negativity Negativity Instrumentality Instrumentality Probability Probability of of success success Confidence Confidence in probability probability Nearness in time Nearness time Time Time available available Causal attribution attribution Causal Sense of of drain Sense

Importance Importance Challenge Challenge Difficulty Difficulty Negative Negative impact impact Positive Positive impact impact Visibility Visibility to others others Progress Progress Challenge Challenge Enjoyment Enjoyment Absorption Absorption Time adequacy adequacy Time Initiation Initiation Stress Stress Control Control Value Value congruency congruency Self-identity Self-identity

Importance Importance Positivity Positivity Probability Probability Difficulty Difficulty Activity Activity Self-effi cacy Self-efficacy Attribution Attribution Clarity Clarity Opportunity Opportunity Progress Progress Challenge/threat Challenge/threat Steps Investment Investment

Commitment Commitment Value Value Probability Probability Difficulty Difficulty Effort Effort Confi dence Confidence Causal Causal attribution attribution Clarity of Clarity of means means Environmental Environmental opportunity opportunity Past Past attainment attainment Satisfaction with with progress progress Satisfaction Social desirability Social desirability Importance Importance Ambivalence Ambivalence Probability of of success success if no action action Probability Impact Impact

ment/investment in the goal, (2) the degree to which the goal is perceived as stressful/challenging, and (3) the anticipated outcome/reward outcome/reward of the goal. stressful/challenging,

3. 3.

Goal Instrumentality Matrix Goal

A number of investigators have pointed to the possibility of of assessing the degree of of inter-goal conflict within the person's goal system. This can be accomplished by constructing for each person a matrix in which both the rows and columns list the person's goals. Respondents are asked to rate the degree of conflict/instrumentality conflict/instrumentality between each pair of goals, until the entire matrix isfilled filled out. Each goal is in effect effect rated twice, in terms of the effect effect that it has on other goals and the effect effect that other goals have on it. While asymmetrical affects affects are potentially potentially possible, in in reality the effects effects are nearly symmetrical, with correlations between conflict conflict generated and contlict conflict received typically close to .8 (Emmons & King, 1988). For the matrix as a whole, the average amount of contlict conflict or instrumentality in the person's goal system is determined and is used as a variable in between-subject between-subject analyses.

E. Psychometric: Psychometric Properties Properties E. differ in the degree to which they have been concerned concerned The personal goal approaches differ with formal psychometric considerations such as reliability and validity. Klinger al, 1981) makes a compelling case for for why traditional reliability (1987; Klinger et aI., estimates are only partially appropriate for for both the content and the dimensions different criteria need to be developed than is typical of personal goals. Somewhat different for traditional measures of personality (i.e., trait inventories). As Klinger (1987)

CHAPTER 20 20

501 501

MOTIVES MOTIVES AND AND GOALS GOALS TABLE IV IV

Goal Dimension Factors Goal Dimension Factors Factors (personal Factors (personal projects)" projects)" Strain Strain

Mastery Mastery

Self-involvement Self-involvement Value Value

Difficulty Difficulty

Progress Progress

Self-identity Self-identity

Stress Stress

Time Time

Absorption Absorption

Challenge Challenge

Enjoyment Enjoyment

Impact

Initiation Initiation

Control Control

Negative Negative

Outcome Outcome

Outcome Outcome

Factors tasks)b Factors (life (life tasks)*

Anxiety/Absorption Anxietyl Absorption Challenge Challenge

Personal Responsibility Responsibility Personal

Rewardingness Rewardingness

Initiative Initiative

Enjoyment Enjoyment

Progress Progress

Absorption Absorption

Control Control

Absorption Absorption

Time Time spent

Progress Progress

Stress Stress

Difficulty Difficulty

Importance Importance Factors tasks y Factors (life (life tasks)*^

Importance/Absorption Importancel Absorption

Stressfulness Stressfulness

Time spent Time spent

Stress Stress

Importance Importance Challenge Challenge Other's view Other's view

Challenge Challenge

Absorption Absorption

Reward/Control Reward/Control Progress Progress

Difficulty Difficulty

Control Control

Initiative Initiative

Enjoyment Enjoyment

Factors strivings)d Factors (personal (personal strivings)** DegreelIntensity Degree/Intensity Value Value

Progress Progress

Commitment

Attainment

Importance

Attainment Attainment

Effort Effort

Internality Internality

Difficulty(-) Difficulty( -)

Clarity Clarity

if no no action Success probability probability if Success action

Confidence Confidence

Effort(-) Effort( -)

Ambivalence(-) Ambivalence( -)

"*»Ruehlman Ruehlmanand andWolchik, Wolchik,1988. 1988.

b * Cantor Cantorand andLangston, Langston, 1989, 1989,Wave Wave 1.1. ^Same Sameas asb,b,Wave Wave3.3. d **Emmons, Emmons,1986. 1986.

Desirability Desirability

Environmental opportunity opportunity Environmental

Probability Probability

Probability Probability

C

Ease Ease

Success

Desirability Desirability Ambivalence(-) Ambivalence( -)

S02 502

ROBERT A. EMMONS EMMONS

and argue, lack need not and Nesselroade Nesselroade (1987), (1987), argue, lack of of stability stability (a (a psychological psychological process) process) need not imply lack lack of (a psychometric imply of reliability reliability (a psychometric situation). situation). Internal Internal consistency consistency estimates estimates are not wholly appropriate appropriate either. Since there is no assumption of of homogeneity homogeneity of of goal content, there goal content, there is is no no reason reason to to expect expect high high internal internal consistencies. consistencies. Although Although such values have such values have been been computed computed for for various various goal goal dimensions dimensions and and have have been been shown shown to be high (Emmons & King, 1989), the meaning of these is not clear-cut. to be high (Emmons King, 1989), the meaning of these is not clear-cut. acknowledging these difficulties, difficulties, some efforts efforts have been made at esti estiWhile acknowledging of these measures. Both the stability of of the goals themselves mating reliabilities of and examined. Emmons 3and the the goal goal dimensions dimensions have have been been examined. Emmons (1986) (1986) computed computed both both 3and 6-month stability coefficients for the 18 striving assessment dimensions. The and 6-month stability coefficients for the 18 striving assessment dimensions. The stabilities scales ranged ranged from .58 to 1-month interval interval stabilities of of the the individual individual scales from .58 to .91 .91 for for the the 1-month of .73) and from from .47 to .70 for for the 3-month period (with a mean of of (with a mean of .60). Social stable, while .60). Social desirability desirability and and importance importance were were the the most most stable, while effort effort and and impact impact were the least stable. Klinger and Cox (1986) administered administered the IntO IntQ to 42 alcoholic inpatients upon intake and 1 month month later. later. The The test-retest test-retest correlations correlations of of 88concerned concerned from .07 to .77, with a mean of of .30. Cantor Cantor and Langston Langston (1989) dimensions ranged from administered the meaning dimensions a year and a half half apart in order to assess unfortunately do not present present correlations correlations between change in task appraisals, but unfortunately the dimensions over this time period. The stability of has also been investigated. In a sample of the goals themselves has of undergraduates, after of 40 40 undergraduates, after 11 year year 82% 82% of of personal personal strivings strivings listed listed at at Time Time 11 were were still (with minor wording changes; 18 months, 45% still present present (with minor wording changes; Emmons, Emmons, 1989c). 1989c). After After 18 months, 45% were were still still present, present, and and aa 3-year 3-year follow-up follow-up yielded yielded aa stability stability of of just just over over 50%. 50%. Thus, Thus, there that strivings people's lives. there is is evidence evidence that strivings reflect reflect enduring enduring concerns concerns in in people's lives. Many Many of of the 50% that the 50% that were were no no longer longer present present were were associated associated with with aa particular particular life life context context (college environment) (college environment) that that was was no no longer longer part part of of the the person's person's life. life. F. Goal Equivalence Classes The units, are, definition, idiographic. idiographic. Indeed, The personal personal goal goal units, are, by by definition, Indeed, the the respective respective propo proponents nents of of each each approach approach have have argued argued that that therein therein lies lies their their advantage. advantage. For For certain certain research desirable. The research purposes, purposes, however, however, aa more more abstract abstract level level of of analysis analysis is is desirable. The categories categories that that have have been been used used for for the the different different units units along along with with percentages percentages of of goals goals falling falling within within each each category category are are shown shown in in Table Table V. V. Categorizing Categorizing goals goals in in this this manner permits comparison manner comparison of of the frequency frequency of of goals in terms of of gender, age, and other social structural variables (Ryff, (Ryff, 1987). The ability to categorize personal personal territory between idiographic and goals will enable researchers to traverse the territory nomothetic levels of of analysis. Conflict and Complexity Complexity G. Goal Properties: Conftict Among all of of the goal attributes previously discussed, perhaps none is more impor imporAmong contended that tant than conflict. This was aptly expressed by Powers (1973), who contended "conflict represents the most serious kind of malfunction malfunction of of the brain short of of "conflict

CHAPTER CHAPTER 20 20

503 503

MOTIVES MOTIVES AND AND GOALS GOALS TABLE TABLE V V

Personal Goal Categories

Personal Goal Categories Current concerns concerns

% %

Pastimes/Travel Pastimesffravel

20 15 11 11 10 10

Family Job/finances Job/finances

Education Religion/politics Love/sex Friends Physical health Miscellaneous

Emotional health

8 8 7 6 4

Personal projects Interpersonal Academic

Recreational Recreational Finance/legal Finance/legal HeaIthfbody HeahhA)ody CulturaUaesthetic Cultural/aesthetic Sports Estate

Intrapersonal Intrapersonal Family Drinking/drugs

Occupational Occupational Boyfriend/girlfriend Boyfriend/girlfriend Hobbies

Reading Sex Vacations/trips Spiritual

% % 14 14 13 13 12 12 7 6 6 5 5 5 4 4 4 3 2 2 2 2 1

Personal Personal strivings strivings Positive Negative Interpersonal Intrapersonal Achievement Achievement Affiliation/intimacy Affiliation/intimacy Power PersonaUGrowth Personal/Growth

Self-presentation Self-presentation Self-sufficiency Self-sufficiency

% % 85 15 15 57 43 18 31 31 12 12 24 12 12 7

Personal Personal goals goals Work/school Social Life Leisure Family Life Life Material/ environment Personal growth Other/general Other/general

% % 20 19 19 118 8 15 14 8 6

physical damage" (p. 253). 253). Similarly, Pervin (1985) concluded his Annual Review chapter by stating that "I am struck by the power of conflicting motivations in my Uterature" (p. 105). patients and the absence of such phenomena in the literature" 105). Theorists and researchers beginning with Freud have stressed the necessity of avoiding inner conflict by maintaining consistency and harmony among aspects of the self and the psychologically injurious consequences of failing to do so. Goal conflict and motivational conflict more generally have had a long history. It has been of central 1927; Homey, Horney, 1945), behavioral (Lewin, 1935; concern in psychodynamic (Freud 1927; Miller, 1959), and cognitive formulations (Lecky, 1945). Until recently, however, little progress had been made in understanding the dynamics of nonpathological forms of conflict. Fortunately, there has been some movement in this area as of late, and a comprehensive review of this literature Uterature can be found in Emmons, King, and Sheldon (1993). Wilensky (1983) discusses the various types of conflict that can occur between goals, at both an intrapersonal and an interpersonal level. At the intrapersonal level, goals may be negatively related to each other (the pursuit of one interferes with the pursuit of another). At the interpersonal level, the goals of two or more individuals may either compete or be in accordance. Wilensky's thorough analysis includes a discussion of the various classes of reasons why goals can come into conflict, and offers suggestions as to how such conflicts might be resolved. Peterson in the context of dyadic relation relation(1989) has also explored interpersonal goal conflict in ships.

S04 504

ROBERT EMMONS ROBERT A. A. EMMONS

Traditionally, conflict has been associated with tension, vacillation, uncer uncertainty, and confusion (Miller, 1959) as well as anxiety, depression, hostility, delu delusions, and hallucinations (Powers, 1973). There is supporting empirical empirical evidence. Palys and Little (1983) found that conflict between personal projects was associated conflict with low life satisfaction, and Emmons and King (1988) reported that conflict between and within personal strivings was related to measures of psychological distress. The pathological effects of conflict conflict have not been limited to the psychologi psychologiconflict in the form cal domain either. As mentioned earlier, chronic motivational conflict of the inhibited power motive syndrome is predictive of of high blood pressure, lowered of immunocompetence levels, and increased risk of coronary heart disease. disease. Emmons Emmons immunocompetence levels, and increased risk of coronary heart and King (1988) found that conflict between personal strivings is associated with aa and King found that conflict between personal strivings is associated with variety of physical symptoms as well as in an increase in health center visits. These variety of physical symptoms as well as in an increase in health center visits. These authors also found that individuals tended to dwell on conflicting strivings but to authors also found that individuals tended to dwell on conflicting strivings but to inhibit acting on on them. them. In In action action theory theory terminology, terminology, it it might might be be said said that that these these inhibit acting individuals have have adopted state orientation. orientation. Pennebaker's Pennebaker's (1985, (1985,1989) inhibition individuals adopted aa state 1989) inhibition model of psychosomatic psychosomatic illness illness has has been been used used to to explain explain why why conflict conflict results results in in model of physical illness. illness. Personal Personal striving conflict may may be be particularly particularly debilitating, debilitating, since it physical striving conflict since it reflects conflicts at at higher higher levels in the the motivational motivational hierarchy 1973). reflects conflicts levels in hierarchy (Powers, (Powers, 1973). So although there has been a scattering of work since Pervin's charge, there is still much work remaining to be done before we fully understand (1) the types of motivational conflicts conflicts that people suffer suffer from; (2) the effects effects of such conflicts conflicts of on cognition, emotion, behavior, and psychosomatic disease; and (3) how such conflicts may be resolved. On the positive side, it does appear that people are aware of and are able to report conflicts between their goals (as well as ambivalence over of a goal) and that such conflict has measurable consequences. The development of conflict matrix, should be alternative measures of goal conflict, in addition to the conflict contaminating a priority. It is possible that ratings on it are susceptible to a host of contaminating influences, such as mood at the time it is filled out. Longitudinal studies on the long-term effects of conflict are also needed, as it has been suggested that conflict may be adaptive in the long run nm (Brim & Kagan, 1980; Emmons et al., 1993). Another goal attribute that has begun to receive some some attention is goal com complexity. Complexity is a structural attribute of goal systems and consists of of two components: differentiation and integration. Differentiation Differentiation refers to the degree of interdependence (low differentiation) differentiation) and independe�ce independence (high differentiation) differentiation) of among the elements (goals) in in the system. Interdependent goals are goals that affect, in an instrumental or conflicting way, other goals within a person's goal of simplicity, whereas system. Greater similarity among one's goals is indicative of differentiation is a sign of complexity in the system. Integration refers to greater differentiation the number number of of linkages linkages between between the the goals. goals. Complexity Complexity also also refers refers to to the the number number the of different different plans, plans, or or means-end connections, that that are are associated associated with with each each goal. goal. of means-end connections, Complex goals goals in in this this sense sense are are goals goals that that can can be be successfully successfully achieved achieved in in aa variety variety Complex of different different ways. ways. Note Note that that this this meaning meaning of of complexity complexity refers refers to to individual individual goals goals of whereas the previous previous use use of of complexity complexity referred referred to to the the goal goal system system as as aa complete complete whereas the

CHAPTER CHAPTER20 20 M011VES MOTIVESAND ANDGOALS GOALS

50S 505

A similar concept is that of of plan reflectivity reflectivity (Cantor et aI., al, 1987), defined in unit. A terms of of richness of plan elaboration for achieving life tasks. differentiation component of self-complexity self-complexity has been examined in an The differentiation of studies by Linville (1982, (1982,1985). intriguing series of 1985). She has demonstrated that the of self-representation self-representation in terms of traits or social roles is inversely related complexity of of affective affective experience. Emmons and King (1989) to the extremity and variability of differentiation to also encompass goal differentia differentiaattempted to expand the notion of differentiation tion, as Linville's work has been limited to the trait and social role domains. They found, in contrast to Linville, that goal differentiation differentiation was related to greater levels arousalof emotional reactivity. The results were interpreted as supporting an arousal regulation theory of of affect affect intensity (Larsen & Diener, 1987). Similarly, Donahue, Robins, Roberts, and John (1993) found found that self-concept self-concept differentiation differentiation was pre predictive of of psychological distress. They suggested that differentiation differentiation reflects patho pathological fragmentation fragmentation rather than flexibiUty flexibility within the self. Clearly, more research is needed in order to settle these contradictory contradictory results. rv. CONCLUSIONS AND FuruRE FUTURE DIRECTIONS IV. DIRECI10NS

In a review such as this, it is the norm to apologize for omitting certain topics. This will be no exception. Topics that would have been covered were it not for space of action (Atkin (Atkinconsiderations include script theory (Tomkins, 1987), the dynamics of self-determination theory (Deci & Ryan, 1985), and the theory son & Birch, 1970), self-determination of relational dynamics (Nuttin, 1984). Fortunately, there are a number of excellent of sources on these approaches that the interested reader can consult. A A natural candidate for future directions will be the exploration of the linkages between motive dispositions and personal goals. These two literatures have tended to go their separate ways. This is not totally surprising, in that some of the personal goal approaches were developed as a reaction against the perceived shortcomings in the motive dispositional approach. Yet motives refer to goals and goal-directed of motives in individual lives. A A action, and goals represent the personalization of of the motives into their constituent constituent goals may be a necessary necessary task decomposition of for settling persistent controversies in the motive literature, such as sex differences differences (Stewart & Chester, (Stewart Chester, 1982) 1982) and and the the difference difference between between affiliation affiliation and and intimacy intimacy motivation (McAdams, 1980). 1980). motivation (McAdams, McClelland McClelland and and his his associates associates (McClelland, (McClelland, Koestner, Koestner, & Weinberger, Weinberger, 1989) 1989) have two forms have distinguished distinguished two forms of of motivation. motivation. One One is is an an affectivelylbiologically affectively/biologically based based system, termed "implicit motives," and the the other is a cognitively/experientiaUy cognitively/experientially based system, system, termed termed "self-attributed based "self-attributed motives," or "explicit motives." These corre correspond to the social motive and idiographic goal approaches described earlier. These independently of of two systems are believed to develop independently, to operate independently different classes of of behavior. The degree to which they each other, and to predict different are independent, however, is disputable. Emmons and McAdams (1991) found significant relations between personal strivings (a form of of self-attributed motives)

S06 506

ROBERT ROBERT A. A. EMMONS EMMONS

and motive and motive dispositions dispositions as as assessed assessed by by aa picture-story picture-story exercise exercise (implicit (implicit motives). motives). King King (1995) (1995) rerported rerported significant significant correlations correlations among among explicit explicit measures measures of of the the same same motive, yet relative independence between explicit and implicit measures motive, yet relative independence between explicit and implicit measures of of the the same motive. same motive. Future Future research research should should aim aim to to identify identify conditions conditions under under which which mea measures sures of of implicit implicit and and explicit explicit motives motives converge converge or or fail fail to to converge. converge. Future Future work work should also include units (both (both explicit implicit measures) should also include multiple multiple units explicit and and implicit measures) as as predictors predictors of McClelland (1981), of relevant relevant outcomes. outcomes. McClelland (1981), Veroff Veroff and and Smith Smith (1985), (1985), and and Biernat Biernat (1989) (1989) have have demonstrated demonstrated the the virtues virtues of of combining combining motives motives with with other other units units such such greater predictive accuracy. Attention Attention should also as values and traits in achieving greater directed toward toward the development of alternative means of of assessing implicit and be directed development of explicit motives, in addition to the picture-story picture-story exercise and free-response free-response listing procedures described described earlier in this chapter. Emmons and King (1992) provide an procedures integrative review of of the relation relation of of thematic measures of of motivation to other affective constructs, such as schemas, scripts, and relational patterns. cognitive and affective psychobiographical The personal goal units can and should be applied in the psychobiographical study of of individuals. Although Although the units claim to be idiographic, the whole person from these approaches. Since a major major task in most psychobiographical psychobiographical is missing from recurrent goal themes, the application of of the personal goal endeavors is to extract recurrent units would seem to be a natural step to take. There may be advantages to framing framing these recurrent recurrent themes in terms of of personal goals. Since careful careful attention has been of these units, it may make it easier to uncover recurrent paid to the assessment of themes in the form form of of these goals in personal documents documents and other other archival material. Perhaps Perhaps the the most most significant significant conclusion conclusion that that can can be be drawn drawn from from aa review review of of that substantial progress is being made in narrowing narrowing the motivational literature is that motithe gap between social-cognitive and traditional personological approaches to moti vation. The goals concept appears to be highly desirable, given its hierarchical concept implies; structure; the flexibility, discriminativeness, and coherence that the concept amenability to measurement measurement and individual differences. differences. We may now now be in and its amenability position to answer G. W. Allport's AUport's (1968) query, query, "What **What units units should we employ?" employ?" a position

REFERENCES AUport, F. (1937). Teleonomic Teleonomic description description in in the the study study of of personality. personality. Character Characterand andPersonal PersonalAllport, ity, 5, 5, 202-214. ity, AUport, G. W. (1968). What units should we employ? In. G. Lindzey (Eds.), Assessment of of Allport, human motives (pp. 239-260). New York: Grove Press. Atkinson, 1. action. New York: Wiley. J. W., & Birch, D. (1970). The dynamics dynamics of of action. effect of experimental arousal Atkinson, J. W., Heyns, R. W., & Veroff, J. (1954). The effect of the affiliation affiliation motive on thematic apperception. Journal of of Abnormal Abnormal and Social of Psychology, Psychology, 49, 49, 405-410. Biernat, M. (1989). Motives and values to achieve: Different Different constructs with different different effects. effects. Journal of Personality,57, 57,69-96. 69-96. of Personality, processes. Hillsdale, NJ:NJ: Erlbaum. (Ed.). (1988). Persons Personsinincontext: context:Developmental Developmental processes, Hillsdale, Erlbaum. Bolger, N. (Ed.).

CHAFfER CHAPTER 20 20

MOTIVES AND ANDGOALS GOALS MOTIVES

507 507

O., & & Kagan, J. (1980). (1980). Constancy and change: A view of the issues. In O. O. Brim & Brim, 0., J. Kagan (Eds.), Constancy human development (pp.(pp. 1-26). Cambridge, Constancyand andchange changeinin human development 1-26). Cambridge, MA: Harvard University Press. C. M. M. (1964). (1964). The course course of of human human life. life. New New York: York: Springer. Springer. Buhler, C. From thought thought to to behavior: behavior: "Having" "Having" and and "doing" "doing" in inthe thestudy studyof ofpersonal personalCantor, N. (1990). From ity and cognition. American 735-750. American Psychologist, Psychologist,45,45, 735-750. Cantor, N., Brower, A., & & Korn, H. (1985). Cognitive bases of personality in a life transition. E. E. E. Roskam Roskam (Ed.), (Ed.), Measurement Measurementand andpersonality personalityassessment assessment 323-331). New In E. (pp.(pp. 323-331). New In York: Elsevier. N., & & Kihlstrom, J. F. F. (1987). (1987). Personality Personality and andsocial socialintelligence. intelligence.Englewood EnglewoodCliffs, Cliffs, Cantor, N., NJ: Prentice-Hall. Cantor, N., & Kihlstrom, Kihlstrom, J. J. F. F. (1989). (1989).Social Socialintelligence intelligence and and cognitive cognitive assessments assessmentsof ofperson personS. Wyer, Jr. Jr. & T. T. K. K. Srull SruU (Eds.), Advances in social cognition cognition(Vol. (Vol.2,2,pp. pp. ality. In R. S. 11-59). -59). Hillsdale, NJ: Erlbaum. & Langston, C. C. A. A. (1989). Ups Ups and downs of life tasks in a life transition. In Cantor, N., N., & L. A. A. Pervin Pervin (Ed.), (Ed.), Goal concepts concepts ininpersonality personalityand andsocial social psychology (pp.87-126). 87-126). L. psychology (pp. Hillsdale, NJ: Erlbaum. J., Niedenthal, P., C, & & Brower, A. (1987). Life tasks, self selfCantor, N., N., Norem, J., P., Langston, c., ideals, and cognitive strategies in a life transition. Journal of Personalityand and concept ideals, of Personality Social Psychology, Psychology,53, 53,1 178-1 1178-1191. Social 191. N., & & Zirkel, S. S. (1990). Personality, cognition, and purposive behavior. In L. A. Cantor, N., of personality personality (pp. 135-164). New Guilford Press. New York: Guilford Pervin (Ed.), Handbook of Carver, C. C. S., S.,& &Scheier, Scheier,M. M.F.F.(1981). (1981).Attention Attention and self-regulation: A control theory approach Carver, and self-regulation: A control theory approach behavior. New NewYork: York:Springer-Verlag. Springer-Verlag. to human behavior. C. S., & Scheier, M. M. F. F. (1982). Control theory: A useful useful conceptual framework framework for Carver, C. S., & Bulletin,92, 92,111-135. Psychological Bulletin, 1 1 1 -135. personality-social, clinical, and health psychology. Psychological S., & & Scheier, M. F. F. (1990). (1990). Origins and functions functions of positive and negative affect: affect: Carver, C. S., A control-process view. Psychological PsychologicalReview, Review,97, 97,19-35. 19-35. Caspi, A. (1987). Personality in the life course. Journal Journal of of Personality Personalityand andSocial SocialPsychology, Psychology, 53, 1203-1213. 53, C. N. N. (1981). The The history of the concept of motivation. Journal of of the of the History of Cofer, C. BehavioralSciences, Sciences,17,17, 48-53. 48-53. Behavioral Deci, E., E., & & Ryan, R. M. M. (1985). (1985). Self-determination Self-determinationtheory. theory.New NewYork: York:Plenum PlenumPress. Press. Donahue, E. M., Robins, R. W., W., Roberts, B. W., & O. P. P. (1993). The divided self: E. M., B. W., & John, O. Concurrent and longitudinal effects adjustment and social roles on effects of psychological adjustment Psychology, 64, 64, 834-846. self-concept differentiation. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 834-846. Deci, E., E., & & Ryan, R. R. M. M. (1985). (1985). Self-determination Self-determinationtheory. theory.New NewYork: York:Plenum PlenumPress. Press. Donley, R. E., & Winter, Winter, D. D. G. G. (1970). (1970). Measuring Measuringthe themotives motivesof ofpublic publicofficials officials at ataadistance: distance: An exploratory exploratory study of American presidents. Behavioral Science, Science, 15, 15,227-236. 227-236. Emmons, R. A. well A. (1986). (1986). Personal strivings: An approach to personality personality and subjective wellbeing. Journal Psychology, 51,51, 1058-1068. Journalof ofPersonality Personalityand andSocial Social Psychology, 1058-1068. Emmons, R. A. A. (1989a). (1989a).The Thebig bigthree, three,the thebig bigfour, four,or orthe thebig bigfive? five?Contemporary ContemporaryPsychology, Psychology, 34, 644-646. 34, Enmions, R. A. A. (1989b). Exploring the relations between traits and motives: The case of Emmons, of narcissism. In D. D. M. M. Buss & & N. N. Cantor (Eds.), Personality Personalitypsychology: psychology:Recent Recenttrends trends and emerging emerging issues issues(pp. (pp.32-44). 32-44).New NewYork: York:Springer-Verlag. Springer-Verlag. Emmons, R. A. (1989c). The The personal personal striving striving approach approach to to personality. personality. In In L. L. A. A. Pervin Pervin (Ed.), (Ed.), Goal concepts conceptsininpersonality personality and social psychology (pp. 87-126). Hillsdale, Erlbaum. and social psychology (pp. 87-126). Hillsdale, NJ:NJ: Erlbaum. .

50S

ROBERT A. A. EMMONS ROIIERT

Emmons, R. A. (1993). Current Current status of of the motive motive concept. concept. In K. Craik, R. Hogan, & R. Wolfe Wolfe (Eds.), Fifty years years of of personality personality psychology psychology (pp. 187-196). New York: Plenum Press. Emmons, R. A. (1996). Striving and feeling: Personal goals and subjective well-being. In GoUwitzer & J. A. Bargh (Eds.), The psychology psychology of of action-linking action-linking cognition cognition and and P. M. Gollwitzer motivation to behavior behavior (pp. 313-337). New York: Guilford Publications, Inc. motivation Emmons, R. A., & Diener, E. (1986). A A goal-affect goal-affect analysis of situational situational choice and avoid avoidJournal of Personality, 20, 309-326. of Research in Personality, ance. Journal Emmons, R. A., & King, L. A. (1988). Conflict among personal strivings: Immediate and Journal of of Personality Personality long-term implications for psychological and physical well-being. Journal Psychology, 54, 1040-1048. and Social Social Psychology, Emmons, R. A., & King, L. A. A, (1989). Personal striving differentiation differentiation and affective affective reactivity. Journal Journal of of Personality Personality and and Social Social Psychology, Psychology, 56, 478-484. Emmons, R. A., & King, L. A. (1992). Thematic analysis, experience sampling, and personal goals. In C. Smith (Ed.), Motivation Motivation and and personality: personality: Handbook Handbook of of thematic content content analysis (pp. 73-86). New York: Cambridge University Press. analysis Emmons, R. A., King, L. A., & Sheldon, K. (1993). Goal conflict conflict and the self-regulation self-regulation of Pennebaker (Eds.), Handbook Handbook of of mental mental control control action. In D. M. Wegner & J. W. Pennebaker (pp. 528-551). Englewood Cliffs, NJ: Prentice-Hall. Emmons, R. A., & McAdams, D. P. (1991). Personal strivings and motive dispositions: Exploring the links. Personality 7, 648-654. Personality and and Social Social Psychology Psychology Bulletin, 117, Foder, E. M. (1984). The power motive and reactivity to power stresses. Journal Journal of of Personality Personality and and Social Social Psychology, Psychology, 47, 853-859. Fontana, A. F., Rosenberg, R. L., Marcus, J. L., & Kerns, R. D. (1987). Type A behavior Journal of of Personality Personality and pattern, inhibited power motivation, and activity inhibition. Journal Social Psychology, Psychology, 52, 177-183. Social Ford, D. H. (1987). Humans Humans as self-constructing self-constructing living living systems. Hillsdale, NJ: Erlbaum. Ford, D. H., & Ford, M. E. (1987). Humans as self-constructing self-constructing living hving systems: An overview. In M. E. Ford & D. H. Ford (Eds.), Humans Humans as self-constructing self-constructing living living systems: systems: Putting Putting framework to work work (pp. 11-46). the framework -46). Hillsdale, NJ: Erlbaum. Motivating humans: Goals, emotions, emotions, and and personal personal agency belief beliefs. News. New Ford, M. E. (1992). Motivating bury Park, CA: Sage. A taxonomy of of human goals and some possible Ford, M. E., & Nichols, C. W. (1987). A Humans as self-constructing self-constructing living living applications. In M. E. Ford & D. H. Ford (Eds.), Humans systems: systems: Putting Putting the framework framework to work work (pp. 289-311). Hillsdale, NJ: Erlbaum. Frese, M. & Sabini, J. (Eds.). (1985). Goal Goal directed directed behavior: behavior: The concept concept of of action action in psychology. psychology. Hillsdale, NJ: Erlbaum. Freud, S. (1927). The ego ego and and the id. London: Institute for Psychoanalysis and Hogarth Press. Motivation, intention, and and volition. volition. Berlin: Springer SpringerHalisch, F., & Kuhl, J. (Eds.). (1987). Motivation, Verlag. Heckhausen, H. (1991). Motivation Motivation and and action. Berlin: Springer-Verlag. Hogan, R. (1983). A A socioanalytic theory of of personality. In M. M. Page (Ed.), Nebraska Nebraska Symposium Symposium on Motivation Motivation (pp. 55-90). Lincoln: University of of Nebraska Press. Hogan, R. (1986). What every student should know about personality psychology. In V. P. Makosky (Ed.), The G. Stanley Hall Hall Lecture Lecture series (Vol. 6, pp. 43-64). Washington, DC: American Psychological Psychological Association. Homey, K. (1945). Our Our inner inner conflicts. New York: Norton. Horney, Hyland, M. E. (1987). Control theory interpretation of of psychological mechanisms of depres depresComparison and integration of of several theories. Psychological Psychological Bulletin, 102, sion: Comparison 109-121.

CHAPTER CHAPTER 20 20

MOTIVESAND ANDGOALS GOALS MOTIVES

509 5W

M. E. (1988). Motivational control theory: An integrative framework. Journal Journal of of Hyland, M. 642-651 . Personality Personalityand andSocial SocialPsychology, Psychology,55,55, 642-651. Jemmott, J. J. B. (1987). Social motives and susceptibility to disease: Stalking individual differ differ267-298. ences in health risks. Journal Journalof ofPersonality, Personality,55,55, 267-298. King, L. L. A. (1995). Wishes, motives, goals, and personal memories: memories: Relations of measures of human motivation. Journal Journalof ofPersonality, Personality,63,63,985-1007. 985-1007. Klinger, E. (1975). Consequences of commitment to and disengagement from incentives. Psychological 223-231. PsychologicalReview, Review,82,82, 223-231. Klinger, E. (1977). Meaning andand thethe incentives in people's lives.lives. Meaningand andvoid: void:Inner Innerexperience experience incentives in people's Minneapolis: Minneapolis: University of Minnesota Press. Klinger, E. E. (1985). (1985). Missing Missinglinks linksin inaction actiontheory. theory.In InM. M.Frese Frese&&J.J.Sabini Sabini (Eds.), Goal directed Klinger, (Eds.), Goal directed behavior: Theconcept concept action in psychology 311-319). Hillsdale, Erlbaum. behavior: The ofof action in psychology (pp.(pp. 311-319). Hillsdale, NJ:NJ: Erlbaum. Klinger, E. E. (1987). (1987). The The interview interview questionnaire questionnaire technique: technique: Reliability Reliability and and validity validityof ofaamixed mixed Klinger, idiographic-nomothetic measure measure of of motivation. motivation. In In J. J. N. N. Butcher Butcher & & c. C. D. D. Spielberger Spielberger idiographic-nomothetic (Vol. 6, pp. 31-48). Hillsdale, NJ:NJ: Erlbaum. (Eds.), Advances Advancesin inpersonality personalityassessment assessment (Vol. 6, pp. 31-48). Hillsdale, Erlbaum. Klinger, E. E. (1989a). (1989a). Emotional Emotional mediation mediation of of motivational motivational influences influences on on cognitive cognitive processes. processes. Klinger, Internationalperspectives perspectivesononachievement achievement In F. F. Halisch Halisch & & J. J. van van den den Bercken Bercken (Eds.), (Eds.), International In task motivation motivation(pp. (pp.317-326). 317-326).Nisse, Nisse,Netherlands: Netherlands:Swets Swets&&Zeitlinger. Zeitlinger. and task Klinger, E. E. (1989b). (1989b). Goal Goal orientation orientation as as psychological psychological linchpin: A commentary commentary on on Cantor Cantor linchpin: A Klinger, and Kihlstrom's "Social intelligence and cognitive assessments of personality. In Jr. & &T. T.K. K.Srull SruU(Eds.), (Eds.),Advances Advancesinin social cognition (Vol. 2, pp. 123-130). social cognition (Vol. 2, pp. 123-130). R. S. Wyer, Jr. Hillsdale, NJ: NJ: Erlbaum. Erlbaum. Hillsdale, & Maxeiner, Maxeiner, M. M.E. E. (1981). (1981). Current Current concerns: concerns: Assessing Assessing therapeuti therapeutiKlinger, E., Barta, Barta, S. S. G., G., & Klinger, cally relevant relevant motivation. motivation. In In P. P. C. C. Kendall Kendall & & S. S. Hollon Hollon (Eds.), (Eds.), Assessment Assessmentstrategies strategies cally cognitive-behavioralinterventions interventions (pp. 161-195). New York: Academic Press. for cognitive-behavioral (pp. 161-195). New York: Academic Press. Klinger, E., & Cox, Cox,W. W.M. M.(1986). (1986).Motivational Motivationalpredictors predictorsofofalcoholics alcoholicsresponses responsestotoinpatient inpatient Klinger, E., & Advances in in Alcohol and Substance SubstanceAbuse, Abuse,6,6,35-44. 35-44. treatment. Advances Kuhl, J. J. (1985a). (1985a). Volitional Volitional mediators mediators of of cognition-behavior cognition-behavior consistency: consistency: Self-regulatory Self-regulatory Kuhl, processes and action versus state orientation. In J. J. Kuhl Kuhl & & J.J.Beckmann Beckmann(Eds.), (Eds.),Action Action control:From Fromcognition cognition behavior 101-128). Berlin: Springer-Verlag. (pp.(pp. 101-128). Berlin: Springer-Verlag. control: to to behavior Kuhl, J. (1985b). From cognition to behavior: behavior: Perspectives for future research on action to to behavior control. In J. Kuhl & & J.J. Beckmann Beckmann (Eds.), (Eds.), Action Action control: control:From Fromcognition cognition behavior (pp. 267-275). Berlin: Berlin: Springer-Verlag. control:From Fromcognition cognition behavior. Berlin: Kuhl, J., & & Beckman, Beckman, J. J. (Eds.). (Eds.). (1985). (1985). Action control: to to behavior. Berlin: Kuhl, Springer-Verlag. Larsen, R. J., & & Diener, Diener, E. E. (1987). (1987).Affect Affect intensity intensity as as an an individual individual difference difference characteristic: characteristic: Larsen, 21,21, 1 -39. A review. Journal Journalof ofResearch ResearchininPersonality, Personality, 1-39. Self-consistency:AAtheory theory personality. New York: Island Press. Lecky, P. (1945). Self-consistency: ofof personality. New York: Island Press. Lewin, K. (1935). A dynamic dynamic theory theoryof ofpersonality. personality.New NewYork: York:McGraw-Hili. McGraw-Hill. Linville, P. W. (1982). Affective consequences of complexity regarding the self and others. In M. S. Clark & & S. T. T. Fiske (Eds.), (Eds,), Affect and cognition cognition (pp. 79-109). Hillsdale, NJ: Erlbaum. NJ: Linville, P. P. W. W. (1985). (1985). Self-complexity Self-complexity and and affective affective extremity: extremity: Don't Don't put put all all of of your your eggs eggs in Linville, one basket. Social Social Cognition, Cognition,3,3,94-120. 94-120. Little, B. B. R. R. (1983). (1983). Personal Personal projects: projects: A A rationale rationaleand andmethod methodfor forinvestigation. investigation.Environment Environment Little, Behavior,15, 15,273-309. 273-309. and Behavior, Little, B. B. R. R. (1987). (1987). Personal Personalprojects projectsand andfuzzy fuzzyselves: selves:Aspects Aspectsof ofself-identity self-identityininadolescence. adolescence. Little, thethe life-span In T. T. Honess & & K. Yardley (Eds.), Self and identity: identity:Perspectives Perspectivesacross across life-span In (pp. 230-245). London: London: Routledge & & Kegan Paul.

510 510

ROBERT EMMONS ROBERT A. A. EMMONS

Little, B. R R. (1989). Personal projects analysis: Trivial pursuits, magnificent magnificent obsessions, and (Eds.), Personality the search for for coherence. In D. M. Buss & N. Cantor (Eds.), Personalitypsychology: psychology: Recent trends trends and emerging emerging issues issues(pp. (pp.15-31). 15-31).New NewYork: York:Springer-Verlag. Springer-Verlag. Maslow, A. H. (1970). Motivation Motivation and personality personality (2nd ed.). New York: Harper & & Row. McAdams, D. P. (1980). A thematic coding system for the intimacy motive. Journal of of Research Personality,14, 14,413-432. 413-432. Research in Personality, The "imago": "imago": A A key key narrative narrative component component of of identity. identity. In In P. P. Shaver Shaver McAdams, D. P. (1985a). The ofpersonality psychology (Vol. (Vol. 6, 6,pp. pp. 1115-141). Beverly Hills, Hills, personality and social psychology 15-141). Beverly (Ed.), Review of CA: Sage. Power, intimacy, intimacy, and the life story: story: Personological Personologicalinquiries inquiriesinto into McAdams, D. P. (1985b). Power, IL: Dorsey Press. identity. identity. Homewood, IL: lives: The emergence of identity as McAdams, D. P. (1990). Unity and purpose in human lives: R. A. Zucker, R R. A. Emmons, & & S. S. Frank (Eds.), (Eds.), Studying Studying a life story. In A. I. Rabin, R 148-200). New York: Springer. persons and lives lives (pp. 148-2(0). stories we welive liveby: by:Personal Personalmyths myths and and the themaking making of of the theself. self McAdams, D. P. (1993). The stories New York: W. Morrow. McAdams, D. D. P. P. (1994). (1994). The Theperson:An introductiontotopersonality personalitypsychology psychology(2nd (2ndEdition). Edition). person: An introduction McAdams, TX: Harcourt Brace. Forth Worth, TX: McAdams, D. D. P., & & Constantian, Constantian, C. C. A. A. (1983). (1983). Intimacy Intimacy and and affiliation affiliation motive motive in in daily daily Journalof of Personality Personalityand andSocial SocialPsychology, Psychology, living. An experience-sampling analysis. Journal 45, 851-861. 45, McClelland, D. C. (1980). Motive dispositions: The merits of operant versus respondent measures. In L. Wheeler (Ed.), Review of of personality personality and social social psychology psychology (Vol. (Vol. 1, 1, pp. 111-41). 1 -41). Beverly Hills, CA: Sage. R. Zucker, & & J. Aronoff Aronoff McClelland, D. C. (1981). Is personality consistent? In A. I. Rabin, R (Eds.), Further Further explorations explorationsininpersonality personality(pp. (pp.87-113). 87-113).New NewYork: York:Wiley. Wiley. McClelland, D. C. (1985). Human motivation. motivation. Glenview, IL: Scott, Foresman. American Psychologist, Psychologist, McClelland, D. C. (1989). Motivational factors in health and disease. American 44, 44, 675-683. 675-683. M cClelland, D. C., McClelland, C, & & Kirschnit, C. (1988). (1988). The effect effect of motivational arousal through films on salivary immunoglobulin immunoglobulin A. Psychology Psychology and and Health, Health,2,2,31-52. 31-52. McOelland, McClelland, D. c., C, Koestner, R, R., & & Weinberger, J. J. (1989). (1989). How Howdo doself-attributed self-attributed and andimplicit implicit motives differ? PsychologicalReview, Review,96, 96,690-702. 690-702. differ? Psychological Miller, N. E. (1959). Liberalization of basic S-R concepts: Extensions to conflict behavior, motivation, and social learning. In S. Koch (Ed.), A study of of psychology as a science science (Vol. 2, pp. pp. 196-292). New York: McGraw-Hill. personality. New NewYork: York:Oxford Oxford University UniversityPress. Press. Murray, H. A. (1938). Explorations in personality. Dialectica, Murray, H. A. (1951). Some basic psychological assumptions and conceptions. Dialectica, 5, 5, 266-292. R. (1987). Some implications of the trait-state distinction for the study of Nesselroade, J. R of development across the life span: The case of personality. In P. B. Baltes, D. L. e-span development and behavior Featherman, & &R R. M. Lerner Lemer (Eds.), Lif Life-span behavior (Vol. 8). New York: Academic Press. lUingworth, K., & & Shaun, S. S. (1993). Strategy-dependent effects of reflecting reflecting Norem, J. K., Illingworth, of on self and tasks: Some implications of optimism and defensive pessimism. Journal of Personality Personality & & Social Psychology, Psychology, 65, 65,822-835. 822-835. & Lazarus, R R. (1990). The The structure structure and and measurement measurement of of personal personal commitments. commitments. Novacek, J., & of Personality, Personality,58, 58,693-715. 693-715. Journal of

CHAPTER CHAPTER 20 20

MOTIVES MOTIVESAND ANDGOAU GOALS

Sl1 511

Nuttin, J. (1957). Personality dynamics. In H. P. David David & &H. H.von vonBracken Bracken(Eds.), (Eds.),Perspectives Perspectives in in personality personalitytheory theory(pp. (pp.183-196). 183-196).New NewYork: York:Basic BasicBooks. Books. Nuttin, J. (1984). Motivation, action. Hillsdale, NJ:NJ: Erlbaum. Motivation,planning, planning,and and action. Hillsdale, Erlbaum. Palys, T. T. S., S., & & Little, Little,B. B.R R.(1983). (1983).Perceived Perceivedlife lifesatisfaction satisfactionand andthe theorganization organizationof ofpersonal personal project systems. Journal Psychology, 44,44, 1221-1230. Journalof ofPersonality Personalityand andSocial Social Psychology, 1221-1230. Pennebaker, J. W. (1985). Traumatic experience and psychosomatic disease: Exploring the roles of behavioral inhibition, obsession, and confiding. Canadian CanadianPsychology, Psychology,26,26, 82-95. Pennebaker, J. W. (1989). Confession, inhibition, and disease. In L. Berkowitz (Ed.), Ad Advances experimentalsocial socialpsychology psychology(Vol. (Vol.22, 22,pp. pp.211-244). 211-244).New NewYork: York:Aca Acavances in in experimental demic Press. flow of behavior: behavior: Toward a theory of goals. In M. Pervin, L. A. (1983). The stasis and flow M. M. Page (Ed.), Nebraska Nebraska Symposium Symposiumon on Motivation Motivation(pp. (pp. 1-53). 1-53). Lincoln: Lincoln: University University of of Nebraska Press. L.A. A. (1985). (1985).Personality: Personality:Current Currentcontroversies, controversies,issues, issues,and anddirections. directions. Annual Review Annual Review Pervin, L. of Psychology, Psychology,36, 36,83-114. 83-114. of conceptsinin Peterson, D. D. R R. (1989). Interpersonal Interpersonal goal goal conflict. conflict. In In L. L. A. A. Pervin Pervin (Ed.), (Ed.), Goal concepts Peterson, personality personality and andsocial socialpsychology psychology(pp. (pp.327-361). 327-361).Hillsdale, Hillsdale,NJ: NJ:Erlbaum. Erlbaum. Powell, A., Royce, J. R., & Voorhes, B. (1982). Personality as a complex information informationR, & processing system. system. Behavioral BehavioralScience, Science,27,27, 338-376. 338-376. processing Powers, W. W. T. T. (1973). Thecontrol controlofof perception. Chicago: Aldine. Powers, (1973). Behavior: Behavior: The perception. Chicago: Aldine. systems: Some spadework at the Powers, W. T. (1978). Quantitative analysis of purposive systems: foundations of of scientific scientific psychology. psychology. Psychological PsychologicalReview, Review,85,85,417-435. 417-435. foundations Roberson, L. L. (1989). Development and and validation vaHdation of of the the Work Work Concerns Concerns Inventory: Inventory: A A Roberson, (1989). Development measure of of employee employee work work goals. goals. Organizational OrganizationalBehavior Behaviorand andHuman Human Decision measure Decision ProProcesses,44, 44,345-367. 345-367. cesses, & Wolchik, S. A. (1988). Personal goals and interpersonal support and Ruehlman, L. S., & Journalof ofPersonality Personality hindrance as factors in psychological distress and well-being. Journal and Social Social Psychology, Psychology,55, 55,293-301. 293-301. Runyan, W. M. M. (1990). (1990). Studying Studying individual individual lives lives and and the the structure structure of of personality personality psychology. psychology. In A. I. Rabin, R. A. Zucker, R. A. Emmons, & & S. Frank (Eds.), Studying Studyingpersons persons and lives (pp. 10-40). New York: York: Springer. lives (pp. Ryff, C. D. (1987). The The place place of of personality personality and and social socialstructure structureresearch researchin insocial socialpsychology. psychology. Journal Psychology, 53,53, 1192-1202. Journalof ofPersonality Personalityand andSocial Social Psychology, 1192-1202. Semmer, N., & & Frese, M. (1985). Action theory in clinical psychology. In M. Frese & & J. Sabini (Eds.), Goal directed concept of of action in psychology (pp.(pp. 296directed behavior: behavior:The The concept action in psychology 296NJ: Erlbaum. 310). Hillsdale, NJ: Smith, C. Handbook ofthematic content analysis. Smith, C. P. P.(Ed.). (Ed.).(1992). (1992).Motivation Motivationand andpersonality: personality: Handbook of thematic content analysis. New York: York: Cambridge University Press. W. D. D.(1992). (1992).Validity Validityof ofquestionnaire questionnaireand andTAT TATmeasures measuresofofneed needfor forachievement: achievement: Spangler, W. Two meta-analyses. meta-analyses. Psychological PsychologicalBulletin, Bulletin,112, 112, 140-154. 140-154. Two & Chester, N. L. (1982). Sex differences in human social motives. In A. Stewart, A. J., & Stewart (Ed.), (Ed.), Motivation Motivationand andsociety society(pp. (pp.172-218). 172-218).San SanFrancisco: Francisco:Jossey-Bass. Jossey-Bass. Stewart Stone, A. A., Cox, D. S., Valdimarsdottir, H., & & Neale, J. M. (1987). Secretory IgA as a measure of of immunocompetence. inmiunocompetence. Journal Journalof ofHuman HumanStress, Stress,13,13, 136-140. measure 136-140. Tomkins, S. S. (1987). Script theory. In 1. J. Aronoff, A. I. Rabin, & &R R. A. Zucker (Eds.), The The emergenceof ofpersonality personality(pp. (pp.147-216). 147-216).New NewYork: York:Springer. Springer. emergence

S12 512

ROBERT EMMONS ROBERT A. A. EMMONS

Veroff, J. (1983). Contextual determinants of personality. Personality andsocial socialPsychology Psychology Personality and Bulletin, Bulletin, 9, 9, 331-343. Veroff, J. (1986). (1986). Contextual factors in the normal personality. In R. L. Rosnow & M. sciences (pp.(pp. 147Georgoudi (Eds.), Contextualism Contextualismand andunderstanding understandingininbehavioral behavioral sciences 147167). New York: Praeger. & Feld, S. (1970). (1970). Marriage Marriageand and work work ininAmerica. America. New New York: York: Van Van Nostrand NostrandVeroff, J., & Reinhold. Veroff, J., & & Smith, D. A. (1985). (1985). Motives and values over the adult years. Advances in Motivation and Achievement, Achievement, 4,4,1-53. 1-53. 120-121. Warren, W. H. (1987). "Behavior" versus "action." Contemporary ContemporaryPsychology, Psychology,32, 32,120-121. classification of human motives. Wicker, F. W. (1984). Categorical goal hierarchies and the classification Journal of of Personality, Personality,52, 52,285-305. 285-305. understanding:AA computational computationalapproach approachtotohuman human (1983). Planning Planning and understanding: Wilensky, R. (1983). reasoning.Reading, Reading,MA: MA:Addison-Wesley. Addison-Wesley. reasoning. Winter, D. G. (1987). Leader appeal, leader performance, and the motive profile ofleaders of leaders and followers: A study of American presidents and elections. Journal of Personality of Personality and Social Psychology, Psychology, 52, 52,196-202. 196-202. (1991). Measuring personality at a distance: Development and validation of of Winter, D. G. (1991). an integrated system for scoring motives in running text. In A. J. Stewart, J. M. Healy, Perspectivesininpersonality: personality: Approaches understanding lives Approaches toto understanding lives Jr., & D. J. Ozer (Eds.), Perspectives 3, pp. (Vol. 3, pp. 59-89). London: Jessica Kingsley Publishers. Personality: Analysis and interpretation of lives. interpretation of lives. New York: Winter, D. G. (1996). Personality: McGraw-Hill. & Carlson, L. (1988). (1988). Using motive scores in the psychobiographical study Winter, D. G., & of an individual: The case of Richard Nixon. Journal of Personality,56, 56,75-102. 75-102. of Personality, & Stewart, A. J. (1978). (1978). The power motive. In H. London & & J. E. Exner, Jr. Winter, D. G., & of personality personality (pp. 391-448). New York: Wiley. (Eds.), Dimensions of of selective selectiveattention attentionininthe theattitude-behavior attitude-behavior relationship. Doctoral (1988). The role of Young, J. (1988). relationship. Doctoral dissertation, dissertation. University of Minnesota, Minneapolis. Zirkel, S., & Cantor, N. (1990). (1990). Personal construal of of life tasks: Those who struggle for independence. Journal of of Personality Personalityand andSocial SocialPsychology, Psychology,58, 58,172-185. 172-185.

21 CHAPTER 21

THE EMOTIONS AN INTEGRATIVE INTEGRATIVE APPROACH AN ApPROACH

JAMES JAMES R. R . AVERILL AVERILL UNIVERSITY OF OF MASSACHUSETTS, MASSACHUSETTS, AMHERST AMHERST UNIVERSIlY

I INTRODUCTION INTRODUCTION I.

If you want to to know what people are like, ask about their loves and hates, hopes and fears, joys and sorrows, and the myriad of other emotions that lend meaning to their lives. But what is is an emotion?

R n. THE THE DOMAIN DOMAIN OF OF EMOTION EMOTION

is—until we are asked to give a definition. Then we We all know what an emotion is-until 600 words (nouns, verbs, verbs, seem to be at a loss for words. Or are we? Roughly 550 to 600 adjectives, and adverbs) in the English language have aa rather clear-cut emotional connotation (Averill, 1975; 1975; Johnson-Laird Johnson-Laird & & Oatley, Oatley, 1989; 1989; Storm Storm & & Storm, Storm, 1987). 1987). Some languages have a larger vocabulary to describe emotions than does English; others have aa much smaller vocabulary (Russell, 1991). For example, Lutz (1982) no general term equivalent to to "emotion" among the found no the Ifaluk, aa people of of 58 words that she she considered unambiguously emotional in in Micronesia, and only 58 countmg words is of little Httle interest. connotation ("about our insides"). But simply counting do the the words signify? And And more generally, what is the relation between What do is the language and emotion? IIANDBOOK OF PERSONA.LITY PSYCHOLOGY

HANDBOOK

OF PERSONALITY PSYCHOLOGY

S13

513

COPYRIGHT © 1997 1997 ,y BY ACADEMIC ACADEMIC Puss PRESS.. CoPYII GHT C ALL IIGH15 OF lEPRODUcnON IN ANY roRM lESERVED.

ALL RIGHTS OF REPRooucnoN IN A^fY FORM RESERVED.

514 514

JAMES R. AVERILL RAvERiLL

A. The The Meaning Meaning of of Emotional Emotional Concepts Concepts A. The class of emotions is held together by what Wittgenstein (1953) called "family ''family resemblances." To illustrate, consider an actual family, the Smiths. Most members family have big ears, but not everyone; most have hazel eyes, but not of the Smith family everyone; and most have stocky builds, but not everyone. No single member of the family need have all these characteristic features (big ears, hazel eyes, and a stocky family forms a recognizable unit. build), yet the family Although opinions vary (cf. Wierzbicka, 1992), emotions, emotions, too, too, form form aa recogniz recognizable unit based on family resemblances (Fehr & Russell, 1984; 1984; Shaver, Schwartz, of Kirson, & O'Connor, 1987). 1987). And like the big ears, hazel eyes, and stocky build of the Smith family, three features are prototypic of emotions, namely, passivity, intentionality, and subjectivity. 1. i. Passivity Passivity

from the ancient Greeks to about the middle of the For most of Western history, from 18th century, what we now refer refer to as emotions were called passions. The term from the Greek pathe via the Latin pati, pari, "passion" and its cognate "passivity" stem from passiones. suffer or undergo change, particu particupassiones. The root meaning of these terms is to suffer larly for the worse. Although it as passions, passions,the theconnota connotait is no longer common to speak of emotions as (of being "overcome") is implicit in emotional concepts. We "fall" tion of passivity (of in love, are "gripped" by anger, "can't help" but hope, and so forth. Because emotional concepts connote passivity, emotions themselves have often often been likened to reflexes and simple sensory experiences, which are also beyond personal control. As will be discussed below, however, people typically have more control over their emotions than the connotation connotation of passivity would suggest. 2. Intentionality Intentionality

Intentionality is another of those terms (like passion) that has a long history in psychological thought, but that can be easily misunderstood if interpreted in its ordinary, everyday meaning. Emotions are intentional in the sense that they are about something. People cannot simply be angry, they must be angry at something, afraid of o/something, hopeful/or something, hopeful for something, and so forth. That "something" is the afraid intentional object of the emotion, the result of an evaluative judgment or appraisal (Solomon, 1993). Intentionality helps distinguish emotions from reflexes, which presume no cognitive intermediary. Intentionality also renders suspect any theory that attempts Jamesto reduce emotional phenomena to simple sensory experiences, such as the James Lange theory. Sensations, whether of a toothache or of bodily arousal, are not about anything; they do not point beyond themselves in the way that emotions do. Intentionality does not, however, distinguish emotions from from rational judgments, which also are about something.

CHAPTER CHAPTER 21 21

THE THE EMOTIONS EMOTIONS

515 515

3. jectivity 3. Sub Subjectivity In everyday discourse, it is common to speak of emotions as subjective and of of rational judgments as objective. These terms (subjective and objective) are ambigu ambiguous, and and hence hence the the contrast contrast between between them them is is easily easily misunderstood. misunderstood. On On one one interpre interpreous, tation, emotions are subjective in the sense of being biased, whereas factual factual judg judgments are impartial (dispassionate). However, not all emotional judgments are misguided, and presumably presumably rational judgments can be, and often often are, self-serving self-serving and and ill-founded. Hence, this this contrast contrast is is not not informative informative of of the the distinction between between emotional and and rational rational judgments. judgments. emotional At relationofofthe theobject object At the the most most fundamental fundamental level, level, subjectivity subjectivity refers refers to to the the relation to the the subject subject of of experience. experience. For For example, if if on on aa rainy rainy day day II see see aa rainbow, rainbow, the the to rainbow is is the the object object of of my experience; I am am the the subject, subject, the theone onewho whosees. sees.Other Other rainbow my experience; persons could could presumably presumably see see the the same same rainbow, rainbow, but but not not exactly exactly the the way way II do. do. When When persons the unique contributions of individual observers are subtracted, what remains is an the unique contributions of individual observers are subtracted, what remains is an "objective" description of the rainbow as, for example, might be found in a book on "objective" description of the rainbow as, for example, might be found in a book on optics. When When the the emphasis emphasis is is on on the the seer seer rather rather than than on on the the seen, as in in an an artist's artist's optics. seen, as rendition of a rainbow, the description is "subjective." Emotions are subjective in this this rendition of a rainbow, the description is "subjective." Emotions are subjective in sense; that is, they are an attribute of the subjective ("inner") pole of experience, sense; that is, they are an attribute of the subjective ("inner") pole of experience, itself. In In many many instances, instances, this this relation relation can can rather than than of of the the external external object object in in and and of of itself. rather itself be be objectively objectively assessed, assessed, for for what what is is beautiful beautiful or or ugly, ugly, beneficial beneficial or or harmful, harmful, is itself is not simply simply aa matter of individual individual judgment. judgment. Society Society and, and, more more indirectly, indkectly, biology biology not matter of stipulate the the kinds kinds of of relations relations that that subserve subserve the the various various emotions. stipulate emotions.

B. The The Organization Organization of of Emotions Emotions B. We have now established some of the (admittedly fuzzy) fuzzy) boundary conditions for distinguishing distinguishing emotional emotional from from nonemotional nonemotional phenomena, phenomena, namely, namely, passivity, passivity, for intentionaUty, and and subjectivity. subjectivity. We We have have yet yet to to map map the the internal internal structure structure of of the the intentionality, emotional domain-how domain—how emotions emotions are are organized organized with with respect respect to to one one another. another. Two Two emotional general approaches approaches have have been been taken taken to to this this issue: issue: categorical categorical and dimensional. dimensional. general

1. Categorical Categorical Approaches Approaches 1. A categorical approach, as its name implies, assumes that emotions can be grouped into into relatively relatively homogeneous homogeneous categories, categories, and and that that these these categories categories are are systematically systematically related to one another, typically in a hierarchical fashion. A zoological taxonomy is a familiar of familiar example of a categorical approach to classification. A number of taxonomies of emotion have been proposed, based primarily on theoretical consider consider1991; Ortony, Clore, Clore, & Collins, Collins, 1988). We We will will ations (e.g., (e.g., de de Rivera, Rivera, 1977; ations 1977; Mees, 1991; consider in in detail detail one one empirically empirically derived derived taxonomy, taxonomy, since since it it comes comes closest closest to to our our consider folk classification classification of of emotion. emotion. folk 590 emotional emotional Storm and and Storm Storm (1987) investigated investigated the the semantic semantic relations relations among among 590 Storm terms. Four Four highly highly educated educated English English speakers speakers served served as as expert expert judges. judges. Through Through a terms. process of of repeated repeated classification, discussion, discussion, and and reclassification, reclassification, they they were were able able to to process organize 525 of of the the words words into into aa taxonomic tree, tree, as as illustrated illustrated in in Figure Figure 1. organize

516 516

JAMES JAMES R. R.AVERILL AVERILL Emotions

Class

Family

Genus

I

Negative

Positive

1

Bad awful

0

A

m

rj.

^^J^t . wonderful

Neutral

Passive

Active

o | x | x §8 ^ | x |

Species S

|

*

*

i

S

S

8

S

g

|

FIGURE11 AAtaxonomic taxonomictree treedepicting depictingthe thehierarchical hierarchicalorganization organizationof of525 525emotional emotionalterms. terms.Only Onlycatego categoFIGURE x represents aa group of 11 to 20 closely ries at the generic level and above are named in the figure. An x

terms; an 0 o represents a covert category (i.e., aa category that has no name in English). Based related terms; on data from Storm and Storm (1987).

Four levels of the taxonomic tree are distinguished in Figure 1. 1. For ease of reference, these levels are labeled (from top to bottom) roughly in the manner of "class," "family," "family,'* "genus," "genus," and "species." "species." At the a zoological taxonomy, namely, "class," highest level of classification, three broad classes are distinguished, consisting of negative, positive, and neutral terms. At the next level, seven families of emotional concepts are identified, although only four of these (bad/awful, good/wonderful, passive and active) are named in ordinary language. These seven families subsume 20 genera of emotional concepts (sadness, pain, anxiety, fear, etc.), which in turn tum encompass 61 61 species. Each species could be further subdivided into smaller units (varieties), but these are not identified in Figure 1.^ 1.1 1^The emotions Thetaxonomy taxonomypresented presentedininFigure Figure11isisbased basedon onclass classinclusion; inclusion;that thatis,is, emotions(or (oremotional emotional concepts) lower in the hierarchy are presumed to be varieties of emotions higher in the hierarchy. alsobe bebased basedon onpart-whole part-wholerelations, relations,asasininanatomical anatomicalschemes. schemes.The The heart, for example, Taxonomies can also heart, for example,

part of, of, not not aavariety varietyof, of, the thecardiovascular cardiovascularsystem. system.Similarly, Similarly,ininsome somerespects respectsanger angermay maybebeconsidered considered is part of jealousy, jealousy, rather than than jealousy being aa subvariety subvariety of anger, anger, as as in in Figure Figure 1. 1. Still other other relations relations a part of cause-effect, similarity) are possible and sometimes sometimes used used in in both both folk folk and and scientific scientific classifications; classifications; (e.g., cause-effect,

may be be linked linkedto to anger angeras asaacontributing contributingcause, cause,or orbecause because in the case of emotions, for example, jealousy may both may result in similar behavior (e.g., aggression). To the extent that the emotions are related to

one another another on on bases bases other other than than class class inclusion, inclusion. Figure Figure 11 gives gives aa misleading misleading picture. picture. Perhaps Perhaps most most one importantly, superordinate-subordinate superordinate-subordinate relations relations may may be be reversed, reversed, depending depending on on the the context. context. For For exam examimportantly, ple, in in one one context context jealousy may may be be considered considered subordinate subordinate to to anger, anger, and and vice vice versa versa in in another another context context ple, (cf. Storm, Storm, & & Jones,

1996). In spite of these considerable qualifications. 1996). qualifications, Figure 11 does illustrate

important points points about about our our folk folk classification classification of of emotion, and and it it will will be be referred referred to to frequently frequently in important subsequent discussion discussion (for (for aa similar similar taxonomy, taxonomy, see see Shaver Shaver et et al, subsequent a!.,

1987). 1987).

CHAPTER CHAPTER21 21 THE THEEMOTIONS EMOTIONS

517 517

Not all categories indicated in Figure 1, particularly particularly at the higher levels of of generality, are explicitly recognized in the English language. This is common in folk taxonomies (Berlin, Breedlove, & Raven, 1968). The unnamed or "covert" categories are indicated by an "O" "0" in Figure 1; 1; named categories are indicated by an "X." Actually, X represents not a single name, but a group (from 1 to 20) of closely related terms. The TTie names actually listed in Figure 1 are those considered most representative of their respective groups. Storm and Storm (1987) emphasize that theirs is a taxonomy of of emotional words and concepts, not of emotions per se. This raises the question: What is the general answers have been offered offered relation between language and emotion? Three general to this question. Emotions—at least certain fundamental fundamental or basic emotions-are emotions—are biologi biologi(a) Emotions-at of emotion cally based and hence independent of language. That is, the language of (cf. Johnson-Laird Johnson-Laird & Oatley, 1989). reflects preexisting divisions "in nature" (cf. (b) Emotions are not divisible into discrete units; whatever distinctions are fundamental about emo emomade in language reflect cultural exigencies, not anything fundamental full range of tions per se. People in all cultures experience the full of possible emotions, differently (cf. subsequent sec secalthough they may conceptualize those experiences differently tion on dimensional approaches). of experiencing experiencing the full range of of emotions; (c) All people are initially capable of culture—as reflected in language-enables language—enables certain possibilities to be real realhowever, culture-as ized, while it closes off off other possibilities. Without special training training (acculturation), emotions of people in one culture can no more experience the emotions of another culture than they can understand the language of another culture.

Some variation of of alternative (c) seems most consistent consistent with current evidence and is the one adopted in this chapter. To adumbrate briefly, human beings have evolutionary history, and the importance of of our biological heritage cannot be an evolutionary of adaptations adaptations that are passed from from one denied. But societies also have a history of biological potentials generation to the next. These social adaptations interact with biological to form the various emotions actually observed in a culture. Language serves as a catalyst for that interaction. Thus, although emotional terms are not to be identified with emotions per se, neither are they simply labels arbitrarily pinned on indepen independently existing entities [alternative (a)]. 2. Dimensional Dimensional Approaches Approaches

Some theorists maintain that emotions do not form discrete categories, but shade of imperceptibly into one another [cf. alternative (b) mentioned above]. Theorists of this persuasion believe that the emotions can be mapped onto an "affective space" of statistical techniques (e.g., factor defined by two or more dimensions. A variety of dimenanalysis, multidimensional scaling) have been used to identify the relevant dimen sions, and the data analyzed have included facial expressions, self-reported self-reported affect, and emotional concepts. The results of many such studies can be summarized briefly (for details, see Larsen & Diener, 1992). Two bipolar dimensions are repeatedly

S18 518

JAMES R. R AVERILL

observed-evaluation observed—evaluation (negative-positive) and activation (aroused-unaroused). Often, a third and even fourth fourth dimension have been found, but such additional dimensions typically account for for a small proportion of the variance, and their nature tends to vary from one study to another. from Figure 2 illustrates the approximate location of 24 emotion terms within the affective affective space defined by the evaluation and activation dimensions. Research by Russell (1983) suggests that the two-dimensional space is consistent across a variety of language groups, including Chinese, Japanese, Gujarati, and Croatian. In spite of the apparent universality of the evaluation and activation dimen dimensions, disagreements remain. For example, Watson and Tellegen (1985) argue for axes, as indicated by the dashed lines in Figure 2. 2. Such a a 45° rotation of the axes, rotation changes the meaning of one dimension to positive aroused aroused affect versus negative unaroused affect (e.g., excited versus bored), and of the other dimension to negative aroused affect versus positive unaroused af fect (e.g., alarmed versus affect serene). Moreover, Haslam (1996) has presented data indicating that emotional concepts are not continuously distributed throughout the affective affective space, as a pure O USED AROUSED AR NEGATIVE NEGATIVE AROUSED AROUSED AFFECT AFFECT

X /

DEFIANT DEFIANT s^, ALARMED AURMED " INCENSED ""s, INCENSED ,

NEGATIVE NEGATIVE

,

,

, , ENVIOUS , , DISGUSTED , , , EMBARRASSED

ASHAMED DISAPPOINTED '"

NEGATIVE UNAROUSED UNAROUSED AFFECT AFFECT

FIGURE 22 FtGIlU

'"

'"

ASTONISHED \ ASTONISHED EXCITED ^'"^ '" EXCITED '" '" EAGER EAGER '"

'"

'"

/

/

'"

'"

,

'"

'"

,

'"

DAZED / DAZED / BORED BORED /' X LETHARGIC LETHARGIC

'" ,

'"

,

'"

,

'"

'"

'"

'"

",

'"

"

'" ,

AMUSED PROUD AFFECTIONATE

" '

'"

POSITIVE AROUSED \ AFFECT

NOSTALGIC HUMBLE SERENE

,

SHY SOLEMN SLEEPY

OSITIVE P POSITIVE

,

,

,

,

,

,

,

POSITIVE POSITIVE UNAROUSED UNAROUSED AFFECT AFFECT

OUSED UNAR UNAROUSED

Organization of of 24 24 emotional emotional terms terms in in aa two-dimensional two-dimensional affective affective space. space. The The horizontal horizontal Organization dimension represents evaluation evaluation (negative (negative versus versus positive positive affect), affect), and and the the vertical vertical dimension dimension activation activation dimension represents (unaroused versus versus aroused aroused affect). affect). The The dashed dashed lines lines represent represent aa 45° 45** rotation rotation of ofthe the axes, axes, as assuggested suggested (unaroused by Watson and Tellegen (1985). The placement of terms may vary depending on context; the placement shown is based on data from Averill (1975) and and Russell Russell (1980). (1980).

519 519

CHAPTER 21 21 THE THEEMOTIONS EMOTIONS CHAPTER

dimensional approach would suggest, but form meaningful clusters within the space. As human beings, we tend to categorize events, and the emotions are no exception. complications aside, we want to ask more generally, generally. What is the But leaving such complications of a dimensional approach as opposed to a categorical approach? value of (aroused-unaroused) di diThe evaluation (negative-positive) and activation (aroused-unaroused) mensions have been central to a number of theories of emotion, either by themselves (e.g., the pleasure-arousal pleasure-arousal theory theory of Reisenzein, 1994) or as elements in more dimeninclusive theories (e.g., Mandler, 1984; Schachter, 1971). The basis of these dimen explanation in terms of cognitive and/or underlying physio physiosions therefore needs an explanation unUkely that a dimensional approach can logical mechanisms. Nevertheless, it is unlikely framework for the analysis of emotion. An analogy will illus illusprovide an adequate framework largetrate the problem. Animals can be arranged along dimensions, for example, large theoretical small and ferocious-tame. Such dimensions (size, ferocity) are of both theoretical aniand practical importance; they cannot, however, substitute for a taxonomy of ani mals based on species, genera, and so forth. Similarly, a dimensional approach to classification of of emotions, although valuable for some purposes, is no substitute the classification for a categorical approach. III. A A FRAMEWORK FOR THE ANALYSIS OF EMOTION HI. of inquiry must start with some assumptions about the appropriate units of Any area of analysis. In the study of emotion, three main possibilities exist, each with somewhat different theoretical implications: (a) emotional syndromes, (b) emotional states, different emotional reactions. To introduce the differences differences among these units, con conand (c) emotional following emotional emotional terms: sider the following A A

B B

C

C

Anger

Angry

Attacking

Fear

Afraid

Grief Grief

Sorrowful

Love

Lustful

Running Weeping Kissing

A are all abstract nouns. They refer to emotional syn synThe terms in column A dromes. dromes, A A syndrome is a theoretical entity; it exists "out there," so to speak, but only as an abstraction or explanatory device. To borrow an analogy from Plutchik (1980), when used to refer to syndromes, emotional concepts are analogous to the of an atom in chemistry. Whether atoms actually exist in the manner concept of of protons and neutrons orbited by electrons) conceptualized (e.g., with a nucleus of is a moot question. Atoms are inferences we make to explain chemical reactions. Similarly, emotional syndromes are inferences we make to explain the reactions of people. of Now consider the terms in column B. They are all adjectives that describe the emotional state state of of the individual. An emotional state is a temporary ((episodic) episodic)

520 520

AVERILL JAMES R. RAvERiLL

disposition on the part of an individual to respond in a manner consistent with a corresponding emotional syndrome. Dispositional variables are quite common in the physical sciences, although they have often been viewed with skepticism by psychologists. Combustibility is a dispositional variable. A combustible material is liable to burst intoflames flames if an appropriate stimulus is applied. Similarly, aa person in an angry state is liable to "explode" if appropriately "triggered." Dispositional material (as (as in in the the case case of of combusti combustiproperties may be enduring characteristics of a material bility) or they may be transitory (as in the "attractiveness" of a reversible magnet). psychological dispositions, emotional states are transitory; other psychological As psychological dispositions (e.g., temperamental traits) are more enduring. The terms in that is,is, thethe actual responses incolumn columnCCrefer refertotoemotional emotionalreactions; reactions; that actual responses an individual might make when in an emotional state. It is important to note that no single response, or type of response, is necessary or sufficient for the attribution of emotion. For example, a person who is in an angry state may act in a variety of of different ways (e.g., physically or verbally attacking the instigator, plotting revenge, talking the incident over with a neutral party, withdrawing affection), all more-ormore-or less consistent with our concept of anger as a syndrome. To conflate emotional syndromes, states, and reactions is to commit what "categorymistake"; mistake";that thatis, is,to tointerpret interpretaavariable variablebelonging belonging Ryle (1949) has called aa "category to one logical category as though it belonged to another. Ryle uses the example of mistake. After observing a young boy watching a parade to illustrate aa category mistake. the marchers, bands, and so forth, the boy wonders, "Where is the parade?" parade?" His mistake is to assume that the parade belongs to the same logical category as the constituents that make up the parade. A particularly common category mistake is to identify emotional syndromes with specific reactions (e.g., physiological arousal, facial expressions, or subjective experience). Such an identification can be grossly misleading if it is assumed (as it often is) that what is true of the reaction is also true of the syndrome; for example, because some emotional reactions, such such as asfacial facialexpressions, expressions,last lastonly onlyaafew fewseconds, seconds, the emotions themselves (i.e., as syndromes) syndromes) must must be be correspondingly correspondingly brief. brief. Return Returning to Ryle's example of the young young boy boyand andthe theparade, parade,one onemight mightjust justasaswell wellassume assume that because some marchers pass quickly by, the parade itself must pass quickly. In addition to the three variables just described (emotional syndromes, states, and reactions), several other types of variables figure figure prominently in the study of trait.Whereas Whereasemotional emotionalstates statesare areshort shortemotion. One of these is the personality trait. term dispositions to respond, traits are relatively long-enduring predispositions. For example, a generally imperturbable individual might be in an anxious state due to unusual circumstances, whereas another be prone prone to to another (trait-anxious) (trait-anxious) individual individual might be frequent bouts of anxiety even in relatively benign circumstances. Many emotional concepts (like anxiety) can refer factthat thathas hasbeen been refer to either state or trait trait variables, variables, aafact the source of considerable confusion in the study of emotion (Spielberger, 1966). Casting our conceptual net even more widely, it is not uncommon to find find emotions linked with biological biological systems systems of behavior, or what used to be called instincts. McDougall (1936), for example, argued that anger is experienced when

CHAPTER 21 21 CHAPTER

THE EMOTIONS

521

the aggressive instinct is aroused, that fear is experienced when the avoidance instinct is aroused, and so forth. Even though the notion of instinct has gone out of vogue, the idea that emotions-particularly of emotions—particularly "basic" emotions-are emotions—are aspects of our biological heritage is is aa very common assumption among contemporary psycholo psychologists (cf. Ekman, 1984; Izard, 1991; 1991; Plutchik, 1980). A resort to biology is one way to account for the experience of passivity, of of being overcome, during emotion. But biology is not the only source of "instinctive" behavior. As Fox (1971) has pointed out, if a species were to lose its dependency on biological systems (as human beings largely have), then a certain class of cultural behavior would have to become like instincts. Such social instincts-or instincts—or what we behavior social systems systemsof of behavior-would behavior—would be be largely largely unconscious unconscious and and automatic, automatic, so so as as call social not to require undue deliberation for their operation, and they would be common of a society. They would also be, we might add, as passionately to large segments of felt as as any any biological biological instinct. instinct. felt

A. Levels Levels of of Organization Organization A. important roles in We have now outlined six kinds of variables that have played imporrant biological psychological theories of emotion: reactions, states, syndromes, traits, and biological and social systems of behavior. Figure 3 organizes these variables into a general framework for for the analysis of emotion. In explaining this figure, it is helpful helpful to framework begin at the top (Level I) and work down. 1. i. Level II

biologiBoth historically and logically, the origins of behavior lie in the individual's biologi potentials. The The sum sum of of aa person's person's biological biological potentials potentials isis his his or or her her cal and social potentials. genotype, i.e., the genetic endowment drawn (through the parents) from the gene of the genotype most relevant relevant to emotions is that pool of the species. The aspect of as discussed above. Biological systems systems responsible for biological systems of behavior, as manifested in responses that contribute to the survival of the species. Examples are manifested forms of aggression, and at atinclude aspects of reproductive behavior, various forms tachment. In a manner analogous to biological potentials, we may speak of social poten potentials or, more specifically, of social systems of behavior-institutionalized behavior—institutionalized patterns of response that help ensure the survival of a society. Whereas biological systems are encoded in the gene pool of of the species, social systems are encoded in the symbols, artifacts, and customs of society. During socialization, the individual is endowed with the potential for for relevant behaviors, a sociotype drawn from from the total repertoire repertoire of of behaviors behaviors available available to to members of of the the society. society. total 2. II 2. Level Level 11

Biological and social potentials interact interact to form, at the next lower level of of organiza organizacapacitiesor orpsychological psychologicalpredispositions. predispositions.This Thisisisthe the fundamental capacities tion, a person's fundamental first level of organization that can be measured directly. It represents the "source

522 522

JAMES R. R AVERILL L E V E L II LEVEL BIOLOGICAL BIOLOGICAL AND AND SOCIAL SOCIAL POTENTIALS POTENTIALS

LEVEL LEVEL IT n FUNDAMENTAL FUNDAMENTAL CAPACITIES CAPACITIES AND AND PREDISPOSITIONS PREDISPOSITIONS

1

Rules of Emotion Emotion

LEVEL LEVEL ill in SPECIFIC SPECIFIC ABILITIES ABILITIES

EMOTIONAL SYNDROMES (Social Roles)

7TT

Initiating and Terminating Terminating Initiating iting and and Termii Conditions Conditions Conditions

/ 1 \

LEVEL rv IV LEVEL EPISODIC EPISODIC DISPOSITIONS DISPOSITIONS

rrn ��

i

EMOTIONAL EMOTIONAL STATES STATES

Situational Constraints

Situational Constraints

LEVEL V LEVEL V COMPONENT COMPONENT RESPONSES RESPONSES

FIGURE fliGURE 3

Rl

R2

R3

Rn

A A framework framework for the analysis of of emotional emotional behavior. behavior.

traits" of personality theory, such as the "Big Five" traits" Five" discussed by Wiggins and Trapnell in Chapter 5 of this volume. We would also include within this category such traits as emotional intelligence, defined as the capacity to monitor one's own & Salovey, Salovey, 1995; 1995; and others' emotions accurately and to respond adaptively (Mayer & & Mayer, 1992), and emotional creativity, defined as the capacity Salovey, Hsee, & effective emotional syndromes (Averill, 1994a; to originate novel yet effective 1994a; Averill & Thomas-Knowles, 1991).

CHAPTER CHAPTER Z1 21

THE THE EMOTIONS EMOTIONS 3. 3.

S23 523

Level lll III

A capacity to A distinction must be made between a capacity to respond respond (Level (Level II), II), on on the the one one hand, and the ability to respond (Level III), on the other. To illustrate, consider a person who has fine pitch discrimination, has the biological potential forfine discrimination, motor motor coordina coordination, and whatever other biological endowment might be necessary to play a musical instrument. Assume also, that this person has been raised in a social environment offering a strong social potential). in which music plays an important role (thus offering Such a person would likely possess the capacity to be a good musician. However, before the person would have the ability to be a classical violinist, say, or a jazz pianist, he or she would have to acquire the rules and skills relevant to that musical genre. In other words, abilities are rule-constituted, and capacities are not, or at least not in the same sense. fully further further on, emotional syndromes can best be As will be explained more fully (from a psychological standpoint) or conceptualized at the level of specific abilities (from in terms of social roles (from a sociological standpoint).

4. Level Level IV IV 4. approGiven the ability to engage in a particular emotional syndrome (e.g., anger), appro priate initiating conditions will result in the activation of of an emotional state. As already described, an emotional state is not a specific kind of response. Rather, it episodic disposition dispositiontotorespond respondininany anyof ofaavariety varietyof ofways ways is a relatively short-term, episodic synconsistent with the rules and expectations that help constitute an emotional syn drome. 5. 5.

Level V Level V

The most concrete level of organization comprises the component responses responsesthat that aa imperson might exhibit when in an emotional state, depending upon constraints im posed by the situation. Shortly, we will discuss in detail some of the issues raised different kinds of response (e.g., physiological change and instrumental acts). by different

B. B. Relations Relations among among Levels Levels One source of confusion confusion among various kinds ofvariables of variables isis the fact fact that, in practice, by the interconnected levels of organization can be bypassed. This is indicated indicated arrows at the left of Figure 3. For example, in certain circumstances, the activation of biological potentials (Level I) can directly elicit component responses (Level V), V), short-circuiting fundamental fundamental capacities (traits), specific abilities (syndromes), and episodic dispositions (states). Similar considerations apply to social potentials and component component responses. The salesperson is expected to smile and be friendly, the bereaved is expected to mourn, and the drill sergeant is expected to be belligerent, regardless of their Because short-circuiting short-circuiting isis comcomemotional state at the moment (Hochschild, 1983). Because

524 524

JAMES AVERIU JAMES R. RAvERiLL

differences among variables at different different levels mon, it is easy to overlook logical differences of organization, thus facilitating the kind of category mistakes described earlier. One implication of the above considerations is that component responses are not inherently emotional, but attain that quality only when interpreted (experi(experi differently, human emotional reenced) as part of an emotional syndrome. Put differently, re sponses are reflexively reflexively related to the higher-order meanings that help constitute syndromes. This is indicated in Figure 3 by the bidirectional curved emotional syndromes. arrows connecting component responses (Level V) to emotional states (Level IV) III). The issue of reflexivity is perhaps most problematic or and syndromes (Level III). As here here conceived, conceived, emotional emotional controversial with reference to conscious experience. As feelings are one kind of component response, the nature of which is examined in more detail shortly. IV. rv. COMPONENT RESPONSES Six Six broad categories of emotional responses can be distinguished: cognitive apprais appraisals, physiological change, expressive reactions, instrumental acts, verbal behavior, empirical and theoretical issues. issues.After After introduc introducand feelings. Each raises important empirical (e.g., cognitive appraisals), several frequently frequently ing a particular kind of response (e.g., asked questions regarding its function and significance are addressed.

A. Appraisals A. Cognitive Cognitive Appraisals A sharp distinction is often made between stimuli and responses. The distinction is, however, misleading. misleading. Before Before an an emotional emotional response response can can occur, occur, the the stimulus stimulus must must be appraised for its relevance to the individual (cf. the criteria of intentionality and itself a kind of response, an imposisubjectivity discussed earlier). The appraisal is itself imposi events. A more formal way of stating this fact fact is to say that tion of meaning on events. the intentional object of an emotion. The The nature nature of of the the appraised appraised appraisal produces the further analysis. object is thus an appropriate starting point for further

/. What WhatIs Isthe theAppraised AppraisedOb Object Emotion? 1. ject ofof anan Emotion? Most emotional objects involve three three aspects-the aspects—the instigation, instigation, target, target,and andobjective. objective. The eliciting condition (as appraised by the person) is the instigation; the person or thing at whom the emotion is directed is the target; and the goal of the emotion is the objective. For example, if Mary is angry at her friend Bill for insulting her and wants to "get even," the object of her anger consists of the perceived insult (the instigation), her friend Bill (the target), and getting even (the objective). Not all emotions emphasize all three aspects of an object equally. More than most emotions, love emphasizes the uniqueness of the target, and hence is not fungible, is, easily transferrable from one person to another as circumstances change. that is, Other emotions emphasize both the instigation and the objective, but leave the target relatively free to vary depending on the circumstances. Thus, abstracting

21 CHAPTER 21

THE EMOTIONS

52S 525

we can say say that the object of anger is revenge for wrongdoing from any specific target, we and the object of fear fear is escape from from danger. Still other emotions focus almost target and the objective unspecified. unspecified. exclusively on the instigation, leaving both the target Joy and grief grief are good examples. 2. Is Appraisal Appraisal Necessary Necessaryfor for the theActivation Activation oof an 2. f an Emotional State? Emotional It is often often assumed that the appraised object is the cause of an emotional state. However, as Dewey (1895) emphasized a century ago, the way a person appraises a situation is part of, not antecedent to, the emotion. For example, to perceive a frightening is already to be in a fearful fearful state. Some contemporary theorists bear as frightening far as to argue that emotions are in principle a kind of appraisal or evaluative go so far judgment: fear fear is the judgment that we are in danger; anger is the judgment that we have been wronged; sadness is the judgment that we have been diminished through loss; and so forth forth (Sartre, 1948; 1948; Solomon, 1993). 3. 3.

I ject, What Activates Iff Not Not the Appraised Appraised Ob Object, Activates an Emotional Emotional State?

There are many potential links in the chain of events leading to an emotional state, Under the the influ influfrom the primarily physical to the primarily cognitive (Izard, 1993). Under an emotional emotional state state can cansometimes sometimesbe be activated activated ence of drugs or hormonal imbalances, an no cognitive intervention intervention (e.g., postpartum depression). depression). Similarly, Similarly,as aswill will with little or no feedback from from bodily responses be discussed in more detail shortly, proprioceptive feedback factors unrelated to the emocan facilitate the occurrence of emotion. Motivational factors emo tional episode can also play a role in how a situation is appraised. Love provides a familiar The target target of of one's one's affection affection is is an an idealization, idealization,partly partly based based on on real realfamiliar example: The Finally, at atthe therisk riskof ofempha emphaity and partly a projection of one's own needs and desires. Finally, sizing the obvious, appraisal is not an all-or-none affair; a variety of cognitive steps, not all of which are emotional in and of themselves, are typically involved in the initiation of an emotional state (Scherer, 1984). Once an appraisal is is made, it can-like can—like any other kind of response-stimulate response—stimulate further activity, including reappraisals. When a reappraisal unrealistically unrealistically short further for self-protective self-protective reasons, we may speak of cognitive circuits the emotional state for defense mechanisms (Lazarus, 1991).

4. Is Is Each Each Emotion Emotion Associated Associated with with aa Diff Different 4. erent Object? Object? With the possible exceptions of "free-floating" "free-floating" anxiety, depression, and mystic-like experiences, each emotion is associated with a distinct object (instigation, target, and/or objective). One of the main tasks of theorists is to identify identify the cognitions that help constitute the appraised objects of emotions. For example, example, isis the the instigation instigation beneficial or harmful? harmful? What is the probability of its occurrence? When did-or did—or might—the indetermimight-the instigation occur (past, present, future), or is its occurrence indetermi nate (ongoing)? With regard to the target, who or what mitiated the instigation (oneself, another person, or chance events)? If the target is a person, is he or she

S26 526

JAMES R. RAvERiLL lAMES AVERJU

responsible (deserving of reward or punishment)? With regard to the objective, what possibilities for for action does the situation afford afford (approach, withdrawal, attack, etc.)? In various combinations, answers to questions such as these help account account for the objects of most commonly experienced emotions (Frijda, 1986; Lazarus, 1991; 1991; Ortony et aI., al., 1988; 1988; Roseman, Antoniou, & Jose, 1996; Smith & Ellsworth, 1985; Weiner, 1985).

B. Peripheral Physiological Change Emotions, as the saying goes, are "gut" **gut" reactions. reactions. This This metaphorical metaphorical way way of of charac characterizing the emotions has a long history, dating back to the ancient Greeks. Its most famous modem formulation formulation is that by William James (1890), who proposed fact, and our "that bodily changes changesfollow follow directly directlythe theperception perceptionofofthe theexciting exciting fact, andthat that our changesas asthey theyoccur occurIS ISthe theemotion" emotion'' (p. (p.449). 449).Although Although James James of the same changes feeling of meant his formulation facial expres expresformulation to include bodily changes of of all kinds (including facial sions, instrumental acts, and even central neural activity), most subsequent research focused on the role of visceral activity during emotion (for (for a and speculation has focused review, see Cacioppo, Klein, Bernston, & Hatfield, 1993). 1993). We may call this focus the "strong" version of James's thesis; it implies that different different emotions are associassoci different patterns of visceral activity, and that people are able to discrimi discrimiated with different nate among those patterns in conscious experience. A somewhat weaker version undifferentiated of James's thesis asserts only that visceral arousal contributes an undifferentiated quale or feeling tone to experience, but that the differentiation quafe differentiation among emotions reflects the way a person interprets the situation (e.g., Schachter, 1971). erent Pattern of of 1. /. Is Each Emotion Associated Associated with a DiJl Different Physiological Physiological Activity? Activity?

The answer to this question is a qualified yes. yes. It It requires requires no no sensitive sensitive instrumentation instrumentation and sophisticated statistical techniques to recognize that the visceral changes that accompany sexual arousal, say, are in important respects different different than those that accompany fear, and that still other reactions are common during grief. However, before different emotions with different before we become too sanguine about identifying identifying different different patterns of physiological activity, several important qualifications qualifications must be kept in mind. First, distinctions such as those just noted (e.g., among sexual arousal, fear, and grief) are between broad categories of emotions, that is, among emotions family levels depicted in Figure 1. defined at the genus or even family 1. Emotions defined different biological potentials at such global levels are probably associated with different 3). For example, sexual arousal is related to (systems of behavior; see Figure 3). grief to the breaking of of attach attachreproductive behavior; fear to avoidance of danger; grief ment bonds; and so forth. As explained earlier, earUer, what is true at one (global) level in a hierarchy need not be true at another (more specific) level. With regard to the differential physiological patterning among issue at hand, there is little evidence for differential specificemotions emotionswithin withinaabroad broadcategory category(e.g., (e.g.,between betweenanger angerand andjealousy). jealousy).Among Among specific

21 CHAPTER 21

THEEMOIIONS THE EMOTIONS

527 527

differentiation is based primarily on the appraised object (instiga (instigaspecific emotions differentiation tion/target/objective) tion/target/objective) of the emotion. Second, physiological reactions (e.g., an increase in cardiac output) occur to support possible overt behavior, whether or not a response is actually made. Since the behavior associated with any given emotion can vary greatly greatly depending depending on the person and the situation, so too will physiological reactions. To take an obvious example, the person who faints in fear is undergoing different different physiological changes than the person who flees; flees; and the angry person who writes a nasty letter to the different physiological state than a person engaged in physical combat. editor is in a different 2. 2. How Sensitive Are People to Physiological Physiological Changes Changes when when

Do Occur? They Do Occur?

of ongoing physiological People are not very sensitive to these changes. Self-reports of responses are typically poorly correlated with actual physiological activity (Reed, 1990). The viscera are not particularly sensitive structures and, Harver, & Katkin, 1990). as noted in response to the previous question, only gross distinctions are observed between broad classes of emotions when actual physiological responses are meamea presents aa paradox: paradox:People Peopledo doreport reportdifferential differential patterns patternsof ofphysiological physiological sured. This presents arousal when asked to describe their emotional experiences. To what do those refer? reports refer? emotion reflect It is possible that self-reports of physiological activity during emotion person "turns "turns red" when when angry and gets "cold feet" when cultural stereotypes (e.g., aa person fearful). However, considerable cross-cultural consistency has also been observed in self-reported physiological responses (Wallbott & Scherer, 1988). This This would would seem seem self-reported to obviate an explanation in terms of cultural stereotypes. Again, however, it should focused on emotions defined be noted that research on this issue has focused defined at a very high affective dimensions (e.g., evaluation evaluation and level of generality, or in terms of broad affective is possible possible that that people are able to to discriminate discriminate and report physiological physiological It is activation). It differences between global categories and dimensions of emotion, especially if they differences focus on very intense episodes. It is also possible, as Rim6, focus Rime, Philippot, and Cisamolo (1990) suggest, that there is symbolic overlap between some emotional and physiophysio associated with certain logical constructs. For example, feelings of joy may become associated physiological processes, not because of any actual commonalities in underlying processes, but because both presumably give rise to "warm" feelings. 3. Is Is Physiological Physiological Change ChangeNecessary Necessaryfor for Emotion? Emotion? 3. Physiological change is not necessary for emotion. This answer might seem unduly (differenconclusive, since most contemporary theories regard physiological change (differen for emotion (e.g., Mandler, 1984; Schachter, tiated or not) as a necessary condition for 1971). But before we accept uncritically the notion that physiological arousal is necessary for for emotion, two facts should be noted: First, the belief that one is aroused, or the meaning of the arousal for the individual, may be more important (Reisenzin determining the experience of emotion than is the actual state of arousal (Reisenz 1983; Valins, 1967); 1967); second, noticeable physiological change is a prominent ein, 1983;

528 528

JAMES R. R AVERILL AVERILL JAMES

feature short-term, relatively intense intense emotional episodes (e.g., sudden feature primarily of of short-term, sudden fright fright as as opposed opposed to to longer longer lasting lasting but but still still episodic episodic fear), fear), and and of of some some emotional emotional syndromes but not (e.g., fear syndromes but not others others (e.g., fear as as opposed opposed to to hope; hope; Averill, Averill, Catlin, Catlin, & Chon, Chon, 1990). 1990). Why, Why, then, then, has has physiological physiological change change become become so so closely closely identified identified with with emo emopsychological tional responses? As alluded to in response to the previous question, psychological phenomena associated with physiological responses on the basis of of phenomena may become associated rather than on any empirically demonstrated functional shared symbolic meaning rather demonstrated functional relationship. Once such an association is made, it can be extremely tenacious, as of theories of of emotion amply illustrates (Averill, 1974, 1974,1990b). the history of 1990b). c. C. Expressive Reactions

Ever pioneering work on The Expression Ever since Darwin's pioneering Expression of of Emotion Emotion in Man Man and Animals (1872/1965), expressive reactions-especially reactions—especially facial expressions-have expressions—have Animals often controversial role in the study of of emotion emotion (Ekman, 1994; played a central and often proposed that some expressive reactions are are Fridlund, 1994; Russell, 1995). Darwin proposed innate and universal within the human species, and, moreover, that such reactions specific emotions. These are actually two separate hypotheses, are associated with specific although they are often treated as one. For For example, according to Izard and Saxton although often treated (1988), "the best available criteria criteria for for the presence of of a particular particular discrete emotion emotion are the muscle movements or changes in appearance appearance that characterize its innate, universal facial facial expression" (p. 631). The discrete, "fundamental" "fundamental" emotions that can be so characterized are presumably joy, surprise, sadness, anger, anger, disgust, disgust, and and can be so characterized are presumably joy, surprise, sadness, fear. Izard and Saxton (1988) go on to note that "the evidence for the innateness fear. Izard and Saxton (1988) go on to note that "the evidence for the innateness and universality universality of of the expression expression of of the fundamental fundamental emotions is sufficiently sufficiently robust robust consider Darwin's hypothesis as an established axiom of of behavioral science" to consider (pp. 651-652). Any hypothesis in the behavioral sciences that merits the status of of an axiom Any Therefore, we will break break Darwin's hypothesis obviously deserves close scrutiny. Therefore, into its two components and evaluate each part separately. We will then consider moderating effect effect that facial facial expression may have on a person's own experience experience the moderating of emotion. of Some Expressive Expressive Reactions Reactions Innate Innate and and Universal? Universal? 1. Are Are Some Research by Eibl-Eibesfeldt Eibl-Eibesfeldt (1970), Ekman (1982), and Izard (1991), among others, Research strongly facial expressions strongly suggests that some facial expressions can be evoked evoked by similar situations because (e.g., danger, separation) across diverse cultures, and even in children who, because they were born bom blind and deaf, have had no chance to learn the expressions from from others. Moreover, homologous expression can be observed in other primate species facial expressions, (Redican, 1982). The issue, then, is less the existence of innate facial significance of of such expressions. but the meaning and significance

CHAPTER 21 21 CHAPTER

THE EMOTIONS

529 529

2. 2. Are Innate Facial Expressions Expressions Associated Associated with with

Specific Emotions? S pecific Emotions?

Innate facial expressions are only indirectly associated with specific emotions. The issue here is similar to that with respect to the relation between physiological response patterns and specific emotions. The demonstration that some facial expres expressions are innate and universal suggests the existence of certain certain biological systems of behavior, for for example, toward aggression, flight, attachment, and the like. These are the biological potentials represented at Level I of Figure 3. As already discussed, there is no direct connection between biological potentials and specific emotional syndromes. An example may serve to reinforce this point. A good deal of evidence suggests that aggressive tendencies are universal in the human species, and that manifested in facial displays and distinct patterns of physiologi physiologisuch tendencies are manifested cal arousal, as well as in direct aggression. However, there is little evidence to specific emotional emotional syndrome syndrome isis always-or always—or even even usually usually— suggest that anger as a specific associated with aggressive tendencies; moreover, aggressive tendencies may enter into a wide variety of emotional and nonemotional syndromes other than anger, such as envy, jealousy, and sadism (Averill, 1982).

3. Can Can Facial Expressions ExpressionsAlter Alter the theExperience Experienceof ofEmotion? Emotion? 3. Facial expressions can modify the experience of of emotion, but only to a modest degree. William James (1890), it may be recalled, proposed that proprioceptive feedback in feedback from from bodily changes accounts for for the experience of of emotion. James included feedback feedback from from the face among the relevant bodily changes, but most early speCUlation of emotional experience. speculation focused on the viscera as the primary source of For reasons reviewed earlier, however, peripheral physiological arousal is inade inadequate to provide the fine discriminations that we we experience among emotions. Some theorists (e.g., Izard, 1991; therefore looked 1991; Tomkins, 1981) have therefore looked to the face as a possible source of emotional experience. The skin and muscles of of the face face are richly innervated and capable of very fine nuances. It is thus not unreasonable to assume that the face might provide the kind of proprioceptive feedback by James. feedback presumed by A meta-analysis of 16 16 studies designed to test the facial-feedback facial-feedback hypothesis suggests that only about 12% 12% of of the the variance variance in in self-reported self-reported emotional emotional experience experience can be accounted for in terms of experimentally manipulated variations in facial Moreover, current data do do not not allow allow any conclusions conclusions expression (Matsumoto, 1987). Moreover, regarding the specificity of the effect, that is, whether an angry expression results in angry feelings, a fearful fearful expression in a fearful fearful feelings, and so forth. About the best that can be said is that facial expressions can have a modest biasing effect effect on a person's mood, for example, along the negative-positive axis depicted in Figure 2 (Izard & Saxton, 1988; 1988; Leventhal & Tomarken, 1986). And even that minor effect effect feedback from the facial musculature (for (for possible need not be due to proprioceptive feedback ahernative mechanisms, see Laird & Bresler, 1992; 1992; Zajonc, Murphy, & Inglehart, alternative 1989).

530 530

RA AVERILL JAMES R. VERJLL

D. Instrumental Instramental Acts Acts D. instrumenIn the psychological literature it is common to distinguish emotional from instrumen tal responses. For example, the two-factor two-factor theory of avoidance learning assumes that fear fear (an emotional response) motivates avoidance (an instrumental response). Similarly, anger presumably motivates aggression, love motivates courtship, and so forth. This is another variation on the distinction between passions and actions. In *'in spite of ourselves"; in action, we move "on purpose." passion, we are moved "in When a distinction becomes embedded in ordinary language, as this one has, it sense, and when imported into scientific discourse, acquires the aura of common sense,' it may be accepted uncritically. That is a mistake. 1. f Emotion Than /. Are Instrumental Acts Less Indicative oof Physiological Expressive Reactions? Reactions? Physiological or Expressive

If anything, instrumental responses are more, not less, important for the identifica identification of emotional states and syndromes than are physiological or expressive reac reactions. This follows, in part, from from our earlier discussion of the appraised object of of objective of the an emotion. One aspect of the appraised object is the goal or objective response. Anger, for for example, is not just a state of physiological arousal or a facial facial grimace occasioned by some appraised wrong; it also involves a desire to correct the wrong. To illustrate the same point in a somewhat different different way, consider that from sadness, anger, joy, or chopping onions. onions. How do we infer infer a person may weep from which emotion is being experienced? In large part, by what the person does (or does not do) in order to change the situation.

E. E. Verbal Behavior Two aspects of verbal behavior can be distinguished, what is said (the content of of speech) and how it is said (e.g., loudness, tone, timing). The latter (expressive) aspect has received most attention from from emotion theorists (Pittam & Scherer, 1993). On a broad conceptual level, however, the expressive aspects of speech involve many of the same advantages and limitations as facial expressions and physiological reactions, which we we have already already considered in some detail. detail. In In this thissection, section,therefore, therefore, we will focus on the instrumental aspect of verbal behavior, that is, on what is said rather than on how it is said. Psychologists tend to distrust self-reports of emotion, and with good reason. It is easier to disguise our emotions with words than with any other kind of of response. However, the distrust of verbal behavior by psychologists is not based simply on the fact that people can, and sometimes do, dissimulate. It is also based on the implicit (and, we believe, mistaken) assumption that the words we use to express emotional concepts are only labels that get attached, sometimes correctly and sometimes incorrectly, to preexisting states of affairs-the affairs—the emotions per se.

CHAYTER CHAPTER 21 21

THE THE EMOTIONS EMOTIONS

531 531

I. Do Emotions Exist Independent o f the /. of the Way They They Are Are Conceptualized? Conceptualized? Emotions do not exist independent of their conceptualization. In order to explain this answer, we must return briefly to an issue discussed earlier, namely, the relation between language and emotion. At the risk of oversimplification, oversimplification, it can be said that the meaning of a word is based on a network of ideas and propositions—a propositions-a kind of "implicit theory" (Murphy & Medin, 1985). 1985). For example, the meaning of of the word "atom" presupposes a (scientific) theory of matter. In a similar manner, the meaning of the word "anger" presupposes a (folk) theory of of emotion. But there is also an important dissimilarity between "atom" and "anger." Physical objects fundamental way as a function function of how they such as atoms do not change in any fundamental are conceived. The same is not true of psychological phenomena. In the human phenomenon named. sciences, the mere act of naming can change the phenomenon differently, emotional concepts are not simply descriptive, they are also Put differently, prescriptive. The network of ideas and propositions that lend meaning to our everyday emotional concepts also help constitute and regulate the behavior denoted by those concepts. Of course, once the meaning of a word has been acquired, verbal from the remainder of of a person's behavior and used to responses can be divorced from conceal as well as to reveal.

2. Are There There Emotions Emotionsffor WhichWe We Have Have No No Words? Words? 2. or Which

Not all concepts find expression in words. We have already noted this fact in connection with the emotional taxonomy presented in Figure 1. It will be recalled al, 1968). that "covert categories" are quite common in folk taxonomies (Berlin et aI., Over time, as individuals and cultures develop, new concepts may emerge for which there are no names. Arieti (1976) has called such emerging cognitions "endocepts," well-formed concepts. Endocepts are particularly impor importo distinguish them from well-formed tant during the early stages of development, whether during childhood, or later during adulthood when a person is grappling with new ideas and experiences. When says, "I know know how I feel, but I cannot put it into words," he or she is a person says, trying to express an emotional endocept. Much poetry, music, and painting can also be viewed as attempts to put into words and other symbolic forms emerging emotional experiences.

3. Do Preverbal In Infants Animals Experience Experience Emotion? Emotion? 3. fants and Animals Only in an extended sense do preverbal infants and animals experience emotion. At birth, the infant's emotional repertoire is limited primarily to the expression of of biological potentials (Level I, Figure 3); 3); socialization and individual experience follows that infants cannot have had little opportunity to exert their influence. It follows experience emotional syndromes which depend, in part, on social norms and rules. fussing of an infant infant is literally a far cry from from the anger of an adult. Similar The fussing. considerations apply to animals. It would be fatuous to deny a continuity between humans and animals in the emotional as in the intellectual domain. This does not

532 532

JAMES JAMES R. R.AVERILL AVERILL

mean, however, animals experience humans, for mean, however, that that animals experience the the same same emotions emotions as as humans, for example, example, that that the the "anger" *'anger" of of aa dog dog on on having having its its bone bone removed removed is is the the same same as as the the anger anger of of aa human human on on being being robbed robbed of of aa prized prized possession. possession. Emotional Emotional continuity, continuity, whether whether ontogenetic or phylogenetic, does not entail sameness (Averill, 1984).

F. Feelings We issue of (the conscious being in We return return now now to to the the issue of feelings feelings (the conscious awareness awareness of of being in an an emotional state), which we introduced introduced earlier in connection with the reflexivity reflexivity of of emotional responses. The terms "feeling" and "emotion" are often often used inter interchangeably. To say, "I feel angry," means much the same as to say, "I am angry." led some theorists to assume that emotions are really feelings, or at least This has led features of emotions (cf. Oatley & Johnson-Laird, 1987; that feelings are essential features Ortony et aI., a l , 1988). Within Within the present framework, framework, feelings are component component re reOrtony they are no more more essential to an emotional state state than are other other kinds of of sponses; they being granted, it is nevertheless true response (e.g., physiological change). But that being that feelings are an important component component of of most emotional states. / . What What Are Are Emotional Emotional Feelings Feelings Feelings Feelings O Of? 1. f?

Emotional Emotional feelings feelings are are not not based based on on any any one one source. source. For For example, example, the the way way we we appraise a situation determines, to a large extent, the way we feel. This fact fact was particularly emphasized by Jung (192111971), (1921/1971), who defined defined feelings as a kind of of evaluative judgment. Thinking, according to Jung, tells us what a thing is (e.g., that generated by hydrogen hydrogen fusion); feeling feeUng tells us what a thing is good for for stars are generated Feedback from from in an evaluative sense (e.g., that stars are like diamonds in the sky). Feedback facial expressions also adds a certain quality peripheral physiological arousal and facial of emotion, as discussed in response to earlier questions. Instru Instruto the experience of whether actual or only desired, make an additional contribution contribution to the mental acts, whether overcome the potential harm? way we feel. Is escape possible? Should I fight to overcome Or is there nothing I can do? Depending Depending on how these questions are answered, a Or feel fear, fear, anger, or depression. depression. person may feel inIn short, cognitive appraisals, physiological and expressive reactions, and in strumental acts all may contribute to emotional feelings. But an emotional feeling feeling strumental not simply a by-product by-product of of other responses; it is a response in its own right. is not pointed out that when people are asked how they feel, they Sarbin (1986) has pointed typically tell a story, a self-narrative. For example, when asked how she feels, the typically respond with a few few uninformative uninformative exclamations, such as Olympic champion may respond "It's incredible," or "It's a wonderful wonderful feeling," and then then launch into a story story about the rigors of of training, the role of of fellow fellow athletes, family family support, the meaning of of the contest, future future plans, and and so forth. forth. 2. Can Can People People Be Be Mistaken Mistaken about about Their Their Own Own Emotional Emotional Feelings? Feelings?

As discussed discussed earlier, emotional feelings are reflexive reflexive (i.e., involve an interpretation interpretation of of one's own experience). It follows follows that people can always be mistaken about the

CHAPTER EMOTIONS CHAPTER 21 21 THE THEEMOHONS

533 533

way they feel. Such mistakes are commonplace in clinical practice. The person who is angry, jealous, envious, afraid, in love, and so forth, is sometimes sometimes the last and not the first to r�cognize the fact. Less common but theoretically recognize theoretically more interesting are occasions when a person feels emotional without being in an emotional state. For example, the ebullience of a person in a hypomanic hypomanic state may be taken, not as a sign of happiness, but of of depression. In some respects, "feeling" an emotion can be compared to "hearing" voices. No matter how vivid and realistic a voice might seem, it will not be considered real unless there is some adequate stimulus to account for for its occurrence. Similar considerations apply to emotional emotional experience. Feelings Feelings for for which there are not adequate instigations are liable to be dismissed as illusory, false, or inauthentic (Averill, 1994b; Morgan & Averill, 1992). No less than other forms of of behavior, feelings can deceive. In fact, if we want to deceive others, it is best first to deceive ourselves through our feelings. V. EMOTIONAL STATES V.

We turn now from component responses (Level V, Figure 3) to emotional states disposi(Level IV, Figure 3). As already described, an emotional state is an episodic disposi syndrome (Level III, tion to respond in a manner consistent with an emotional emotional syndrome Figure 3). That, however, does not say a great deal. Dispositions must be explained in of underlying mechanisms, as when combustibility combustibility (a dispositional property) is terms of of molecular structure. explained in terms of temptaWhen it comes to the mechanisms that underlie emotional states, the tempta processes (e.g., an "emotion system" and tion appears strong to postulate distinct processes a "cognitive system") to correspond to the distinction between passions and actions. The bias of the present chapter is to resist that temptation. An analogy may help clarify the basis for this resistance. Diseases have traditionally traditionally been conceived conceived of of as passions of the body, things that the body "suffers." At At one time it was common to postulate special processes to account for diseased states (including, for example, witchcraft and possession possession by the devil). However, most diseases can be explained in terms of normal physiological reactions to abnormal abnormal conditions (e.g., the presence of pathogens, injury, genetic mutations). There are no special "disease processes" or "laws of disease" in physiology. Similarly, it is most most parsimonious to assume that emotional states can be accounted for in the way ordinary psychological pro psychological processes are organized to meet unusual circumstances. We need not rely on analogy to make a case against the "special process" view of of emotional states. From a cognitive cognitive perspective, the emotions involve all those processes (e.g., attention, perception, memory, concept formation, language, nonemotional behavior. True, some forms of of and problem solving) that enter into nonemotional cognition may be more common during emotional emotional than during more deliberate, processes inherently "rational" states (Epstein, 1994), but that does not make such processes reference to the central neural emotional. Similar observations may be made with reference

S34 534

JAMES JAMESR. R AVERILL AVERILL

mechanisms that mediate emotions. Many of the same brain circuits that are in involved in emotional processes are also involved in "higher-order" thought processes (Davidson, 1994). 1994). At most, one might distinguish neural circuits that have some specificity specificity for for biological systems of behavior (e.g., aggression, attachment, sexual arousal) and for for such broad dimensions of behavior as evaluation (negative (negativepositive) and activation (aroused-unaroused). However, for for reasons already disdis affective dimensions cannot be equated with specussed, biological systems and affective spe cific emotions. In short, an adequate discussion of the mechanisms that mediate emotional states would encompass nearly all of psychology. We must leave the matter at that. VI. EMOTIONAL SYNDROMES SYNDROMES VI.

Fortunately, we need not wait for for an adequate understanding of the mechanisms before we attempt to understand understand the principles by that mediate emotional states before which emotional syndromes are organized. In this respect, we may draw an analogy between emotions and language. The psychological and physiological mechanisms that enable a person to speak are poorly understood. However, that does not particular language is organized organized— prevent us from trying to understand the way a particular English, say, as opposed to Chinese. Languages are organized according to rules rules— the rules of grammer. Emotional syndromes are also organized according to rules rules— the rules of emotion.

A. Emotional Emotional Rules Rules A. Three aspects of rules may be distinguished: constitutive, regulative, and procedural. For ease of discussion, it is easier to speak, not of aspects, but of different different types of rules, depending on which aspect predominates. To illustrate the difference difference among constitutive, regulative, and procedural rules, consider a game such as chess. Some rules (e.g., with regard to the nature of of the pieces and way they may be moved) help constitute the game as a game of the pieces and way they may be moved) help constitute the game as a game of chess, as opposed, say, to a game of checkers. Other rules (e.g., limiting the time chess, as opposed, say, to a game of checkers. Other rules (e.g., limiting the time between regulate how how the the game game isisplayed played on on any any given given occasion. occasion. between moves) moves) help help to to regulate Still other rules, called procedural, help determine strategy-how well the the game game Still other rules, called procedural, help determine strategy—how well is played. Like the game of chess, an emotional syndrome is also constituted constituted and regu regulated by rules, and its successful successful enactment requires skills based on procedural rules. The nature of the various rules is best illustrated by what happens when they are broken or misapplied. Take anger. If a constitutive rule of anger is broken, the response may be interpreted interpreted as a manifestation manifestation of some other condition (a neurotic syndrome, perhaps); it will not be considered true anger. If a regulative rule is be considered inappropri inappropriviolated, the response may be accepted as anger, but it will be ate or delinquent. Finally, if a procedural rule is broken, the failing will be treated

CHAFJER EMOnONS CHAPTER 21 21 THE THEEMOHONS

535 535

more a matter is, in a sense, matter of ineptitude than inappropriateness. The individual is, an emotional klutz. The The acquisition acquisition of of appropriate appropriate procedural procedural rules rules is is aa common common goal goal in in psycho psychotherapy also recognize recognize the therapy (e.g., (e.g., skills skills training). training). All AH emotion emotion theorists theorists also the importance importance of regulative rules in determining the way emotions . are displayed, particularly of across cultures (Ekman, 1982). The role of constitutive rules has been less frequently frequently emotional syndrome of the emotional recognized. Change a constitutive rule, and the nature of itself changes, not just its display or enactment. If If the change is for for the worse, we itself for the better, of emotional speak of emotional disorders (Averill, 1988); if it is for creativity (Averill & Nunley, 1992; creativity 1992; Averill & Thomas-Knowles, 1991). B. Emotional Roles The rules of emotion help to establish a corresponding corresponding set of emotional emotional roles; these roles are the various emotional syndromes that we recognize in ordinary ordinary language. emotional role, like any other other social role, can be analyzed analyzed in terms of privileges, An emotional following discussion we restrictions, obligations, and entrance requirements. In the following will use as examples the roles of anger, romantic love, and grief, the details of which found elsewhere (Averill, 1982, 1982,1985; can be found 1985; Averill & Nunley, 1993, respectively). / . PrIvileges Privileges 1. that would would be Emotional roles (syndromes) allow a person to engage in behavior behavior that discouraged in ordinary circumstances. The specific specific behavior behavior allowed or "excused" "excused" discouraged varies depending depending on the emotion. For example, when angry a person person can literally when angry from murder murder to get away with murder (i.e., have a crime of homicide mitigated mitigated from manslaughter—a "crime of passion"). When in love, a person may engage in sexual manslaughter-a passion "). grievbehavior that otherwise might be viewed as socially inappropriate. And when griev ing, a person is exempted from from obligations related to work and entertainment, at work least for a period of time. 2. Restrictions Restrictions 2.

There are limits to what a person can do when emotional and "get away with it." should be appropriate to the situation: They They should should not be Emotional responses should too mild or too strong, too short or too idiosyncratic. For example, short or too prolonged, or a a plea of anger will not be accepted in a court of law if the crime is committed in too manner. Similarly, are expected to be be discrete and too cruel cruel or or unusual unusual aa manner. Similarly, lovers lovers are expected to discrete and bereaved wife wife begins dating honorable in their affairs. And, in the case of grief, if a bereaved after the death of her husband, the genuineness of grief may be cast of her grief too soon after into doubt. Obligations 3. Obligations Whereas there are some things a person cannot do while emotional (restrictions), there are other things that that must be done (obligations). An angry person, for for example, is expected to take action to correct the wrong, or else the sincerity of his anger, anger.

S36 536

JAMES JAMES R. R AVERILL AVERILL

or or even even his his character, character, will will be be questioned. questioned. The The same same is is even even more more evident evident in in the the case case of of love, love, which which carries carries a a commitment commitment to to the the well-being well-being of of the the loved loved one. one. Among Among the is, the examples examples we we have have been been using, using, it it might might seem seem that that grief grief is is an an exception. exception. It It is, however, however, the the exception exception that that proves proves the the rule. rule. In In all all societies societies the the bereaved bereaved are are expected to mourning practices. practices. These obligations can expected to perform perform certain certain mourning These obligations can be be neglected neglected only only at at considerable considerable risk; risk; the the individual individual who who fails fails to to comply comply with with societal societal expecta expectafollowing bereavement often subject subject to severe sanction. sanction. tions following bereavement is often Entry Requirements Requirements 4. Entry

Most, Most, though though not not all, all, social social roles roles have have entry entry requirements; requirements; that that is, is, they they can can be be occupied occupied only only by by persons persons of of a a certain certain age, age, sex, sex, training, training, and/or and/or social social position. position. No No matter matter how how skillfully skillfully enacted, enacted, something something more more is is required required to to distinguish distinguish between between *'in the role." That That "something more" merely "playing a role" and actually being "in subjective involvement. A A delusional schizophrenic might is not simply sincerity or subjective skillfully play the role of of a judge, say, following following all the relevant relevant rules, and he might skillfully sincerely believe that he is a judge. Yet, he would not be a judge. Authentic role sincerely enactment requires social recognition and legitimacy. The same is true of of emotional enactment recognition roles. (Recall from from our our earlier earlier discussion that a person can feel feel emotional without being in in an an emotional emotional state.) state.) being The The entry entry requirements requirements for for anger anger are are vague vague but but nevertheless nevertheless discernable. discernable. For For example, persons higher higher in authority (e.g., example, persons in authority (e.g., parents) parents) are are afforded afforded more more right right to to become angry angry than than are are persons persons lower lower in in authority authority (e.g., (e.g., children), children), and and similarly similarly become with certain certain occupations (e.g., a drill sergeant sergeant as opposed opposed to a minister). The entry requirements for for love are more explicit. Below a certain age, a person is not fall in love, and if if sexual infatuation infatuation should lead to sexual behavior, "allowed" to fall statutory rape may be charged. Grief, too, has its entry requirements. In many statutory grief is to be experienced experienced may be stipulated stipulated according traditional societies the way grief of the bereaved, as well as to the nature of of the relationship of of to the age or sex of bereaved to the decreased-spouse, decreased—spouse, parent, child, cousin, and so forth. forth. the bereaved addition to the above features features (privileges, restrictions, obligations, and In addition entry entry requirements), requirements), emotional emotional roles roles can can be be distinguished distinguished from from other other social social roles roles (such (such as as that that of of judge, judge, ballet ballet dancer, dancer, or or father) father) in in that that emotional emotional roles roles are are transitory transitory or transitional. That is, for short-term short-term occupancy occupancy only; they or transitional. That is, emotional roles are for afford a means of of transition transition from from one state of of affairs affairs to another, when more afford "normal" modes of of responding are not sufficient sufficient (Averill, 1990a; Oatley, 1992). Vn. EMOTIONAL POTENTIALS AND AND CAPACITIES VH.

Emotional roles Emotional roles emphasize emphasize the the "objective" "objective" nature nature of of emotional emotional syndromes. syndromes. That That is, unlike emotional responses, emotional is, unlike emotional states states and and component component responses, emotional syndromes syndromes (roles) (roles) exist exist independent independent of of the the experiencing experiencing individual. individual. A A definition definition of of emotional emotional syn synimportance of of society in determining determining dromes as social roles also emphasizes the importance presume emotional behavior; it does not, however, ignore biology. Most social roles presume

CHAPTER CHAPTER 21 21 THE THEEMOTIONS EMOTIONS

537 537

relevant biological as well as social potentials (see Level I, Figure 3). The role of judge, for example, presumes a degree of native intelligence; the role of ballet dancer, musical talent and motor dexterity; and the role of of father, certain sexual characteristics. The same is no less true of emotional roles. The etc.) are The emotional emotional reactions reactions of of infancy infancy (fussing, (fussing, cooing, cooing, withdrawal, withdrawal, etc.) are primarily could not primarily biological. biological. It It could not be be otherwise. otherwise. But But almost almost immediately, immediately, socialization socialization begins begins to to occur occur through through the the mediation mediation of of parents parents and and other other caregivers. caregivers. The The tempera temperamental traits (Level II, Figure 3) that eventually emerge from this interaction of coeffigenetic endowment and environmental influence typically have heritability coeffi ± .10, depending on the population studied. And, as explained cients of around .50 ± earlier, temperamental traits are still only predispositions to emotion. Before an individual is able to enter into an emotional role, and hence to experience and express a specific emotion, the rules and beliefs that help constitute that emotion A person who has internalized the relevant rules and who is must be internalized. A adroit in their application may be said to be emotionally intelligent inteUigent (Mayer & once-andSalovey, 1995; 1995; Salovey et al., aI., 1992). But internalization is not a static, once-and for-all affair. affair. Rules Rules can can be be changed, changed, aa fact fact that that makes makes possible possible emotional emotional creativity creativity for-all on both both the the individual and the the social social levels levels (Averill (Averill & & Nunley, Nunley, 1992; Averill & & on individual and 1992; Averill Thomas-Knowles, 1991). 1991). Thomas-Knowles,

vm. Vin. CONCLUDING OBSERVATIONS The emotions play a central role in most theories of of personality. Yet, personality personaUty psychologists have shown a remarkable tolerance for ambiguity with regard to the psychologists often than not, emotional concepts have been nature of emotions per se. More often undifferentiated fashion, or else in a manner unique to each theory. used in a global, undifferentiated of the psychoanalyst, psychoanalyst, for example, is different different from the anxiety of the The anxiety of different from the anxiety of the behaviorist. A A fuller existentialist, and both are different of emotion is necessary for appreciation of the variety, organization, and principles of the advancement and possible integration of personality theory.

REFERENCES Arieti, S. (1976). Creativity: synthesis. New York: Basic Books. Creativity:The Themagic magic synthesis. New York: Basic Books. Averill, J. R. (1974). An analysis of psychophysiological symbolism and its influence on Journalffor theTheory TheoryofofSocial SocialBehavior, Behavior, 4, 147-190. theories of emotion. Journal or the 4, 147-190. Catalog of of Selected Selected Averill, J. R. (1975). A semantic atlas of emotional concepts. JSAS Catalog DocumentsininPsychology, Psychology,5,5,330. 330.(Ms. (Ms.No. No.1 103) 1103) Documents R. (1982). (1982).Anger Angerand andaggression: aggression: essay emotion. New York: Springer-Verlag. AnAn essay onon emotion. New York: Springer-Verlag. Averill, J. R. R. (1984). The acquisition of emotions during adulthood. In C. Z. Z. Malatesta & Averill, J. R. & C. Izard (Eds.), Affective Affective processes processesininadult adultdevelopment development(pp. (pp.23-43). 23-43).Beverly BeverlyHills, Hills, CA: Sage. CA:

538 538

JAMES AVERILL JAMES R. RAvERiLL

Averill, J. R R. (1985). The social construction of emotion: With special reference to love. In K. Gergen & K. Davis (Eds.), The social socialconstruction constructionof ofthe theperson person(pp. (pp.90-109). 90-109).New New York: Springer-Verlag. Averill, J. R R. (1988). Disorders of emotion. Journal of of Social and Clinical Clinical Psychology, Psychology, 6, 6, 247-268. Averill, J. R R. (199Oa). (1990a). Emotions as episodic dispositions, cognitive schemas, and transitory social roles: Steps toward an integrated theory of emotion. In D. Ozer, J. M. Healy, & & A. J. Stewart (Eds.), Perspectives Perspectivesininpersonality personality(Vol. (Vol.3a, 3a,pp. pp.139-167). 139-167).London: London:Jessica Jessica Kingsley Publishers. J, R R. (199Ob). (1990b). Inner feelings, works of the flesh, the beast within, diseases of the Averill, J. mind, driving force, and putting on a show: Six metaphors of emotion and their theoretical extensions. In D. E. Leary (Ed.), Metaphors Metaphors in the history history of of psychology (pp. 104-132). New York: Cambridge University Press. Averill, J. R R. (1994a, July). Emotional Creativity CreativityInventory: Inventory:Scale Scaleconstruction construction and validation. and validation. Paper presented at the meetings of the International Society for Research on Emotion, Cambridge, England. R. (1994b). I feel, therefore therefore I am-I am—I think. In P. Ekman & &R R. J. Davidson (Eds.), Averill, J. R nature of of emotion: emotion: Fundamental Fundamental questions questions (pp. 379-385). New York: Oxford The nature University Press. Averill, J. R., & Chon, K. K. (1990). Rules of of hope. hope. New York: Springer-Verlag. R, Catlin, G., & & Nunley, E. P. (1992). Voyages Voyages of of the heart: heart:Living Livingan anemotionally emotionallycreative creative Averill, J. R., R, & life. New York: Free Press. life. R., & & Nunley, E. P. (1993). Grief Grief as an emotion and as a disease. In M. S. Stroebe, Averill, J. R, W. Stroebe, & R R. O. Hansson (Eds.), Handbook Handbook of of bereavement bereavement (pp. (pp. 77-90). 77-90). New New York: Cambridge University Press. Averill, J. R, R., & & Thomas-Knowles, Thomas-Knowles, C. C. (1991). (1991). Emotional Emotionalcreativity. creativity.In In K. K.T. T.Strongman Strongman (Ed.), (Ed.), International International review reviewof ofstudies studieson onemotion emotion(Vol. (Vol.1,1,pp. pp.269-299). 269-299).London: London:Wiley. Wiley. Berlin, B., Breedlove, D., & & Raven, P. (1968). (1968). Covert categories and folk taxonomies. American Anthropologist, Anthropologist, 70, 70, 290-299. 290-299. Cacioppo, J. T., Klein, D. J., Bernstom, G. G., & Hatfield, E. (1993). The psychophysiology psychophysiology of of emotion. In M. Lewis & J. Haviland (Eds.), Handbook of of emotions emotions (pp. 119-142). New York: Guilford Guilford Press. Darwin, C. (1965). (1965). The expression expressionof of the theemotions emotionsin inman manand andanimals. animals.Chicago: Chicago:University University of Chicago Press. (Original work published 1872) of R. J. (1994). Complexities in the search from emotion-specific physiology. In P. Davidson, R &R R. J. Davidson (Eds.), The nature nature of of emotion: emotion: Fundamental Fundamentalquestions questions(pp. (pp. Ekman & Oxford University Press, 237-242). New York: Oxford Press. structuraltheory theoryof ofthe theemotions. emotions.New NewYork: York:International InternationalUniversi Universide Rivera, J. (1977). A structural ties Press. Dewey, J. J. (1895). (1895). The The theory theory of of emotion. II: The significance significance of of emotions. emotions. Psychological Psychological II: The Review, Review, 2, 2, 13-32. Eibl-Eibesfeldt, I. (1970). Ethology: Ethology: The The biology biologyof ofbehavior behavior(E. (E.Klinghammer, Klinghammer,Trans.). Trans.).New New York: Holt, Rinehart & Winston. Emotion in the human face (2nd ed.). Cambridge, England: Cam CamEkman, P. (Ed.). (1982). Emotion bridge University Press. Ekman, P. (1984). Expression and the nature of emotion. In K. Scherer & & P. Ekman (Eds.), Approaches to emotion (pp. 319-343). Hillsdale, NJ: Erlbaum.

21 CHAPTER 21

THE EMOll0NS EMOTIONS THE

539 539

Ekman, P. (1994). Strong evidence for universals in facial expressions: A reply to Russell's PsychologicalBulletin, Bulletin,115, 115, 268-287. 268-287. mistaken critique. Psychological (1994). Integration Integration of ofthe thecognitive cognitiveand andthe thepsychodynamic psychodynamicunconscious. unconscious.American American Epstein, S. (1994). Psychologist, Psychologist,49, 49,709-724. 709-724. Fehr, B., B., & Russell, J. A. (1984). Concept of emotion viewed from from a prototype perspective. Journal General, 113,113, 464-486. Journalof ofExperimental ExperimentalPsychology: Psychology: General, 464-486. Fox, R. (1971). The cultural animal. In J. F. Eisenberg & W. S. Dillon (Eds.), Man and (pp. 273-296). Washington, DC: Smithsonian Insti beast: beast:Comparative Comparativesocial socialbehavior behavior (pp. 273-296). Washington, DC: Smithsonian Institution Press. Fridlund, A. J. (1994). Human Human facial expression: expression:An Anevolutionary evolutionaryview. view.San SanDiego, Diego,CA: CA: Academic Press. Frijda, N. H. ((1986). 1986). The emotions. emotions, Cambridge, England: Cambridge University Press. (1996). The The discreteness discreteness of of emotion emotion concepts: concepts:Categorical Categoricalstructure structurein inthe theaffective affective Haslam, N. (1996). 21,21, 1012-1019. circumplex. Personality Personalityand andSocial SocialPsychology PsychologyBulletin, Bulletin, 1012-1019. Hochschild, A. R. (1983). The managed managed heart. heart Berkeley: University of California California Press. of emotions. emotions. New New York: York: Plenum Press. Press. Izard, C. E. (1991). The psychology of proIzard, C. E. (1993). Four systems for emotion activation: Cognitive and noncognitive pro PsychologicalReview, Review,100, 100, 68-90. 68-90. cesses. Psychological Izard, C. E., & & Saxton, P. M. (1988). Emotions. In R. C. Atkinson, R. J. Herrnstein, Hermstein, G. Lindzey, and R. D. Luce (Eds.), Handbook of of experimental experimentalpsychology psychology(2nd (2nd ed., ed..Vol. Vol. 1, pp. 627-676). New York: Wiley. 1, James, W. (1890). The principles principles of of psychology psychology (2 (2 vols). vols). New New York: York: Henry Henry Holt. Holt. Johnson-Laird, P. N., & & Oatley, K. (1989). (1989). The language of emotions: An analysis of a semantic field. field. Cognition Cognitionand andEmotion, Emotion,3,3,81-123. 81-123. Psychologicaltypes: types:Collected Collectedworks works (Vol.6).6).Princeton, Princeton,NJ: NJ:Princeton Princeton Jung, C. G. (1971). Psychological (Vol. University Press. (Original work published 1921) & Bresler, C. (1992). The process of emotional experience: A self-perception self-perception Laird, J. D., & theory. In M. S. Clark (Ed.), Review personality and andsocial socialpsychology psychology(Vol. (Vol.13, 13, Review of of personality pp. 213-234). Newbury Park, CA: Sage. & Diener, E. (1992). Promises and problems with the circumplex model of Larsen, R. J., & of emotion. In M. S. Clark (Ed.), Review Review of of personality personalityand andsocial socialpsychology psychology(Vol. (Vol.13, 13, pp. 25-59). Newbury Park, CA: Sage. Lazarus, R. S. (1991). Emotion and adaptation. adaptation. New NewYork: York:Oxford OxfordUniversity UniversityPress. Press. Leventhal, H., & & Tomarken, A. J. (1986). (1986). Emotion: Today's problems. Annual Review of of Psychology, Psychology, 37, 37,565-610. 565-610. 13-128. Lutz, C. (1982). The domain of emotion words on Ifaluk. American Ethnologist, Ethnologist,9,9,1113-128. Mandler, G. G. (1984). (1984). Mind and and body: body: Psychology Psychologyof ofemotion emotionand andstress. stress. New York: Norton. New York: Norton. Mandler, Matsumoto, D. (1987). The role of facial response in the experience of emotion: More Psy methodological problems and a meta-analysis. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 52, 52,769-774. 169-11A. chology, Mayer, J. J. D., D., & &Salovey, Salovey,P. P.(1995). (1995).Emotional Emotionalintelligence intelligenceand andthe theconstruction constructionand andregulation regulation Mayer, & Preventive PreventivePsychology, Psychology,4,4,197-208. 197-208. of feelings. Applied & McDougall, W. W. (1936). (1936). An An introduction introduction to social psychology psychology (23rd (23rd ed.). ed.). London: London: Methuen. Methuen. McDougall, Mees, U. U. (1991). (1991). Die Struktur der Emotionen. Emotionen. Gottingen: Gottingen: Hogrefe. Hogrefe. Mees, Morgan, C, «fe Averill, Averill, J.J. R. R. (1992). (1992).True True feelings, feelings, the theself, self, and and authenticity: authenticity: A A psychosocial psychosocial Morgan, c., & perspective. In D. Franks & & V. Gecas (Eds.), Social Socialperspectives perspectivesininemotions emotions(Vol. (Vol.1,1, CT: JAI Press. pp. 95-123). Greenwich, CT:

S40 540

JAMES JAMES R. R AVERILL AVERILL

Murphy, G. L., & & Medin, D. L. (1985). (1985). The role of theories in conceptual coherence. Psychological PsychologicalReview, Review,92, 92,289-316. 289-316. Oatley, K. (1992). Best laid schemes: schemes: The psychology of of emotions. emotions. Cambridge, England: Cambridge University Press. Oatley, K., & Johnson-Laird, P. P. N. N. (1987). (1987).Towards Towardsaacognitive cognitivetheory theoryof ofemotions. emotions.Cognition Cognition and Emotion, 1, i, 29-50. Ortony, A., Clore, G. L., & Collins, A. (1988). (1988). The cognitive cognitive structures structuresofofemotions. emotions,Cam Cambridge, England: Cambridge University Press. Pittam, J., & & Scherer, K. R. (1993). Vocal expression and communication of emotion. In M. Lewis & J. Haviland (Eds.), Handbook Handbook of of emotions (pp. (pp. 185-197). New York: Guilford Guilford Press. synthesis. New NewYork: York:Harper Harper &&Row. Row. Plutchik, R. (1980). Emotion: A psychoevolutionary synthesis. Redican, W. K. (1982). (1982). An evolutionary perspective on human facial displays. In P. Ekman (Ed.), Emotion in the human face (2nd ed., pp. 212-280). Cambridge, England: Cam Cambridge University Press. Reed, S. D., Harver, A., & & Katkin, E. S. S. (1990). Interoception. In J. T. Cacioppo & Principlesof ofpsychophysiology: psychophysiology:Physical, Physical, social, inferential L. G. Tassinary (Eds.), Principles social, andand inferential elements (pp. 253-294). New York: Cambridge University Press. elements Psychological Reisenzein, R. (1983). The Schachter theory of emotion: Two decades later. Psychological Bulletin, 94, 94, 239-264. Bulletin, Reisenzein, R. (1994). Pleasure-arousal theory and the intensity of emotions. Journal of of Personalityand andSocial SocialPsychology, Psychology,67, 67,525-539. 525-539. Personality Rim6, B., & Cisamolo, D. (1990). Social schemata of peripheral changes in Rime, B., Philippot, P., & emotion. Journal of of Personality Personalityand and Social SocialPsychology, Psychology,59,59,38-49. 38-49. Roseman, I. J., Antoniou, A. A., & & Jose, P. E. (1996). (1996). Appraisal determinants of emotions: Cognition and Emotion, Emotion, Constructing a more accurate and comprehensive theory. Cognition 10, 10, 241-277. Russell, J. A. (1980). A A circumplex circumplex model model of of affect. affect. Journal Journalof ofPersonality Personalityand andSocial SocialPsychol Psychology, ogy, 39, 39, 1161-1178. Russell, J. A. (1983). Pancultural aspects of the human conceptual organization of emotions. Journal of of Personality Personalityand andSocial SocialPsychology, Psychology,45, 45,1281-1288. 1281-1288. Russell, J. A. (1991). Culture and the categorization of emotions. Psychological Psychological Bulletin, Bulletin, 110, 110, 426-450. 426-450. Russell, J. A. (1995). Facial expressions of emotion: What lies beyond minimal universality? Psychological PsychologicalBulletin, Bulletin,118, 118,379-391. 379-391. of mind. mind. London: Hutchinson. Ryle, G. (1949). The concept of P., Hsee, C. C. K., K., & &Mayer, Mayer,J.J.D. D.(1992). (1992).Emotional EmotionalIntelligence Intelligenceand andthe theself-regulation self-regulation Salovey, P., affect. In D. M. Wegner & & J. W. Pennebaker (Eds.), Handbook of of mental control of affect. (pp. 258-277). Englewood Cliffs, NJ: Prentice-Hall. Sarbin, T. R. (1986). (1986). Emotion and act: Roles and rhetoric. In R. Harre Harr6 (Ed.), The social social construction constructionof ofemotions emotions(pp. (pp.83-97). 83-97).Oxford, Oxford, England: England:Basil BasilBlackwell. Blackwell. emotions: Outline of of a theory (B. (B. Frechtman, Trans.). Sartre, J. P. (1948). The emotions: Trans.). New York: Philosophical Library. Schachter, S. (1971). Emotion, obesity, obesity, and crime. crime. New NewYork: York:Academic Academic Press. Press. Scherer, K. R. R. (1984). (1984).On Onthe thenature natureand andfunction function of ofemotion: emotion:AAcomponent componentprocess processapproach. approach. In K. R. Scherer & P. Ekman (Eds.), Approaches to emotion (pp. 293-317). Hillsdale, NJ: Erlbaum.

CHAPTER CHAPTER 21 21

541 541

THE THE EMOTIONS EMOTIONS

Shaver, P., Schwartz, J., Kirson, D., & O'Connor, C. (1987). Emotion knowledge: Further exploration of of a prototype approach. Journal of of Personality Personality and and Social SocialPsychology, Psychology, exploration 52, 1061-1086. Smith, C. A., & (1985). Patterns of cognitive appraisal in emotion. Journal & Ellsworth, P. C. (1985). of of Personality Personalityand andSocial SocialPsychology, Psychology,48,48,813-838. 813-838. Solomon, R. C. (1993). Thepassions: passions:Emotions Emotionsand andthe themeaning meaningofoflife life(Rev (Reved.). ed.).Indianapolis, Indianapolis, IN: Hackett. Spielberger, C. D. (1966). Theory Ed.) AAnxiety nxiety Theory and and research research on on anxiety; anxiety. In In C. C. D. D.Spielberger Spielberger ((Ed.), and behavior (pp. 3-20). New York: Academic Press. & Storm, T. (1987). A taxonomic study of the vocabulary of emotions. Journal c., & Storm, C, of Personalityand andSocial SocialPsychology, Psychology,53,53,805-816. 805-816. of Personality Storm, C., C, Storm, T., & & Jones, C. (1996). Aspects of meaning in words related to happiness. Cognition and and Emotion, Emotion, 10, 10,279-302. 279-302. Cognition Tomkins, S. S. (1981). The role of facial response in the experience of emotion: A reply to Journalof of Personality Personalityand andSocial SocialPsychology, Psychology, Tourangeau and Ellsworth. Journal 40,40,355-357. 355-357. Valins, S. (1967). Emotionality and information information concerning internal reactions. Journal of of Personalityand andSocial SocialPsychology, Psychology,6,6,458-463. 458-463. Personality & Scherer, K. R. (1988). How How universal universal and and specific specific is is emotional emotional experience? experience? Wallbott, H. G., & Facetsof ofemotion emotion from 27 countries onfive five continents. In K. R. Scherer (Ed.), Facets Evidence from (pp. 31-56). Hillsdale, NJ: Erlbaum. & Tellegen, A. (1985). Toward a consensual structure of mood. Psychological Psychological Watson, D., & Bulletin, 98, 98, 219-235. 219-235. Bulletin, Weiner, B. (1985). An attributional theory of achievement motivation and emotion. PsychoPsycho logical Review, Review,92, 92,548-573. 548-573. logical Cognitive Science, Science,16, 16,539-581. 539-581. Defining emotion concepts. Cognitive Wierzbicka, A. (1992). Defining (1953). Philosophical Philosophical investigations. investigations.Oxford, Oxford,England: England:Basil BasilBlackwell Blackwell&& Wittgenstein, L. (1953). Mott. T., & & Inglehart, M. (1989). Feeling and andfacial facial efference: efference: Implications Implications Zajonc, R. B., Murphy, S. T., (1989). Feeling PsychologicalReview, Review,96, 96,395-416. 395-416. of the vascular theory of emotion. Psychological ,

This Page Intentionally Left Blank

22 CHAPTER 22

PSYCHOLOGICAL DEFENSE CONTEMPORARY THEORY AND RESEARCH CONTEMPORARY

DELROY L. L. PAULHUS UNIVERSITY OF BRITISH COLUMBIA UNIVERSITY OF BRITISH COLUMBIA BRAM FRIDHANDLER FRIDHANDLER BRAM UNIVERSITY OF OF CALIFORNIA CALIFORNIA,, SAN SAN FRANCISCO UNIVERSITY FRANCISCO

SEAN HAYES SEAN HAYES PEPPERDINE UNIVERSITY PEPPERDINE UNIVERSITY

With some trepidation, we tackle the monumental task of reviewing the contempocontempo rary literature on psychological defense. The task is particularly daunting because (a) the literature is vast and scattered, (b) the empirical basis of the topics varies dramatically, and (c) many psychologists remain skeptical of the very notion of defense, often because of a general distrust of things psychoanalytic. For these reasons, we have given priority to claims with empirical support and to work conducted since the last handbook chapter on this topic (Eriksen & Pierce, 1968). In its broadest sense, psychological psychologicaldefense defenserefers referstotothe theprocess processofofregulating regulating anxiety, depression, depression,and and loss lossof ofself-esteem. self-esteem.Defense Defensemecha mechapainful emotions such as anxiety, nisms are are usually usually defined defined more more narrowly narrowly as as mental mental processes processes that that operate operate uncon unconnisms sciously to reduce some painful emotion. In the classical sense, the latter have been further restricted to threats aroused by the individual's thoughts and wishes, particularly psychological conflict over issues of sex and aggression. Our decision to address the larger topic of psychological defense is less a virtue than a necessity given that the term "defense mechanism" has been used so HANDBOOK HANDBOOK OF OF PERSONAUTY PERSONALITY PSYCHOLOGY PSYCHOLOGY

543 543

CoPYRIGHT COPYRIGHT C © 1997 BY BY ACADEMIC PlESS PRESS.. ALL IIGtrIS OF REPRODUcnON IN ANY FORM RESERVED.

ALL RIGHTS OF REPRODUCHON IN ANY FORM RESERVED.

S44 544

PAULHUS, FRlDHANDLER, AND PAULHUS,FRIDHANDLER, AND HAVES HAYES

widely and liberally as to cover virtually all forms of psychological defense. For conscious mech example, some influential theorists have argued for the inclusion of conscious mechanisms. Others have called for an expansion to cases where external realities (e.g., physical danger) are the source of threat. In contrast to relaxing the traditional definition, some theorists have proposed further restrictions. Some, for example, favor restricting defense mechanisms to effectiveoperations with maladaptive outcomes. Others have responded that the effective ness of a defense cannot be determined prior to its use or independent of the context. Still others argue that certain defenses are a priori adaptive. Another condiproposed restriction requires intentionality of reducing distress as a necessary condi tion. This This restriction restrictionisis rejected, rejected,in in turn, turn,by bytheorists theoristswith withaamechanistic, mechanistic,information informationprocessing view of defense. Unfortunately, we cannot offer a solution to this lack of definitional consensus. Instead, we will refer to the broad topic of this paper as psychological defense while reminding the reader at times of the stricter classical definition of defense mechanisms. We have organized the literature into digestible chunks by grouping together all the authors who share a conceptual framwork (e.g., Vaillant, Bond, Perry, do justice to each writer's work. At best, we can hope Cooper). We cannot hope to do to provide a theoretical overview of each approach, a sampling of relevant research, and references to any reviews. The chapter will culminate with an attempt to link the common theoretical strands. I CONTEMPORARY PSYCHOANALYSIS L

Although the concept of a defense mechanism originated with Sigmund Freud, his daughter Anna Freud Freud (1936) was was particularly particularly influential influential in in establishing establishing the the processes processes and distinctiveness of various defenses. defenses. Modern Modern psychoanalytic writers writers have have contin continued to dissect the concept of defense (S. 1989; Gero, 1951; (S. H. Cooper, 1989; 1951; Schafer, 1968, 1976; 1976; Wallerstein, 1983). Only a handful, however, have had a significant 1968, impact on the psychoanalytic view of defense. One of these is Otto Kernberg (1976, 1984), who was strongly influenced in this regard by the "object relations" theory of Melanie Klein. Kernberg has defined a new diagnostic category, the "borderline personality organization," organization,'' which differs from normal and neurotic organizations by virtue of its dependence on a particular variety of defense, namely, "splitting" and related defenses. These defenses, he argues, operate not by limiting awareness of offending wishes or ideas but by "dissociating them in consciousness." In other words, people with with such such personality personality organizations organizations do do experience their unacceptable unacceptable or intolerable and feelings, unlike people with a neurotic organization who mtolerable thoughts and "borderline" repress these mental contents. The essence of defensive activity in the "borderline" befallen personality is recalling the tumultuous experiences almost as they had befallen someone else, blaming them on "bad" "bad" aspects of self or on irredeemably "bad" others, neither of which are felt to have substantial connections with the "good" self and others. self

CHAPTER22 22 PSYCHOLOGICAL PSYCHOLOGICALDEFENSE DEFENSE CnAnER

545 545

Renewed interest in psychoanalytic defense has led, somewhat ironically, to the rehabilitation of of Freud's competitor, Pierre Janet (e.g., (e.g.. Bowers & Meichenbaum, 1984; C. Perry & Laurence, 1984). Janet saw defense defense as a constriction of conscious consciousness rather than a shunting to the unconscious. A failure to integrate certain experi experiences within the personality personality was said to induce detached psychological automatisms. Although considerably less elaborate, Janet's view of defense defense provides a conception that is eminently compatible with modem information-processing information-processing theories. Note, in conclusion, that many psychoanalysts eschew the notion of of discrete defense defense mechanisms operating for limited periods of of duration. Instead, they argue that defense defense is ubiquitous and pervasive: virtually every mental act involves a trade tradeoff off of of anxiety anxiety and awareness. Some analysts go further to argue against any standard set of of "defense "defense mechanisms"; instead, any mental process or capacity can be used toward defensive ends (Brenner, 1982). This view plays down the idea that some individuals are more defensive than others: instead, individuals differ in their style of of defense. Indeed, this defensive defensive style is so broad ranging that it represents a fundamental fundamental component of of character. In recent times, this view has been well articulated by David Shapiro (1965, (1965,1981). 1981). Given space limitations, we cannot elaborate on these issues; we can only refer the reader to some key volumes. For discussion of of other defenses and more refer elaborate discussions of theoretical issues, we recommend S. H. Cooper (1989), (1985,1990), Erdelyi (1985, 1990), Sjoback (1973), and Wallerstein (1983). We also recommend Kline (1972) and Fisher and Greenberg (1977) for more extensive treatment and reviews of early research on defenses.

n. n. REPRESSION In many ways, repression represents the flagship in the psychoanalytic fleet of defense defense mechanisms. Freud considered it so central that he labeled it the cornerstone of psychoanalysis. Subsequent analysts viewed repression as, if not a cornerstone, at least the prototype prototype of of defense mechanisms because it incorporated such central elements as emotional conflict, unconscious motivation, signal anxiety, and long longterm unaccounted-for unaccounted-for distress. Here, our simple working definition of repression is the shunting of of distressing emotions into the unconscious. Although inaccessible, the repressed emotions into the unconscious. Although inaccessible, the repressed emotion emotion can can create chronic distress. create chronic distress. In In their their handbook handbook chapter chapter on on psychological psychological defense, defense, Eriksen Eriksen and and Pierce Pierce (1968) (1968) provided a thorough review of the early research. Those studies sought provided a thorough review of the early research. Those studies sought to to show show that that memories memories associated associated with with threat threat or or distress distress were were more more difficult difficult to to recall. recall. The The studies however, were later interpreted studies that that did did show show such such effects, effects, however, were later interpreted as as interference interference effects effects due due to to stress, stress, not not repression. repression. A A particularly particularly scathing scathing review review by by Holmes Holmes (1974) virtually brought a halt to this era of research. Recently, however, the topic of repression has attracted renewed attention under the label "repressed memories." Unfortunately, LFnfortunately, the documentation of this

S46 546

PAULHUS,FRIDHANDLER, FRIDHANDLER,AND ANDHAYES HAYES PAULHUS,

phenomenon consists primarily of of unsystematic reports by therapists. The critiques leveled by experimental psychologists (e.g., Loftus, 1993) have centered around two pieces of evidence. First, laboratory research demonstrates unequivocally that nonemotional events. Second, when untrue emotional events are better recalled than nonemotional suggestions are implanted in experimental subjects, they may be recalled with a concluded certainty equal to that of true memories. Not surprisingly, many observers concluded ^'repressed memories" were, in fact, false memories implanted in the that so-called "repressed (Ofshe minds of their clients by certain "true believer" therapists (Of she & Watters, 1994). In support of the scattered reports by therapists, however, are some recent prospective studies. For example, Williams 129 women prospective studies. For example, Williams (1994) (1994) interviewed interviewed 129 women with with previously documented histories of sexual victimization in childhood. A large propro portion of the women (38%) did not recall the abuse that had been reported 17 years earlier. The problem with such studies for our purposes, is that they only address the question of whether these events can be forgotten, not whether they are repressed. repressed. To substantiate the latter, it must be shown that the memories can be recovered, thereby demonstrating that they were present in some form all along. Note that harmful effects of of child sexual abuse are not at issue; the issue is whether those the harmful forget the trauma are still distressed because a repressed conflict conflict children who forget reremains to fester. Until such studies are forthcoming, this topic will surely re main controversial.

m. DENIAL m. Although overlapping with other defenses, denial refers primarily to defense defense against painful aspects of external reality (e.g., Goldberger, 1983). It is usually considered a primitive defense, both in the sense of developing early and in the sense of crudeness and simplicity (A. Freud, 1936). Nonetheless, detailed analysis by Breznitz (1983) distinguished seven (1983) and and Spence Spence (1983) (1983) has has distinguished seven kinds kinds of of denial denial varying varying in in subtlety subtlety and stage of analysis. and stage of analysis. Recent theoretical treatments include several dealing with the denial of of death (Becker, 1973; Kubler-Ross, 1969; Lifton, 1968; see also terror management in the next next paragraph). paragraph). Other Other comprehensive comprehensive theoretical theoretical treatments treatments of of denial denial include include those by Moore and Rubinfine (1969), Sjoback (1973), and Dorpat (1985). Empirical treatments include those by Breznitz (1983) and Spence (1983). The most comprehensive experimental treatment is that by Greenberg, Pyszczynski, (TM)." Based on the and Solomon (1986) on what they term "terror management (TM)." ideas of Ernest Becker (1973), the theme is that awareness of one's mortality creates the potential for overwhelming terror. This pervasive force is said to mold cultural beliefs to provide philosophies or religions that preclude the terror. In general these beliefs require (a) a means whereby individuals can escape their fate, for example, '*If I behave morally, I will have everlasting life," and (b) a sense that one is "If satisfying the requirements for escape, that is, being a good person. The latter satisfying

CHAPTER CHAPTER 22 22

PSYCHOLOGICAL PSYCHOLOGICALDEFENSE DEFENSE

547 547

involves preserving one's self-esteem. In short, maintaining high self-esteem self-esteem preserves one's death-exempt status thereby precluding the terror of certain death. The authors have conducted a series of of experiments to validate TM theory (Greenberg, 1986; Rosenblatt, Greenberg, Solomon, Pyszczynski, & Lyon, 1989). One robust finding is that exposing subjects to symbols of of death increases their tendency to affirm cultural beliefs. The theory was also supported by Paulhus and Levitt (1987), who found that distracting subjects with death-related words increased of TM TM theory appeared their tendency to claim socially desirable traits. An extension of al., in press). recently (Pyszczynski et aI.,

SELF-DECEPTION IV. SELF.DECEPTION The term self-deception self-deception has been used in at least three distinguishable ways. In one usage, self-deception self-deception is a distinct form of defense in which the individual shows moral weakness in disavowing some unpleasant truth (e.g., Eagle, 1988; Fingarette, self-deception is not a type of defense, but 1969; Sarbin, 1988). In another usage, self-deception a state of affairs inherent in defense mechanisms: mechanisms: it is a motivated unawareness of of two conflicting conflicting representations of the same target (Paulhus, 1988; one of 1988; Sackeim & Gur, 1978). Finally, self-deception self-deception is also used as a generic term to cover mispercep misperceptions about oneself oneself (e.g., Gilbert & Cooper, 1985; Goleman, 1986). Although often claimed to be paradoxical (Gergen, 1985; 1985; Sartre, 1943/1956), of self-deception self-deception has commanded increased attention in recent years. the notion of fullThis interest may be traced to a few seminal works. In philosophy, the first full fledged treatise was published in 1969 by Herbert Fingarette. In sociobiology, a commentary by Trivers (1976) was followed quickly by more thorough treatments by Lockard (1978) and Alexander (1979). In In psychology, psychology, earlier earlier empirical empirical work work by by Frenkel-Brunswik Frenkel-Brunswik (1939) (1939) and and Murphy Murphy (1975) preceded the first rigorous Sackeim (1979). (1975) preceded thefirst rigorous experiment experiment by by Gur Gur and and Sackeim (1979). Increasing Increasing confidence confidence in in the the scientific scientific merit merit of of this this enterprise enterprise is is evidenced evidenced by by the the recent recent spate spate of empirical reports Sackeim, 1979; of empirical reports on on the the subject subject (e.g., (e.g., Gur Gur & Sackeim, 1979; Jamner Jamner & Schwartz, Schwartz, 1987; 1987; Monts, Monts, Zurcher, Zurcher, & Nydegger, Nydegger, 1977; 1977; Paulhus, Paulhus, 1984; 1984; Quattrone Quattrone & Tversky, Tversky, 1984; Sackeim & Gur, 1979). Finally, books with 1984; Sackeim Gur, 1978, 1978,1979). Finally, books with "self-deception" "self-deception" in in the the title title have appear (Goleman, Martin, 1985; 1985; have begun begun to to appear (Goleman, 1986; 1986; Lockard Lockard & Paulhus, Paulhus, 1988; 1988; Martin, Murphy, 1987). Murphy, 1975; 1975; Sloan, Sloan, 1987). In the most general sense of of false self-beliefs, the concept of of self-deception self-deception has been applied in human ethology as well as social, clinical, and personality psychology. In ethology, for example, evidence for the adaptive value of of limited self-knowledge self-knowledge in lower organisms implies an evolutionary basis for human self selfdeception (Alexander, 1979; D. T. Campbell, 1983; Lockard, 1978; Trivers, 1985). of In social psychology, too, self-deception self-deception has sometimes been defined as a lack of awareness of of internal psychological psychological processes (e.g., Wilson, 1985).

S48 548

PAUuru8, FRmHANDLER, AND PAULHUS,FRIDHANDLER, AND HAYES HAYES

self-deception is being applied more selectively in the sense Increasingly, self-deception of motivated motivated unawareness. Among social psychologists, an increasing interest in motivated biases and distortions has has legitimized the study of formerly formerly taboo concepts such as self-deception self-deception (e.g., Gilbert & Cooper, 1985; 1985; Krebs, Denton, & Higgins, 1988; Snyder, 1985). 1985). Information-processing Information-processing theories have also been brought to 1988; self-deception (Greenwald, 1988; 1988; Paulhus & Suedfeld, 1988; 1988; Sackeim, 1988). bear on self-deception self-deception has been widely In clinical psychology and psychiatry, the concept of self-deception 1985; Eagle, 1988; 1983; Schafer, 1976). applied (Dorpat, 1985; 1988; Sackeim, 1983; The term, self-deception, has also been applied appUed to a series of personality measures. Sackeim Sackeim and and Gur Our (1978) (1978)originally originally developed developed self-report self-report measures measures of of both both self- and other deception. The former, labeled the Self-Deception Self-Deception Questionnaire, comprised accusations about threatening but common beliefs and feelings (e.g., "Have you ever worried that you might be a homosexual?"). homosexual?")- Respondents giving a high number of extreme denials are scored as self-deceptive. These two scales refined and validated to form Paulhus's (1984) Balanced Inventory of Desirable Desirable were refined of the BIDR, self-deception self-deception has Responding (BIDR). In the most recent version of been partitioned into enhancement and denial (Paulhus & Reid, 1991). Research confirmed that thefirst using the BIDR has confirmed first component of social desirability variance self-deceptive enhancement (e.g., Paulhus, Paulhus, found in self-reports is best interpreted as self-deceptive 1984,1986,1991). 1984, 1986, 1991). theme—that self-deception self-deception Many of the writings on this topic share a common theme-that benefits is a normal and generally positive force in human behavior. The claimed benefits from perpetuating the genetic structure of the individual, to improving the vary from interference with self-preserving self-preserving individual's ability to deceive others, to minimizing interference behaviors, and to contributing to the psychological health, health, stability, stability, and and performance performance of the individual as well as society as a whole (Lockard & Paulhus, 1988). V. TRAIT TRAIT AND AND TYPE TYPE ApPROACHES APPROACHES V.

defenses is often often considered considered to be a continuous individual The tendency to use defenses difference difference variable within the normal range of personality.1 personality.^ This concept underlies such measures as the Marlowe-Crowne scale (for a review, see Crowne, 1979), the (for a review, see Paulhus, 1991), and the Self-Deception Self-Deception Question QuestionMMPI K-scale (for naure (reviewed above). naire A select few warrant special attention here because they have a psychoanalytic flavor in combination with a high level of empirical scrutiny. Normal individuals ftavor are measured on a single trait or type (combination of traits) and studied intensively. A. A. Byrne Byrne Although not fully consistent with the psychoanalytic meaning of repression, Byrne repressivestyle styleto to describe describe the the trait-like trait-liketendency tendency to to defend defend (1961) used the term repressive Inthe therealm realmofofquestionnaire questionnaireresearch, research,such suchgeneral generaltendencies tendenciesare areoften oftencalled calledresponse responsestyles styles 1^In (for a review, see see Paulhus, 1991).

CHAPTER CHAPTER 22 22

PSYCHOLOGICAL PSYCHOLOGICAL DEFENSE DEFENSE

549 549

against threats. To measure this style, Byrne Byrne developed developed the the widely widely used used Repression Repressionreflect a clear theoretical assumption: Sensitization (R-S) scale. Items on this scale reflect Individuals reporting reporting no anxiety must have a repressive cognitive style, that is, a tendency to minimize the existence or potency of threats, whereas those reporting tendency excess distress must have a sensitizing cognitive style. Thus Byrne postulated a bipolar trait wherein both poles represent represent defensive styles and the midpoint repre represents good adjustment. After After examining the accumulated evidence, however, Bell and Byrne (1978) concluded that pathology could be present in any of these three groups. Originally a conceptually driven 156-item scale (Byrne, 1961), the final R-S winnowed down to a 127-item version using part-whole correlations (Byrne, was winnowed Barry, & Nelson, 1963). The ensuing flood of published research using the scale differences in defen defensuggests a widespread acceptance of the notion of individual differences siveness (see review by Bell & Byrne, 1978). Among the most cited studies using the R-S scale is that conducted by M. S. Schwartz, Krupp, and Byrne (1971). (1971). They followed up the medical status of 50,000 completed the R-S scale. One striking finding was that repressors patients who had completed suffered primarily from from organic problems whereas sensitizers suffered suffered from psychosuffered psycho logical problems. Nonetheless, a number of the later approaches to defense (discussed below) begin with a criticism of Byrne's approach. One major was Byrne's theoretical theoretical major target was inflexible (e.g., assumption that defenses are trait-like, that is, cross-situational and inflexible 1965; Lazarus Lazarus & & Folkman, Folkman, 1984). 1984).The The major major criticism criticism of of the the R-S R-S scale scale itself itself Haan, 1965; its conceptual similarity and embarrassingly high correlation with measures has been its 1969; Golin, Herron, & Lakota, & Reineck, 1967). of anxiety (Gleser (Gieser & Ihilevich, 1969; diagnosed as repressors on the The resulting paradox was that the same subjects diagnosed well-adjusted) on standard anxiety R-S were diagnosed as truly low-anxious (i.e., well-adjusted) scales. B. B. Weinberger Weinberger and and Schwartz Schwartz

conWeinberger and Schwartz and their colleagues have tackled the problematic con founding of the R-S scale with anxiety measures. Beginning with D. A. Weinberger, Schwartz, and Davidson (1979), they they explained explained that that measures measures based based on on self-reported self-reported distress were incapable of distinguishing repressors from the truly low-anxious. The problem is that some subjects (truly low-anxious) accurately accurately report low anxiety; others (repressors) defensively defensively disavow their anxiety. Instead, following Boor and Schill (1967), Weinberger and associates devel developed a typology measure of repressive style and supported its validity with both experimental and correlational studies (for a review, see D. A. Weinberger, self-report 1990). Repressors were one of four groups identified by crossing a self-report of trait anxiety (e.g., the Taylor Manifest Anxiety Scale) with a measure measure of of defensiveness, the Marlowe-Crowne Social Desirability Scale (Crowne & Marlowe, 1964). Among low-defensives, both high- and low-anxious subjects are

550 550

PAULHUS, FRIDHANDLER, AND AND HAVES PAULHUS,FRIDHANDLER, HAYES

taken at their word. word. The low anxiety scores of subjects scoring high on defensive defensiveness (repressors), however, are assumed to result from their avoidance of distressing knowledge about themselves, including knowledge of their levels of of anxiety. The fourth group-high-defensive/high-anxious-are group—high-defensive/high-anxious-are rare and, therefore, ignored (see Davis, 1990, 1990, p. 391). New data as well as reinterpretation of previous studies have supported the validity of the typology index of repressive style. For example, D. A. Weinberger et a!. behav al. (1979) found repressors to be higher than low-anxious subjects on six behavioral and physiological measures of anxiety, despite the fact that the repressors scored lower on the self-report anxiety measure. Asendorpf Asendorpf and Scherer (1983) replicated and extended thesefindings. findings. Newton and Contrada (1992) also replicated the same verbal-autonomic response dissociation for repressors but only in a public publicevaluation situation. Other behaviors of repressors indicating a fundamental sensitivity to negative perforaffect include avoidance of negative implications of new information, greater perfor mance impairments under stress, and less empathy (reported in D. A. Weinberger, 1990). Weinberger offers several possible explanations for this pattern, including 1990). and somatic cues, and and attribu attribuselective attention, altered construal of environmental and (1990) emphasized repressors' perfor perfortional biases. Although D. A. Weinberger (1990) mance impairments under stress, their tendency to avoid processing sources of negative affect might actually benefit performance in certain tasks (Bonanno, Davis, & Schwartz, 1991). Singer, & 1991). 1990; Davis & & Further exploring the repressor's character, Davis (1987, 1990; 1987) found that repressors not only lack awareness of their current Schwartz, 1987) difficulty gaining access to emotional memories.^ emotional states, but also have difficulty memories.2 A series of studies has suggested that repressors have a comparable range of mem emotional memories but are less able to access them, particularly when the memories involve anger, fear, and self-consciousness (Davis, 1990). Less elaborate processing of emotional experiences and disruptions in "indexing" of emotional memories were proposed as explanations. In the search for consequences, G. E. Schwartz (1990) found evidence that the repressive style impairs physical health and he offers a "psychobiological" model of repressive style. Under some circir cumstances (e.g., conjugal bereavement), however,' however, repressive style has actually proved beneficial for physical health (Bonanno, Keltner, Holen, & & Horowitz, 1995). A nagging psychometric problem with the typological model is the confound confoundx 22 table. As a result, ing created by scoring only three groups of a conceptual 2 X main effects may be responsible for some findings findings attributed to the interaction al, 1989). of anxiety and defensiveness scores (e.g., Warrenburg et aI., 1989). Therefore, ^This Thisfinding findingisisconsistent consistentwith withBonanno Bonannoand andSinger's Singer's(1990) (1990)conclusion conclusionfrom fromaaliterature literaturereview review 2 that repressive style is associated with a preoccupation with relatedness and intimacy.

CHAPTER CHAPTER 22 22

PSYCHOLOGICAL PSYCHOLOGICALDEFENSE DEFENSE

sst 551

rather than a typology based on arbitrary cutoff cutoff points, we recommend a multiple multiplefuture research.3 research.^ regression approach for future Note that this emerging picture of the "repressor" construct construct appears to differ definition of in two important respects from the traditional definition of repression. First, represrepres sion has traditionally been defined as avoiding awareness of of impulses or other of awareness of mental contents, rather than avoidance of of emotions as in the present personality style paradigm. Second, as D. A. Weinberger (1990) notes, repressive personality has traditionally been strongly associated with "hysterical" personality traits, which are in sharp contrast with the dour, phlegmatic, excessively excessively rational subjects in the present studies. For these reasons and others, some critics have questioned the appropriateness of the label "repressor" for such individuals (Holmes & McCaul, 1989; Vaillant, 1992). A A related typology was recently suggested by Shedler, Mayman, and Manis defensiveness scale, however, the adjustment (1993). Instead of being crossed with a defensiveness self-report is crossed with adjustment ratings by clinical judges. Thus the group self-report self-reported and low in clinician-reported cUnician-reported adjustment is assumed to be high in self-reported defensively denying denying their maladjustment. Results from a series of defensively of studies confirmed that, compared to the truly adjusted group, the defensive defensive subjects showed more defensive word associations and higher levels of defensive of physiological physiological responsivity to stress. of adminis adminisClearly, the strong suit of trait and typology measures is their ease of scoring—qualities that facilitate research. Their current status, for better tration and scoring-qualities or for worse, is exemplified in a recent study by Turvey and Salovey (1994): They different origins and rationales, all converged found that extant measures, despite different of trait defensiveness. Nonetheless, they remain empirically on one underlying factor of difficult to untangle from measures of anxiety (p. 288). difficult

VI. IlAAN HAAN AND AND COLLEAGUES VI. influence on contemporary The work of Norma Haan has had the single strongest influence work—on both defense mechanisms and coping. Beginning with a model similar work-on (1963,1965,1977) to that of Kroeber (1963), Haan (1963, 1965, 1977) defined ego actions as processes that accommodate, assimilate, and maintain organization under conflict. The pro processes themselves are not tied to any specific psychoanalytic content (sexual or aggressive impulses). Nor are the ego functions inherently conscious or unconscious; their operation is best described as preconscious—available preconscious-available to discovery, but not persistently within awareness. 3 within the more recent typology; rather than dividing subjects ^This Thisproblem problemseems seemstotobeberesolved resolved within the more recent typology; rather than dividing subjects on defensiveness defensiveness using using the the Marlowe-Crowne Marlowe-Crowne scale, scale, they they are are divided on restraint, restraint, using using aa new new self selfon divided on report measure. measure. Thus Thus repressors repressors are are now now defined as high high restraint/low A. Weinberger Weinberger & & report defined as restraintllow distress distress (D. (D. A. Schwartz, 1990). Schwartz, 1990).

552 552

YES PAULHUS, PAULHUS, FJuDHANDLER, FRIDHANDLER,AND ANDlIA HAYES

Hann and Kroeber reasoned that a taxonomy of of ego actions should not be restricted to defenses involving involving maladaptive distortion of truth and reality, as previ restricted previous work had. Instead, they proposed two independent, parallel modes of of expres expression-coping sion—coping and defense. Coping reflects purposeful, adaptive, conscious, flexible, flexible, and present-oriented behavior choices that adhere to reality and logic and are morally superior (Haan, 1985, 1986). Defenses are compelled, maladaptive, rigid, distorted, and past-oriented behaviors that distort reality. reaUty. If either of of these modes fail, then fragmentation may be invoked. The latter is characterized by irrationally psychopathological symptoms such as ritualistic and automatic behavior, expressed psychopathological of reality. Haan (1977) summarizes this triad by stating, and involves clear violations of "A person will cope if he can, defend if he must and fragment if he is forced forced to do so" (p. 79). Within each of of these modes of expression, the taxonomy was further subdi subdivided into 10 generic subcategories of ego processes. For conceptual convenience, the 10 generic processes are clustered into those which are primarily cognitive, affective regulating, reflexive-introspective, reflexive-introspective, and attention focusing. The choice of affective of expression are dependent upon the situational demands and! and/ process and its mode of of life situations that predispose the individual to idiosyncratic strategies. or a series of Q-sort interview technique composed of of 60 items, Haan (1965) developed a O-sort of 20 coping and defense defense ego processes. The scales can be used by 3 for each of selfclinicians to assess a client's defensive profile. Haan (1965) also developed self of the ego-process scales by administering the MMPI and CPI scales. report versions of identify for each scale those items which differentiated differentiated the subjects She could then identify Although who were rated by clinicians as highly defensive from those who were not. Although self-report versions were used in more than 30 studies, neither the interview or the self-report ever cross validated (Morrissey, 1977). In response to criticisms of of the original scales, Joffe and Naditch developed improved versions of of the ego process scales (J-N; Joffe Joffe & Naditch, 1977). The authors selected those items from the CPI and the MMPI which predicted clinicians ratings of of the 20 coping and defense processes outlined by the model. Those items which correlated the highest with the criterion ego ratings were then cross validated. Factor analyses of of the defense and coping mechanisms (Haan, 1963; Joffe & Naditch, 1977) indicate that the items can be categorized into Controlled Coping, Defense scores. The Haan Expressive Coping, Structured Defense, and Primitive Defense Defense Mecha Mechascale and the J-N have shown some convergent validity with the Defense Inventory (GIeser (Gleser & Ihilevich, 1969; Vickers & Hervig, 1981). There is also nisms Inventory families coping with seriously seriously ill evidence for the predictive validity of the J-N in families 1981; Kupst et aI., al., 1984), and as an antecedent antecedent to children (Kupst & Schulman, 1981; seeking psychotherapy (Gurwitz, 1981). of defenses, as measured by the J-N, has been shown to be positively Use of of Type A A behavior, whereas use of related to Speed and Impatience subscores of coping mechanisms was positively related to Job Involvement (Vickers, Hervig, Rahe, & Rosenman, 1981). Intemal Internal locus of control is associated with the use of coping, and external locus correlated with defenses (Vickers, Conway, & Haight,

CHAPTER22 22 PSYCHOLOGICAL PSYCHOLOGICALDEFENSE DEFENSE CHAnER

55) 553

1983). Haan (1985, 1986) investigated the relations between the Haan and J-N scales and moral development. She concluded that coping strategies facilitate, while defensiveness hinders, mature moral behavior. coping-defense-fragmentation triad represents an expansion In sum, Haan's coping-defense-fragmentation and elaboration of the theretofore narrow focus of of defense defense mechanisms. In addition, the associated ego-process scales represent the first comprehensive battery designed defenses."* to tap defenses.4

VB. Vn. VAILLANT AND COLLEAGUES George Vaillant continues to be a leader in the naturalistic study of of defenses defenses from a psychodynamic point of view (e.g., Vaillant, 1971, 1971,1976,1992; 1976, 1992; Vaillant, Bond, & primarily on three longitudinal Vaillant, 1986). His empirical work has been based primarily psychosamples from which copious data were available. His position has been that psycho logical defenses become clear only over long spans of of time as the individual "adapts '^defensive style" to refer to enduring to life." Vaillant (1971) has used the term "defensive of a tendencies to employ either specific mechanisms or, more often, mechanisms of developmental view of particular level of "maturity." He has taken a developmental of defenses, adherence to reality reality in terms of of maturity describing their degree of adaptiveness and adherence (Vaillant, 1971) and seeking evidence that such maturity correlates with chronologi chronological age and with various measures of mental health. Vaillant's methods, although always empirical in a broad sense, have a novelis novelistic aspect, particularly in his early work (Vaillant, 1971). His approach relies, to a greater degree than most psychological research, on his own impressions of of subjects. This has led to rich description and lucid theory, but has to some extent undermined Neverhis argument that individuals' defensive styles can be objectively identified. Never betheless, he has demonstrated robust and theoretically important associations be defensive style and several measures of of successful adaptation. tween maturity of defensive of varied biographical and evaluative materials gathered longitu longituVaillant's use ofvaried dinally over the course of decades has permitted a direct assessment of of defenses defenses in significant significant life contexts. It is also possible through this method to distinguish enduring defenses from ones that recede once a particular circumstance has passed. Valliant codes defenses by first condensing life history materials into vignettes that demonstrate aa subject's demonstrate subject's techniques techniques of of managing managing life life problems problems and and then then grouping grouping these these vignettes vignettes according according to to which which defense defense they they seem seem to to represent. represent. A A pair pair of of raters raters then theoretical hierarchy then categorizes categorizes each each group group of of vignettes vignettes within within Vaillant's Vaillant's theoretical hierarchy of of 18 defenses, 18 defenses, ranging ranging from from psychotic psychotic to to mature mature (Vaillant, (Vaillant, 1971). 1971). The The raters raters also also code the prominence of each defense in the subject's overall style. of the degre6 degree to which This coding procedure does not permit an assessment of reliable ratings can be made directly from the life historical materials, because all of the materials. (See ratings are mediated by Vaillant's own categorization of *Recently, Recently,another anotherQ-sort Q-sortapproach approachtotodefenses defenseshas hasappeared appeared(Davidson (Davidson&&MacGregor, MacGregor,1996). 1996).

4

SS4 554

PAULHUS, FRIDHANDLER, AND AND HAYES PAULHUS,FRIDHANDLER,

McCuUough, 1992, for a more stringent test of reliability using the same materials McCullough, and defenses). Nonetheless, raters agreed in 70% 70% of cases on the exact label assigned to a cluster. Interrater reliability of ratings of defense prominence ranged widely, -.01 (.72~.84), however, for the more from .01 to .95. .95. Reliability was consistently high (.72-.84), critical dimension of defensive maturity (Vaillant, 1976). with adjustment adjustment and and mental health, Defensive maturity was found to correlate with both cross sectionally and prospectively (Vaillant, 1992). 1992). Nevertheless, maturity of al., 1986). There is some defense constitutes an independent dimension (Vaillant et ai., evidence that the the use of mature styles increases increases over overthe the course courseof ofadult adultdevelopment development (Vaillant, 1976). Specific evidence for a causal role of defensive maturity in adult adjustment is provided by the moderating influence of childhood environment: defense maturity made a difference in adult adjustment primarily for subjects with bleak childhood environments (Vaillant et al., ai., 1986). A related pattern has been found for social class. Correlations with Bond's Defense Style Questionnaire (DSQ) have provided some validation of the ratings of individual defenses. The DSQ is a self-report instrument in which subjects are queried about response styles related to defense (Bond, Gardner, Christian, & & Sigal, 1983). In his influential book, Vaillant (1992) reviewed theory and research in the psychoanalytic and psychiatric approach to the study of defense mechanisms. He summarized the history of these concepts in Freud's work, advocated a renewed emphasis on defenses in diagnosis and treatment planning, and surveyed several prominent nomenclatures (of which his his own ownhas has been been the the most most influential). influential). Contrib Contributed chapters by other authors presented current research. Vaillant reviewed data from his three longitudinal samples and concluded that gender, SES, and culture style; however, the validity of this conclusion do not significantly affect defensive style; is limited by limitations in the method of rating defenses, as discussed previously. Finally, Vaillant presented a revised version of Haan's Q-sort as a new method of obtaining objective and statistically convenient ratings of defensive style from life data. Reliability Reliabilityand andvalidity validitydo donot notyet yetappear appearadequate. adequate.(How (Howhistorical or interview data. [1992] for for aa successful successful application application of of Vaillant's Vaillant's Q-sort.) Q-sort.) ever, see Hart and Chmiel [1992] Perry, a former student of Vaillant, has, together with Cooper, studied the relation of defenses to psychiatric diagnosis and other aspects of functioning functioning (J. C. & Cooper, 1989, Perry & 1989, 1992). 1992). These authors developed the Defense Mechanisms Rating Scales (DMRS), which differ from Vaillant's in using videotaped interviews instead of extensive longitudinal materials and in providing more rigorous assessassess ment of interrater reliability. Significant associations have been found between classes of defenses and psychological symptoms, both subjectively and objectively evaluated. Level of functioning has also shown significant associations with the & Cooper, 1989, defense classes (J. C. Perry & 1989, 1992). However, associations with •diagnosis have not been demonstrated in all of their studies (e.g., Bloch, Shear, .diagnosis & Perry, 1993). Markowitz, Leon, & Led by Perry, a group including Vaillant, Horowitz, Fridhandler, Cooper, and Bond contributed the Defensive Functioning Scale to the latest official psychiatric diagnostic system, the DSM-IV DSM—IV (American Psychiatric Association, 1994, pp. 7511994, pp. 757). This scale-drawn scale—drawn mainly from Vaillant's and Perry's systems with elements 757).

CHAPTER CHAPTER 22 22

PSYCHOLOGICAL PSYCHOLOGICALDEFENSE DEFENSE

555 555

of "axis" of DSM-IV defines of Horowitz's-is Horowitz's—is a proposed additional "axis" of diagnosis. (The DSM—IV defines a diagnostic axis as a "domain of information that may help the clinician cUnician plan treatment and predict outcome [po [p. 25]"; currently, a full diagnosis requires assess assessment on five axes.) The Defensive Defensive Functioning Scale consists of of 31 defined defenses, grouped into seven "levels" according to their mode of action and their degree of adaptiveness or maturity. As part of of the diagnostic evaluation, the patient's most prominent current defenses and overall defensive defensive level are rated. These ratings are of the patient's based on the clinician's observations and his or her interpretation of inhistory; that is, no interview protocol or operationalized rating procedure are in of diagnosis, cluded in the scale at this time. Although currently an optional part of and though its field reliability and validity are unknown, this scale represents the officially sanctioned sanctioned rating rating of of defenses defenses in inpsychiatric psychiatric diagnosis diagnosis and and as assuch such may may first officially ultimately have a significant significant impact on mental health diagnosis and treatment. Vin. HOROWITZ AND AND COLLEAGUES vm.

Mardi Horowitz and his colleagues (Horowitz, 1986; Horowitz, Markman, Stinson, Horowitz & & Stinson, Stinson, 1995) 1995) have have advanced advanced aa theoret theoretFridhandler, & Ghannam, 1990; Horowitz of defense that originated in a psychodynamic framework framework but gradually ical model of of cognitive psychology. The fundamental fundamental assumption in incorporated principles of prothis model is that there exists an unconsciously operating system of "control pro cesses" that govern what is represented in conscious awareness and determine what form this awareness takes. Following the structure of of Haan's (1977) theory, several form of which may have adaptive, maladaptive, or "dysreg "dysregprocesses are posited, any one of ulatory" outcomes depending on the context in which it is employed. defenses are retained-29 retained—29 are specifically cited The traditional psychoanalytic defenses 1988)—but are not considered considered to be fundamental fundamental processes. Instead, (Horowitz, 1988)-but following the distinction made by Wallerstein (1983), the traditional defenses defenses are 1990). For regarded regarded as as outcomes outcomes of of more more basic basic cognitive cognitive processes processes (Horowitz (Horowitz et et aI., al., 1990). For example, the psychoanalytic mechanism "undoing" "undoing" is thought to be accomplished through the process of "sequencing ideas through switching concepts," that is, is, rapidly shifting one's one's beliefs beUefs or point of view (Horowitz, 1988, 1988, p. p. 202). 202). The The traditional traditional psychoanalytic "control" operates operates unconsciously psychoanalytic assumptions assumptions that that defense defense or or "control" unconsciously for for the the purpose purpose of of keeping keeping conflictual conflictual material material out out of of awareness awareness are are retained. retained. In ap In addition addition to to this this synthesis synthesis of of cognitive cognitive psychology psychology with with psychoanalytic psychoanalytic approaches, Horowitz's proaches, Horowitz's view view of of defense defense is is distinguished distinguished by by the the assertion assertion that that defense defense can 1988; can be be accomplished accomplished through through activation activation of of certain certain person person schemas schemas (Horowitz, (Horowitz, 1988; Horowitz aI., 1990). Horowitz et et al., 1990). In In this this view, view, individuals individuals have have multiple multiple images images of of self self and and others, others, as as well well as as mUltiple multiple images images of of typical typical interactions, interactions, termed termed "role "role relationship relationship models." models." One One method method of of defense defense is is to to change change the the image image or or schema schema that that is is currently currently active, which in turn controls those specific aspects of self and other that are currently perceived. Aspects of of this theory were drawn from the object-relations perspective, increasingly influential influential in psychoanalysis (Kernberg, 1976). which is increasingly

PAULHUS, FRlDHANDLER, AND PAULHUS^FRIDHANDLER, AND HAYES

SS6 556

This approach approach originated originated in in the the study study of of psychological psychological coping copingstrategies strategiesfollow following traumatic stress (Horowitz, 1986). Observing a regular pattern of ''intrusions" "intrusions" ''omissions" in consciousness following trauma, Horowitz concluded that unand "omissions" un conscious control processes must be present. Examination of videotaped psychopsycho theratherapies of individuals with unresolved posttraumatic reactions revealed that thera pists helped these persons to modify certain control operations, and the controls identified in this way became the basis for the developing taxonomy. Three levels of regulation were specified: regulation of mental set, regulation of person schemas, and, at the most fine-grained fine-grained level, regulation of conscious representations and sequencing (Horowitz, 1988). Although the system is primarily theoretical, it is being applied in the intensive study of single subjects (Horowitz et al, al., 1994).

IX. YI AND IX. ERDEL ERDELYI AND COLLEAGUES Erdelyi has offered the most sustained and detailed cognitive treatment of defense. has proposed a broad theoretical framework subsuming the major experimental He has literatures on defense, namely, perceptual defense (Erdelyi, 1974) and repression & Goldberg, 1979). He argues (Erdelyi, 1985) that effects on memory (Erdelyi & can be be selective information rejection laboratory evidence demonstrates that "there can from awareness" (p. 259), although it has not been experimentally established that this capacity is in fact used toward defensive ends. He has claimed further that defensive bias influences the entire sequence of information processing and that there is no reason to believe that a single mechanism is responsible for all this defensive activity. For example, perceptual defense may involve processes ranging from ocularfixation fixation to selectivity in transfer from raw storage (e.g., iconic storage) into short-term memory. In his more recent work (Erdelyi, 1990), he has proposed a mechanism for repression that he he argues is fully consistent with with Sigmund Sigmund Freud's Freud'sviews, views,particularly particularly his earlier ones. Erdelyi argues that "repression" "repression" refers refers to to purposefully purposefully not not thinking thinking of a target memory (or thinking of something else), which results in the target being overtaken by "oblivescence" (or, more simply, forgetting), resulting in amnesia. He cites evidence that there are two contrary tendencies in memory, reminiscence and oblivescence/forgetting. Thinking about a memory-"rehearsing memory—"rehearsing the remembered pur and searching for the inaccessible" (1990, p. p. 4)-promotes 4)—promotes the former, and if pursued in aa sustained sustained way wayresults resultsin inhypermnesia hypermnesiaor orgreater greateraccessibility accessibilityof ofthe thememory. memory. "Not thinking" thinking" yields the opposite results, namely, amnesia. This amnesia is not necessarily permanent, however; it is often reversible by thinking of the target memory. In other words, due to the properties of human memory, deliberate "not sufficient to remove a memory from awareness. This process thinking" over time is sufficient is potentially reversible, in keeping with therapist reports of recovery of rere pressed memories. Erdelyi (1990) cites experimentalfindings findings to support this view. He compares & classic findings findings on forgetting over time with one of his own studies (Erdelyi &

CHAPTER 22 22 CHAmR

PSYCHOLOGICALDEFENSE DEFENSE PSYCHOLOGICAL

557 557

Kleinbard, 1978) showing increasing memory over time. This difference, he suggests, is due to thinking versus "not thinking." He He also also cites cites clinical clinical findings with neurologi neurologirelated procedural cal amnesics that declarative declarative facts may be forgotten while related procedural aspects childof memory remain. This is consistent with Freud's contention that repressed child of actions and emo form of emohood events are not remembered but are "repeated," in the form tions. Although psychoanalytic writers and many others (cf. (cf. Heilbrun & Pepe, 1985) have generally assumed that people are unaware of of their defensive defensive activity, Erdelyi asserts that there is nothing about repression that must be unconscious. In addition, he argues that repression need not be defensive at all. In In this this definition, definition, repression is process that is simply simply not not thinking thinking in in order order to to produce produce amnesia, amnesia, aa process that may may serve serve defensive defensive or nondefensive nondefensive ends. Although Erdelyi focuses on repression, he offers framework for understand offers a framework understanding other defenses as well. This framework is based on Bartlett's view of of memory as a reconstructive reconstructive process. Such reconstructions might well be subject to defensive defensive A 9-year-old 9-year-old girl was asked bias. Erdelyi (1990) provides an example of such bias. A to read a brief story and then was asked to reproduce it on several successive of the frightening of the story occasions. In her early recall efforts, most of frightening aspects of experimental studies, form the were absent. Such case studies, along with a few experimental basis of Erdelyi's well-articulated theory of defense. X. ImLEVICH AND GLESER GLESER X

Goldine GIeser Gleser and David Ihilevich (1969; Ihilevich Ihilevich & & GIeser, Gleser, 1986) have developed developed a theory of responses to conflict and an instrument to assess them-the them—the Defense Defense classification Mechanisms Inventory (DMI). The authors proposed a two-tiered classification system to describe possible responses to threat. At the first level, they distinguished between problem-solving efforts that are directed at changing oneself oneself and those efforts aimed at changing the surrounding context. Second, they classified all possible responses as problem-solving, coping, or defense strategies, although they admitted that they were unclear as to how these three levels interact. Utilization of of any of of the three modes of of response to conflict is assumed to reduce anxiety and enhance self-esteem. Problem-solving strategies (changing Problem-solving strategies oneself, the threat. oneself, changing changing the the environment) environment) have have aa direct direct impact impact upon upon the threat. Coping Coping strategies hope) have strategies (assertiveness, (assertiveness, trust, trust, stoicism, stoicism, responsibility, responsibility, and and hope) have an an influence influence on upon the the threat, threat, not not upon upon the the threat on the the individual's individual's mood mood or or motivation motivation expended expended upon threat itself, and are based in trust and and reality. reality. Defenses Defenses (aggression, projection, intellectual intellectualization, intrapunitive, and repressive actions) actually remove the problem from awareness or distort it, thus falsifying falsifying reality and offering an illusion of of control. Coping and defensive strategies arise only when problem-solving strategies are ineffective or inappropriate. In addition, the appropriateness and the adaptiveness of the choice of response are dependent upon the context and the individual's of characteristic style.

558 558

PAULHUS, FRlDHANDLER, PAULHUS, FRIDHANDLER,AND ANDHAYES HAYES

The DMI (Ihilevich & Gieser, Gleser, 1986) was designed to assess five categories of response to threat: Projection (PRO), Principalization (PRN), Turning Against Object (TAO), Turning Against Self Self (TAS), and Reversal (REV). Subjects respond to the 10 scenarios by selecting their most and least likely reactions, in terms of of their actual behavior, affect, thoughts, and fantasy behavior. To date, the scale is self-report measure of of defense mechanisms in research the most extensively utilized self-report see Ihilevich Ihilevich & & Gieser, Gleser, 1986, 1986, for for aa compre compre(Cramer, 1988; Vickers & Hervig, 1981; 1981; see hensive review). Generally, the DMI has found to have high retest and interitem reliability, but there are some problems with the conceptual status of of TAO as a vahdity for the PRN scale (Cramer, 1988). Concurrent validity defense and weak validity has been shown between the DMI and the MMPI defensive scales, the Byrne Repression-Sensitization scale, Haan's Q-sort Technique, Joffe Joffe and Naditch's Ego Repression-Sensitization Process scale, Coping Operations Operations Preference Preference Enquiry, Enquiry, the the Blacky Defense Process scale, Schutz's Schutz's Coping Blacky Defense Preference Inquiry, Inquiry, and and the the denial denial subscale subscale of of the the Marlowe-Crowne Marlowe-Crowne Social Social Desir DesirPreference ability scale. ability scale. interrelationships among the five scales indicate It has been suggested that the interrelationships that the DMI represents a single continuum. For example, males score higher on PRO and TAO, and TAO has been related to masculine orientation. Females score TAS, higher on T AS, which is related to feminine orientation. Assertive individuals endorsed PRN, whereas nonassertives endorsed TAO and TAS. As such, the DMI has been conceptualized as a continuum with acting-out defenses at one end and inwardly focused defenses at the other (Cramer, 1988; 1988; Juni & Yanishefsky, 1983). of aggression It has also been argued that this continuum represents an expression of of aggression (REV and PRN) at (TAO and PRO) at one pole, and an inhibition of of the DMI was provided by the findings that TAO and the other. Clinical validity of TAS PRO are higher in psychiatric patients, and T AS is higher in suicidal patients. of interpersonal Finally, the DMI scales have been linked to a promising new system of StoUak, & Loraas, 1994). defenses (Woike, Aronoff, Stollak, APPROACH XI. PERCEPT-GENETIC ApPROACH XI.

Over a period of of four decades, a group of of psychologists centered at Lund University dynamic-constructivist view of of personaUty perin Sweden have advocated a dynamic-constructivist personality called per cept genesis (PG). The theory concerns how perceptions evolve, how they are cept of external interlinked with behavior, and how they form the individual's conception of reality. For a recent collection of studies, see Hentshel, Smith, and Draguns (1986). Percept genesis was derived from the microgenesis work of the 1920s (e.g., Sander, 1928/1961), which presumed that unattended processes preceded conscious perception. In effect, the perception of of external reality is an an outcome of of internal processes prone to subjective influences. Microgenesis assumes that the immediacy of percepts is of percepts is in in fact fact an an illusion illusion arising arising from from the the perceiver's perceiver's focus focus on on correct correct recogni recognition of of the stimulus. Analysis of of the preparatory phase through repetition of brief presentations can reveal the processes leading up to recognition. PG theorists

CHAPTER CHAPTER22 22

PSYCHOLOGICAL PSYCHOLOGICALDEFENSE DEFENSE

559 559

emphasize that that these emphasize these processes processes are are highly highly influenced influenced by by the the individual's individual's personality personality developand experiences. Thus, percepts are deeply rooted in the perceiver's develop mental history. The PG model further suggests that perception recapitulates ontogeny; that is, perception reflects the sequence of life experiences. One cannot examine a of the process without reference to earlier stages. "PG *TG sees the perceptual section of subjective prestages to the final intersubjective intersubjective act as a process moving from the sUbjective of the stimulus" (Smith & Westerlundh, 1980, p. 109). meaning of Tachistoscopic methods form the basis for a number of of PG tests designed to measure defenses. Two of the most widely cited tests are the Defense Mechanisms Meta-Contrast Technique (MCT; Kragh & Smith, Test (DMT; Kragh, 1960) and the Meta-Contrast Smith & & Henriksson, 1956). 1956).These These methods methods entail entail an an examination examination of a subject's subject's 1970; Smith stimulus—a focally central person. Initially below changing interpretations of a hero stimulus-a perceptual threshold, the exposure time is gradually increased until the subject identification of the hero stimulus. Next, a threatening stimulus reaches a stable identification (e.g., a monstrous figure sneaking up) is gradually introduced to provoke anxiety. interVarious defensive reactions are evaluated by analyzing the subject's changing inter of the hero and threatening stimuli, and/or the temporal delay to correct pretations of of the stimuli. Subjects often often show defensive interpretations of the recognition of before the threat has been accurately identified. stimulus before of associated methodologies, the PG researchers have provided Using an array of different perceptual stages evidence for their distinctive patterns of defenses during different different life stages, and in different different forms of psychopathology. In Sweden, the and different DMT is in standard use by the military for pilot selection. The tests have also been utilized to predict soldier and attack-diver abilities, and to distinguish decorated veterans from nondecorated veterans. These techniques are also regularly applied in clinical research. One finding was that repression is the central defense defense mechanism in hysterics, and isolation the central strategy for compulsives. (For aa review review of of in hysterics, and isolation the central strategy for compUlsives. (For research, see Smith & & Westerlundh, 1980). ScanAlthough research with PG methodologies appears to be widespread in Scan dinavian nations and growing in other European countries (e.g., Cline, 1987; 1987; C. KiessUng, 1990), for some reason it is still minimal in Cooper, 1988; Hentschel & Kiessling, North America. XII. . PLUTCIDK PLUTCHIK AND AND COLLEAGUES XII

Plutchik and colleagues view defenses defenses as more accessible and changeable than the Kellerstrictly unconscious process assumed by other theoreticians (Plutchik and Keller man, 1980). On the basis of Plutchik's (1962; Schaefer & Plutchik, 1966) theoretical and empirical work on emotions, the defenses are said to vary in their degree of similarity to each other, show a circular configuration, and lie on a continuum of most to least primitive. The 8 basic defenses are said to derive from 8 basic emotions within an evolutionary framework. framework.

560 560

PAULHUS, YES PAULHUS, FRmHANDLER, FRIDHANDLER,AND ANDIIA HAYES

The authors identified 16 defense processes from psychiatric and psychoana psychoanalytic literature, and assembled them in The Life Style Index. On the basis basis of of the responses of samples, extreme and ambiguous items were eliminated, of several normal samples, and only those items that differentiated differentiated the highest scorers from from the lowest scorers defense mechanisms pro prowere retained. Clinicians' matchings of the items to the defense A comprehensive review of vided further refinement. A of findings with the scale is listed in Conte and Plutchik (1995). Clinical CUnical validity was supported by the finding that schizophrenics scored sig significantly higher than college students on all scales. High-self-esteem High-self-esteem individuals score lower on regression, compensation, projection, and intellectualization, and highly anxious individuals showed a reverse pattern plus low scores on denial. Rim (1987,1989) (1987, 1989) showed subject gender and age to be moderating factors to Plutchik's scales, and found that Extraversion was positively correlated with minimization, neurotimapping, and reversal, but was negatively correlated with blame, whereas neuroti cism was positively related to minimization and suppression. of defenses defenses is supported in that: In short, the construct validity of this set of (1) psychiatric patients use defenses more than do normals, and (2) among normals, those who use defenses tend to have lower self-esteem. Most recently, Plutchik and of eight coping variables colleagues have extended the model to a parallel set of (Plutchik & Conte, 1989). Xin. . STRESS AND AND COPING XIII difficult to draw a clear line between the literature on coping and Although it is difficult that (a) active, that on on defense, defense, coping coping concepts concepts place place more more emphasis emphasis on on (a) active, conscious conscious efforts measurement, and efforts at at managing, managing, (b) (b) process process rather rather than than trait trait measurement, and (c) (c) variability variability across are Haan across situations. situations. The The major major predecessors predecessors of of current current work work are Haan (1963), (1963), Moos Moos (1974), (1974), and and Pearlin Pearlin and and Schooler Schooler (1978). (1978). In theory and research on In recent recent years, years, the the most most active active program program of of theory and research on coping coping has Project, summarized has been been that that of of the the Berkeley Berkeley Stress Stress and and Coping Coping Project, summarized in in Lazarus Lazarus and postulates three distinguishable stages stages as and Folkman Folkman (1984). (1984). The The approach approach postulates three distinguishable as critical critical mediators of relations and immediate and mediators of stressful stressful person-environment person-environment relations and their their immediate and long longrange outcomes: primary appraisal, secondary and coping. coping. Through pri range outcomes: primary appraisal, secondary appraisal, appraisal, and Through primary person evaluates or she has anything mary appraisal, appraisal, the the person evaluates whether whether he he or she has anything at at stake stake in in the anything, can the encounter. encounter. Secondary Secondary appraisal appraisal involves involves evaluating evaluating what, what, if if anything, can be be done. cognitive and done. Finally, Finally, coping coping is is defined defined as as the the constantly constantly changing changing cognitive and behavioral behavioral efforts resources (Lazarus (Lazarus & Folkman, efforts to to manage manage demands demands that that exceed exceed the the person's person's resources Folkman, 1984). An of coping 1984). An individual individual may may use use any any one one of of aa wide wide variety variety of coping responses, responses, depending on the situation and recent events, as well as individual coping style. Several standardized measurement instruments arose from this research propro gram, including the Hassles and Uplifts scale (DeLongis, Coyne, Dakof, Folkman, & Lazarus, 1982) and the Ways of Coping scale (WOC; Folkman & Lazarus, 1985).

CHAFfER CHAPTER 22 22 PSYCHOLOGICAL PSYCHOLOGICALDEFENSE DEFENSE

561 561

The WOC tool. The w o e is perhaps the best-known and most widely utilized research tool respondent is asked to think of supplied with a specific of a recent stressful event or is supphed specific example. The subject then responds by checking the applicable responses from a list ("I tried list composed composed of of aa broad broad range range of of cognitive cognitive strategies strategies ("I tried to to forget forget the the whole whole thing") and behavioral strategies ("I got professional professional help") that people use to manage internal or external demands in stressful encounters. The original WOC (Folkman & Lazarus, 1980) contained 68 items in a yes/no format indicating whether whether the respondent respondent had or had not used each strategy. In the revised version containing 67 items (Lazarus & Folkman, 1984), the response format format was changed from yes/ four-point Likert scale (0, does not apply or not used; 1, used somewhat; no to a four-point 2, used quite a bit; 3, used a great deal). Early factor analyses of the WOC indicated problem-focused and emotion-focused emotion-focused strategies (Lazarus & two general factors: problem-focused confrontFolkman, 1984). More recent factor analyses have indicated eight factors: confront responsibilive coping, distancing, self-controlling, seeking social support, accepting responsibil ity, escape-avoidance, escape-avoidance, planful planful problem solving, and positive reappraisal (Folkman, Lazarus, Dunkel-Shetter, DeLongis, & Gruen, 1986). McCrae and Costa (1986) have shown that, by changing the referent from a specific situation to "what you different strategies to generally do," the WOC becomes a trait measure, linking different each of of the the Big Five dimensions dimensions of of personality. personality. each Big Five Studies by the Lazarus group tend to emphasize changes in strategies across situation. In a typical study, stress and coping among students were examined at of a university examination (Folkman & & Lazarus, 1985). Hopefulness, Hopefulness, three stages of whereas eagerness, worry, and fear were most common during preparation period whereas happiness, relief, disgust, and disappointment were more common after grades had after the exam but before before grades been announced. During the intermediate stage, after had been announced, high levels of of all the above emotions were reported. model to incorporate an DeLongis and O'Brien (1990) have extended the model interpersonal factor. And the model has stimulated research by other groups. At least two other instruments have been developed that expand on the WOC scale (Carver, Scheier, & Weintraub, 1989; McCrae, 1984). Other researchers have fo focused on the processes underlying coping. For example, the notion of of affect affect regula regulation has provided a homeostatic, hydraulic model for the process of coping with stress (e.g., G. E. Schwartz, 1977). Finally, Carver and Scheier (1981) have developed the most comprehensive process model, labeled attention/self-regulation attention/self-regulation theory. Several independent programs of research on coping warrant brief brief mention here. Seymour Epstein has postulated a construct of of "constructive" thinking that links Hnks all adaptive coping responses to a single global factor (Epstein & Meier, 1989). Suzanne Kobasa proposed the construct of "hardiness" to capture the psychological effects in those groups (e.g., executives) factors that minimize the physical health effects particularly subject to prolonged stress (Kobasa, Maddi, & Kahn, 1982). Finally, the Plutchik and Conte (1989) coping constructs were derived from Plutchik's of emotion. evolutionary theory of

562 562

PAULHUS, FRlDHANDLER, AND PAULHUS,FRIDHANDLER, AND HAYES HAYES

XIV. XIV. INFORMATION.PROCESSING INFORMATION-PROCESSING ApPROACHES APPROACHES

informationIn tune with the cognitive revolution, Erdelyi (1974) called for an information processing approach to studying defenses. Issues of consciousness, repression, and threat became issues of attention, memory, andfiltering. Since then a number of filtering. Since theoretical pieces have havebeen beencouched couchedininsuch suchinformation-processing information-processingterms terms(Grzego (Grzegolowska, 1976; 1976; Hamilton, Hamilton, 1983; 1983;G. G.E.E.Schwartz, Schwartz,1977). 1977).Indeed, Indeed, impact cognitive thethe impact ofof cognitive psychology may be seen in most modern analyses of defense either in terms of of theory (e.g., Horowitz) or in terms of method (e.g., percept-genetic). Before we return to these broad theories, we will consider some smaller domains that have been subject to particular scrutiny. A. Individual Individual Differences Differences A. Heilbrun's recent work is a good example of the direct application of cognitive (1985), for psychology to the measurement of defenses. In Heilbrun and Pepe (1985), defenses are are assessed assessedby byexamining examiningthe thecognitive cognitiveprocessing processingofofself-descrip self-descripexample, defenses tions under various motivational conditions. conditions. Discrepancies between various condicondi tions provide measures of projection, repression, rationalization, and denial. denial. The authors concluded that unconscious utilization of projection and rationalization was related to successful control of stress, stress, whereas unconscious repression was related to excessive stress. stress. The conscious use of denial was related to a low level of stress.

B. B. Attention Attention and and Defense Defense The dynamics dynamicsof ofdefense defense can canbe bestudied studiedby byexamining examiningthe theinterplay interplaybetween betweenselective selective attention and attentional breakdowns known as "intrusions," '"intrusions," that is, is, the partial interference of threatening thoughts in in some someongoing ongoingthought thoughtprocess. process.Sophisticated Sophisticated cognitive methodology and analyses (e.g., signal signal detection) detection) are are necessary necessary to to capture capture such phenomena. phenomena. Spence Spence(1983), (1983),for forexample, example,showed showedthe theindirect indirecteffects effectson onspeech speech patterns of weakly defended beliefs. Nielsen and Sarason (1981) examined disrup disruptive effects effects of sexual and achievement-related distractors on a dichotic shadowing task. Bonanno and Wexler (1992) also found selective perception effects effects as a func function of stimulus affective valence. valence. Finally, Blum and his colleagues used hypnotic inductions to condition affect to arbitrary words words (e.g., (e.g.,Blum, Blum,1986; 1986;Blum Blum&&Barbour, Barbour, 1979): de 1979): the disruptive effects faded over time as selective inattention gradually developed. Wegner's recent research (1989) has suggested that intrusions actually result suppressunwanted unwantedthoughts. thoughts.Subjects Subjectsinstructed instructedtotoavoid avoidaaparticu particufrom attempts to suppress (e.g., white bears) were later reported to have more intrusions of such lar thought (e.g., thoughts than a group of subjects actually instructed to think about white bears. colleagues have have demonstrated demonstrated aa link link between between disrup disrupFinally, Paulhus and his colleagues & Van Selst, 1989). 1989). For example, Paulhus and tion and defense (Paulhus, Graf, & Levitt (1987) (1987) found found that, that, in in the thepresence presence of of threatening threatening distractors, distractors, subjects subjectsshowed showed

CHAPTER CHAPTER ZZ 22

PSYCHOWGICAL PSYCHOLOGICALDEFENSE DEIENSE

S63 563

a temporary temporary increase in the positivity of self-descriptions. This sequence provides provides an automatic mechanism for for defending defending the individual under under stress (Paulhus, 1993). 1993). As a whole, this body research points to a dynamic attentional substrate for body of of research for defense. psychological defense. c. C. Subliminal Impact

validity of Many cognitive psychologists have also come to accept the vaUdity of subliminal form somewhat somewhat different different from from early models (Bargh, 1984; perception, albeit in a form Dixon, 1981; Marcel, 1983; Zajonc, 1980). The work of of Silverman and his colleagues Activation (SPA) of of un un(e.g., Silverman, 1983) on the Subliminal Psychodynamic Activation fantasies warrants some acknowledgement acknowledgement in a review of of defense defense mecha mechaconscious fantasies support the hypothesis nisms literature. Some 100 articles and doctoral dissertations support efficacy of of the methodology methodology 4 to 11 (for (for reviews, see Hardaway, 1990; and the efficacy 1990). This This extensive body body of literature literature Silverman, 1983; J. Weinberger & Silverman, 1990). influenced the work of of others, particularly particularly the recent percept perceptis well cited and has influenced genesis theoreticians (discussed later). defenses per se, and research on individual differences, differences, have not However, defenses focus of of the SPA investigations. Potentially Potentially,, SPA could be used to stimulate stimulate been the focus defense in the laboratory laboratory (Geisler, 1986). However, the theoretical and empirical empirical defense foundations of of SPA have recently recently been the subject subject of of trenchant trenchant critiques (for (for foundations Brody, 1988; 1988;for for reply, reply, see see J. J. Weinberger, Weinberger, 1989). 1989). reviews, see Balay & Shevrin, 1988; 1988; Brody, Psychophysiology of of Defense Defense D. Psychophysiology influence of of cognitive psychology has also prompted prompted wider use of of psychophysio psychophysioThe influence logical measurements (e.g., Epstein & Clarke, 1970; Shevrin, 1988). For example, been used to indicate repression (e.g., (e.g.. Hare, 1966; D. A. Wein Weinskin response has been berger et al., Repression-prone individuals have also shown increased increased evoked evoked berger aI., 1979). Repression-prone for unacknowledged unacknowledged threats (Shevrin, Smith, & Fritzler, 1970). Finally, potentials for Aronoff, and Messe (1986) studied the role of of defensiveness defensiveness in impression Assor, Aronoff, formation using physiological arousal as a dependent dependent measure. In the study of of formation of psychopathology psychopathology as a whole, a dual rationale for for studying defense, as in the study of psychophysiological responses is that they are commonly seen as an indicator of of psychophysiological psychological damage (Davidson, 1993) as well as defensive defensive activity (Gerin et aaI., l , 1995). XV. SOCIAL PSYCHOLOGY xv. of defense defense and the unconscious were virtually taboo in For many years, the topics of social psychology. psychology.^5 This rejection rejection peaked with the advent of of attribution attribution theory, 5 core concepts, forfor example, cognitive dissonance, were ^At Atthe thesame sametime, time,ititseemed seemedthat thatcertain certain core concepts, example, cognitive dissonance, were simply simply euphemisms euphemisms for for the the study study of of defense defense mechanisms. mechanisms.

S64 564

PAULHUS, PAULHUS, FRlDHANDLER, FRIDHANDLER,AND ANDHAYES HAYES

where the tendency was to explain explam all mental processes in terms of "cold **cold cognition," that is, cognition devoid of affect (e.g., Greenwald, 1980; 1980; Nisbett & Ross, 1979). Although traditional terminology is still eschewed, the 19808 19908 have seen 1980s and 1990s defense—motivation, the unconscious, and an active interest in the elements of defense-motivation, even the possible benefits of bias. Indeed, recent reviews of social cognition now accept the importance of these elements (e.g., Fiske & Taylor, 1991; 1991; Showers & Cantor, 1985). In social psychology, the concept of a motivation has emphasized maintaining selfself-esteem rather than warding off anxiety. Typically, threats to self or enhancing self-esteem esteem are induced by fabricating academic and social failures (in contrast to psychoanalytic threats) that may be studied in the laboratory. In the revised theory of cognitive dissonance (Aronson, 1969), for example, a threat to self-esteem self-esteem is of considered necessary for dissonance reduction. The most comprehensive of these self-esteem maintenance: Four motivational models is Tesser's (1986) theory of self-esteem closefactors (maintenance, relative performance, importance of the domain, and close self-esteem and, ness of the comparison other) interact to determine threat to self-esteem and, therefore, subsequent subsequent behavior behavior (Tesser (Tesser & Campbell, Campbell, 1982). therefore, 1982). Greenwald's (1980) seminal article extended the notion of defense to cognitive defense conservation. Indeed, a sustained program of research by Swann has that defense of self-esteem is less important than shown defense defense of identity (e.g., Swann, 1992). of Baumeister (1993) went further to cite the motivation to escape the self self to explain defensive phenomena. C. C. R. Snyder's elaboration of "excuse "excusea wide range of defensive making" (e.g., Snyder & Higgins, 1988) 1988) also broadened the range of defensive defensive processes to include protection of self-image and sense of control. encouraged The evidence for "depressive realism" (Mischel, 1979) has also encouraged social psychologists to consider possible positive consequences consequences of inflated self selfperceptions (e.g., Kruglanski, 1989). This view is best represented represented in the influential influential review by Taylor and Brown (1988). They lay out the benefits benefits of positive illusions for mental health. They also distinguish these beneficial beneficial positive illusions from traditional traditional defenses, defenses, which which they they view view as as maladaptive. maladaptive. The The Taylor Taylor and and Brown Brown review review was followed by up an entire issue of the Journal of Social and Clinical Psychology titled "Self-Illusion: When Are They Adaptive? (Snyder, 1989). Colvin and Block (1994) countered with data suggesting that self-enhancement self-enhancement illusions are funda fundamentally detrimental. socialOver time, the term "defense" has gradually crept into a variety of social psychological terms such as "defensive attribution," "defensive "defensive self-presentation," self-presentation," and "defensive pessimism" (Norem & Cantor, 1986). At least one active topic has term—projection (e.g., J. D. Campbell, 1986; Holmes, 1981; retained the traditional term-projection while distinguishing distinguishing between between attribution attribution and and defensive defensive Paulhus & Reynolds, 1995) while forms (Sherwood, 1980). sum, it appears that social psychologists have begun to address virtually In sum, the full gamut of psychoanalytic defenses, albeit with different different labels. Many would argue that this delay was necessary because, only now, with improved laboratory

CHAPTER 22 22 PSYCHOLOGICAL PSYCHOLOGICALDEFENSE DEFENSE CHAPTER

S6S 565

psychoanalytic community, can technology and with less pressure from a dominant psychoanalytic such phenomena be studied effectively. effectively. XVI. DEVELOPMENTAL ANALYSES XVI.

A A number of developmental psychologists (Chandler, Paget, & Koch, 1978; Cramer, 1983; Feldman & & Custrini, Custrini, 1988) 1988) have have proposed proposed that that defenses defenses can can be conceptualized conceptualized along a developmental continuum, according to their complexity and degree of maturity. Following Piaget's stage model of cognitive development, and based on the belief belief that defensive strategies vary in their complexity, these writers have different stages and (b) there are argued that (a) the various defenses appear at different identifiable stages of of development for each specific defense. identifiable Unked to an infant's lack For example, denial occurs early in childhood and is linked of muscular ability abiUty to remove itself itself from anxiety-arousing situations. Sleep is thus of noxious a common behavioral manifestation. Later, a child physically acts to exclude noxious of danger. stimuli (hands over eyes), and finally uses language to deny the existence of More advanced defensive strategies, such as projection, emerge later in childhood, tied again to physical and cognitive developments. Intellectualization, asceticism, identification appear still later, typically, in adolescence. Vestiges of of all mecha mechaand identification of the later-developing later-developing nisms can and do exist into adulthood, but a preponderance of defenses is presumed to exist in the healthy adult. defenses emerge. There is some consensus about how the more advanced defenses Feldman and Custrini (1988) argue that as children mature, they gain an increased ability to perceive when others are being deceptive, and should better understand their own self-deceptive self-deceptive activities, such as when they utilize denial. Thus, a broader of more effective effective defensive defensive strategies is required to better deceive oneself range of off anxiety. A A child is forced forced to abandon an earlier, simpler defense defense in and ward off of an increasing awareness of its operation: a conscious defense is an the light of ineffective ineffective defense (Cramer, 1983, 1991). Thus, it must be replaced with a more complex strategy that remains out of of awareness, and therefore is effective. effective. Chandler et al. (1978) found found evidence to support this developmental developmental sequence. Preoperational defensive strategy. At Preoperational children are incapable of of comprehending comprehending any defensive a slightly older age, concrete operational children are capable of inverse (repression, denial) and later reciprocal (displacement, reaction formation, rationalization) de defenses. Finally, formal operational operational children can employ all types, including the most complex defenses, projection and introjection, which deal with statements about statements and second-order propositions. Cramer (1991) (1991) has confirmed confirmed a develop developmental sequencing of of denial, projection, and identification. Feldman, Jenkins, and Popoola (1979) indirectly validated these findings in a study on the development of self-deception self-deception techniques in children. of of defenses in children and adolescents, the For comprehensive treatments of referred to recent books by Cramer (1991) and Smith and Danielsson reader is referred (1982).

566 566

PAULHUS, FRIDHANDLER, AND PAULHDS,FRIDHANDLER» AND HAYES HAYES

XVII. RELATIONS AMONG THEORETICAL TAXONOMIES A number of the theoretical systems noted above include taxonomies of defense defense— some even describe the structural relations among them. Unfortunately, these taxotaxo We note four nomies differ dramatically both in terminology and in organization. We their cognitive cognitive complexity complexity common criteria for categorizing defensive processes: (a) their al., 1978; 1978; Cramer, 1983), (b) their internal internaland level of development (Chandler et aI., & Ihilevich, 1969; 1969; Ihilevich & & Gleser, external orientation (Gleser (GIeser & GIeser, 1986), (c) their 1956, 1969, 1969, 1977; and (d) (d) their level maturity-immaturity (Haan, 1956, 1977; Vaillant, 1971), and 1977; Lazarus & Folkman, Folkman, 1984; 1984; Vaillant et aI., al., 1986). of conscious awareness (Haan, 1977; eye. A closer examination However, there is less diversity than meets the eye. reveals that these taxonomies have some fundamental similarities. By pointing out the similarities in their organizational principles, we may help reconcile apparently diverse systems. and OMI DMI models involve a distinction For example, the Lazarus, Epstein, and between problem-focused responses (those altering the troubling transaction) and and affect regulation). This dichotomy, emotion-focused responses (those directed at affect and avoidance avoidance strategies strategies (Suls (Suls & & to some extent, parallels that between attentional and Fletcher, 1985; 1985; Taylor, 1990). 1990). Miller's (1989) distinction between monitoring and and blunting has has a similarflavor. Such theorists theorists suggest suggest that thatavoidant avoidant or or emotion-focused emotion-focused flavor. Such are superior in managing short-term or uncontrollable stress whereas strategies are may be more effective for long-term or attentional or problem-focused strategies may controllable stressors (Lazarus, 1986; 1986; Suls & Fletcher, 1985; 1985; Taylor & Clark, 1986). Another growing theme distinguishes defensive from enhancement processes: pro one form minimizes negative information about the self, and the other form promotes positive information (e.g., Sackeim, 1983). 1983). Some writers have argued further that, ultimately, a good offense can can have defensive value, that is, it can can buffer buffer the individual from subsequent threats. Examples of enhancement processes include Taylor and and Brown's (1988) positive illusions and Paulhus and Reid's (1991) self selfdeceptive enhancement. Although those writers see offensive and defensive propro repre as independent, Baumeister, Tice, and Hutton (1989) argue that they reprecesses as sent default strategies of high-versus low-self-esteem low-self-esteem individuals. Using another common organizing principle, Cramer and the DMI DMI theorists argue that certain defenses are are internally oriented (for example, turning against self) while others can can be placed on an an externally oriented pole (projection). Thus, & GIeser, Gleser, 1986) 1986) attempts are made to relate in empirical work (e.g., Ihilevich & and locus of control. Starting with with Cohen Cohen (1964), (1964), defensive styles tofield field articulation and the articulation articulation of of the the defensive defensive styles styles of ofthose those with with guided the a similar distinction has guided and low self-esteem (projectives). high self-esteem (defensives) and maturity: Haan's Haan's coping copingAnother useful organizing principle is a hierarchy of maturity: parallel Vaillant's Vaillant'sfour-tiered four-tieredmature-immature/ mature-immature/ defense-fragmentation trio closely parallel and Fienberg (1973) also proneurotic-psychotic defenses. Semrad, Grinspoon, and pro classification system of ontogenetic maturity. Similarly, the 29 defenses posed a classification

22 CHAPTER 22

PSYCHOLOGICALDEFENSE DEFENSE PSYCHOLOGICAL

567 567

outlined by Horowitz (1988), the 28 described by J. C. Perry and Cooper (1989), and the 12 of Hauser (Jacobson et al., aI., 1992) can be ordered along this mature-immature mature-immature continuum. By contrast, Ihilevich and Gieser's Gleser's OMI DMI mechanisms all fall at the same level of of the hierarchy, namely, the neurotic/immature neurotic/immature level. One can also order the defensive processes in terms of of the degree of conscious consciousness involved. The mature (Vaillant) or coping (Haan, Plutchik) processes (e.g., sublimation, suppression, humor) and some of the higher level neurotic defenses lower(e.g., intellectualization, isolation) are assumed to be more conscious than the lower level psychotic or fragmented mechanisms (e.g., delusional projection). As with Haan's coping processes, those tapped by the Ways of of Coping scale are held to be conscious. Thus, for example, Haan's or Vaillant's suppression suppression resembles the WOC's self-control. However, when they become automatized and no longer require atten attentional resources, they lose status as coping processes (Lazarus & & Folkman, 1984, p. 131). Plutchik and Conte (1989) are the most explicit in explaining how, as a defense becomes more conscious, it develops into a parallel coping process that is far more adaptive. traditional requirement Note that some recent theorists have challenged the traditional that defenses be fully unconscious (A. Freud, 1936). They emphasize instead the defenses with coping (Erdelyi, 1990; Plutchik & & Keller Kellerflexible interplay of all defenses man, 1980). of these dimensions-mature-immature, dimensions—mature-immature, conscious consciousIt is instructive that three of primitive-complex—are assumed to be closely connected: That unconscious, and primitive-complex-are is, to the extent that defense is conscious and complex, it tends to be viewed as mature. Thus a central theme runs through these ostensibly different different theoretical implicaorderings. Unfortunately, this theme is burdened with evaluative and moral implica tions. Moreover, despite accumulating evidence to the contrary, the hierarchy is often assumed to correspond to increasing adaptiveness. often After After conducting this review, we cannot accept the claim for for a single dimension of adaptiveness for defenses. There are too many reasonable yet incommensurate criteria for adaptiveness: short-term distress, long-term distress, task performance, reproductive success, social adjustment, and so forth. In our view, the adaptiveness of defenses can be evaluated only locally-that locally—that is, only after after specifying specifying a precise criterion as well as a precise point in time.6 time.^

XVin. CONCLUSION xvm. Apropos the topic of psychological defense, this chapter required the balancing of two conflicting goals. We hoped to demonstrate the diversity of of current theories and operationalizations of psychological psychological defense. At the same time, we hoped to

' Kruglanski Kruglanski (1989) (1989) makes makes aa similar similar point point about about evaluating evaluating accuracy accuracy in in general. general.

6

S68 568

PAULHUS, PAULHUS, FRmHANDLER, FRIDHANDLER,AND ANDHAYES HAYES

integrate integrate the the literature. literature. We We suspect suspect that that we we have have been been more more successful successful at at the the former former goal goal than than the the latter.7 latterJ There There are are already already aa number number of of useful useful integrative integrative schemes schemes currently currently available available (e.g., 1995; Vaillant, (e.g., S. S. H. H. Cooper, Cooper, 1989; 1989; Horowitz Horowitz et et aaI., l , 1990; 1990; Conte Conte & Plutchik, Plutchik, 1995; Vaillant, 1992). literature reviewed 1992). None of these taxonomies, however, can subsume all the literature terminology for for various here until there is more consensus on the terminology various defenses. Even fatally incommensurate incommensurate with others. some theoretical models remain fatally Nonetheless, psychological defense defense must Nonetheless, all all psychologists psychologists interested interested in in psychological must agree agree that that the the current current lack lack of of consensus consensus is is aa far far cry cry better better than than the the peremptory peremptory dismissal dismissal of of the the very very notion notion of of defense defense heard heard only only aa few few years years ago. ago.

ACKNOWLEDGMENTS Work on this chapter was supported by the Social Science and Humanities Research Council of Canada as well as the John D. and Catherine T. MacArthur Foundation.

REFERENCES (1994). Diagnostic Diagnostic and and Statistical StatisticalManual: Manual:(4th (4thEdition). Edition). American Psychiatric Association (1994). Washington, DC: American American Psychiatric Association. affairs, Seattle: University of Washington Alexander, R. D. (1979). Darwinism Darwinism and human affairs. Press. (1969). Cognitive dissonance: A current current perspective. In L. Berkowitz (Ed.), Aronson, E. (1969). Advances in experimental Vol. 4, pp. 1-34). New York: Aca experimental social psychology ((Vol. Academic Press. The discrepant discrepant repressor: repressor: Differentiation Differentiation between between Asendorpf, J. B., & Scherer, K. R. (1983). The autonomic-facial-verbal patterns low anxiety, high anxiety, and repression of anxiety by autonomic-facial-verbal of behavior. Journal of of Personality Personalityand andSocial SocialPsychology, Psychology,45, 45,1334-1346. 1334-1346. Assor, A., A, Aronoff, J., & Messe, L. A (1986). An experimental A. (1986). experimental test of of defensive processes in impression formation. Journal of of Personality Personality and andSocial SocialPsychology, Psychology,50, 50,644-650. 644-650. Balay, J., & Shevrin, H. (1988). (1988). The subliminal psychodynamic activation method: A critical Psychologist,43, 43,161-174. 161-174. review. American Psychologist, A. (1984). Automatic and conscious processing. In R. S. Wyer Wyer & T. K. Srull (Eds.), Bargh, J. A Handbook Handbook of of social cognition cognition (Vol. 3, pp. 1-43). Hillsdale, NJ: Erlbaum. Baumester, R. F. (1993). Escape Escape from from self. self New York: Guilford Guilford Press. Baumeister, R. R. F., P., Tice, Tice, D. M., & & Hutton, Hutton, D. D. G. G. (1989). Self-presentational Self-presentational motivations motivations and and Baumeister, personality differences differences in self-esteem. Journal of of Personality, Personality,57, 57,547-579. 547-579. ^Holmes Holmes (1974, (1974,1981; Holmes&&McCaul, McCaul,1989) 1989)deserves deservesparticular particularmention mentionasasaacritic criticofofthe the 7 1981; Holmes evidence for for defense defense mechanisms. mechanisms. He He continues continues to to argue argue that that no no convincing evidence for for the the existence existence evidence convincing evidence of of these these defenses defenses has has ever ever been been produced. produced. His His reviews reviews appear appear to to have have been been highly highly influential; influential; as as he notes himself himself (Holmes (Holmes&&McCaul, McCaul,1989), 1989),the the volume laboratory work repression declined drastically volume ofof laboratory work onon repression declined drastically notes after 1974 review. 1992; Erdelyi, have after his his 1974 review. Equally Equally careful careful reviewers reviewers (S. (S. H. H. Cooper, Cooper, 1992; Erdelyi, 1985), 1985), however, however, have drawn drawn much much more more favorable favorable conclusions conclusions from from the the same same literature. literature.

CHAPTER CHAPTER 22 22

PSYCHOLOGICAL PSYCHOLOGICALDEFENSE DEFENSE

569 569

denial of of death. death.New New York: York: Free Free Press. Press. Becker, E. (1973). The denial A., & & Byrne, D. D. (1978). (1978). Repression-sensitization. Repression-sensitization. In In H. H.London London & &J.J.E. E.Exner Exner (Eds.), (Eds.), Bell, P. A., Dimensions of of personality personality (pp. 449-485). New York: Wiley. Bloch, A. L., Shear, M. K., Markowitz, J. C c.,, Leon, A. C c.,, & & Perry, J. C. (1993). An defense mechanisms in dysthymia. American Journal of of Psychiatry, empirical study of defense Psychiatry, 150, 1194-1198. 1194-1198. 150, computer model modelfor forunconscious unconsciousinteractions interactions offect affect potentials Blum, G. S. (1986). A computer of af potentials and and cognitiveprocesses processes(Tech. (Tech.Rep. Rep.No. No.8611). 8611).Santa SantaBarbara: Barbara:University UniversityofofCalifornia. California. cognitive Blum, G. S., & Barbour, J. S. (1979). Selective inattention inattention to anxiety-linked stimuli. Journal Journal of 108,108, 182-224. of Experimental ExperimentalPsychology: Psychology:General, General, 182-224. Bonanno, G. A., Keltner, D., Holen, A., & & Horowitz, Horowitz, M. M. J.J.(1995). (1995).When Whenavoiding avoidingunpleasant unpleasant emotions might not be such a bad thing: Verbal-autonomic response dissociation and bercdivemQntJournal Journal Personality
570 570

PAULHUS, PAULHUS, FRlDHANDLER, FRIDHANDLER,AND ANDHAYES HAYES

Chandler, M. M. J., mystification ofofpsycho logi J., Paget, Paget, K. K. F., F., & &Koch, Koch,D. D.A. A.(1978). (1978).The Thechild's child'sde demystification psychological defense mechanisms: mechanisms: A A structural structural and and developmental developmental analysis. analysis. Developmental DevelopmentalPsy Psychology, chology, 14, H 197-205. Cline, P. (1987). The scientific British Journal JournalofofMedical MedicalPsychology, Psychology, scientific status of the DMT. British 60, 53-59. 60, Cohen, A. A. R R. (1964). Attitude change changeand andsocial socialinfluence. influence.New New York: Basic Books. Cohen, York: Basic Books. Colvin, C. C. R., & Block, Block, J.J. (1994). (1994). Do Do positive positiveillusions illusions foster foster mental mental health? health? An An examination examination R, & Colvin, PsychologicalBulletin, Bulletin,116, 116, 3-20. of the Taylor and Brown formulation. Psychological 3-20. New Conte, H. R, R., & & Plutchik, R R. (Eds.) (1995). Ego defenses: defenses: Theory Theoryand andmeasurement. measurement New York: Wiley. Cooper, C. (1988). Predicting susceptibility to short-term stress with the Defence Mechanism Test. Work and Stress, Stress,2,2,49-58. 49-58. mechanisms: A A compara comparaCooper, S. S. H. H. (1989). Recent Recent contributions contributions to to the the theory theory of of defense defense mechanisms: Cooper, tive view. view. Journal Journalof ofthe theAmerican AmericanPsychoanalytic Psychoanalytic Association, 865-891. tive Association, 37, 37, 865-891. Cooper, S. S. H. H. (1992). The The empirical empirical study study of of defensive defensive processes: processes: A A review. review. In In J. J. W. W. Barron, Barron, Cooper, M. N. N. Eagle, Eagle, & & D. D. L. L. Wolitzky Wolitzky (Eds.), (Eds.), Interf Interface psychoanalysis and psychology (pp. ace ofofpsychoanalysis and psychology (pp. M. Washington, DC: DC: APA APA Press. 327-346). Washington, Cramer, P. (1983). Children's use of defense mechanisms in reaction to displeasure caused by others. Journal Journalof of Personality, Personality,51,51,79-94. 79-94. by Cramer, P. P. (1988). The The Defense Defense Mechanism Mechanism Inventory: Inventory: A A review review of of research research and and discussion discussion Cramer, Journalof of Personality PersonalityAssessment, Assessment, 142-164. of the the scales. scales. Journal of 52,52, 142-164. Cramer, P. P. (1991). (1991). The Thedevelopment developmentofofdefense defensemechanisms: mechanisms: Theory, research, and assessment Cramer, Theory, research, and assessment. York: Springer-Verlag. New York: Crowne, D. P. P. (1979). The The experimental experimentalstudy studyofofpersonality. personality. Hillsdale, Erlbaum. Hillsdale, NJ:NJ: Erlbaum. Crowne, Crowne, D. P., & & Marlowe, D. A. (1964). The The approval approvalmotive. motive.New New York: York: Wiley. Wiley. Davidson, K. K. (1993). Suppression Suppression and and repression repression in in discrepant discrepant self-other self-other ratings: ratings: Relations Relations Journalof ofPersonality, Personality,61,61, 669-691. with thought thought control control and and cardiovascular cardiovascular reactivity. reactivity. Journal 669-691. with & MacGregor, M. W. (1996). Reliability of an idiographic Q-sort measure of of Davidson, K., & 66,66, 624-639. defense mechanisms. Journal Journal of ofPersonality PersonalityAssessment, Assessment, 624-639. Davis, P. J. (1987). (1987). Repression and the inaccessibility of affective memories. Journal Journal of of Personality 585-593. Personalityand andSocial SocialPsychology, Psychology,53,53, 585-593. Davis, P. J. (1990). Repression and the inaccessibility of affective memories. In J. L. Singer forfor personality theory, psychopathol (Ed.), Repression Repressionand anddissociation: dissociation:Implications Implications personality theory, psychopathology, and health health (pp. (pp.387-404). 387-404).Chicago: Chicago:University Universityof ofChicago ChicagoPress. Press. Davis, P. J., & & Schwartz, Schwartz, G. G. E. E. (1987). (1987). Repression Repression and and the the inaccessibility inaccessibility of of affective affective memo memoPsychology, 52,52, 155-162. ries. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 155-162. DeLongis, A., Coyne, J. C, c., Dakof, G., Folkman, S., & Lazarus, R R. S. (1982). Relationship of daily hassles, uplifts, and major life events to health status. Health Health Psychology, Psychology, 1, 119-136. 1, DeLongis, A., & & O'Brien, T. (1990). An interpersonal framework for stress and coping: An application to the families Crowther, families of of Alzheimer's Alzheimer'spatients. patients.In InM. M.A. A.P.P.Stephens, Stephens,J.J.H.H. Crowther, S. E. Hobfoll, & D. L. Tennenbaum (Eds.), Stress Stress and andcoping copingininlater-life later-lifefamilies. families. New York: Hemisphere Publishing. York: Wiley. Dixon, N. (1981). Preconscious Preconsciousprocessing. processing.New New York: Wiley. Dorpat, T. L. (1985). The cognitive arrest hypothesis of denial. International InternationalJournal Journalofof Psychoanalysis, Psychoanalysis,64, 64,47-58. 47-58. Eagle, M. (1988). Psychoanalysis and self-deception. In J. S. Lockard & & D. L. Paulhus (pp. 78-98). Englewood Cliffs, NJ:NJ: (Eds.), Self-deception: Self-deception:An Anadaptive adaptivemechanism? mechanism? (pp. 78-98). Englewood Cliffs, Prentice-Hall.

22 CHAPTER 22

PSYCHOLOGICALDEFENSE DEFENSE PSYCHOLOGICAL

571 571

& Clarke, S. (1970). Heart-rate and skin conductance during experimentally experimentallyEpstein, S., & Effects of anticipated intensity of noxious stimulation and experience. induced anxiety: Effects Journal 84,84, 105-112. Journalof ofExperimental ExperimentalPsychology, Psychology, 105-112. Epstein, S., & Meier, P. (1989). (1989). Constructive thinking: A broad coping variable with specific 57,57, 332-350. components. Journal Journalof of Personality Personalityand andSocial SocialPsychology, Psychology, 332-350. Erdelyi, M. H. (1974). A new look at the new look: Perceptual defense defense and vigilance. Psychological PsychologicalReview, Review,81,81,1-25. 1-25. Erdelyi, M. cognitive psychology. New York: Freeman. M. H. (1985). Psychoanalysis: Psychoanalysis:Freud's Freud's cognitive psychology. New York: Freeman. of Erdelyi, M. H. (1990). (1990). Repression, reconstruction, and defense: History and integration of psychoanalytic and experimental frameworks. In J. L. Singer (Ed.), Repression Repression and and dissociation: andand personality styles: Current theory and research dissociation:Defense Defensemechanisms mechanisms personality styles: Current theory and research (pp. 1-31). Chicago: University of Chicago Press. Erdelyi, M. H., & Goldberg, G. (1979). Let's not sweep repression repression under the rug: Toward repression. In J. J. F. F. Kihlstrom & & F. J. Evans (Eds.), Functional Functional a cognitive psychology of repression. disordersof of memory memory (pp. (pp.355-402). 355-402).Hillsdale, Hillsdale, NJ: NJ:Erlbaum. Erlbaum. disorders & Kleinbard, J. (1978). Has Ebbinghaus decayed with time? The growth of Erdelyi, M. H., & of recall (hypermnesia) (hypermnesia) over over days. days. Journal Journalof ofExperimental Experimental Psychology: Human Learning recall Psychology: Human Learning Memory, 4,4,275-289. 275-289. and Memory, W., & & Pierce, Pierce,J.J. (1968). (1968).Defense Defensemechanisms. mechanisms.In InE. E.F. F.Borgatta Borgatta&&W. W.W. W.Lambert Lambert Eriksen, C. W., (Eds.), Handbook of personality personality theory theory and research research (pp. (pp. 1007-1040). Chicago: Chicago: (Eds.), Rand McNally. & Custrini, Custrini, R. R. J. J. (1988). (1988). Learning Learningto tolie lieand and self-deceive: self-deceive: Children'S Children'snonverbal nonverbal Feldman, R. S., & & D. L. Paulhus (Eds.), Self-deception: Self-deception: communication of deception. In J. S. Lockard & adaptive mechanism? mechanism?(pp. (pp.40-53). 40-53).Englewood EnglewoodCliffs, Cliffs,NJ: NJ:Prentice-Hall. Prentice-Hall. An adaptive & Popoola, L. (1979). Detection of deception in adults and Feldman, R. S., Jenkins, L., & Development, 50, 50,350-355. 350-355. children via facial expressions. Child Development, Self-deception.New NewYork: York:Humanities HumanitiesPress. Press. Fingarette, H. (1969). Self-deception. Fisher, S., & & Greenberg, Greenberg, R. R. (1977). The The scientific scientificcredibility credibility Freud's theories therapy. Fisher, ofof Freud's theories and and therapy. New York: Basic Books. Fiske, S. T., & & Taylor, S. E. (1991). (1991). Social Social cognition. cognition.New NewYork: York:McGraw-Hill. McGraw-Hill. & Lazarus, R. S. (1980). (1980). An analysis of coping in a middle-aged community Folkman, S., & 21,21, 219-239. sample. Journal Journalof ofHealth Healthand andSocial SocialBehavior, Behavior, 219-239. Folkman, S., & Lazarus, R. S. (1985). If it changes, it must be a process: Study of emotion and coping during three stages of a college examination. Journal of of Personality Personalityand and Social Social Psychology, Psychology,48, 48,150-170. 150-170. Folkman, S., Lazarus, R. S., Dunkel-Shetter, C, C., DeLongis, A., & Gruen, R. J. (1986). Dynamics of a stressful stressful encounter: Cognitive appraisal, coping, and encounter out outcomes. Journal Psychology, 50,50, 992-1003. Journalof ofPersonality Personalityand andSocial Social Psychology, 992-1003. Frenkel-Brunswik, E. (1939). Mechanisms of self-deception. Journal of Social Psychology, Psychology, of Social 10, 10, 409-420. Freud, A. (1936). The ego International egoand andthe themechanisms mechanismsofofdefense. defense.New NewYork: York: InternationalUniversi Universities Press. Geisler, C. (1986). The The use use of of subliminal subliminal psychodynamic psychodynamic activation activation in the study study of of repression. repression. Journal Psychology, 51, 51, 844-851. Journalof ofPersonality Personalityand andSocial Social Psychology, 844-851. Gergen, K. J. (1985). f (1985). The ethnopsychology of of self-deception. In M. W. W. Martin (Ed.), Sel Selfdeception deceptionand andself-understanding self-understanding(pp. (pp.228-243). 228-243).Lawrence: Lawrence:University UniversityPress PressofofKansas. Kansas. Gerin, W., Litt, M. D., Deich, J., & & Pickering, T. G. (1995). Self-efficacy Self-efficacy as a moderator of perceived control effects effects on cardiovascular reactivity: Is enhanced control always beneficial? Psychosomatic 390-397. PsychosomaticMedicine, Medicine,57,57, 390-397.

572 572

PAULHUS, PAULHUS, FRJDHANDLER, FRIDHANDLER,AND ANDHAYES HAYES

Gero, G. (1951). The concept of defense. Psychoanalytic Psychoanalytic Quarterly, Quarterly,20,20,565-578. 565-578. Gilbert, D. T., & Cooper, J. (1985). (1985). Social psychological strategies of of self-deception. In Self-deception and self-understanding self-understanding (pp. (pp. 75-94). 75-94). Lawrence: Lawrence: M. W. Martin (Ed.), Self-deception University Press of Kansas. Gieser, Gleser, G. C, C, & Ihilevich, D. (1969). (1969). An objective instrument for for measuring defense defense mecha mechaPsychology, 33, 33, 51-60. nisms. Journal Journalof of Consulting Consultingand andClinical Clinical Psychology, 51-60. Goldberger, L. (1983). (1983). The concept and mechanisms of denial: As selective overview. In denialof ofstress stress(pp. (pp.83-101). 83-101).New NewYork: York:International InternationalUniversi UniversiS. Breznitz (Ed.), The denial S. ties Press. psychology ofof self-deception. New York: Vital lies, lies, simple simpletruths: truths:The The psychology self-deception. New York: Goleman, D. (1986). Vital Simon & Schuster. R, & Reineck, L. (1967). Factor analytic study of the Golin, S., Herron, E. W., Lakota, R., Con Manifest Anxiety, Extraversion, and Repression-Sensitization Repression-Sensitization scales. Journal of of ConPsychology, 31, 31,564-569. 564-569. sulting Psychology, Greenberg, J., Pyszczynski, T., & Solomon, S. (1986). The The cause cause and and consequences consequences of of a need need for self-esteem: A terror management theory. In R. R F. Baumeister (Ed.), Public self for and private self self (pp. 189-212). New York: Springer-Verlag. The authoritarian authoritarian ego: ego: Fabrication Fabrication and and revision revision of ofpersonal personal history. history. Greenwald, A. G. (1980). The American American Psychologist, Psychologist,33, 33,603-618. 603-618. Greenwald, A. G. (1988). (1988). Self-knowledge and self-deception. In J. S. Lockard & & D. L. Paulhus (Eds.), Self-deception: Self-deception:An An adaptive adaptivemechanism? mechanism?(pp. (pp. 113-131). 113-131).Englewood Englewood Cliffs, NJ: Prentice-Hall. Grzegolowska, H. J. (1976). An activation theory approach to defense mechanisms. Polish Psychological PsychologicalBulletin, Bulletin,7, 7,225-233. 225-233. Gur, R R. C C,, & Sackeim, H. A. (1979). Self-deception: A concept concept in search of a phenomenon. Journal 37,37, 147-169. Journalof of Personality Personalityand andSocial SocialPsychology, Psychology, 147-169. Gurwitz, P. M. (1981). Paths to psychotherapy in the middle years: A longitudinal study. Social 67-76. Social Science Scienceand andMedicine, Medicine,15E, 15E, 61-16, Haan, N. (1963). Proposed model of ego functioning: Coping and defense defense mechanisms in relationship to 1.0. PsychologicalMonographs, Monographs,77, 77,1-23. 1-23. I.Q. change. Psychological Haan, N. (1965). Coping and defense mechanisms related to personality inventories. Journal of of Consulting ConsultingPsychology, Psychology,29,29,373-378. 373-378. Haan, N. (1969). A tripartite model of ego functioning functioning values and clinical and research 148, 14-30. applications. Journal Journal of ofNervous Nervousand andMental MentalDisease, Disease, 148, 14-30. defending. New NewYork: York:Academic Academic Press. Press. Haan, N. (1977). (1977). Coping and defending. Processes of of moral moral development: development: Cognitive Cognitive or social disequilibrium? disequilibrium? Devel DevelHaan, N. (1985). Processes opmental Psychology, Psychology,21, 21,996-1006. 996-1006. defined in two Haan, N. (1986). Systematic variability in the quality of moral action, as defined Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, 1271-1284. 50,50, 1271-1284. formulations. Journal V. (1983). (1983). The Thecognitive cognitivestructures structures and processes of human motivation personalofhuman motivation and and personal Hamilton, V. and processes ity. New York: Wiley. ity. R. A. (1990). Subliminally activated symbiotic fantasies: Facts and artifacts. Hardaway, R PsychologicalBulletin, Bulletin,107, 107, 177-195. Psychological 177-195. R. D. (1966). Denial Denial of of threat threat and and emotional emotional response response to impending painful painful stimulation. stimulation. Hare, R Journalof of Consulting ConsultingPsychology, Psychology, 359-361. Journal 30,30, 359-361. Hart, D., & Chmiel, S. S. (1992). Influence of defense mechanisms on moral judgment develop development—a longitudinal study. Developmental Psychology, Psychology, 28, 28,722-730. 722-730. ment-a

22 CHAPTER 22

PSYCHOLOGICALDEFENSE DEFENSE PSYCHOWGICAL

573 573

Heilbrun, A. B., Jr., Jr., & & Pepe, Pepe, V. V. (1985). (1985). Awareness Awareness of of cognitive cognitive defences defences and and stress stress manage management. British 58,58, 9-17. BritishJournal JournalofofMedical MedicalPsychology, Psychology, 9-17. perforHentschel, U., & Kiessling, M. (1990). Are defense mechanisms valid predictors of perfor S. E. Hampson, J. Reykowski, & mance on cognitive tasks? In G. I. van Heck, S. J. Zakrzewski (Eds.), Personality Personalitypsychology psychologyininEurope Europe(Vol. (Vol.3,3,pp. pp.203-223). 203-223).Lisse, Lisse, Amsterdam: Swets & Zeitlinger. Hentschel, U., Smith, G., & Draguns, J. G. (Eds.). (1986). The roots roots of of perception: perception:Individual Individual differences awareness.New New York: York: differences in information information processing processingwithin withinand and beyond beyond awareness. North-Holland. Holmes, D. S. (1974). Investigations of repression: Differential Differential recall of material experimen experimentally or naturally associated associated with ego threat. Psychological PsychologicalBulletin, Bulletin,81, 81,632-653. 632-653. Holmes, D. S. (1981). Existence of classical projection and the str�ss-reducing of stress-reducing function function of PsychologicalBulletin, Bulletin,90, 90,460-466. 460-466. attributive projection: A reply to Sherwood. Psychological & McCaul, K. D. (1989). Laboratory research on defense mechanisms. In Holmes, D. S., & investigationof ofpsychological psychologicalstress stress(pp. (pp. R. W. W. J. J. Neufeld Neufeld (Ed.), (Ed.), Advances Advances in in the the investigation R 161-192). New York: Wiley. Stress response responsesyndromes syndromes(2nd (2nded.). ed.).Northvale, Northvale,NJ: NJ:Jason JasonAronson. Aronson. Horowitz, M. J. (1986). Stress psychodynamics. New York: Basic Books. Introduction to psychodynamics. Horowitz, M. J. (1988). Introduction C, Stinson, c., C, Fridhandler, B. M., & & Ghannam, J. H. (1990). Horowitz, M. J., Markman, H. c., classification theory of defense. In J. L. Singer (Ed.), Repression and and dissociation: dissociation: A classification (Ed.), Repression Implicationsfor forpersonality, personality,psychopathology psychopathology and health 61-84). Chicago: UniverImplications and health (pp.(pp. 61-84). Chicago: Univer sity of Chicago Press. Horowitz, M. J., J., Milbrath, Milbrath, c., C, Jordan, Jordan, D., D., Stinson, Stinson, c., C, Redington, Redington, D., D., Ewart, Ewart, M., M.,Fridhandler, Fridhandler, B., Reidbord, S., & Hartley, D. (1994). Expressive and defensive defensive behavior during of Personality, Personality,62, 62,527-563. 527-563. discourse on unresolved topics. Journal of Horowitz, M. J., & Stinson, C. H. (1995). Defense as aspects of person schemas and control R. Plutchik & H. Conte (Eds.), Ego defenses: defenses: Theory Theoryand andmeasurement measurement processes. In R (pp. 79-97). New York: Wiley. & Gieser, Gleser,G. G. C. C.(1986). (1986).Defense Defensemechanisms: mechanisms: Their classification, correlates, Ihilevich, D., & Their classification, correlates, and and measurement withthe theDefense Defense Mechanisms Inventory, Owosso, Associates. Owosso, MI: MI: DMIDMI Associates. measurement with Mechanisms Inventory. S. T., T.,Noam, Noam, G., G.,Powers, Powers,S. S.I., I.,& &Gelfand, Gelfand, E. E. (1992). (1992). Jacobson, A. M., Beardslee, W., Hauser, S. Ego defense defense mechanisms manual. In G. E. Vaillant (Ed.), Ego mechanisms mechanismsof of defense: defense: A for clinicians clinicians and researchers researchers (pp. A guide for (pp.261-278). 261-278).Washington, Washington, DC: DC:American American Psychiatric Press. Jamner, L. D., & Schwartz, G. E. (1987). (1987). Self-deception Self-deception predicts self-report self-report and endurance 21 1211-223. -223. of pain. Psychosomatic PsychosomaticMedicine, Medicine,48,48, Joffe, Joffe, P., & & Naditch, Naditch, M. M.P. P.(1977). (1977).Paper Paperand andpencil pencilmeasures measuresof ofcoping copingand anddefense defenseprocesses. processes. ofof self-environment organization In N. Haan (Ed.), Coping Coping and anddefending: defending:Processes Processes self-environment organization (pp. 280-297). New York: Academic Press. style: State or trait? Journal of Personality Juni, S., & Yanishefsky, D. S. (1983). Defensive style: of Personality Assessment, Assessment,47, 47,536-538. 536-538. Kemberg, O. F. (1976). Object New York: Object relations relations theory theoryand andclinical clinicalpsychoanalysis. psychoanalysis. New York: Jason Aronson. Kemberg, O. disorders: Psychotherapeutic strategies. New New Haven, O. F. F.(1984). (1984).Severe Severepersonality personality disorders: Psychotherapeutic strategies. Haven, CT: CT: Yale University Press. Kline, P. (1972). Fact theory. London: Methuen. Factand andfantasy fantasyininFreudian Freudian theory. London: Methuen. Kobasa, S. C., C, Maddi, S. R, R., & Kahn, S. (1982). (1982). Hardiness and health: A prospective study. Journal Psychology, 42, 42, 168-177. Journalof ofPersonality Personalityand andSocial Social Psychology, 168-177.

574 574

PAULHUS, PAULHUS, FRlDHANDLER, FRIDHANDLER,AND ANDHAYES HAYES

for diagnosis and personnel Kragh, U. (1960). The Defense Mechanism Test: A new method for Psychology, 44, 44,303-309. 303-309. selection. Journal of of Applied Psychology, Kragh, U., & Smith, G. J. W. (Eds.). (1970). Percept-genetic Percept-geneticanalysis. analysis,Lund: Lund:Gleerups. Gleerups. Krebs, D., Denton, K., K, & Higgins, N. C. (1988). On the evolution of self-knowledge. In K K. MacDonald (Ed.), Sociobiological perspectiveson onhuman humandevelopment development(pp. (pp.105-139). 105-139). Sociobiologicalperspectives New York: Springer-Verlag. Kroeber, T. C. (1963). The coping functions of the ego mechanisms. In R. R White (Ed.), The study of of lives (pp. 179-198). New York: Atherton. Kruglanski, A A. W. (1989). The psychology of being "right": The problem of accuracy in social perception and cognition. Psychological PsychologicalBulletin, Bulletin,106, 106,395-405. 395-405. Kubler-Ross, E. (1969). On death death and dying. dying. London: Macmillan. Kupst, M. J., & Schulman, J. L. (1981). (1981). The CPI subscales as predictors of parental coping with childhood leukemia. Journal of of Clinical ClinicalPsychology, Psychology,37, 37,386-388. 386-388. Kupst, M. J., Schulman, J. L., Maurer, H., Honig, G., Morgan, E., & Fochtman, D. (1984). A two-year two-year follow-up. follow-up. Journal Journalof ofPediatric PediatricPsychology, Psychology, Coping with pediatric leukemia: A 9, 149-163. 9, R. S. (1986). The costs and benefits of denial. In S. Breznitz (Ed.), Denial of of stress Lazarus, R stress (pp. 1-30). New York: International Universities Press. Lazarus, R New York: Springer. Stress, appraisal, appraisal,and andcoping. coping. New York: Springer. R. S., & Folkman, S. S. (1984). (1984). Stress, Lifton, R. (1968). life: Survivors Survivorsof of Hiroshima. Hiroshima.New NewYork: York:Random RandomHouse. House. (1968). Death in life: NewsS. (1978). On the adaptive significance of self-deception. Human Ethology NewsLockard, J. S. letter, 21, 21,4-7. 4-7. letter, (1988). Self-deception: Self-deception:An An adaptive adaptivemechanism? mechanism? Lockard, J. S., & Paulhus, D. L. (Eds.). (1988). Englewood Cliffs, NJ: Prentice-Hall. The reality reality of of repressed repressed memories. memories. American AmericanPsychologist, Psychologist,48, 48,518-537. 518-537. Loftus, E. F. (1993). The A. J. (1983). Conscious and unconscious perception: Experiments on visual masking Marcel, A and word recognition. Cognitive Cognitive Psychology, Psychology,15, 15,197-237. 197-237. Martin, M. W. (Ed). (1985). Self-deception Self-deception and self-understanding. self-understanding.Lawrence: Lawrence:University University Press of Kansas. McCrae, R. R R. (1984). Situational determinants of coping responses: Loss, threat, and chal challenge. Journal of of Personality Personality and andSocial SocialPsychology, Psychology,46, 46,919-928. 919-928. McCrae, R. R, R., & Costa, P. T. (1986). Personality, coping, and coping effectiveness effectiveness in an adult sample. sample. Journal of of Personality, Personality,54, 54,385-405. 385-405. McCullough, McCuUough, L. (1992). Toward Toward reliability in identifying identifying ego defenses: Clinical techniques. and research In G. E. Vaillant (Ed.), Ego mechanisms mechanismsof of defense: defense:AAguide guidefor forclinicians clinicians and researchDC: American Psychiatric Press. ers ers (pp. 171-179). Washington DC: Miller, S. S. M. (1989). Cognitive informational informational styles in the process of coping with threat and frustration. Advances in Behavioral Behavioral Research Researchand andTherapy, Therapy,11,11,223-234. 223-234. Mischel, W. (1979). On the interface of cognition and personality: Beyond the person personsituation debate. American Psychologist, Psychologist,34, 34,740-754. 740-754. Monts, J. K, A, & Nydegger, R K., Zurcher, L. A., R. V. (1977). Interpersonal self-deception self-deception and personality correlates. Journal of of Social Psychology, Psychology,103, 103,91-99. 91-99. Moore, B. B. E., & Rubinfine, D. L. (1969). The mechanism of of denial (Monograph Series). New York: York: International International Universities UniversitiesPress, Press,Kris KrisStudy StudyGroup Groupof ofthe theNew NewYork YorkPsycho Psychoanalytic Institute. Moos, R R. H. (1974). (1974). Psychological techniques in the assessment of adaptive behavior. In G. V. Coelho, D. A A. Hamburg, & J. E. Adams (Eds.), Coping and adaptation adaptation (pp. 101-111). New York: Basic Books.

22 CHAPTER 22

DEFENSE PSYCHOLOGICAL DEFENSE

575 575

Morrissey, R. F. (1977). The Haan model of ego functioning: An assessment of of empirical self-environment organ i research. In N. Haan (Ed.), Coping Coping and defending: defending: Processes Processes of of self-environment organization (pp. 250-279). New York: Academic Press. zation Murphy, G. (1975). Outgrowing Outgrowing self-deception. self-deception. New York: Basic Books. & Contrada, R. J. (1992). Repressive coping and the verbal-autonomic Newton, T. L., & Journal of of Personality Personality and Social response dissociation: The influence of social context. Journal Psychology, Psychology, 62, 62, 159-167. Nielsen, S. L., & Sarason, l. I. G. (1981). Emotion, personality and selective attention. Journal Journal of of Personality Personality and and Social Social Psychology, Psychology, 41, 41, 945-960. Human inference: Strategies Strategies and and shortcomings of social social judg judgNisbett, R., & Ross, L. (1979). Human shortcomings of ment. Englewood Cliffs, NJ: Prentice-Hall. Norem, J. K., & Cantor, N. (1986). Defensive pessimism: Harnessing anxiety as a motivation. Journal 1, 1208-1217. Journal of of Personality Personality and Social Psychology, Psychology, 5 51, Ofshe, Making monsters: monsters: False memories, memories, psychotherapy, psychotherapy, and and sexual Of she, R., & Watters, E. (1994). Making hysteria. New York: Charles Scribner's Sons. Two-component models of of socially socially desirable responding. Journal Journal of of Paulhus, D. L. (1984). Two-component Personality and Social Social Psychology, Psychology, 46, 598-609. Personality Self-deception and impression managment in test responses. In Paulhus, D. L. (1986). Self-deception A. Angleitner & J. S. Wiggins (Eds.), Personality Personality assessment assessment via questionnaire questionnaire (pp. 143-165). New York: Springer-Verlag. Paulhus, D. L. (1988). Self-deception: Where do we stand? In J. S. Lockard & D. L. Paulhus Self-deception: An An adaptive adaptive mechanism? mechanism? (pp. 251-255). Englewood Englewood Cliffs, Cliffs, NJ: (Eds.), Self-deception: Prentice-Hall. Paulhus, D. L. (1991). Measurement and control of response bias. In J. P. Robinson, P. R. of personality personality and and social psychological social psychological Shaver, & L. S. Wrightsman (Eds.), Measures of attitudes (pp. 17-59). San Diego, CA: Academic Press. attitudes Paulhus, D. L. (1993). Bypassing the will: The automatization of of affirmations. In D. M. Handbook of of mental mental control control (pp. 573-587). Wegner & J. W. Pennebaker (Eds.), Handbook Englewood Cliffs, CUffs, NJ: Prentice-Hall. Englewood Attentional load increases the positivity Paulhus, D. L., Graf, P., & Van Selst, M. (1989). Attentional Social Cognition, 7, 7, 389-400. of self-presentation. Social Paulhus, D. L., & Levitt, K. (1987). Desirable responding triggered triggered by affect: Automatic Automatic egotism? Journal Journal of of Personality Personality and Social Social Psychology, Psychology, 52, 245-259. Paulhus, D. L., & Reid, D. (1991). Attribution and denial in socially desirable responding. Journal Journal of of Personality Personality and Social Social Psychology, Psychology, 60, 60, 307-317. Paulhus, D. L., & Reynolds, S. (1995). Enhancing validity and agreement in personality impression: Putting the person back into person perception. Journal Journal of of Personality Personality and and Social Social Psychology, Psychology, 69, 69, 1233-1242. Paulhus, D. L., & Suedfeld, P. (1988). A dynamic complexity model of of self-deception. In adaptive mechanism? mechanism? (pp. J. S. Lockard Lx)ckard & D. L. Paulhus (Eds.), Self-deception: Self-deception: An An adaptive 132-145). Englewood Cliffs, NJ: Prentice-Hall. Peariin, L. I., & Schooler, C. (1978). The structure of of coping. Journal of Health and and Social Pearlin, Journal of Behavior, 19, 2-21. Behavior, Perry, c., C , & Laurence, J. R. R. (1984). (1984). Mental Mental processing processing outside outside of awareness: awareness: The The contributions contributions of Freud and Janet. In K. S. Bowers & D. Meichenbaum (Eds.), The unconscious unconscious of reconsidered reconsidered (pp. 9-48). New York: Wiley. & Cooper, S. H. (1989). An empirical study of Perry, J. C c.,, & of defense defense mechanisms: I. Clinical Archives of of General General Psychiatry, interview and life vignette ratings. Archives Psychiatry, 46, 444-452.

576 576

PAULHUS, FRlDHANDLER, AND HAYES PAULHUS,FRIDHANDLER, AND HAYES

Perry, 1. J. c., C, & Cooper, S. S. H. H.(1992). (1992).What Whatdo docross-sectional cross-sectional measures measuresof ofdefense defensemechanisms mechanisms predict? In G. E. forforclinicians E. Vaillant (Ed.), Ego mechanisms mechanismsof ofdefense: defense:AAguide guide clinicians and researchers researchers(pp. (pp.195-216). 195-216).Washington, Washington,DC: DC:American AmericanPsychiatric PsychiatricPress. Press. Plutchik, R., & Conte, H. R. (1989). Measuring emotions and their derivatives: Personality and coping styles. styles. In In S. S.Wetzler Wetzler& &M. M.M. M.Katz Katz(Eds.), (Eds.),Contemporary Contemporary traits, ego defenses, and approachestotopsychological psychologicalassessment: assessment: Clinical experimental psychiatry (Vol.(Vol. I , 1, approaches Clinical andand experimental psychiatry pp. 239-269). New York: Brunner/Mazel. New York: Theory, research, research,and andexperience. experience. New York: (1980). Emotion: Theory, Plutchik, R., & Kellerman, H. (1980). Academic Press. Pyszczynski, T., Greenberg, J., & Solomon, S. S. (in press). Why do we need what we need? PsychologiA Terror Management perspective on the roots of human social motivation. Psychologi Inquiry, cal Inquiry. (1984). Causal versus diagnostic contingencies: On self selfQuattrone, G. A., & Tversky, A. (1984). Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, deception and the voter's illusion. Journal 46,46, 236-248. Rim, Y. (1987). A comparative study of two taxonomies of coping styles, personality and Personalityand andIndividual IndividualDifferences, Differences, 8, 521-526. sex. Personality 8, 521-526. (1989). Coping styles and use of means of influence influence in marriage. Personality Personality and and Rim, Y. (1989). Individual Differences, Differences,10,10,87-91. 87-91. Individual Rosenblatt, A., Greenberg, J., Solomon, S., Pyszczynski, T., & Lyon, D. (1989). Evidence The effects effects of of mortality mortality salience salience on on reactions reactions to to those those for Terror Management Theory I: The who violate violate or or uphold uphold cultural cultural values. values. Journal Journalof ofPersonality Personality and Social Psychology, and Social Psychology, 57, 57, who 681-690. Sackeim, H. A. (1983). Self-deception, self-esteem, and depression: The adaptive value of of Empiricalstudies studiesof ofpsychoanalytic psychoanalytictheories theories (pp. (Ed.), Empirical (pp. lying to oneself. In J. Masling (Ed.), 101-157). Hillsdale, NJ: Erlbaum. Sackeim, H. A. (1988). (1988). Self-deception: A synthesis. In J. S. Lockard & D. L. Paulhus Self-deception:An An adaptive adaptivemechanism? mechanism?(pp. (pp.146-165). 146-165).Englewood EnglewoodCliffs, Cliffs,NJ: NJ: (Eds.), Self-deception: Prentice-Hall. & Gur, R. C. (1978). (1978). Self-deception, other-deception other-deception and consciousness. In Sackeim, H. A., & G. E. Schwartz & D. Shapiro (Eds.), Consciousness Advances in in Consciousnessand andself-regulation: self-regulation: Advances research research(Vol. (Vol.2,2,pp. pp.139-197). 139-197).New NewYork: York:Plenum Plenum Press. Press. (1979). Self-deception, other-deception, and self-reported self-reported Sackeim, H. A., & Gur, R. C. (1979). Journalof of Consulting Consultingand andClinical Clinical Psychology, 213-215. Psychology, 47, 47, 213-215. psychopathology. Journal Sander, F. (1928). Experimentelle Ergebnisse der Gestaltpsychologie. In Bericht uber den by by 10, Kongress fur experimentelle experimentellePsychologie. Psychologic,(Reprinted (ReprintedGanzheitspsychologie, Ganzheitspsychologie, 10. Kongress C. C. F. Sander & H. Volkelt, Eds., 1961, Munchen: Beck) Sarbin, T. R. (1988). Self-deception Self-deception in the claims of hypnosis-subjects. In J. S. Lockard & D. L. Paulhus (Eds.), Self-deception:An An adaptive adaptivemechanism? mechanism?(pp. (pp.199-112). 199-112).Engle Engle(Eds.), Self-deception: wood Cliffs, NJ: Prentice-Hall. J.-P. (1956). Being and nothingness nothingness (H. Barnes, Trans.). New York: Washington Sartre, J.-P. Square Press. (Original work published 1943) Schaefer, E. S., & Plutchik, R. (1966). (1966). Interrelationships of emotions, traits, and diagnostic Schaefer, PsychologicalReports, Reports,18,18,399-410. 399-410. constructs. Psychological 49,49,49-62. 49-62. (1968).Mechanisms Mechanismsof ofdefense. defense.International InternationalJournal Journalof ofPsychoanalysis, Psychoanalysis, Schafer, R. (1968). Haven, CT: Yale University Press. (1976).A A new newlanguage languagefor forpsychoanalysis. psychoanalysis.New New Haven, CT: Yale University Press. Schafer, R. (1976). Schwartz, G. E. (1977). Psychosomatic disorders and biofeedback: A psychobiological model of disregulation. In J. D. Maser & M. E. P. Seligman (Eds.), Psychopathology: Psychopathology:Experi Experi(pp. 49-60). San Francisco: Freeman. mental models (pp.

CHAPTER 22 22

PSYCHOLOGICAL DEFENSE DEFENSE PSYCHOWGICAL

577 577

Schwartz, O. 1990) . Psychobiology G. E. ((1990). Psychobiology of repression and health: A A systems perspective. In J. L. Singer (Ed.), Repression Repression and and dissociation: dissociation: Defense Defense mechanisms mechanisms and and personality personality styles: ) . Chicago: University of 405-434). of Chicago Press. styles: Current Current theory theory and and research (pp. 405-434 Schwartz, M. S., Krupp, N. E., & Byrne, D. ((1971). 1971) . Repression-sensitization Repression-sensitization and medical Journal of of Abnormal Abnormal Psychology, Psychology, 78, 78, 286-291. diagnosis. Journal 1973) . Development Grinspoon, L., & Fienberg, S. E. ((1973). Development of of an ego profile scale. Semrad, E. V., Orinspoon, Archives of of General General Psychiatry, Psychiatry, 28, 70-77. Archives Shapiro, D. ((1965). 1965) . Neurotic Neurotic styles. New York: Basic Books. Autonomy and and rigid character. New York: Basic Books. 1981 ) . Autonomy Shapiro, D. ((1981). of mental health. American American 1993) . The illusion of Shedler, J., Mayman, M., & Manis, M. ((1993). Psychologist, 48, 1117-1131. Psychologist, 1 1 17-1 131. G. O. G. ((1980). Self-serving biases in person perception: A A re-examination of Sherwood, O. 1980). Self-serving Psychological Bulletin, 90, 445-459. projection as a mechanism for defense. Psychological Shevrin, H. ((1988). convergent psychodynamic psychodynamic and electrophysiologi electrophysiologi1988). Unconscious conflict: A A convergent Horowitz (Ed.), Psychodynamics Psychodynamics and and cognition cognition (pp. 117-167 117-167). cal approach. In M. J. Horowitz ). of Chicago Press. Chicago: University of Shevrin, H., Smith, W. H., & Fritzler, D. ((1970). 1970) . Subliminally stimulated brain and verbal responses of twins differing Abnormal Psychology, differing in repressiveness. Journal Journal of of Abnormal Psychology, 76, 39-46. 1985) . Social cognition: A look at motivated strategies. Annual Annual Showers, C c.,, & Cantor, N. ((1985). Review of of Psychology, Review Psychology, 36, 275-305. subliminal psychodynamic activation method: Overview Overview and Silverman, L. H. ((1983). 1983) . The subliminal comprehensive comprehensive listing of of studies. In J. Masling Masling (Ed.), (Ed.), Empirical Empirical studies studies of of psychoanalytic psychoanalytic theories ) . Englewood Cliffs, theories (pp. 41-61 41-61). Qiffs, NJ: Erlbaum. 1973) . The psychoanalytic Sjoback, H. ((1973). psychoanalytic theory theory of of defensive defensive processes. processes. New York: Wiley. Sloan, T. S. ((1987). 1987) . Deciding: Deciding: Self-deception Self-deception in life choices. London: Methuen. 1984) . Stabilization and automization Smith, O. G. J. W. ((1984). automization of of perceptual activity over time. In W. D. Frohlich, O. G. Smith, J. O. G. Draguns, & & U. Hentschel Hentschel (Eds.), (Eds.), Psychological Psychological processes processes cognition and personality (pp. 135-142 135-142). and personality ) . Washington, DC: Hemisphere. in cognition G. J. W., & Danielsson, A. ((1982). Anxiety and defense defense strategies in childhood childhood and 1982) . Anxiety Smith, O. adolescence. New York: International Universities Press. adolescence. G. J. W., & Henriksson, M. (1956). effect on an established established percept percept of a perceptual ( 1956) . The effect Smith, O. process beyond awareness. Acta Acta Psychologica, Psychologica, 11, 346-355. Smith, O. G. J. W. & Westerlundh, B. (1980). A process perspective on ( 1980) . Perceptgenesis: A perception-personality. In L. Wheeler (Ed.), Review Review of of personality personality and and social social psychol psychology ogy (pp. 94-124 94-124). ). London: Sage. Snyder, C. R 1985) . Collaborative companions: The relationship of self-deception R. ((1985). self-deception and excuse-making. In M. W. Martin (Ed.), Self-deception Self-deception and and self-understanding self-understanding (pp. 35-51). of Kansas. 35-51 ) . Lawrence: University Press of Snyder, C. R R. (Ed.). ((1989). 1989) . Self-illusions: When are they adaptive? (Special issue). Journal of Social Social and and Clinical Clinical Psychology, Psychology, 8, 1-221. of R., & Higgins, R R. L. ((1988). effective role in the negotiation of of Snyder, C. R, 1988) . Excuses: Their effective reality. Psychological Psychological Bulletin, 104, 23-35. of denial. In S. Breznitz (Ed.), The denial denial of of stress (pp. Spence, D. P. ((1983). 1983) . The paradox of 103-124). 103-124 ) . New York: International Universities Press. efficacy of of avoidant and nonavoidant coping Suls, J., & Fletcher, B. ((1985). 1985) . The relative efficacy strategies: A meta analysis. Health Psychology, (3), 249-288. Psychology, 4 4(3),

578 578

PAULHUS, PAULHUS,FRlDHANDLER, FRIDHANDLER,AND ANDHAYES HAYES Swann, B. (1992). Seeking truth: Swann, W. B. truth: Finding Finding despair: despair: Some Some unhappy consequences consequences of of aa negative Science, 1, 15-18. Current Directions DirectionsininPsychological Psychological Science, i, 15-18. negative self-concept. self-concept. Current Taylor, Psychogist, Taylor, S. S. E. E. (1990). Health Health psychology: psychology: The The science science and and the the field. American Psychogist,

45, 45, 40-50.

Taylor, (1988). Illusion Taylor, S. S. E., & & Brown, Brown, J. J. (1988). Illusion and and well-being: well-being: A A social social psychological psychological perspective perspective 193-210. on on mental health. health. Psychological PsychologicalBulletin, Bulletin,103, 103, 193-210.

Taylor, Taylor, S. S. E., E., & Clark, Clark, L. L. F. F. (1986). (1986). Does Does information information improve improve adjustment adjustment to to noxious noxious events? events? social psychology psychology (Vol. (Vol. 3, 3, pp. pp. In M. M. J. J. Saks Saks & & L. L. Saks Saks (Eds.), (Eds.), Advances in applied social In 11-28). -28). Hillsdale, Hillsdale, NJ: NJ: Erlbaum. Erlbaum. Tesser, A. A. (1986). Some Some effects effects of of self-evaluation self-evaluation maintenance maintenance on on cognition cognition and and action. action. In In Tesser, R R. M. M. Sorrentino Sorrentino & & E. E. T. T. Higgins Higgins (Eds.), (Eds,), The The handbook handbookof ofmotivation motivationand andcognition: cognition: of social social behavior behavior (pp. (pp. 435-464). 435-464). New New York: York: Guilford Guilford Press. Press. Foundations of of Tesser, Tesser, A., A., & & Campbell, Campbell, J. D. (1982). Self-evaluation Self-evaluation maintenance maintenance and and the the perception perception of friends and and strangers. Journal of of Personality, Personality,50, 50,261-279. 261-279. friends

selfishgene. gene.New NewYork: York:Oxford OxfordUniver UniverTrivers, R R. L. L. (1976). Foreward. Foreward. In In R R. Dawkins, The selfish Trivers, sity Press. sity Trivers, R R. L. L. (1985). Social evolution. evolution, Menlo Menlo Park, Park, CA: CA: Benjamin/Cummings. Benjamin/Cummings. Trivers, Turvey, C, & Salovey, Salovey, P. P. (1994). Measures Measures of of repression: repression: Converging Converging on on the the same same construct? construct? Turvey, C., & Imagination,Cognition Cognition& & Personality, 279-289. Imagination, Personality, 13, 13, 279-289.

Vaillant, G. G. E. E. (1971). Theoretical Theoretical hierarchy hierarchy of of adaptive adaptive ego ego mechanisms: mechanisms: A A 30-year 30-year follow followVaillant, General Psychiatry, Psychiatry,24, 24, up of of 30 men selected for for psychological psychological health. health. Archives of of General up 107-118.

Vaillant, G. G. E. E. Vaillant,

Natural history history of of male male psychological psychological health: health: The The relation relation of of choice choice (1976). Natural of General General Psychiatry, Psychiatry, Archives of 33, 535-545. Vaillant, G. G. E. E. (1992). Ego mechanisms mechanismsof ofdefense: defense:A Aguide guide clinicians researchers. Vaillant, forfor clinicians andand researchers. Washington, DC: DC: American American Psychiatric Psychiatric Press. Press. Washington, Vaillant, G. G. E., Bond, M., M., & & Vaillant, C. C. O. O. (1986). An An empirically empirically validated validated hierarchy hierarchy of of defense defense mechanisms. mechanisms. Archives Archives of of General GeneralPsychiatry, Psychiatry,43,43,786-794. 786-794. Vickers, R R. R., Jr., Conway, T. T. L., & & Haight, Haight, M. M. A. A. (1983). Association Association between Levenson's Levenson's Vickers, R, Jr., Dimensions of of Locus of of Control and measures of coping and defense defense mechanisms. Psychological PsychologicalReports, Reports,52, 52,323-333. 323-333. Vickers, R R, Jr., & Hervig, L. K. (1981). Comparisons of three psychological defense R. R., defense 45,45, 630-638. mechanism questionnaires. questionnaires. Journal Journalof ofPersonality PersonalityAssessment, Assessment, 630-638. Vickers, R R, Jr., Hervig, L. K., Rahe, R R. R., R. H., & Rosenman, R R. H. (1981). Type A behavior PsychosomaticMedicine, Medicine,43, 43,381-396. 381-396. and coping coping and and defense. defense. Psychosomatic and Wallerstein, R R. (1983). Defenses, Defenses, defense defense mechanisms, mechanisms, and and the the structure structure of of the the mind. mind.Journal Journal Wallerstein, of the American AmericanPsychoanalytic Psychoanalytic Association, Supplement, 31, 201-225. of Association, Supplement, 31, 201-225. Warrenburg, R, & Warrenburg, S., Levine, Levine, J., Schwartz, Schwartz, G. G. E., Fontana, Fontana, A. A. F., F., Kerns, Kerns, R R. D., Delaney, Delaney, R., & Mattson, R R. (1989). Defensive Defensive coping coping and and blood blood pressure pressure reactivity reactivity in in medical medical patients. patients. Mattson, Journalof ofBehavioral BehavioralMedicine, Medicine, 407-424. Journal 12,12, 407-424. Wegner, Wegner, D. M. M. (1989). White bears bears and andother otherunwanted unwantedthoughts. thoughts.New NewYork: York:Viking VikingPress. Press. Weinberger, D. A. A. (1990). The The construct construct validity validity of of the the repressive repressive style. style. In In J. J. L. L. Singer Singer (Ed.), (Ed.), Weinberger, Repressionand anddissociation: dissociation: Implications personality theory, psychopathology, Repression Implications forfor personality theory, psychopathology, and and health (pp. (pp. 337-386). Chicago: Chicago: University of Chicago Press. health Weinberger, & Schwartz, Schwartz, G. G. E. E. (1990). Distress Distress and and restraint restraint as as superordinate dimenWeinberger, D. D. A., A., & superordinate dimen sions of of self-reported self-reported adjustment: adjustment: A A typological typological perspective. perspective. Journal of of Personality, Personality, sions 58, 381-417. 58, of ego ego mechanisms of defense defense to to adult adult adjustment. adjustment. of mechanisms of

CHAPTER 22 CHAPTER

PSYCHOLOGICAL DEFENSE

579 S79

Weinberger, D. A., A., Schwartz, Schwartz, G. G. E., E., & & Davidson, Davidson, R. R. J. J. (1979). (1979). Low-anxious Low-anxious high-anxious, high>anxious, Weinberger, and and repressive repressive coping coping styles: styles: Psychometric Psychometric patterns patterns and behavioral behavioral and and physiological physiological responses to stress. stress. Journal of of Abnormal Abnormal Psychology, Psychology, 88, 88,369-380. 369-380. Weinberger, J. (1989). (1989). Response to Balay & Shevrin: Constructive critique critique or misguided attack? American Psychologist, Psychologist,44, 44,1417-1419. 1417-1419. Weinberger, J., J., & Silverman, L. H. (1990). Testability Testability and and empirical empirical verification verification of of psychoana psychoana-

lytic dynamic propositions through subliminal psychodynamic activation. Psychoana Psychoanalytic lytic Psychology, Psychology, 7,7,299-339. 299-339. Williams, Williams, L. L. M. M. (1994). (1994).Recall Recallof ofchildhood childhoodtrauma: trauma:AAprospective prospectivestudy studyof ofwomen's women'smemories memories 1 167-1176. of of child child sexual abuse. Journal of of Consulting Consulting &&Clinical ClinicalPsychology, Psychology,62,62, 1167-1176. states: Wilson, T. T. D. D. (1985). (1985). Self-deception Self-deception without without repression: repression: Limits Limits on on access access to to mental mental states: Wilson,

In M. M. W. W. Martin Martin (Ed.), Self-deceptionand andself-understanding self-understanding(pp. (pp.95-116). 95-116).Lawrence: Lawrence: In (Ed.), Self-deception University Press Press of of Kansas. Kansas. University Woike, B. B. A., Aronoff, Aronoff, J., Stollak, StoUak, G. G. E., & & Loraas, Loraas, J. A. A. (1994). (1994). Links Links between between intrapsychic intrapsychic Woike, and interpersonal defenses in dyadic interaction. Journal of of Research ResearchininPersonality, Personality, 28, 101-113. 28, 101-1 13. Psy Zajonc, Zajonc, R. R. B. B. (1980). (1980). Feeling Feeling and and thinking: thinking: Preferences Preferences need need no no inferences. inferences. American Psychologist,35, 35,151 151-175. chologist, -175.

This Page Intentionally Left Blank

CHAPTER 23 CHAPTER 23

INTERNAL INHIBITIONS iNfflBinoNS AND CONTROLS L MEGARGEE MEGARGEE EDWIN I. THE THE FLORIDA FLORIDA STATE STATE UNiVERSITY UNIVERSITY

I. INTRoDuCfloN INTRODUCTION I.

If you tum turn back to to the the Table of Contents and review the the topics covered in in this Handbook, you youwill willsee see that thatmost mostpersonality personalitytheorists theoristsare areinterested interestedin in what whatpeople people Handbook, do and why they do it. This chapter is different. We are going to be discussing what do not not do doand and why why they they do donot notdo doit. it. Our Our focus focus will will be beon onthe theinternal internal people do inhibitions, controls, and restraints that cause people to refrain from behaviors that they consider to to be be wrong wrong or or inadvisable. inadvisable. on what II shall call the the "positive" Most personality theorists concentrate on causes of behavior: the traits, habits, motives, and attitudes that lead an artist to to be be competitive, or or that cause aa neurotic create, that drive aa "Type A" person to individual to to become anxious. However, as as Robert Robert Frost Frost pointed pointed out out in in the the "The "The Not Taken" any decision to to perform one one act act Road Not Taken" (Untermeyer, 1955, p. p. 54), 54), any to do do something something else. else. To To also involves the decision, conscious or unconscious, not to we must study the steering and brakes understand how an automobile functions, we as the the engine and power train. To To understand human behavior, we we must must as well as examine controls and inhibitions as well as motives and drives. the brakes and steering on on our our automobiles, we we generally do do not not As with the think much about inhibitions and controls until they fail. My interest in inhibition my professional career studying antisowas born of necessity: I have spend most of my antiso cial behavior and violence. In contrast with most of my fellow contributors to this Handbook, the walls surrounding my my research laboratories have often been topped Handbook, IlANDBOOK PERSONA.Lfl'Y PSYCHOLOGY PSYCHOLOGY HANDBOOK OF OF PERSONALITY

581 581

CoPYII GHT C COPYRIGHT © 1997 BY ACADEMIC PRESS. ALL liGHTS OF lEPRODUcnON IN ANY FORM IESEiVED. A a RIGHTS OF REPRODUCHON IN ANY FORM RESERVED.

582 582

EDWIN EDWIN L L MEGARGEE MEGARGEE

with barbed barbed wire rather than ivy, and featured featured gun towers instead of of bell towers. Although Although I could go home at night, my subjects had to stay behind because society had decided that that they required required stringent external external controls because they lacked adequate internal internal inhibitions and restraints. Naturally, in this setting I developed an interest in controls, and internal inhibitions are a major major construct construct in my theoretical framework for for the study of of aggressive behavior. behavior. framework In this chapter, internal inhibitions and controls will be discussed discussed from from a broad broad perspective, emphasizing emphasizing concepts and theories from from a range of of disciplines. The offered will be my own and should be regarded regarded as hypotheses to stimulate views offered thought and discussion rather rather than as "revealed "revealed truth." In the course of of this discus discussion, we shall address the following following topics: defining what we mean by internal controls or inhibitions, 1. Problems in defining semantically semantically and operationally operationally interdisciplinary perspectives perspectives 2. Philosophical issues and interdisciplinary 3. How How inhibitions and controls are acquired acquired or enhanced How inhibitions and controls are diminished, lost, or overcome overcome 4. How 5. Methodological Methodological problems in doing empirical research on controls controls for research, theory, and practice practice 6. Implications for ISSUES.- WHAT Do D O WE W E MEAN BY n. DEFINITIONAL IsSUES: D. INTERNAL CONTROLS? of our our first problems is the fact that internal controls and inhibitions are One of difficult to define. This is because they are "negative" constructs that must be difficult defined defined by exclusion. Moreover, since inhibitions are linked to values which vary from person person to person, society to society, and period period to period, the specific specific behaviors behaviors from inhibited will also differ. differ. Let us examine each of of these problems. that are inhibited Defining a Negative Negative Constmct Construct A. Defining personality constructs are adduced adduced to explain why people engage in certain certain Most personality behaviors. Terms such as 'leadership," "leadership," "anxiety," and "achievement" all connote connote constellations of of attributes attributes and observable observable behaviors. These can be used to construct construct operational definitions of exemplify these operational definitions of these constructs or to identify identify people who exemplify traits. However, "inhibitions," "taboos," "internal constraints," "superego," and other such terms in the thesaurus are used to explain why certain behaviors do all other not occur. It is obviously obviously much easier to construct construct an operational operational definition definition of of not rather than suppressed suppressed behaviors. observable rather Variability of of Values B. VarlabDity

This problem problem is compounded compounded by the fact that specific behaviors that are inhibited vary as a function function of of each person's values. Since values differ differ from from one individual, individual.

CHAPTER CHAPTER 23 23

INmBmoNs .... INHIBITIONS AND AND CONTRO CONTROLS

583 583

social subgroup, or culture to the next, it is difficult difficult to stipulate the behavioral omissions that suggest taboos are operating. At a fast food restaurant, you may of a hamburger. Are observe a number of patrons choose the salad bar instead of they normally opposed to eating meat, or do they simply prefer prefer vegetables? This leads us to the most difficult difficult aspect of defining inhibitions, the need to rule out alternative explanations for failures to act.

c. C. Definition by by Exclusion Exclusion

Before we can conclude that someone's failure to perform perform some act results from internal prohibitions, we must exclude all the other reasons why that person might refrain refrain from from that act. Here are a few alternatives that must be ruled out: /. Lack Lack oof Motivation 1. f Motil'ation If an individual lacks the appropriate motivation, drive, incentive, or desire to perform the act in question, then his or her failure to respond cannot be attributed to inhibitions. This means researchers must establish that an appropriate drive failure to perform perform some act results state exists before they can conclude that a failure from inhibitions. from

2. Inability Inability to to Perform 2. Peiform

It may be that the response in question is not in the person's repertoire or that some external constraint prevents the person from from engaging in the behavior in question. A recent film depicted a high school computer hacker who broke into his school's computer system and changed all his F's to A's. This film may have inspired thousands of students to emulate his behavior, but most were unable to do so because they lacked his technical expertise and/or his access to a computer.

3. 3.

Anticipation of of a Negatil'e Negative Outcome Anticipation

People will also refrain refrain from from behavior if it is not likely to be successful successful or if bad things are likely to happen to them as a result. Arthur Arthur Bremer stalked President Nixon but never shot at him because he was unable to penetrate the Secret Service screen ((Institute Institute of Medicine, 1984). 1984). His lack of of internal inhibitions was amply demonstrated when he shot Governor George Wallace instead. People will also refrain refrain from behavior that is likely to result in unpleasant consequences. If several large and apparently vicious dogs raid my backyard barbe barbecue and start eating the steaks, I will not attempt to retrieve my property from their slavering jaws. I can replace a Delmonico easier than my hand. 4. Response Competition

At any given moment, there may be a number of of different different responses competing which are mutually incompatible. You are presently reading this book. You are probably not solving the New York Times crossword puzzle, making love to your sweetheart, or mowing the lawn. The fact that you are reading this book does not

584 584

EDWIN EDWIN L I. MEGARGEE MEGARGEE

imply that you have any internal inhibitions against engaging in any of against engaging of these other behaviors. the world, behaviors. It It simply simply means means that that at at this this point point in in the the history history of of the world, for for whatever whatever reason, you have chosen to read this book. reason, you have chosen to read this book. To recapitulate the problems associated with excluding other other explanations, a person's failure failure to perform perform some act that we would have expected expected them to perform perform at that time and place implies internal constraints or inhibitions only if if we can be sure that (a) the appropriate motivational motivational state was present, (b) the person person was of performing performing the response, (c) the external external situation situation did not not indicate the capable of response would fail fail or result in a negative outcome, and and (d) we can be reasonably sure that the person did not simply prefer prefer to do something something else.

D. Other Problems difficulties listed thus far far should surely give pause to those who would attempt The difficulties formulate a thoroughly thoroughly satisfactory satisfactory operational definition of internal inhibitions. to formulate definition of However, there are other problems as well. So far, we have been discussing occasions when some expected expected behavior behavior failed failed to occur. Can we at least infer infer that internal constraints were absent absent if if the behavior did take place? No. There may have been internal inhibitions that that were simply inadequate in the face face of of strong temptation. It is no accident accident that, unlike any other item on the menu, many restaurants put the high-calorie desserts on a cart which is wheeled directly to the patron's table to maximize the temptation. Another Another problem is that inhibitions and taboos can vary as a function function of of the time, the place, the object, and the specific specific act. It is all right for for a football football player to tackle an opposing player during a game Saturday Saturday but but he should should refrain refrain from from President of of the University at the prayer prayer breakfast decking the President breakfast Sunday. Internal constraints must also be differentiated differentiated from from other other similar constructs. that the forbidden forbidden drive One is repression. Might not our inhibitions be so strong that from awareness? I may think the reason that I never purchased the is blocked from controversial novel Satanic Satanic Verses in 1989 is because I had controversial had no (conscious) interest psychoanalyst might argue that of interest in reading that book. A psychoanalyst that my apparent apparent lack of actually stemmed from Ayatollah Kho from an unconscious unconscious need not to offend offend the late AyatoUah Khofather figure, figure, because I had not completely resolved resolved my meini, an obvious father not yet completely conflicts will be beyond beyond the scope of of this chapter. Oedipal conflict. Such unconscious conflicts Suffice it to say that any repressions that influence influence behavior life Suffice behavior will simply make life difficult for for researchers researchers and theorists. that much more difficult

m. PHILOSOPHICAL ISSUES m . PIDLOSOPIDCAL As we have noted, internal controls and constraints constraints are closely closely associated associated with from a variety variety of of disciplines have discussed discussed them from from values. As a result, people from different perspectives. Indeed, many of the issues and concerns were were a number of different raised centuries before before the first psychologist drew breath.

CHAPTER CHAPTER 23 23 INJuBmoNS iNHiBrnoNsAND ANDCONTROLS CONTROLS

585

Why such interest in this particular aspect of personality? Many people assume that internal controls and constraints necessarily involve ethical or moral prohibi prohibitions against pedorming performing acts that are disapproved of by society. Thus "internal constraints" are viewed as being synonymous with "conscience" or "morality," and the the theoretical theoretical and and empirical empirical issues issues raised raised by by psychologists psychologists are are regarded regarded as as simply simply one one more more attempt attempt to to explore explore the the age-old age-old dilemma dilemma of of good good versus versus evil. evil. Actually, Actually, these these concepts concepts are are not not congruent. congruent. As As the the trials trials of of the the Nazi Nazi war war criminals criminals at at Nuremberg Nuremberg demonstrated, demonstrated, some some people people have have well-developed well-developed value value systems systems that that are are at at odds odds with with the the moral moral codes codes espoused espoused by by the the larger larger society. society. Nevertheless, there is considerable overlap, and much of the thinking and research that have been done on such topics as moral development, social conformity, social deviance, psychopathy, cultural relativism, and, yes, good versus evil, are relevant to the issues we will be discussing. Your theoretical perspective on the origin of internal controls is probably influenced influenced by your basic view of human nature. Reduced to its essence, the basic fundamentally good or evil. philosophical question is whether people are fundamentally Those who maintain the people are basically good blame a corrupt society before the for human misery and evil. This is the allegory of the Garden of Eden before of the foremost foremost proponents of this philosophy philosophy was Jean Jacques snake intruded. One of Emile (1762), "Everything is good when it Rousseau (1712-1788), who wrote in Emile of the Creator; everything degenerates in the hands of man" (Beck, (Beck, leaves the hands of 1980, p. 264). This viewpoint underlies nondirective and humanistic approaches to impediments psychotherapy that view the therapeutic task as removing acquired impediments self-actualization and growth. to self-actualization philosophy is the belief belief that people are basically Diametrically opposed to this philosophy evil and that left left to their own devices they will exploit and prey upon one another. of Original Sin following the Fall. Thomas Hobbes (1588-1679) (1588-1679) This is the Doctrine of of the major advocates of of this position. Whereas Rousseau, whose favorite was one of book was Robinson Robinson Crusoe, extolled the virtues of of people in a state of nature, Hobbes, in Leviathan Leviathan (1651), described a state of of nature as "No arts; no letters; no society; and, which is worst of of all, continual fear and danger of of violent death; and the life of of man, solitary, poor, nasty, brutish, and short" (Beck, 1980, p. 264). Latter-day proponents of this view include Freud and his followers who maintained that a major goal of of child rearing is to civilize and control the primitive id impulses and needs that are present from birth; i.e., to develop the ego. A A third view, developed somewhat more recently, holds that people have no innate good or bad tendencies, but instead are products of their environments. John Locke (1632-1704) used the analogy of a blank slate or tabula rasa to convey the notion that experience determines our character. More recently behaviorists espoused and social learning theorists such as B. F. Skinner and Albert Bandura have espoused this view. Theorists who believe in the innate goodness of of humankind have no need to account account for the development of of values or controls. Their basic position is that the

S86 586

EDWIN EDWIN L L MEGARGEE MEGARGEE

human equipped with human animal animal comes comes equipped with these these attributes. attributes. Instead, Instead, their their theoretical theoretical task task is is to to account account for for evil evil and and explain explain what what went went wrong. wrong. The people who The people who adopt adopt the the latter latter two two positions, positions, namely, namely, that that people people are are born born without innate controls, must explain without innate controls, must explain how how we we acquire acquire inhibitions. inhibitions. Obviously, Obviously, one's one's academic academic discipline disciphne will will influence influence one's one's theory. theory. Religions Religions often often cite cite some some form form of of redemption. Anthropology redemption. Anthropology focuses focuses on on the the transmission transmission of of culture, culture, and and sociology sociology on values and loyalties. Psychiatry on the the conflict conflict of of group group values and loyalties. Psychiatry tends tends to to seek seek signs signs of of psychopathology, and psychology psychology investigates investigates individual individual personality personality characteristics. characteristics. psychopathology, and of these views, but we should remain aware of the fact fact We shall examine some of influence that cutting across disciplines, these three basic philosophical perspectives influence adopt. the positions theorists adopt. the next next section, section, we we shall shall examine examine some some of of the the hypotheses hypotheses that that have have been been In the advanced advanced to to explain explain how how internal internal controls controls and and inhibitions inhibitions originate, originate, and and how how they they are overcome. IV, ORIGINS OF INTERNAL INHIBITIONS IV. INHIBmONS of reasons refrain As we noted noted in Section Section II, II, there are all sorts of reasons why people people may refrain from behaving in certain ways. Appropriate or adequate motivation may be lacking, from repertoire or beyond their capabilities, or external the response may be outside their repertoire sanctions may be imposed. These factors, as important as they are in predicting predicting behavior, are outside the purview of of this section. focus on internal internal inhibitions and controls that deter us In this section we will focus from behaviors that we would otherwise perform. As we shall see, these taboos from general, while while some specific. Some Some are variable. Some Some inhibitions inhibitions are are general, some are are quite quite specific. are are variable. lasting, temporary. The denominator is lasting, and and some some are are temporary. The common common denominator is that that we we are are referring referring to internal impediments rather than than external external constraints. to internal impediments rather constraints. Physiological Mechanisms Meclianisms A. Physiological Given the the fact fact that that we we are are biological biological organisms, organisms, it it is is obvious obvious that that all all our our behavior behavior Given has has aa physiological physiological basis. basis. What What we we do, do, say, say, or or think think depends depends on on our our neurons, neurons, hormones, organs, and tissues. Nevertheless, Nevertheless, partly partly because because of of the the difficulties difficulties involved involved in in doing doing physiological physiological research research on on humans humans and and partly partly because because of of an an American American bias bias in in favor favor of of environmen environmental paid only tal explanations, explanations, most most theorists theorists traditionally traditionally paid only lip lip service service to to the the physiological physiological bases bases for for personality personality functioning. functioning. This This situation situation is is changing. changing. With With the the recent recent technological technological advances advances in in our our ability study the ability to to study the basic basic genetic genetic material material and and to to conduct conduct nonintrusive nonintrusive investigations investigations of of the the central central nervous nervous system, system, we we have have come come to to recognize recognize the the role role of of physiological physiological and and hereditary hereditary factors factors in in major major mental mental disorders disorders previously previously thought thought to to have have aa purely purely functional basis. Concomitant Concomitant with these clinical advances is a greater appreciation appreciation functional of the importance of of physiological factors in normal personality functioning functioning as well. of

CHAPTER CHAPTER 23 23 INlDBmONS INHIBITIONSAND ANDCONTROLS CONTROLS

587 587

In a broad sense, there are many physical reasons why a person may refrain from doing something. Diabetes and circulatory problems can cause male sexual impotence, a broken leg will interfere with ski jumping, and acute nausea will sap one's motivation to do just about anything except retch and hope for an early demise. Ultimately, Ultimately, we all stop behaving behaving because we die. die. However, these physical demise. inhibitions are not what we mean by "internal *Hnternal controls," even though they may be very important in predicting predicting behavior. Physiological research on internal inhibitory mechanisms is still relatively primitive. We have already noted the difficulty difficulty of of defining internal controls opera operaof physiological physiological experi experitionally. This problem is compounded by the limited range of perform with humans. Even though animal studies permit ments we can ethically perform greater precision, many will question their relevance. In this section we shall touch on possible hereditary influences and the central nervous system substrate for internal controls. Genetic Mechanisms Mechanisms 1. Genetic homo sapiens sapiens is a notoriously difficult genetiUnfortunately, homo difficult species for behavioral geneti cists to investigate investigate given our propensity for assortative mating, our low reproduction rate, our long maturation period, and our hopelessly heterogeneous gene pools. Consequently relatively little is known about the genetic bases of personality func functioning in general and restraints in particular. In this section, most of our speculations inferences from other areas of research. are based on inferences Criminals, The first empirical research relevant to possible a. Research on Criminals. of internal controls and inhibitions was performed by scientists genetic determinants of whose primary interest was in people who apparently lacked adequate controls, namely, convicted criminals. In the 19th century, Cesare Lombroso (1835-1909) reverhypothesized that criminals' savage behavior suggests that they are "atavistic rever of human evolution. Such atavism, sals" or throwbacks to a more primitive stage of manifested by physical signs or "stigmata," such Lombroso maintained, would be manifested as low sloping foreheads. In the 20th century extensive studies were carried out by Goring (1913) and Hooton (1939) to test Lombroso's hypothesis, but sampling and methodological flaws rendered their results inconclusive (Rosenquist & & Megar Megarmethodological gee, 1969). Somewhat more convincing are studies showing a higher rate of concordance for criminality among monozygotic than dizygotic twins (Christiansen, 1977), and of criminality among adoptive children whose biological parents were higher rates of criminals than among those whose biological parents were noncriminals, irrespective of the criminality criminaUty of of the adoptive parents (Mednick, Gabrielle, & & Hutchings, 1984; of Mednick & Volavka, 1980). However, any research using convicted criminals is of the heritability of internal controls. only tangentially related to the question of Heritability of of Personality Personality Factors. Factors, Using factor analysis, some personality b. Heritability researchers have suggested that five or six fundamental dimensions underlie our

S88 588

EDWIN EDWIN L L MEGARGEE MEGARGEE

perceptions and descriptions of of one another (Hogan, 1986). One of of these dimen dimensions, "conscientiousness," bears a passing resemblance to our concept of of internal controls. According to Hogan, "conscientiousness "conscientiousness contrasts people who are depend dependable and conforming with those who are undependable undependable and nonconforming" (1986, p. 58). Using personality test scores as operational definitions definitions of of these dimensions, researchers such as Bouchard (1984) and Loehlin and Nichols (1976) have compared the scores of of monozygotic and dizygotic twins to estimate the heritability of of these traits and concluded that about half the variance can be attributed to genetic factors.

c. The Evolutionary Evolutionary Perspective. Perspective. Some scientists have suggested that inner controls are innate because they have selective value for the survival of of the species. In his paper on "Moral Conduct and Character," Robert Hogan "assumes that morality is a natural phenomenon, an adaptive response to evolutionary pressure, of man's and that an understanding of moral behavior is relative to our knowledge of biological and social nature" (1973, p. 218). This evolutionary perspective has not only been used to explain general ethical tendencies, but also to account for specific taboos such as those against incest or of the incest taboo, Gardner Lindzey (1967), homicide. Noting the near universality of in his presidential address to the American Psychological Association, theorized that it must be genetically based. Lindzey noted that the literature on interpersonal attractiveness showed that people, like most creatures, are most attracted to one another on the basis of similarity, familiarity, and proximity. Without an incest taboo, these tendencies of inbreeding, for it is the members of of one's would ordinarily lead to a high level of immediate family family who best fit these specifications. Since inbreeding would be delete deleterious for the species, Lindzey (1967, p. 1056) argued that "the biological necessity of of outbreeding led to the evolution of a set of prohibitions against this powerful powerful tendency t e n d e n c.y .. . .. "" The noted ethologist Konrad Lorenz (1963/1966) maintained maintained that inhibitions against aggressive behavior also have a genetic basis. In the normal course of evolution, animals that had the physical capacity to kill other members of of their own species evolved inhibitions against the use of of their deadly weapons when combating one another. Rattlesnakes, for example, fight by wrestling one another and never use their fangs because the venom would be lethal. Because our ancestors were less ferocious, such inhibitions had little selective modern humans did not inherit strong inhibitions against homicide. How value, so modem However, once we invented weapons and passed this knowledge down from generation to generation, we rapidly became a lethal species as the quick growth of of technology outstripped the slow course of evolution. This imbalance, according to Lorenz, accounts for our high homicide rates. rf. Selective Selective Breeding. Breeding, Another line of evidence comes from selective breed breedd. ing of animals. Although animal husbandrymen have been more interested in breed-

CllAFrE R 23 CHAPTER 23 INlUBmONS iNmernoNSAND ANDCONTROLS CONTROLS

589 589

of fighting bulls, cocks, dogs, and fish, it is well documented ing aggressive strains of that animals can also be bred for docility and tractability. The monks at the Hospice of St. Bernard recognized recognized that the large dogs they were breeding to assist travelers lost in the snowy mountain passes would be of little Uttle help if they had the feisty temperament of a terrier or the aggressiveness of a pit bull. Hence they deliberately selected the gentlest animals as well as those those 1942,1954). with the most stamina, endurance, and intelligence (S. E. Megargee, 1942, 1954). Of course, it is questionable whether a dog's high threshold for aggressive behavior is equivalent to a person's internal controls, or whether being gentle and docile is equivalent to being controlled or inhibited. conceptually equivalent

e. Evidence Evidence on Genetic Genetic Mechanisms. Mechanisms, Thus it can be seen that there is no evidence for the inheritance of generalized sets of of values or specific definitive evidence inhibitions such as the incest taboo. Given the ethical constraints governing genetic unUkely that any research with humans, as well as the technical difficulties, it is unlikely definitive studies will be forthcoming forthcoming in the foreseeable future. Nevertheless, despite definitive of internal controls, there is enough the bias toward environmental explanations of evidence to allow us to entertain genetic hypotheses. suggestive evidence 2. Central Central Nervous Nervous System System 2.

The best evidence for a neurological substrate for internal inhibitions stems from clinical studies showing that various types of CNS impairment can lead to impulsivity and diminished ethical constraints. Beyond this broad observation, our knowledge of specific inhibitory mechanisms in the human brain is rather vague. Specific centers have been identified that act to inhibit very basic forms of of immense rats that have gorged behavior such as eating and drinking. Photographs of themselves into obesity after removal of the hypothalamic centers that signal satiety introductory psychology texts, and scientists have "turned are a standard feature of introductory off' aggressive behavior in some animals by electrically stimulating areas of the off" brain believed to inhibit aggression (Johnson, 1972; 1972; Mark & Ervin, 1970). These of primary importance to clinicians attempt attemptcentral nervous system mechanisms are of ing to diagnose the possible causes of impulsive or poorly controlled behavior. turn to a consideration of the neurological neurological basis of higher ethical When we tum principles, much less is understood. It seems definite that the cortex is involved, and research on the development of moral and ethical sensitivity makes it clear that a certain level of cortical development is required for children to make "mature" ethical choices. Again, this is most relevant for examining those cases in which there has been a breakdown in ethical behavior. As we shall see, toxic substances interfere with the cortex can have the effect of diminishing one's or diseases that interfere internal inhibitions. 3. Other Other Physiological Physiological Factors Factors

Although their exogenous origin makes them outside the purview of "internal inhibitions" in the usual sense, it should be noted that drugs can be used to reduce

S90 590

EDWIN L EDWIN L MEGARGEE MEGARGEE

undesirable behavior reduce drive strength, strength, as in the appetite behavior patterns. Some act to reduce suppressants dieters or suppressants used used by by dieters or the the so-called so-called "chemical **chemical castration" castration*' sometimes sometimes for sex offenders. Others strengthen strengthen a person's ability to inhibit undesir undesiradvocated for behavior that is contrary contrary to the individual's own code of of able behavior, that is, behavior recurrent problem problem in clinical settings are patients who are able able to control values. A recurrent their behavior while on medication in an inpatient setting but who discontinue their after discharge. medication and act out after

B. B* Psychological Sources of of Inhibitions Rouseau (1758), "The **The first of of all laws is to respect the law," but According to Rouseau cometh this respect for for law and order? Developmental Developmental psychologists such whence cometh Robert Selman (1980) have as Jean Piaget (1932), Lawrence Kohlberg (1981), and Robert comprehension and understanding understanding charted distinct patterns and stages in children's comprehension of moral issues. Do these stages in moral development development stem from from maturation maturation of from the growth of of the child's cognitive abilities from from a concrete, simplistic or from understanding to a more abstract and complex appreciation appreciation of of the world? Those who, like Rousseau, believe in the innate goodness of of humankind humankind do not have to explain how most children children become socialized socialized to their their particular particular culture's culture's values. For them, the unfolding unfolding of of nature's plan plan is a sufficient sufficient explanation. Their task is to explain how society interferes interferes with this normal process. unformed However, those who agree with Locke that we are born essentially unformed and are shaped by experience, or who subscribe to Hobbes' view that we are inherently selfish and amoral, must explain how it is that most of of us are more or reach adulthood. To account for for this process, theorists less civilized by the time we reach of our our particular must answer two basic questions: (1) How do we learn the rules of from society? (2) Why is it that we obey them? We shall discuss these questions from the viewpoints of behaviorism, cognitive social learning theory, and psychoanalysis. 1. Socialization: Learning Leamlng the L Socialization: the Rules Rules o off the the Culture Culture

of development, everyone acquires a sense of of values, a Somehow in the course of moral code that specifies specifies what we should and should should not do. Broadly Broadly speaking, of acceptance" and "latitudes "latitudes they learn behavior can be divided into "latitudes of of rejection" (E. L Megargee, 1973). of I. from one The behaviors included in these latitudes vary from one society society to to the the next. next. Most Americans repudiate the practice of different religious of killing people who have different beliefs, of butchering butchering cattle beliefs, but but have have no no compunctions compunctions about about the the custom custom of cattle for for food. food. In India, however, Hindus who would be aghast at the thought thought of of killing a cow slaughtered thousands of of Muslims in the early 1950s (Luckenbill & & Sanders, 1977). slaughtered of acceptance and and rejection can change In any given society, the latitudes of rejection can over time. In America, for homosexual relationships between consenting for example, homosexual adults have become more accepted, while dueling, which was once de riguer in rejected. certain situations, is now rejected.

CHAPTER 23 23 CHAPTER

iNHiBrnoNS AND AND CONTROLS CONTROLS INmBmoNs

591 591

Not only do developing children learn the broad latitudes of accepted and rejected behavior in their cultures and subcultures, but with increasing years and sophistication they come to understand the subtleties within these latitudes. In the latitude of acceptance, certain behaviors are prescribed, prescribed, but others are actually preferred, while in the latitudes of rejection, some behaviors are proscribed but preferred, others are ate permitted (E. !. I. Megargee, 1973). These These distinctions, distinctions, too, too,vary vary from from place place to place and change over time. In recent years, political dissent has shifted from being proscribed to permitted in the USSR while the opposite trend has taken place in China. At the outset, children's behavior is guided and controlled by their caretakers, caretakers, but in time, whether it is the Code of Hammurabi, the Ten Commandments, or the Analects of Confucius, children learn what is regarded as right and wrong in their culture. How does this come about?

Conditioningvia viaPunishment. Punishment, With Withregard regardtotolearning learningthe thebasic basic"do's **do'sand and a. Conditioning a. don'ts" of the culture, there is broad general agreement that a system of rewards and punishments punishments for for "bad" "bad"behavior behaviorare arenecessary. necessary.Theorists Theorists for approved behavior and disagree regarding whether these contingencies alone are a sufficient explanation. sufficient "rewards" are used to promote and encourage appropriate Reinforcement or "rewards" behavior and to foster positive role models. Punishments, on the other hand, are used to discourage disapproved behavior. Since we are focusing on inhibitions and restraints, we will concentrate on punishment. proverbialbelief belief that thatpunishment punishmentinduces inducesinternal internal inhibi inhibiIt is, quite literally, aa proverbial (13:24), tions: according to the Book of Proverbs (13 : 24), "He that spareth his rod hateth son; but he that loveth him chasteneth him betimes," his son; betimes," while a Chinese proverb states, "Beat your child once a day. If you don't know why, he does" (Tripp, 1970, p. 759). Many, perhaps most, psychologists agree that punishment can foster internal inhibitions. Discussing how we develop inhibitions against aggression, for example, Dollard, DoUard, Doob, Miller, Mowrer, and Sears (1939, p. 33) wrote, "The basic variable that determines the degree to which any specific act of aggression will be inhibited appears to be anticipation of punishment. .. .. .. [T]he principle derives from the law of effect; those acts cease to occur which, in the past, have been followed by punishment." To this they added that injury to a love love object also constitutes constitutes punish punishment and that anticipation of failure is equivalent to anticipation ment and that anticipation of failure is equivalent to anticipation of of punishment punishment (Dollard et al., aI., 1939, 1939, p. 34). There is considerable popular support for these "common sense" notions. Since the days of Jeremy Bentham (1748-1832), modern penology has been based on the theory that properly administered punishment deters crime; "specific deter deterrence" rence" means that once we have been punished we are less likely to repeat our transgressions, while "general deterrence" refers to inhibitions fostered in others who may observe our penalty. There is no doubt that the immediate prospect of punishment can suppress suppress behavior; witness how many cars reduce speed at the sight of aflashing flashing blue blue light. light.

592 592

EDWIN MEGARGEE EDWIN L L MEGARGEE

once they are out of But the rapidity with which they resume speeding once of sight of of the police suggests that no lasting inhibitions, as we have defined defined them, were fostered. Numerous studies attest to the fact that for for punishment to be an effective effective deterrent it must be swift, sure, and sufficiently sufficiently strong to outweigh the pleasures pleasures counderived from the sanctioned act (Wilson & Herrnstein, 1985). In totalitarian coun tries, a mere accusation may result in immediate execution, execution, a policy policy which one one would suppose would be very effective effective in suppressing unwanted unwanted behavior. However, effective we should recall that in World War II, the French mounted an extremely effective underground resistance against the German Occupation Occupation forces despite the fact that Gestapo and SS killed tens of of thousands of of French civilians in reprisal. the Gestapo In the United States, many advocate advocate increasing the severity severity of of punishment to reduce crime. As A s long as less than 2% 2% of of the crimes committed result in imprison imprisonafter the offense, offense, simply ment, and sentences are typically imposed months or years after increasing sentences will have little impact. But what of of the developing child? How H o w well does the conditioned conditioned anticipation of of punishment account for socialization? Behaviorists, particular the so-called "radi ''radifinding fundamental fundamental laws governing governing the relation relationcal behaviorists," concentrate on finding stimuU and responses, laws that apply to to ship between observable events such as stimuli intervening variables variables or hypothetical all organisms and which should not depend on intervening constructs such as "traits" **traits" or "cognitions" (Skinner, 1971). Since punishment ac accounts for avoidance behavior in rats and pigeons as well as people, it is tempting tempting to use it, coupled with rewards for appropriate behavior, as a sufficient sufficient explanation for human ethical judgment. How H o w adequate is this explanation? Let us return to Dollard DoUard et al.'s (1939) example of of aggressive behavior. Over the years, I have developed developed and refined a theoretical theoretical framework framework for the analysis of of aggressive behavior behavior in which I balance the the factors factors favoring favoring an an aggressive aggressive response response against against those those that that oppose oppose it it (E. 1. I. Megargee, 1993). If Megargee, 1982, 1982,1993). If the the factors factors opposing opposing it it are are stronger, stronger, then then that that response response blocked, but stronger, then that response response is is is blocked, but if if the the motivating motivating factors factors are are stronger, then that is possible. possible. The instigation") and and the fact that The motivating motivating factors factors include include anger anger ("intrinsic ("intrinsic instigation") the fact that aggression aggression may may be be aa means means to to some some end end ("extrinsic ("extrinsic instigation"). instigation"). (As (As AI A l Capone Capone once noted, "You word and than you once noted, "You can can get get much much farther farther with with aa kind kind word and aa gun gun than you can can with a kind word alone" [Peter, 1977, p. 141].) In addition, I include habit strength, with a kind word alone" [Peter, 1977, p. 141].) In addition, I include habit strength, which which comes comes from from having having been been rewarded rewarded for for aggressive aggressive behavior behavior in in the the past. past. Bal Balanced against the elements which increase the likelihood of an aggressive anced against the elements which increase the Ukelihood of an aggressive response response are are those those that that inhibit inhibit aggression, aggression, both both internal internal (conscience) (conscience) and and external external (the (the pres presence ence of of parent parent or or aa policeman). policeman). Dollard aggressive behavior DoUard et et al. al. (1939) (1939) argued argued that that punishment punishment for for aggressive behavior creates creates internal inhibitions. inhibitions. As anxiety anxiety about about performing performing forbidden forbidden acts acts that that foster foster internal A s we we have have noted, as noted, as long long as punishment punishment is is swift swift and and certain certain it it can can be be effective, effective, but but few few parents parents are are all-knowing all-knowing or or ever-present. ever-present. Sometimes Sometimes the the child's child's misbehavior misbehavior may may be other times be punished, punished, but but other times not. not. In In the the absence absence of of punishment, punishment, aggression aggression will will be be rewarded, so gradually its habit strength increases. Moreover, Moreover, since it was learned learned via partial reinforcement, the aggressive behavior will be very resistant resistant to extinction.

CHAPTER 23 23 INlHBmONS iNmBmoNsAND ANDCONTROlS CONTROLS CHAPTER

593 593

The most likely outcome is a discrimination will be learned rather than a moral absolute; for example, only hit your sibling when your mother is not looking. From this standpoint, it seems unlikely that externally imposed rewards and punishments sufficient to account for our elaborate rules regulating the expression of of aggres aggresare sufficient sion, much less our overall moral codes.

b. Social Learning. Learning, Unlike radical behaviorists, social learning theorists are different from other animals (Feshbach & Weiner, willing to treat people as being different 1986). While not denying denying the importance of direct rewards and punishments, Band Bandphenomena resulting from ura (1969, p. 118) argued that "virtually all learning phenomena direct experience can occur on a vicarious basis through observation of of other persons' behaviors and its consequences to them." In kindergarten, if the first child disciphned, the rest of of his classmates quickly learned to throw a spitball was severely disciplined, behavior—at least when the teacher was looking. to inhibit this behavior-at modeUng and observational learning do not account for the broad range Even modeling of rules and prohibitions that we acquire. Cognitive Cognitive social learning theory goes of beyond personal experience and observation and emphasizes the importance of stating the rules and stipulating the behaviors that will be rewarded or punished (Feshbach & Weiner, 1986). Moreover, cognitive social learning theorists stipulate that the child is not dependent on externally externally imposed rewards and punishments. Once they have incorporated values, they can and do reward themselves when they behave appropriately appropriately (Bandura, 1977) or feel badly when they do wrong or fail to live up to expectations (Feshbach & Weiner, 1986, p. 149f). This leads us to our next question: Why is it that people adopt these value systems and try to live by them? 2. Acquiring jecting Values Acquiring Controls Controls and and lntro Introjecting Values

Given the fact that it is impossible to have an external reinforcer watching our every move, it is essential that we develop internal systems of of control. But why is it that people come to reward themselves for doing well ("self-efficacy" ("self-efficacy" in Bandura's [1977] terminology), or punish themselves for doing wrong ("guilt" in the jargon of both clergymen and psychologists). of This question has intrigued theologians and philosophers for centuries. Some religions maintain that some form of direct intervention by the deity is responsible form of for people developing consciences. Philosophers such as Thomas Hobbes, David Hume, John Locke, and Jean Rousseau developed the theory of of the "social con contract" in which people agreed to trade the individual freedoms they enjoyed under anarchy for the security of of an organized and lawful lawful society. Psychologists tend to ascribe the development development of values and morality to events that take place within the family family in early childhood which make children want to please their parents. However, their specific specific explanations differ.

a. Behaviorist Watsonian tradition tradition attribattrib Behaviorist Explanations. Explanations. Behaviorists Behaviorists in in the the Watsonian uted children's identification identification with their parents to classical conditioning. While the

594 594

EDWIN L EDWIN I. MEGARGEE MEGARGEE

infant is suckling, the mother's presence becomes associated with the reduction of hunger and with all sorts of of pleasurable sensations. Through conditioning, she becomes a "secondary reinforcer" whose presence and approval are sought in their own right. To gain this reinforcement, the child learns to please her and live up to of the father. Even when her expectations, and, presumably to a lesser extent, those of the children have grown to adulthood and become parents themselves, they may still evaluate their behavior according to whether it measures up to their mothers' values.

b. Social Social Learning Learning Theory. Social learning theorists go beyond simple op operant conditioning to explain the acquisition of complex behavior behavior patterns (Bandura, 1977; Mischel & Mischel, 1976). They make a distinction between the acquisition performance of of behavior. and the performance relaMany habits are first acquired by imitating models. Modeling requires a rela tionship between the child and the figure being imitated, although that relationship of the imitator. Teachers, peers, television characters, may exist only in the mind of athletes, and literary or religious figures may serve as examples, but in early child childhood, the parents are usually the primary models. Through imitation, very complex patterns of of behavior can be rapidly acquired, and no direct reinforcement reinforcement is regarded as necessary (Feshbach & Weiner, 1986, p. 147). This contrasts with the slow shaping of behavior through direct rewards and punishments that behavior theory says of is required. faciUtated by verbal processes, explicit As we have noted, social learning is facilitated rules, and explanations of contingencies. Once a behavior pattern is acquired and reinforced if it is to be maintained. The agents of of the child performs it, it must be reinforced acculturation, who may or may not be the original models, must reward the behavior pattern, or the child must find it intrinsically satisfying satisfying and enjoyable (Feshbach & Weiner, 1986). The social learning explanation is better at explaining how we learn "positive" behaviors restraints. It It is is easier behaviors than than it it is is at at accounting accounting for for inhibitions inhibitions and and restraints. easier to to imitate imitate something that is done than something that is not. Still, there are forceful forceful models for international hero for inhibitions inhibitions and and controls; controls; Martin Martin Luther Luther King King became became an an international hero and and exemplar self-control and exemplar for for thousands thousands by by advocating advocating and and exemplifying exemplifying self-control and nonvio nonviolence lence in in the the face face of of the the most most extreme extreme provocation. provocation. Psychoanalytic theory. Psychoanalytic theory preceded social learning c. Psychoanalytic theory by a half century, so Freud and his followers did not have the advantage of drawing on as rich a base of empirical research. Indeed, most of of their observations were made on men and women who had been raised in Europe during the sexually repressed Victorian period and who were seeking treatment for serious neuroses. Many manifested serious sexual conflicts dating back to early childhood. Today, it seems likely that many had been abused. felt no need As clinicians treating patients, Freud and other psychoanalysts felt to confine their theories to externally observable behavior. Instead their primary focus was on intrapsychic events as inferred from the verbal reports of of their patients

CHAPTER 23 23 CHAPTER

iNHiBrnoNS AND AND CONTROLS CONTROLS INlDBmoNs

595 595

during treatment. From this rather skewed sample, Freud formulated a comprehen comprehensive theory of personality that has shown amazing vitality over the decades. More than other approaches, psychoanalytic theory recognizes that stressful necessarily involved in moral moral and andethical ethicaldecisions, decisions, approach/avoidance conflicts are necessarily clashes between what we want to do and what we should do, or, in analytic jargon, between the demands of the id, which which is is concerned concerned only onlywith withhedonism hedonismand andoperates operates according to the pleasure principle, and "the strict super-ego, which lays down definite standards for .. .. .. conduct and, which, if those standards are not obeyed, punishes it with tense feelings of inferiority and of guilt" (Freud, 1933/1965, p. 78). According to Freud's formulation of the structure of personality, an aspect not typically included in other approaches, this rivalry is mediated by the ego, which operates according to the reality principle. As Freud wrote, "The poor ego . . . serves three severe masters and does what it can to bring their claims and demands into harmony with one another. These claims are always divergent and often seem incompatible. No wonder the ego fails so often in its task. Its three tyrannical p. 77). masters are the external world, the super-ego, and the id" id" (1933/1965, p. At age three or four, during the the "phallic stage" stage" of ofdevelopment, development,when whenchildren children sourceof of excitement excitement and andstimulation, stimulation,they theydevelop develop are focused on their genitals as aa source for an an exclusive relationship with their opposite sex parent and are a yearning for jealousy of of the the same same sex sex parent parent who who demands demands so much much of the other consumed with jealousy parent's time and attention. These feelings are frightening in their intensity and in their implications. During the "Oedipal conflict," a boy fears his presumably omnipotent father will discover the son's incestuous yearning for his mother and castrate him to take revenge and prevent their union. This fear is no doubt exacerbated by the sight of little girls, whose external organs indicate that something of this nature must have taken place (Munroe, 1955). Repressing his desire for the mother, the boy identifies with his father, vicariously obtaining satisfaction by striving to be like him in every possible way. This includes introjecting his values, thereby creating the super-ego (Munroe, 1955). Ukewise repress their desire for the Girls undergoing the "Electra complex" likewise father and come to identify with the mother. This explains why boys and girls develop gender-specific values. Since girls had nothing to fear from castration, Freud asserted that identification and super-ego formation are never quite complete in women, women,aanotion notionwhich whichHelen HelenBee Bee(1985, (1985,p.p.326) 326)castigates castigatesasas"totally "totallyunsupported unsupported by later research . . . . .. "" This introjection of values does not take place in a vacuum. As Freud wrote, in a passage that could easily have been written by a contemporary social learning theorist, "The child is brought up to a knowledge of his social duties by a system of loving rewards and punishments, he is taught that his security in life depends on his parents (and afterwards other people) loving him and on their being able to believe that he loves them" them" (1933/1965, p. 164). Quite apart from his hypotheses regarding the structure of personality, Freud is one of the few personality personaHty theorists who specifically acknowledged the important

S96 596

EDWIN MEGARGEE EDWIN L L MEGARGEE

role religion can play in human human behavior fostering and and maintaining behavior in general general and and in fostering maintaining internal maintained that internal inhibitions inhibitions in in particular. particular. Although Although he he himself himself maintained that "religion ''religion is is an an illusion" illusion" (Beck, (Beck, 1980, 1980, p. p. 679), 679), Freud Freud noted noted that that people people transfer transfer their their familial familial relations "unaltered into religion. Their parents' prohibitions relations and and values values "unaltered into their their religion. Their parents' prohibitions and and demands With the the help demands persists persists within within them them as as aa moral moral conscience. conscience. With help of of this this same same system world of men" (193311965, system ofrewards of rewards and and punishments, punishments, God God rules rules the the world of men" (1933/1965, p. p. 164). 164). Unlike virtually ignores ignores the Unlike Freud, Freud, contemporary contemporary American American psychology psychology virtually the influ influence of of the 19 19 current of religion religion on behavior. Surveying Surveying the indices of current introductory introductory psychology texts that happen happen to be on my shelves, I found only 22 had found only had any entry for "religion"; one mentioned mentioned religion as a coping response response on one page page and for "religion"; discussed religious conversions. Not one mentioned religion as a factor factor the other discussed mentioned religion influencing influencing values or ethical decision making. Nevertheless, for for many many people people relireli gious beliefs and practices play an important role in fostering fostering ethical behavior, and a religious conversion or loss of faith can greatly alter a person's values. Some Factors Factors In Influencing Development o off Inhibitions Inhibitions 3. Some fluencing the Development

Although theorists disagree on how we become socialized and learn learn to control our Although socialized and agreement in psychology psychology and other disciplines on the behavior, there is general agreement environmental conditions that are most apt to foster foster a stable sense of of values and environmental behavior in accordance accordance with those values. the ability to regulate one's behavior A A warm nurturing nurturing environment environment in which the children children form form a close bond bond of of affection and respect with their caregivers combined with fair, consistent discipline affection fair, consistent is most conducive to developing internal controls. The more the agents of of socializa socializaneighborhood, the church, and and the school, share tion, first the parents, and later the neighborhood, enforce a strong common set of of values, and the more they live up to and and enforce consistently exemplify exemplify these principles in their everyday everyday lives, the more likely it is consistently that the children will incorporate them (Hogan, 1973; Hogan, Hogan, Johnson, & Emler, that 1978; W. McCord 1985). 1978; McCord & J. McCord, 1959; 1959; Wilson & Herrnstein, 1985). On the other hand, as we shall see, parental absence and disharmony, On the other hand, as we shall see, parental absence and disharmony, inconsis inconsistency, models are are associated tency, rejection, rejection, abuse, abuse, and and poor poor role role models associated with with problems problems in in developing Glueck, 1950; developing values values and and controlling controlling behavior behavior (S. (S. Glueck Glueck & E. E. Glueck, 1950; W. W. McCord 1971; Miller, 1958; McCord & J. J. McCord, McCord, 1959; 1959; E. E. I. I. Megargee, Megargee, Parker, Parker, & Levine, Levine, 1971; Miller, 1958; Nye, 1958; Rosenquist & Megargee, 1969; Wilson & Herrnstein, 1985). 4. f Inhibitions 4. Other Other Sources Sources o of Inhibitions

In this section we have been concentrating concentrating on how we develop the sorts of of internal proper perspective, it should inhibitions that fit our definition. To gain proper should be noted of the factors that were excluded excluded from from this strict definition nevertheless that some of strict definition inhibitory function. function. serve an inhibitory of bad things happening happening as a result of some action. A A One is the prospect of result of against some illegal act such as insider trading person who has no moral scruples against deterred by the prospect of of a prison term term (Wilson (Wilson & Herrnstein, 1985). might be deterred DoUard et al. (1939) noted, the person may not to act act if if it Similarly, as Dollard may decide not

CHAPTER CHAPTER23 23 INmBmoNS INHIBITIONSAND ANDCONTROI3 CONTROLS

597 597

seemed likely that the scheme would fail in its objective or that the act might bring pain or disgrace to loved ones. These latter considerations involve calculating the risk or the odds that some something bad will happen. As we have noted, the certainty of punishment has long been been aa major major factor factor in in deterrence deterrence theory. theory. Recently Recently Don Don and and Steve Steve Gottfredson Gottfredson have have been been developing developing the the parallel parallel notion notion of of "stakes," "stakes," arguing arguing that that the the amount amount one one has has to to lose lose in in terms terms of of reputation, reputation, property, property, and and other other considerations considerations should should also also be be included included in in the the equation. equation. Recapping Recapping the the psychological psychological explanations explanations for for the the development development of of internal internal controls is evident controls and and inhibitions, inhibitions, it it is evident that that there there are are many many theories theories and and much much relevant relevant information, information, but but few few definitive definitive data. data. It It is is possible possible to to describe describe the the conditions conditions favor favorable development, but able to to and and the the stages stages of of moral moral development, but the the specific specific mechanisms mechanisms are are largely a matter for conjecture. Interestingly enough, many hypotheses about the obser"psychological" factors that foster the development of values were based on obser vations of people who were conspicuous for their lack of of controls or restraint. We will tum turn now to a discussion of of factors that inhibit the development of values and principles inhibitions and which may be used to diminish or overcome those ethical principles that have been acquired.

iNfflBinoNS V. OVERCOMING INlDBmONS off from the important developmental Sometimes preschool children take time off tasks of learning values and resolving their Oedipal conflicts to build towers with construct castles of sand. Doing so they may learn another of wooden blocks or to construct life's lessons—it lessons-it is much easier to tear down a structure than it is to build it. So, too, with ethical codes. Human nature is such that we are much more interested in moral lapses than moral moral triumphs. triumphs. Examine Examine the the offerings offerings in in your your television television viewing viewing guide guide or or the the titles titles in in the the VCR VCR rental rental store. store. How How many many deal deal with with the the lives lives of of saints saints and and how how many many with with sinners? sinners? Before Before you you blame blame the the low low taste taste of of the the "mass "mass audience," audience," conduct conduct a a similar similar survey survey on on your your bookshelf bookshelf and and see see what what you you selected. selected. If If the the titles titles dealing dealing with with human human failings failings prevail, prevail, do do not not be be distressed. distressed. It It has has always always been been thus. thus. While While you you are are surveying surveying your your bookshelf, bookshelf, take take down down the the Old Old Testament Testament and and tum turn to to the the second second and and third third chapters chapters of of Genesis. You You will will find find that that only only 33 verses verses are are used used to to describe describe Adam Adam and and Eve's Eve's life life together together in in the the Garden Garden of of Eden, Eden, but but 24 24 verses verses are are devoted devoted to to their their temptation temptation and and fall fall from from grace. grace. Social scientists are no exception; we are more likely to deal with the Cain's of of this world than the Abel's. Anthropologists are especially interested in those of the tribe that somehow fail to share its cultural values, sociologists members of have an entire subarea devoted to the study of "social deviance," and criminology is a discipline unto itself. The number of psychologist who deal with abnormal or deviant behavior vastly exceeds the number who focus on healthy functioning.

598

EDWIN L I. MEGARGEE

As we noted above, many of our theories about the origins of inhibitions stemmed from of from studies of people whose behavior is characterized by a lack of restraints. In general, the factors factors that are associated with a failure to develop adequate controls are the obverse of those conducive to positive socialization. Unfortunately it appears there are many more ways to diminish or overcome Unfortunately foster them. inhibitions than there are to foster

A. Physiological Mechanisms As we noted earlier, all behavior is physiologically mediated. Thus whether we from the unfolding of an regard internal prohibitions and controls as resulting from innate genetic pattern or as resulting from conditioning or learning, physiological factors must have an impact. factors 1. Mechanisms /• Genetic Mechanisms

personaUty researchers have obtained data In the previous section, we noted that personality consistent with the hypothesis that individual differences differences in controls and that are consistent differ regarding restraints are at least partially determined by heredity. Authorities differ differences the mechanisms. Eysenck (1964, 1981) maintained that these genetic differences differences in brain physiology, whereas Buss and are mediated through actual differences inherited are temperamental differences differences that pre prePlomin (1984) suggested what is inherited dispose people toward being more or less restrained (Hogan, 1986). influence If Eysenck's view is correct, then we might infer that factors which influence further on, might overcome the innate the central nervous system, as described further pattern. On the other hand, if it is predispositions that are innate, then experiential factors factors become more important in determining whether or not these predispositions are realized. 2. Central Central Nervous Nervous System System 2. defined them depend on the proper functioning functioning of the Internal controls as we have defined brain. In order to make an ethical decision, I must first examine a proposed course of action and decide whether to classify classify it as "right" or "wrong" according to my unique set of values; if it is "wrong," then then II must must decide decide whether whether or or not not II will willsuccumb succumb confronted by to the temptation anyway. Recall the existential crisis of the dieter confronted the chocolate mousse. for specific specific inhibitory inhibitory centers in the hypothalamus that govern consum consumExcept for conmatory behavior and, possibly, certain types of aggressive behavior, moral con straints and inhibitions seem to be cortical functions. It is, of course, the cerebral cortex that is the most recently evolved area of the human brain and the area functioning. associated with what we regard as "higher" functioning. As we shall soon see, all sorts of things can go wrong with the central nervous effects vary with the system in general and the cortex in particular. The specific effects nature and the location of the damage, but a safe rule of thumb is that while brain damage may impair internal inhibitions, it virtually never augments them. Indeed,

CHAPTER 23 23 CHAPTER

INmBmoNs iNfflBmoNS AND AND CONTROUi CONTROLS

599 599

impulsive behavior and diminished ethical sensitivity are often among the first behavioral symptoms of cortical malfunctioning. malfunctioning.

a. Traumas, Vascular Disorders. Disorders, Although Although itit isisencased encased in in the the Traumas, Tumors, and Vascular skull, the brain is subject to injury injury from from exogenous causes such as blows to the head or gunshot wounds. From within, cerebral vascular infarctions, aneurysms, arteriovenous malformations, and tumors can all create lesions. Diffuse Diffuse generalized cortical damage can be associated associated with a general lessening of ethical sensitivity, increased irritability, impulsive behavior, and impaired judg judgeffects of more focal lesions, such as those caused by tumors, wounds, ment. The effects and strokes, depend on the area that is damaged. Temporal lobe tumors and hypothalamic lesions are sometimes associated with aggressive acting out. b. Disease and and Infection. Infection, A A number number of of diseases diseases can can diminish diminish cortical cortical func funcb. Disease tioning. These include disorders which apparently have a genetic basis, such as disease. Pick's disease, and the senile psychoses, as well as infectious infectious Alzheimer's disease, Memory loss is is the the primary primary characteristics characteristics diseases such as encephalitis and syphilis. Memory former group, but with both the innate and the infectious disorders, cortical of the former impairment can be accompanied by a loss of ethical sensitivity and moral constraints. from cognitive impairment, since knowing the difference difference This may stem in part from between right and wrong is a cognitive function. But it also seems clear that there is a reduced capacity to control one's behavior.

ChemicalSubstances. Substances, Earlier, Earlier,we wenoted noted that that certain certain psychotropic psychotropic medica medicac. Chemical c. tions may be prescribed to assist neuropsychiatric patients in controlling their behavior. Such substances can be remarkably effective; without them the widespread deinstitutionalization de institutionalization of the mentally ill would not have been possible. However, of these patients to control their when these medications are withdrawn, the ability of behavior may decrease. Other drugs, most notably alcohol, act to diminish inhibitions. Indeed, this is one reason they are so widely used. As Ogden Nash wrote, "Candy is dandy, but liquor is quicker." 3. Endocrinological System System

The hormones secreted by the endocrinological system can have powerful powerful effects effects on behavior. When the sympathetic portion of the autonomic nervous system is effect is aroused, adrenalin is released and the body is prepared for for action. The effect similar to bringing a military unit to a full state of of combat readiness, with weapons off, and senses alert for for the first sign of enemy action. loaded and locked, safeties off, Given such a state of activation, people are prepared to respond instantaneously, and the influence of internal inhibitions is minimized. Hormones secreted by the gonads play an important role in stimulating sexual desire or libido, especially especially in males. males. The TheMiddle MiddleEastern Eastern potentates potentates who whohad hadeunuchs eunuchs understood the guard their harems may not have been endocrinologists, but they understood

EDWIN MEGARGEE EDWIN L L MEGARGEE

600 600

effects of of castration. Younger Younger readers are probably probably well aware aware of of how sexual arousal arousal effects can overcome moral prohibitions; older readers can probably probably remember. Testosterone Testosterone also stimulates aggressive behavior behavior and dominance dominance in a variety variety of species. Along with thyroxine and progesterone, excessive testosterone testosterone can cause of endocrinological system are irritability. The autonomic nervous system and the endocrinological complexly complexly intertwined with environmental environmental and personality personality factors; the point to remember is that these factors factors are among those that can mitigate internal inhibitions inhibitions against various forms forms of of behavior. behavior. against 4. Other Other Physiological Physiological Factors Factors 4.

A A number number of other physiological factors factors have been associated associated with diminished diminished inhibitions. Some have speculated speculated that psychopathy, psychopathy, which is characterized characterized by a underdeveloped set of of inhibitions, may have a physiological basis. Eysenck severely underdeveloped speculated that psychopaths psychopaths have an innate deficit deficit in their ability to be (1964) speculated conditioned or to learn from from their their mistakes so that punishment punishment is relatively ineffec ineffecconditioned McCord and W. McCord McCord (1964) speculated that brain damage, possibly to tive. J. McCord the hypothalamus, in combination combination with with parental rejection might be responsible. responsible. the hypothalamus, in parental rejection might be Quay (1965, p. p. 181) suggested suggested that that perhaps perhaps "basal "basal reactivity to stimulation stimulation is is lowered lowered reactivity to Quay so that that more more sensory sensory input input is is needed needed to to produce produce efficient efficient and and subjectively subjectively pleasur pleasurso able cortical cortical functioning." functioning." Because of this this presumably presumably innate deficit, the the psychopath psychopath able Because of innate deficit, is driven driven to to seeking seeking additional additional sensory sensory stimulation. stimulation. is We should also note that physical illness can reduce external inhibitions. A A person who has a terminal illness may feel feel that he or she has "nothing "nothing to lose" person would not otherwise have allowed themselves. and engage in behavior that they would something antisocial or reprehensi reprehensiThis does not necessarily mean that they will do something For example, one hardworking individual who had never allowed allowed himself himself to to ble. For take a vacation put · work aside and went on a cruise when he learned he was suffering from from an untreatable life-threatening life-threatening illness. suffering Psychological Fadon Factors B. Psychological from the many physical factors than can diminish inhibitions, we will find fluid Turning from Before that there is an even more varied array of "psychological" mechanisms. Before anthropologists, sociologists, psychiatrists, or theologians take umbrage, let me stipulate that the term "psychological" is being used in its broadest broadest sense hasten to stipulate to mean "nonphysiological." As we shall see, theorists theorists from from a variety of of disciplines have contributed contributed a heterogeneous array array of of explanations. Surveying these notions, it appears that by and large theorists have been appear addressing two distinct issues. The first is why some people in every society appear to have values that diff er from differ from that society's norms. The second is more concerned concerned with determining determining why so many of of us do not live up to our respective respective codes of of turn. values. We shall discuss each issue in tum.

CHAPTER 23

iNHntrnoNS AND CONTROLS AND CONTROL'l INmBrnoNs 1. 1.

601 601

Problems in Value Development

Almost everybody who has a reasonably adequate central nervous system develops some code of values. Whether or not this code is adequate is, in the truest sense of the word, a value judgment. Nevertheless, in virtually every society throughout history there have been some individuals whose values were deemed inadequate by their fellow citizens. How can this come about?

a. Deficient Values. Values, In Ingeneral, general, the thefactors factors associated associated with withdefective defective value value a. Deficient development are the obverse of those conducive to good socialization. The most serious deficiencies are likely to be observed in children reared in situations which prevented basic bonding, the resolution of thefirst first developmental crisis described by Erik Erikson (1950) as "trust vs. mistrust." This might occur, for example, in children growing up under conditions of extreme deprivation such as the famine famineravaged areas of the Sahara or the war-torn sections of Lebanon. Children raised in totally impersonal institutions or in homes characterized by severe rejection and abuse might also be included, but probably to a somewhat lesser extent. Studying the development of juvenile delinquency some years ago, Sheldon and Eleanor Glueck (1950) noted that a lack of cohesiveness and nurturance in the parental home was associated with delinquency, a finding finding that has oft been & J. McCord, 1959; 1959; Rosenquist & & repeated in a variety of cultures (W. McCord & 1969; Wilson Wilson & &Herrnstein, Hermstein, 1985). 1985). Coupled Coupled with with this this was wasinappropriate inappropriate Megargee, 1969; discipline; that is, discipline that is either lax or excessive, inconsistent or unfair. if chastisement or punishment helps condition values, erratic schedules Obviously, if of negative reinforcement will interfere with such learning. age, familial situations situations that thatinterfere interferewith withthe the process processof ofidentification identification At a later age, and introjection are detrimental to value formation. In broken homes, the process of identification might be subverted by parental absence or by the efforts of one parent to diminish the other in the child's estimation (Bee, 1985, p. 332f). Writing from a psychoanalytic perspective, Adelaide Johnson (1949) noted that some apparently well-socialized parents might obtain unconscious gratification from their children's acting out. In a family the writer was seeing in therapy, the father, who was overtly outraged over his son's auto thefts, was noted to whisper, "Gee that took real guts," guts," when his son described the high-speed police chase that had had ensued. Of Of course course the the father father vehemently vehemently denied denied his hissotto sottovoce voceremark. remark.Such Suchdual dual messages, according to toJohnson Johnson (1949), (1949),lead leadto towhat whatshe she termed termed"superego "superegolacunae." lacunae." Previously we noted that value development is abetted by growing up in a milieu in which all the agents of socialization work together to foster a consistent set of values. Obviously, a situation in which this is not the case, in which the child is exposed to differing values or in which adults say one thing and do another, is less conducive to good moral development. These conditions may well yield values differ from those prescribed by the larger society. We shall now turn to a that differ discussion of such deviance.

602 (02

EDWIN EDWIN L L MEGARGEE MEGARGEE

b. fe. Deviant Deviant Values. We often think of of people who engage in socially repre reprehensible behavior as being immoral or amoral. The problem, however, may not be inadequate values but values that differ differ from those of the society at large or from the laws of that society. Sociologists interested in deviance have been especially of ways that it might interested in this phenomenon and have proposed a number of come about. Within a heterogeneous heterogeneous country like the United States, we will find many subgroups and subcultures with somewhat differing views of what constitutes accept acceptable and unacceptable unacceptable behavior. It is not surprising that cultural conflict conflict was one of of the first explanations offered offered for deviant values (Sellin, 1938). While these subcul subcultural gaps were most evident in the United States when immigration was at its peak, of communication appear to have lessened the disparities somewhat mass means of (Rosenquist & Megargee, 1969; Velez-Diaz & Megargee, 1971). In Israel, however, some scholars have attributed deviant behavior to cultural conflicts conflicts between Euro European and Sephardic immigrants (Shoham, 1962). A number of sociologists have pointed to "anomie" **anomie" or normlessness as a A factor in producing deviant values. Robert Merton (1938, 1957) noted that the factor Horatio Alger myth requires that everyone, no matter what his or her prospects, should strive for status and material success. However, we are also supposed to be conflicting honest and upright. Many people may have to choose between these conflicting values, because, given their circumstances and abilities, there is no realistic way for them to do both. As Cloward and Ohlin (1960, p. 86) described it, "Faced with of access of of these goals, and unable to revise limitations on legitimate avenues of their aspirations downward, they experience intense frustration; the exploration of nonconformist alternatives may be the result." nonconformist This "differential "differential opportunity" or "strain" theory dovetails neatly with Earnest Sutherland's (1939) "differential "differential association" theory which emphasizes the influence of of deviant role models. In an urban ghetto, the role models for success are people who made it are rarely rarely people who made it from from the the streets streets to to the the corporate corporate boardrooms; boardrooms; instead instead is the pimps and pushers pushers with their gold chains and expensive cars who are it is conspicuous. period of conspicuous. If, If, during during this this period of exploring exploring "nonconformist "nonconformist alternatives," alternatives," youths youths are 1959) or are recruited recruited by by aa gang gang (Salisbury, (Salisbury, 1959) or have have the the opportunity opportunity to to serve serve as as runners runners for for neighborhood neighborhood crack crack dealers, dealers, they they may may be be inducted inducted into into aa very very lucrative lucrative life life of of crime values which crime while while still still very very young young and and become become socialized socialized with with street street values which are are antithetical of the antithetical to to the the moral moral codes codes of the larger larger society. society. Deviant subcultures are not only found in the streets and ghettos; they also exist at the upper end of of the financial and political ladder (Clinard & Quinney, Void, 1958; VoId Void & Bernard, 1986). Faced with a strain between ethical and 1973; VoId, legal restrictions and a get-rich-quick mentality, some Wall Street brokers recently adopted a deviant set of of values and made money the truly old fashioned way-they way—they stole it. A A strain between ends and means can occur in the political arena as well. Of those convicted convicted of of wrongdoing wrongdoing in connection connection with the Watergate Watergate break-in and the Iran-Contra affair, both G. Gordon Liddy and Oliver North appeared to have

CHAPTER CHAPTER 23 23 INlDBmoNS INHIBITIONSAND ANDCONTROU CONTROLS

603 603

strong, well-developed, but deviant value systems which dictated that they should engage in illegal behavior to accomplish goals that they felt felt were more important than abiding by the law. Although other theorists have also discussed deviant subcultures and culture conflict, let us turn to the interactionist or labeling perspective as an explanation Garfinkel (1956), Lemert for the development of deviant values. Becker (1964), Garfinkel (1967), and other interactionists maintained that society creates social deviance by thereby labeling labeUng formulating rules and applying sanctions to people who break them, thereby them as deviants. According to Garfinkel (1956), one consequence of of a "degradation ceremony" such as suspension from school, a criminal conviction, or a commitment to a mental hospital is that the stigmatized individual may accept the label and adopt deviant values consistent with this new identity. Thus a person who is regarded as being "immoral," "crazy," or "bad" "bad" in some respect may start associating with other Of course this people who are similarly labeled and emulating their behavior. Of convinces the labelers of of the correctness of their initial appraisal, and additional stigmatization may be applied that solidifies the deviant self-concept. self-concept. Although early childhood is when our basic values are formed, people continue reinforcement to learn and to develop throughout their lives. Circumstances and reinforcement schedules may change, so the values we learned as a child may not equip us for the challenges we face as adults. A youthful idealist may find that the Golden Rule of business and that Charles Dickens does not work well in the competitive world of was nearer to the mark when he wrote, "Do other men for they would do you. That is the true business precept" (Beck, 1980, p. 547). Military training is designed Psychoin part to help personnel overcome the taboo against killing other humans. Psycho therapy may be required to help adults overcome strong sexual inhibitions ingrained into them as children. Overcoming Controls Controls 2. Overcoming without having acquired a code It is virtually impossible for you to reach adulthood without of of values. It may not agree with everyone else's code of of values, and it may not anyone else's code, but it is yours and you will usually try, with even agree with anyone greater or lesser success, to abide by it. Whether you succeed depends on your ability to control your behavior. Internal controls operate only when we are tempted to do something that is contrary to our code of of values. If there was no temptation, there would be no need for restraint. Scruples get us involved in internal conflicts. These conflicts conflicts can be stressful unfortunate effect effect of of either and occasionally anxiety provoking. They also have the unfortunate preventing us from doing things we would like to do, like eating the chocolate mousse when we are dieting, or making us feel guilty if we do succumb. Given of us have devised ingenious ways these circumstances, it is not surprising that most of of overcoming our moral inhibitions.

604 6M

EDWIN I. MEGARGEE EDWIN MEGARGEE

a. Rationalization. Rationalization. Rationalization can be used to justify acts and thereby effective circumvent the injunctions against them. Rationalizations are especially effective in situations in which the moral boundaries are fuzzy. As we have noted, we must first categorize an act as belonging to that class of of acts we regard as "wrong" before moral prohibitions come into play. In the current policy debates over capital punishment and abortion, all the participants classify executions andlor and/or agree that murder is wrong. The problem is that some classify abortions as murder and others do not. Rationalization can be used to convince us that an act that appears to be wrong actually is not. This makes it permissible. For example, most politicians probably would agree that it is wrong for public officials to accept bribes. But some legislators might reason that if they have already decided to vote for a piece of legislation it does no harm to accept a contribution from a contractor who will official might reason that turning down the benefit from the project. Indeed, the official tantamount to denying contractors their rights to participate contribution would be tantamount Amendment! in the political process. Why, it might even be contrary to the First Amendment! One common form of of rationalization is to concede the general principle, but to classify classify the present case as an exception to the general rule. The key word "but" is a good sign of of this sort of of rationalization: "I know she said 'no,' *no,' but she really didn't mean it"; "Sure, dealing drugs is bad, but if I don't sell them, someone else will and they might sell my customers bad stuff." As the old country preacher stated, "A lot of of sinners slide into Hell on their 'buts.' "

b. 6. Value Conflict. Conflict When two or more values conflict, they tend to neutralize one another (Sykes & Matza, 1957). If If the values are not deeply held or evenly balanced, this does not pose a great conflict. Indeed we may be able to use this conflict conflict to allow us to do what we want without feeling gUilty. guilty. (After all, if if the forbidden behavior was not fun, we would not be tempted in the first place.) People can also use value conflicts to manipulate others into abandoning their scruples. Thus, a dieter who might refrain from ordering a piece of of cake at a restaurant may acquiesce at a wedding if the bride and groom insist that they will be offended offended by a refusal. Maintaining their friendship seems more important than avoiding calories. Of course, the more one wants to do the forbidden act, the more effective effective value conflicts are in overcoming scruples. The basic issue in many values conflicts is whether the end justifies the means. Politicians may feel that they need to get elected for the good of the country, even even if it requires negative campaigning. The principles in the Iran-Contra affair felt felt that maintaining the security of of their convert operation justified lying to Congress. Hogan's (1970, 1973) theory of moral behavior postulates an ethical continuum (1970,1973) from considerations of "personal conscience" to "social responsibility" that can different solutions to certain ethical dilemmas such as whether one should dictate different do something one regards as personally wrong to benefit the overall social group. of Shakespeare's tragedy Julius Caesar in Such a moral dilemma is at the heart of After which the cunning Cassius used Brutus' patriotism to turn him against Caesar. After

CHAPTER CHAPTER 23 23 INmBmoNs bsHiBrnoNSAND ANDCONTROLS CONTROLS

60S 60S

the assassination, an agonized Brutus attributed his participation to value conflict, explaining, "Not that I loved Caesar less, but that I loved Rome more" (III,ii,22). As Antony noted, "Brutus is an honorable man; So are they all, all honorable men" (III,ii,88). When the competing values involve deeply deeply held convictions, we are placed in a double approach-avoidance approach-avoidance conflict, the type that creates the most stress and conflict is resolved by attempting attempting a compromise. Suppose anxiety. Sometimes such conflict that a young man's buddies ask his help in robbing a store. Which is worse, disloyalty steaUng? He may offer to help by being the lookout or driving the get-away car or stealing? but not actually going into the store.

c. Suspension Suspension of of Individual Individual Values. In the wake of of World War II, a number of social psychologists began investigating how the Holocaust could occur. Studies of authority showed by Asch (1956) on conformity and Milgram (1974) on obedience to authority that people will often often suspend their individual values and instead let others dictate their behavior, even when they feel it is wrong. Similarly, studies of of bystander intervention by Darley and Latane (1968) demonstrated that people who would be inclined to assist someone someone in distress refrain from doing so when there are other people present who are not helping. Suspension of individual values is particularly strong in a group setting. Studies of group decision making have demonstrated that the decisions made by groups are apt to be riskier or more extreme than the decisions made by the individual participants. Janis' (1972) research on the phenomenon he dubbed "groupthink" of Pigs invasion showed that, to maintain a consensus, a group following the Bay of of people will agree with decisions that they individually think are incorrect and/ of and! or morally wrong. in Anonymity assists in this deindividuation. One of of the first major steps in of laws curbing the power of the Ku Klux Klan in the South was the passage of forbidding people to wear masks in public. From Kent, Ohio, to Beijing, China, it is easier to open fire on unarmed students if if you are an anonymous soldier who is "following orders" rather than an individual dressed in civilian clothes acting on your own. d. d, Dehumanization Dehumanization of of the Victim. Victim, The more empathy we have for someone, the more difficult difficult it is for us to hurt or injure them, physically physically or emotionally. Our appreciation of that person's humanity serves to activate our internal inhibitions against doing wrong to our fellows. By the same token, anything that dehumanizes a potential victim makes it easier for us to suspend these values. Oddly, it seems less reprehensible for someone of thousands of to give an order that may result in the deaths of of strangers in a far away land than it is for that person to strangle a single individual. Some kidnapped in tents or covered their heads hostages have reported that their captors kept them in development of human ties that might inhibit the captors with bags to prevent the development from killing their victims.

606 606

EDWIN EDWIN I. L MEGARGEE MEGARGEE

Racial, class, and Racial, religious, religious, social social class, and gender gender differences differences increase increase the the emotional emotional distance distance between between people people and and decrease decrease their their irihibitions inhibitions against against harming harming one one another. another. The The ordinary ordinary German German was was better better able able to to tolerate tolerate the the Holocaust Holocaust as as long long as as it it was was happening happening to to the the Jews, Jews, the the Gypsies, Gypsies, and and the the mentally mentally defective, defective, just just as as the the ordinary ordinary American AIDS when American was was less less concerned concerned about about AIDS when it it appeared appeared to to be be aa disease disease confined confined to to homosexuals homosexuals and and drug drug addicts. addicts. In In preparation preparation for for combat, combat, it it is is aa standard standard tactic tactic to to tell tell the the troops troops how how different different and and reprehensible reprehensible the the enemy enemy forces forces are are in in order order to diminish inhibitions that that might might hinder hinder their their effectiveness. effectiveness. diminish inhibitions

e. Psychopathology. Functional Psychopathology, Functional as as well well as as organic organic mental mental disorders disorders can can also also decrease Daniel M'Naghten's decrease inhibitions inhibitions and and controls. controls. Daniel M'Naghten's inhibitions inhibitions against against shooting shooting Sir Sir Robert Robert Peel Peel were were overcome overcome by by his his paranoid paranoid delusion delusion that that Peel Peel was was persecuting persecuting him. Actually M'Naghten belief that he was acting in M'Naghten shot the wrong man, but his belief self-defense led to his acquittal and the formation formation of of the M'Naghten M'Naghten test of of legal self-defense sanity in 1843. Other disorders can also lead to a diminution diminution of of inhibitions. A A profoundly profoundly Other depressed person may feel there is nothing to live for for so that ordinary inhibitions inhibitions ineffective. Extreme Extreme guilt may lead to self-punitive self-punitive or suicidal behavior behavior that are ineffective. psychogenic fugue fugue or multiple personality, a would otherwise be inhibited. In a psychogenic repressed behavior behavior patterns patterns without without being aware of of person may act out various repressed doing so. Although certain disorders can lead lead to diminished diminished controls, it would be an Although error to equate mental disorders with uncontrolled uncontrolled and and possibly dangerous behavior. behavior. error A catatonic catatonic stupor, stupor, for for example, example, is is perhaps perhaps the the ultimate ultimate in in inhibited inhibited behavior. behavior. A

VI. MEmODOLOGICAL VI. METHODOLOGICAL PROBLEMS In the the preceding preceding pages pages we we reviewed reviewed aa diverse diverse broad broad array array of of theories theories regarding regarding In the origins origins of of internal internal inhibitions and how how such such inhibitions inhibitions may may be be overcome. overcome. As As the inhibitions and scientists, most most psychologists psychologists would would prefer prefer to to test test divergent divergent theories theories empirically. empirically. scientists, Unfortunately, number of of conceptual, ethical, and methodological problems make Unfortunately, a number difficult to conduct conduct definitive definitive experiments on controls and inhibitions. it difficult A. Acquisition Acquisition of of Values differing theories on the acquisition of of values presents us with our most Testing differing difficult challenge. Children Children obviously cannot be randomly assigned to various difficult of child rearing to test the effects effects of of various disciplinary disciplinary practices or to patterns of different families families to determine the effects effects of of single-parent single-parent homes. Although Although some different Harlow's (1971) studies of of monkeys reared reared by wire wire animal experiments, such as Harlow's relevant to issues in child development, development, by and large large mesh surrogate mothers, are relevant infrahuman subjects are not suitable for for research on the development development of of values. infrahuman

CHAPTER 23 23 CHAPTER

iNHiBmoNS AND AND CONTROLS CONTROLS INlDBmONS

607 607

the absence absence of of experimentation, experimentation, we we have have to to resort resort to correlational correlational methods, methods, In the nature." Since Since such such studies studiesare areinevi inevinaturalistic observations, and "experiments of nature." tably confounded, we must adopt a variety of strategies and use many different subject populations in the hope that the variables that confound one study will be absent in the next. Since values are culture specific, cross cultural research is vitally important. Gradually over time, a core of reliable associations should emerge. As indicated earlier, there is general agreement that a warm, stable family setting is most conducive to developing values and controls, at least in our culture. Oddly enough, much of this consensus has come about through research on people who appear to have deficient values and/or inadequate controls, namely, juvenile delinquents and adult criminals. When examined more closely, even these studstud ies show some cultural specificity. Comparing the family patterns associated with delinquency in three different cultures, Rosenquist and E. I. Megargee (1969) Mexicannoted differences in the effects of the father-son relationship in Mexican, Mexican American, and Anglo-American families. More research is needed on the antecedents of positive socialization in a variety of cultures. While longitudinal research is desirable, other approaches can also be used. By identifying people differing in socialization, we can test hypotheses about their upbringing and antecedents. For example, we found that middle class college students who differed in socialization, as measured by the Socialization scale of the California Psychological Inventory (Gough, 1969), came, as predicted, and cohesiveness (E. (E. I.I. Megargee Megargee from families that differed significantly in stability and al., 1971). But would the stability of the nuclear family be as important in a et aI., culture that relied on communal child rearing? This is an empirical question that needs to be answered before we overgeneralize from our own society. One of the bright bright spots spotsin inthe thearea areaofofacquisition acquisitionofofvalues valueshas hasbeen beenthe theresearch research on moral development by psychologists psychologists such such as asPiaget, Piaget,Selman, Selman,and andKohlberg. Kohlberg.Their Their studies have shown that children in Western countries progress through a series of distinct stages in their moral judgments. Are these stages universal or specific to Western culture? One test of moral judgment (Kohlberg, 1981) creates hypothetical ethical dilemmas by juxtaposing carefully chosen antithetical values. In one oft-cited example, the respondent is asked to decide whether "Heinz" was wasright right or or wrong wrong when when he he stole stole an an exorbitantly exorbitantly priced medicine he could not afford to purchase in order to save his dying wife. Western children "progress" "progress" from answering that Heinz was wrong because he broke the law to responding that human values supersede property values or laws. laws. But is this sequence culturally universal? What would be found in a country such as North North Korea Korea in inwhich whichpositive positivesocialization socializationconsisted consistedof ofunquestioning unquestioningobedience obedience to the state and loyalty to the late Marshall Kim II Sung, known and revered as the "Great Leader"? As we noted, the physiological foundations for socialization and restraints remain largely unexplored. Perhaps the paradigms devised by developmental psypsy chologists could be used to investigate the role of the central nervous system by

608

EDWIN L MEGARGEE

applying them to to people with various various types types of of CNS CNS impairment impairment or or developapplying them people with develop mental mental disabilities. Comparative Comparative anthropological anthropological and and ethnographic ethnographic studies studies including including data data on child-rearing patterns patterns and and values values can can be be used used to to help help test test the the generality generality of of our our child-rearing observations. Social Social histories histories and and treatises treatises on on child child rearing rearing from from various various eras eras and observations. and cultures also also provide provide aa useful useful perspective perspective on on our our ethnocentric ethnocentric assumptions. assumptions. Fortu Fortucultures nately data collection is is easy easy and and inexpensive, inexpensive, involving involving only only aa trip trip to to the the library, library, nately data collection with, stop at at the the Anthropology Anthropology Department for aa quick quick consult. consult. Aries with, perhaps, perhaps, aa stop Department for Aries and Duby's Duby's histories histories of of private private life life in in ancient ancient Rome, Rome, Byzantium, Byzantium, and and medieval medieval and (1934) and and Mead's Mead's(1961, (1961,1975) anthropological Europe (1987-1988) and and Benedict's Benedict's (1934) Europe (1987-1988) 1975) anthropological observations are are good good starting starting points. observations points. As As we we have have noted, noted, studies studies of of people people with with apparently apparently deficient deficient socialization socialization have have suggested suggested that that certain certain patterns of of living living and and early early life life experiences experiences are are crucial to moral development. development. These These hypotheses hypotheses can can be be partly partly tested tested by by prospective prospective studies studies in which which people people who who do do and and do do not not have have these deficits are identified, predictions predictions in these deficits are identified, are made, and then then the thesubjects subjects are arefollowed followed up upover overtime timetoto determine determinewhether whether made, and the predicted predicted patterns patterns emerge. emerge. This This approach, approach, which which is is much much more more powerful powerful than than the the more more common common retrospective retrospective design, design, can can be be used used to to explore explore the the sequelae sequelae of the of such factors factors as as parental parental absence absence due due to to death death or or dissension, dissension, of of being being exposed exposed at such at an early early age age to to differing differing sets sets of of values values and and mores, mores, and and even even the the long-term long-term effects effects an of severe and abuse. of severe deprivation deprivation and abuse. B. Restraint B. Exercising Exercising Control Control or or Restraint It considerably easier on It is is considerably easier to to conduct conduct empirical empirical studies, studies, especially especially experiments, experiments, on when and whether people choose to behave according to their values than it is on how their values developed. Indeed, researchers in a variety of allied areas have already accumulated a number of relevant empirical findings even though they were primarily interested in studying different different phenomena. As we have noted, social psychologists psychologists interested interested in in conformity, conformity, obedience obedience to to authority, authority,group groupdecision decision making, and bystander intervention have contributed considerable information on those those situations situations in in which which people people do do and and do do not not behave behave in in accordance accordance with with their their values. values. Similarly, Similarly, clinical clinical psychologists psychologists studying studying the the effects effects of of alcohol alcohol on on aggression aggression and and other other social social behaviors behaviors have have provided provided us us with with relevant relevant findings findings and, and, perhaps perhaps more of more important, important, research research designs designs that that can can be be used used to to study study directly directly the the effects effects of alcohol alcohol and and other other substances substances on on controls controls and and restraints. restraints. Still, Still, there there are are numerous numerousproblems. problems.First Firstand andforemost foremostisisthe theproblem problemwe wecited cited at at the the beginning beginning of of this this chapter, chapter, namely, namely, the the fact fact that that inhibitions inhibitions and and controls controls are are negative concepts concepts that must be be defined defined by by exclusion. exclusion. How How can can one one be be certain certain that negative that must that inhibitions prevented prevented certain certain expected expected behaviors behaviors from from occurring? occurring? Researchers Researchers may may inhibitions have to to induce induce the the appropriate appropriate motivational motivational state state at at the the outset outset of of the the experiment, experiment, have include aa manipulation manipulation check check to to make make sure sure the the procedure procedure was was effective, effective, and and go include go through aa variety variety of of contortions contortions to to make make sure sure that that various various alternative alternative explanations explanations through for restrained restrained behavior behavior are are eliminated. eliminated. for

CHAFfER CHAPTER 23 23

iNHiBmoNS AND AND CONTROlS CONTROLS iNHIBmoNs

609 609

Fortunately, these sorts of dilemmas typically involve approach-avoidance conflicts and there may be verbal utterances and other indications of the existential struggles that are taking place. ("Oh, it looks so good but I really shouldn't. Are you positive that the Black Forest tort doesn't have any calories?") Similarly, long response latencies, vacillation, and signs of anxiety or gUilt guilt may help the observer to infer that inhibitions or constraints are operating. Problems in studying controls and restraints are compounded by the fact that many many taboos taboos are are very very specific specific with with respect respect to to situations situations and and targets. targets. The The classic classic research by Hartshorne and May (1928), for example, found that children cheat in some situations but not in others. A man that would never dream of hitting his mother might beat his wife, but only in private after he has been drinking. Hogan's (1970, 1973) research indicates that, in addition to socialization, we must also consider empathy, autonomy, and whether a person is guided more by his or her personal conscience or a sense of social responsibility. A potentially fruitful area for research is on the situational factors that influinflu ence whether or not we act on our inhibitions and controls. In a series of studies, the present writer and his colleagues have investigated the situational factors that determine whether people high in Dominance, as assessed by the CPI Dominance & Carbonell, 1988). A similar scale, actually assume leadership (E. I. Megargee & utiUzed to study the circumstances in which people with varying paradigm could be utilized self-control inhibit inhibit their theirbehavior behavioror orsuccumb succumbto totemptation. temptation. levels of socialization or self-control Is the presence of others conducive or detrimental to self-control? Are we better behaved in the presence of some people than we are with others? What are the effects of prosocial or antisocial models? Field research on actual examples of people exercising constraint or control because of value judgments they have made is a largely unexplored but potentially an allied alliedresearch research area, area,in inthis thiscase casebehavioral behavioral important source of data. Once again, an medicine, could be a valuable valuable source source of of data data and and designs. designs. The The literature literature on on compli compliance, dieting, and smoking cessation involves exercising restraints and self-control in real life hfe situations with important contingencies involved. In order for ethical values to come into play, it is necessary that a particular behavior be classified as "right" "right" or "wrong." "wrong." As we have noted, psychologists interested in moral development have constructed a number of hypothetical dilemdilem mas to investigate moral reasoning. These studies of artificial situations could be supplemented by research on "real life" ethical dilemmas. For many young men in the 1960s, participation in the Vietnam war presented such a crisis; crisis; for young women in the 1990s, abortion can be a similar issue (Gilligan, 1977). These problems often involve agonizing taking public public stands. stands.From Fromaaresearch researchstandpoint, standpoint,the theparticu particupersonal decisions and taking lar issue and its resolution are not as important as the process by which the person decided whether or not the choice was congruent with his or her particular code of values, and the subsequent problems encountered in living with the decision. Research with people who have had to grapple with such dilemmas and their

610 610

EDWIN MEGARGEE EDWIN I. I. MEGARGEE

consequences could help us determine the generality of of findings findings based on hypotheti hypothetical situations. c. C. Building Building for the the Future

Psychological research on internal inhibitions and constraints and their influence influence on behavior is still in its infancy. Although we have several theories and a number number of assumptions regarding regarding the origins and operations of internal controls, relatively relatively of little has been established with certainty. For various reasons, personality researchers have not been eager to study controls and inhibitions. Those studies that have been done have often often involved failed to develop what others regarded regarded as adequate or appropriate appropriate people who failed values or who failed failed to live up to them and behaved in an antisocial fashion. It is, perhaps, significant significant that the editors of this Handbook Handbook recruited recruited a psychologist whose focused on criminal behavior behavior and violence to write this chapter. chapter. research has focused Of Of course, compared compared with other disciplines, we psychologists are new to the of internal inhibitions and restraints. The first stages in any scientific inquiry inquiry study of are to review the literature, make preliminary observations, and form form hypotheses. of the primary theses of of this chapter is that psychologists interested in studying One of internal inhibitions and controls should cast off off our disciplinary disciplinary blinders and con conof anthropologists, criminolo criminolosider the observations, speculations, and theories of theologians—even playwrights and gists, sociologists, ethicists, philosophers, and theologians-even poets-in poets—in short, of of all the scholars who have struggled with these issues over over the years. Controls involve behaving according to one's values, and individual individual values values Controls differ differ from from culture to culture and from from era to era. Therefore Therefore another another thesis has been that observations and investigations of the origin and application of of ethical values in our society must be replicated in other cultures, and that our studies in the present should be enriched by a consideration of how people behaved in the past. A A third thesis has been that a great great deal of research relevant to these issues has been conducted conducted by psychologists who were primarily interested in other other questions. physiologiAmong the areas cited were social psychology, behavioral medicine, and physiologi A number number of of important important observations and hypotheses can be derived derived cal psychology. A ascertain their relevance to controls controls by reviewing the literature in these areas to ascertain and restraints. of formulating formulating a theory of of aesthetics supported supported Finally, no one would dream of of people who are color blind or tone deaf. Yet much, perhaps only by observations of of our observations and theories about inhibitions and controls have been most, of based on investigations of people who are poorly socialized or impulsive. In the future we should investigate people with positive as well as negative value systems future functioning as well as deficient deficient systems of controls. and well functioning

CHAPTER CHAPTER 23 23

iNHiBrnoNS AND CONTROI.'i CONTROLS INlDBmONS

611 611

ACKNOWLEDGMENT Preparation of of this this chapter chapter was was supported in in part part by by Grant No. No. 88-IJ-CX-0006 88-IJ-CX-0006 from from the National Institute Institute of of Justice to to Edwin Edwin I.I. Megargee and and Joyce Joyce L. L. Carbonell. National

REFERENCES Aries, Aries, P., & & Duby, Duby, G. G. (Eds.). (Eds.). (1987-1988). (1987-1988). A history historyo/private ofprivate life life (A (A.Goldhammer, Goldhammer,Trans.) Trans.) (Vols. (Vols. 11 & & 2). 2). Cambridge, Cambridge, MA: MA: Belknap Belknap Press Press of of Harvard Harvard University University Press.

Asch, S. S. E. (1956). (1956). Studies of of independence independence and and conformity: conformity: A A minority minority of of one one against against a unanimous (Whole 416). PsychologicalMonographs, Monographs,70,70, (WholeNo. No. 416). unanimous majority, majority. Psychological Principlesof ofbehavior behaviormodification. modification. New York: Holt. Bandura, A. A. (1969). (1969). Principles New York: Holt. Bandura, (1977).Self-efficacy: Self-efficacy: Toward Towardaa unifying unifying theory theory of of behavioral behavioral change. change.PsychologiPsychologiBandura, A. A. (1977). Bandura, 84,191-215. 191-215. cal Review, Review, 84, Bartlett'sfamiliar familiarquotations quotations(15th (15th ed.). Boston: Little, Brown. Beck, E. E. M. M. (1980). (1980). Bartlett's Beck, ed.). Boston: Little, Brown. The other otherside: side:Perspectives Perspectivesonon deviance. New York: Free Press. Becker, H. H. S. S. (1964). (1964). The Becker, deviance. New York: Free Press. The developing developingchild. child.New NewYork: York:Harper Harper&&Row. Row. Bee, H. (1985). (1985). The Bee, H. Patternsof ofculture. culture.Boston: Boston:Houghton-Mifflin. Houghton-Mifflin. Benedict, R. R. (1934). (1934). Patterns Benedict, Jr. (1984). (1984). Twins Twins reared reared together together and and apart: apart: What What they they tell tell us us about about human human Bouchard, T. T. J., Bouchard, J., Jr. Individualityand and determinism determinism(pp. (pp.147-184). 147-184).New NewYork: York: diversity. In In S. S. W. W. Fox Fox (Ed.), diversity. (Ed.), Individuality Plenum Press. Plenum Buss, A A. H., & & Plomin, Plomin, R. R. (1984). (1984). Temperament: Temperament:Early Early developing personality traits. Hillsdale, developing personality traits. Hillsdale, NJ: Erlbaum. Erlbaum. NJ: Christiansen, K. K. O. O. (1977). (1977). A A preliminary preliminary study study of of criminality criminality among among twins. twins. In In S. S. A A. MedMed Christiansen, Biosocialbases basesof ofcriminal criminalbehavior behavior (pp. 89-108). nick & & K. K. O. O. Christiansen Christiansen (Eds.), Biosocial nick (pp. 89-108). New York: York: Wiley. Wiley. New & Quinney, Quinney, R. R. (1973). (1973). Criminal Criminalbehavior behaviorsystems: systems: typology (2nd ed). New Clinard, M. M. B., B., & A Atypology (2nd ed). New Clinard, York: Holt, Holt, Rinehart Rinehart & & Winston. Winston. York: A, & ofof delinquent Cloward, Delinquencyand and opportunity: opportunity:AAtheory theory delinquent Cloward, R. R. A., & Ohlin, Ohlin, L. L. E. E. (1960). (1960). Delinquency gangs. gangs. Glencoe, Glencoe,IL: IL:Free FreePress. Press. Darley, J. M., & of & Latane, Latan^, B. (1968). Bystander intervention in emergencies: Diffusion Diffusion of Psychology, 8, 377-383. responsibility. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 8, 377-383. Dollard, Frustration DoUard, J., Doob, L. L. W., W., Miller, N. E., Mowrer, O. H., & & Sears, R. R. (1939). Frustration and aggression. aggression.New NewHaven, Haven,CT: CT:Yale YaleUniversity UniversityPress. Press. Erikson, E. H. (1950). Childhood Childhood and andsociety. society.New NewYork: York:Norton. Norton. Eysenck, H. J. Crime and and personality. personality.Boston: Boston:Houghton-Mifflin. Houghton-Mifflin. J. (1964). (1964). Crime Eysenck, H. J. personality.New NewYork: York:Springer-Verlag. Springer-Verlag. J. (1981). A model for personality. Feshbach, S., & Personality(2nd (2nded.). ed.).Lexington, Lexington,MA: MA:D. D.C. C.Heath. Heath. & Weiner, B. (1986). Personality Freud, (J. (J. Strachey, Trans.). New introductory lectures lecturesononpsychoanalysis psychoanalysis Strachey, Trans.). New Freud, S. S. (1965). (1965). New introductory York: York: Norton. (Original work work published published 1933) 1933) AmericanJournal Journalof of Garfinkel, H. H. (1956). (1956). Conditions Conditions of of successful successful degradation degradation ceremonies. ceremonies. American Garfinkel, Sociology, 61, 61,420-424. 420-424. Sociology, C. (1977). (1977). In In aa different different voice: voice:Women's Women'sconception conceptionof ofthe theself selfand and morality. morality.Harvard Harvard Gilligan, C. Gilligan, Educational Review,47, 47,481-517. 481-517. Educational Review,

612 612

EDWIN MEGARGEE EDWIN L L MEGARGEE Glueck, S., & Glueck, E. (1950). (1950). Unraveling Unravelingjuvenile juvenile delinquency. delinquency.New New York: York: Common Commonwealth Fund. Goring, C. (1913). (1913). The English English convict: convict: AA statistical statisticalstudy. study,London: London:His HisMajesty's Majesty'sStatio Stationery Office. Office. Inventory. Palo Alto, CA: Gough, Manualfor for the theCalifornia CaliforniaPsychological Psychological Inventory. Palo Alto, CA: Gough, H. H. G. G. (1969). (1969). Manual Consulting Psychologists Press. Harlow, H. (1971). (1971). Learning to love. love. San Francisco: Albion. Vol.Vol. 1. Studies in in Hartshorne, Hartshorne, H., & & May, May, M. M. A A. (1928). (1928). Studies Studies ininthe thenature natureof ofcharacter: character: 1. Studies deceit. deceit. New NewYork: York:Macmillan. Macmillan. Clinical Hogan, R R. (1970). (1970). A dimension of moral judgment. Journal of of Consulting Consulting and Clinical

Psychology,35, 35,205-212. 205-212. Psychology, Psycho Hogan, R R. (1973). (1973). Moral conduct and moral moral character: A psychological perspective. PsychoHogan, logical logical Bulletin, Bulletin,79,79,217-232. 217-232.

Hogan, R R. (1986). (1986). What What every every student student should should know know about about personality personality psychology. psychology. In In V. V. P. P. Hogan, Makosby (Ed.), (Ed.), The G. G. Stanley Stanley Hall lecture lecture series series(Vol. (Vol. 6,6, pp. pp. 43-64). 43-64). Washington, Washington, Makosby

DC: American American Psychological Psychological Association. Association. DC: Hogan, R, R., Johnson, Johnson, J. J. A, A., & & Emler, Emler, N. N. P. P. (1978). (1978). A A socioanalytic socioanalytic theory theory of of moral moral developHogan, develop ment. New Directions Directionsfor for Child ChildDevelopment, Development,2,2,1-18. 1-18. ment. anan anthropological study: Vol.Vol. 1. The native Hooton, The American Americancriminal, criminal, anthropological study: 1. The native Hooton, E. E. A A. (1939). (1939). The white criminal criminal of of native nativeparentage. parentage.Cambridge, Cambridge,MA: MA:Harvard HarvardUniversity UniversityPress. Press. white Institute of of Medicine Medicine Committee Committee on on Research Research and and Training Training Issues Issues Related Related to to the the Mission Mission Institute of Secret Service: Behavioral of the the Secret Secret Service. Service. (1984). (1984). Research Researchand andtraining trainingfor forthethe Secret Service: Behavioral science andmental mentalhealth healthperspectives. perspectives.Washington, Washington,DC: DC:National NationalAcademy AcademyPress. Press. science and

Janis, I. I. L. L. (1972). (1972). Victims Victims of of groupthink. groupthink. Boston: Boston: Houghton-Mifflin. Houghton-Mifflin. Johnson, A A. (1949). (1949). Sanctions Sanctions for for super-ego super-ego lacunae lacunae of of adolescence. adolescence. In K. K. Eissler Eissler (Ed.), (Ed.), Searchlights Press. Searchlightson ondelinquency delinquency(pp. (pp.225-245). 225-245).New NewYork: York:International InternationalUniversities Universities Press. animals. Philadelphia: Philadelphia: Saunders. Saunders. Johnson, R R. (1972). (1972). Aggression Aggression in man and animals. Johnson, Kohlberg, L. (1981). development. New NewYork: York:Harper Harper &&Row. Row. (1981). Essays Essays on moral development. problems, and social control. Englewood Cliffs, Lemert, Lemert, E. M. M. (1967). (1967). Human Humandeviance, deviance,social social problems, and social control. Englewood Cliffs, NJ: Prentice-Hall. Prentice-Hall. Lindzey, G. (1967). (1967). Some remarks concerning incest, the incest taboo, and psychoanalytic theory. theory. American American Psychologist, Psychologist,22,22,1051-1059. 1051-1059. Loehlin, J. C., C, & Nichols, R R. C. (1976). (1976). Heredity, Heredity, environment and personality. personality. Austin: University of Texas Press. Lorenz, K. (1966). (1966). On aggression aggression(M. (M.Latzke, Latzke,Trans.). Trans.).New NewYork: York:Bantam BantamBooks. Books.(Original (Original work published 1963) Luckenbill, D. F., F., & Sanders, W. B. B. (1977). (1977). Criminal Criminal violence. In E. Sagarin & F. Montanino (Eds.), Deviants: Deviants:Voluntary Voluntaryactors actors a hostile world. Morristown, General Learnin in a hostile world. Morristown, NJ:NJ: General Learn ing Press. Press. ing Mark, V. V. H., & Ervin, S. S. R R. (1970). (1970). Violence Violence and and the thebrain. brain.New New York: York: Harper Harper & & Row. Row. McCord, . New McCord, J., & & McCord, McCord, W. W. (1964). (1964). The The psychopath: psychopath:An Anessay essayon onthe thecriminal criminalmind mind. New York: Van Van Nostrand-Reinhold. Nostrand-Reinhold. York: McCord, ofof thethe Cambridge McCord, W., & & McCord, McCord, J. J. (1959). (1959). Origins Originsof of crime: crime:AAnew newevaluation evaluation CambridgeSomerville study. York: Columbia Columbia University Press. study. New New York: University Press. Mead, M. Mead, M. (1961). (1961). Coming Coming of of age age in in Samoa: Samoa: AA psychological psychologicalstudy studyofofprimitive primitiveyouth youthfor for WesternCivilization. Civilization.New NewYork: York:W.W.Morrow. Morrow. Western Mead, M. M. (1975). (1975). Growing up in in New Guinea: Guinea:AA comparative comparativestudy studyofofprimitive primitiveeducation. education. New York: York: W. W. Morrow. Morrow. New

CHAPTER CHAPTER 23 23

INlUBmONS iNHiBmoNS AND AND CONTROLS CONTROLS

613 613

Mednick, S. S. A., A., Gabrielle, Gabrielle, W. W. F., Jr., & Hutchings, Hutchings, B. (1984). Genetic Genetic influences influences in in criminal criminal F., Jr., B. (1984). -894. convictions: convictions: Evidence Evidence from from an an adoption adoption cohort. cohort. Science, Science, 224, 224,891 891-894. Mednick, S. S. A., & Volavka, J. (1980). (1980). Biology and crime. In N. Morris & M. Tonry (Eds.), Crimeand andjustice: justice:An Anannual annual review of research (Vol. 2, pp. 85-158). Chicago: UniverCrime review of research ( Vol. 2, pp. 85-158). Chicago: Univer sity sity of of Chicago Chicago Press. Press. Megargee, E. E. I. I. (1973). (1973). The The heuristic heuristic value value of of the the concept concept of of "social "social deviance" deviance" for for psycholoMegargee, psycholo gists. Representative RepresentativeResearch Research Social Psychology, 4, 67-81. gists. in in Social Psychology, 4, 67-81. Megargee, E. E. I. I. (1982). (1982). Psychological Psychological correlates correlates and and determinants determinants of of criminal criminal violence. In In Megargee, M. M. Wolfgang Wolfgang & N. N. Weiner Weiner (Eds.), (Eds.), Criminal Criminal violence violence (pp. (pp. 81-170). Beverly Beverly Hills, CA: CA: Sage. B. Sutker (Eds.), Megargee, E. I. (1993). (1993). Aggression and violence. In H. E. Adams & P. B. Comprehensive Comprehensivehandbook handbookofofpsychopathology psychopathology(2nd (2ndEd., Ed.,pp. pp.617-644). 617-644).New NewYork: York: Plenum Press. Press. Plenum Megargee, E. E. I., I., & Carbonell, Carbonell, J. J. C. C. (1988). (1988). Evaluating Evaluating leadership with with the the CPI. CPI. In In C. C. D. Megargee, Spielberger & J. J. N. N. Butcher (Eds.), (Eds.), Advances in personality assessment (Vol. 7,7, pp. pp. assessment (Vol. Spielberger 203-219). Hillsdale, Hillsdale, NJ: NJ: Erlbaum. Erlbaum. 203-219). Megargee, E. E. I., Parker, Parker, G. G. V. V. c., C, & & Levine, Levine, R. R. V. V. (1971). (1971). The The relationship of of familial familial and and Megargee, social factors factors to to socialization socialization in in middle middle class class college college students. Journal of of Abnormal social Psychology, 77, 77,76-89. 76-89. Psychology, Megargee, S. S. E., E., Jr. Jr. (1942). (1942). Dogs. Dogs. New New York: York: Harper Harper & & Brothers. Brothers. Megargee, Megargee, S. S. E., E., Jr. Jr. (1954). (1954). The dog dictionary. dictionary.New NewYork: York:World. World. Megargee, Merton, R. R. (1938). (1938). Social Social structure structure and and anomie. anomie. American Sociological SociologicalReview, Review,3,3,672-682. 672-682. Merton, social structure. structure.Glencoe, Glencoe,IL: IL:Free FreePress. Press. Merton, R. R. (1957). (1957). Social theory and social Merton, authority.New NewYork: York:Harper. Harper. Milgram, S. S. (1974). (1974). Obedience Obedience to toauthority. Milgram, Miller, W. W. B. B. (1958). Lower class class culture culture as as aa generating generating milieu milieu of of gang gang delinquency. delinquency. Journal Miller, (1958). Lower Social Issues, Issues,14(3), 14(3),5-19. 5-19. of Social Mischel, W., W., & & Mischel, Mischel, H. H. N. N. (1976). (1976). A A cognitive cognitive social social learning learning approach to to morality morality and and Mischel, self-regulation. In In T. T. Lickona Lickona (Ed.), (Ed.), Moral Moraldevelopment developmentand andbehavior. behavior. Theory, research Theory, research self-regulation. issues. New NewYork: York:Holt. Holt. and social social issues. Munroe, R. R. L. L. (1955). (1955). Schools of of psychoanalytic psychoanalytic thought. thought.New NewYork: York:Dryden DrydenPress. Press. Munroe, Family relationships relationshipsand and delinquent behavior. New York: Wiley. Nye, F. F. I. I. (1958). (1958). Family Nye, delinquent behavior. New York: Wiley. Peter, L. L. J. J. (1977). (1977). Peter's Peter*squotations: quotations:Ideas Ideas our time. New York: Morrow. Peter, forfor our time. New York: W.W. Morrow. child. New NewYork: York:Macmillan. Macmillan. Piaget, J. J. (1932). (1932). The moral judgment of of the child. Piaget, Quay, H. H. C. C. (1965). (1965). Psychopathic Psychopathic personality personality as as pathological pathological sensation sensation seeking. seeking. American Quay, Journalof ofPsychiatry, Psychiatry,122, 122,180-183. 180-183. Journal Rosenquist, C. C. M., M., & & Megargee, Megargee, E. E. I.I. (1969). (1969). Delinquents Delinquentsininthree threecultures. cultures. Austin: University Rosenquist, Austin: University of Texas Texas Press. Press. of Salisbury, H. H. E. E. (1959). (1959). The shook-up generation. generation. New NewYork: York:Harper Harper &&Row. Row. Salisbury, Sellin, T. T. (1938). (1938). Culture Culture conflict conflictand andcrime. crime.New NewYork: York:Social SocialScience ScienceResearch ResearchCouncil. Council. Sellin, of interpersonal interpersonalunderstanding. understanding.New NewYork: York: AcademicPress. Press. Selman, R. R. (1980). (1980). The growth of Academic Selman, Shoham, S. S. (1962). (1962). The The application application of of the the "culture "culture conflict" conflict" hypothesis hypothesis to to the the criminality criminality Shoham, of and Police Science, of immigrants immigrants in in Israel. Journal Journalof of Criminal CriminalLaw, Law,Criminology, Criminology, and Police Science, 53, 207-214. 207-214. 53, Skinner, B. F. (1971). (1971). Beyond freedom and dignity. dignity. New NewYork: York: Knopf. Knopf. Skinner, B. F. Sutherland, E. E. H. H. (1939). (1939). Principles Principlesof ofcriminology criminology(3rd (3rded.). ed.).New NewYork: York:Lippincott. Lippincott. Sutherland, Sykes, G., G., & & Matza, Matza, D. (1957). Techniques Techniques of of neutralization: A theory theory of of delinquency. delinquency. Sykes, D. (1957). neutralization: A American 664-670. American Sociological SociologicalReview, Review,22,22, 664-670. Tripp, A. A. H. H. (1970). (1970). International International thesaurus thesaurusofofquotations. quotations.New NewYork: York: ThomasY.Y. Crowell. Thomas Crowell. Tripp,

EDWIN EDWIN L I. MEGARGEE MEGARGEE

614 614

Untermeyer, L. L. (1955). Modem Modem American and modern New York: York: Har HarUntermeyer, modem British poetry. New court, Brace. udgments Velez-Diaz, A., & Megargee, E. I. (1971). (1971). An investigation investigation of of differences differences in value jjudgments between youthful youthful offenders offenders and and nonoffenders nonoffenders in in Puerto Puerto Rico. Rico. between

of Criminal Journal of Criminal Criminology and and Police Police Science, Science,61, 62,549-553. 549-553. Law, Criminology Theoreticalcriminology. criminology.New NewYork: York:Oxford OxfordUniversity UniversityPress. Press. Void, G. (1958). Theoretical Void, Void, G., & & Bernard, T. T. (1986). (1986). Theoretical Theoreticalcriminology criminology(3rd (3rded.). ed.).New NewYork: York:Oxford OxfordUniver University sity Press. Press. Wilson, Wilson, J. J. Q., & & Herrnstein, Herrnstein, R. R. J. J. (1985). (1985). Crime Crime and human nature. nature. New NewYork: York: Simon Simon && Schuster.

PART VII PARTVH

PERSONALITY AND THE THE SELF

This Page Intentionally Left Blank

CHAPTER 24 CHAPTER 24

CONGRUENCE OF OTHERS' AND SELF.JUDGMENTS AND SELF-JUDGMENTS OF PERSONALITY DAVID C. FUNDER UNIVERSITY OF CALIFORNIA, RIVERSIDE UNIVERSITY OF CALIFORNIA, RIVERSIDE

C. RANDALL COLVIN NORTHEASTERN UNIVERSITY NORTHEASTERN UNIVERSITY

I. I. THE THE QUESTION QUESTION OF OF SELF-OrnER SELF-OTHER AGREEMENT AGREEMENT A. A, Two Two Reasons Reasons for for Being Being Interested Interested in in the tiie Question Question Do other other people view you you the the same way way you you view yourself? Most people find this to be be an an interesting question, for for two two basic reasons. First, the the self self that a person presents to others, and and the the way way that that self self isis perceived perceived by by others, others, importantly importantly him influences how those others treat him him or her her and and how how the the person views himinfluences how or herself. Someone viewed as as incompetent incompetent will not be be given a job, and and someone someone as dishonest dishonest will not not be be lent money, but but someone viewed as as warm warm will viewed as have many friends; the may the amount of self-esteem self-esteem the the individual develops may depend upon the degree to which he or she accepts each of these characterizations. one is is viewed positively or negatively, it is is probably strategi strategiwhether one Moreover, whether to have an an accurate idea of how how one one is is regarded by by the the others in useful to cally useful one's social world. IIANDBOOK HANDBOOKOF OFPERSONALITY PERSONAUTYPSYCHOLOGY PSYCHOLOGY

617 617

COPYRIGHT COPYRIGHT C © 1997 1997 BY BY ACADEMIC ACADEMIC PRESS. PRESS. ALL A L LlImns RIGHTSOFOFIEnODUCIlON REPRODUCTIONININANY ANYFORM FORMRESERVED. RESERVED.

FuNDER AND COLVIN FUNDERANDCOLVIN

618 618

Second, the opinions of of others are a useful useful source of information information about what a person might really be like. This may be a subtle point but it is an important one. If If you want or need to know whether whether you have musical talent, it makes sense to expert musician who has had a chance to observe observe you perform perform musically. musically. consult an expert If If you want to know whether whether you have athletic talent, it makes sense to consult an expert coach coach who has seen you perform perform athletically. If If you want to know what kind expert of personality you have, the situation is even simpler. Many individuals have seen of perform interpersonally, and everybody everybody is an "expert." "expert." you perform B. Sociological and Psychological Psycliologicai Perspectives on the Question Empirical research on self-other classified according to its self-other congruence can be classified relevance to one or the other of of the two reasons just mentioned mentioned for for finding such continuum which we will congruence important. Most research can be placed along a continuum label as ranging from from "sociological" through "social psychological" to "personality" "personality" perspectives. Sociological approaches approaches to self-other self-other congruence emphasize emphasize the social construction and consequences consequences of of the self-image, self-image, whereas whereas psychological perspec perspecconstruction tives are more more likely to attend attend to the possibility possibihty that self self and others' views might actually characterize what a person person is like. From the sociological perspective, the self self does not exist outside of of the From of those who behold it; it is little more than an arbitrary arbitrary social construct. construct. minds of perspective yields theorizing about the "looking-glass self" and symbolic This perspective interactionism (Blumer, 1937; interactionism 1937; Mead, 1934; 1934; Shrauger & Schoeneman, 1979; Stryker Stryker & Gottlieb, 1981), which which focuses upon the self-image that arises out of of ongoing social interaction interaction between between a person and his or her social world. Empirical research on self-other self-other congruence emphasizes how the perceptions others have us, and the assessments we make of of of of others' perceptions of of us, affect affect what 1979). Such research regards we think of of ourselves (Shrauger (Shrauger & Schoeneman, 1979). regards emergent emergent social consensus and the processes by which it comes about to the phenomena of of interest, and pays little attention to what, if anything, might be correctly sayable about about the person person who is judged. In fact, in the view of of this tradition, questions about accuracy are not really meaningful. Some writers even about complain about how "language usage and convenience make it almost impossible about make to avoid writing as though the self self were being conceived conceived as a concrete entity" (Stryker (Stryker & Gottlieb, 1981, 1981, p. 453)-which, 453)—which, in sociological theorizing, is generally regarded as a grave mistake. regarded Some theorizing within social psychology shares part of of the outlook of of the sociological approach. Modern-day "constructivists," exemplified by Kruglanski Modern-day exemplified Kruglanski (1989, who sometimes also uses the term "phenomenal"), emphasize emphasize how reality reality and perceptions perceptions of of reality reaUty are not easily separated, if they are separable separable at all. Therefore, this perspective also avoids regarding the self self as a "concrete entity." entity." As Funder Funder and and West (1993) note, the constructivist viewpoint focuses on the way of the "same" stimulus can vary according judgments are made and how judgments of

CHAPTER 24 CHAnER

SELF-OTHER CONGRUENCE SELF-OrnER

619 619

to point of view, while eschewing evaluation of any particular judgment as inacinac curate. This focus on the process of judgment is a general attribute of other social self-other congruence. For example, Swann's (1984) psychological approaches to self-other work on identity negotiation investigates the tension between self verification, verification, confirmawhere a target strives to confirm his or her self-concept, and behavioral confirma tion, where a target is led to behave in ways that confirm to the expectations and desires of others. Here we find some-a some—a little-concern little—concern with how the targets of judgment actually behave (as opposed to the sociological and constructivist approaches noted previously, which manifest no such concern), and with the extent to which their behavior is accurately characterized by self and others' self-fulfilling prophecies and personality descriptions (albeit mostly as a result of self-fulfilling the like). Standing in strong contrast to all of the approaches considered so far is that of personality psychology. This approach is based on the assumption that manifest differences differences in social behavior that may reflect actual differ differindividuals manifest 1995). Investigators who share this perspective ences in their personalities (Funder, 1995). regard the personality of the individual who is judged to be the phenomenon of central importance and are interested in the task of personality assessment. frequently used, is simply to ask a person One possible method of assessment, frequently for a description of the individual's personality. and his or her acquaintances for self and others' judgments of personality personaUty then becomes a Congruence between self matter of convergent validity, judgments that agree with each other being (e.g.. Cheek, presumed to be more accurate than judgments that do not agree (e.g., 1982; Funder, 1980a; Funder Funder & & Colvin, 1988; 1988; Funder & & Dobroth, 1987; 1987; Funder, 1982; 1995; McCrae, 1982). 1982). Personality, from from this perspective, is Kolar, & Blackman, 1995; regarded as a relatively stable and coherent structure residing with the person, and in in that that sense sense is is viewed viewed as as aa "concrete "concrete entity." entity." This This structure structure is is not not directly directly and visible, but is is (partially) (partially) revealed revealed through through social social behavior and can, can, in in turn, turn, be be visible, but behavior and inferred by the other people in an individual's social world. (imperfectly) inferred

€• Two Two Ways Ways to to Ask Ask the the Question Question c. course, that that the the self self and and his his or or her her self-other congruence requires, of course, Examination of self-other The precise precise way way this this question question is for personality descriptions. The acquaintances be asked for asked reveals much about the interests of the investigator, and has important for interpreting the answers that will be obtained. consequences for One way is to ask the target's acquaintances to describe, not the personality of the target, but how the target perceives or will describe his or or her own own personality. personality. 1960s, this thiswas was the the traditional traditional method method for for research research From about the 1930s to the early 1960s, Hastorf (1950), in a typical early study, asked target on accuracy. Bender and Hastorf self-descriptions on three measures: measures: one of behavior in social students to provide self-descriptions abiUty. Another group situations, another on dominance, and another on emphatic ability. of students, who were acquaintances of the targets, then completed the same three

620 620

FVNDER FUNDER AND AND COLVIN COLVIN

measures as they they thought thought the targets targets had answered answered them. them. For two of the three measures, significant agreement was found between targets' targets' and acquaintances' responses. Early researchers regarded this sort of ability to predict targets' self selfjudgments as an indicator of the acquaintances' degree of social sensitivity or empathy (e.g., Gage & & Cronbach, 1955; 1955; Taft, 1966). When accuracy research began to reappear during the middle 1980s, for some reason the opposite approach became common: Investigators asked the target to describe, not his or her own personality, but rather how the target believed he or she would be described by others. It is not clear why this change occurred, but notice how it constitutes a subtle shift in focus from a concern with what other people are like, to a concern with how one appears to others. Such a shift seems consistent with what some commentators have seen as an increasingly "narcissistic" cultural tone to the 1980s as opposed to the 1950s (e.g, Fine, 1986). An advantage of these two approaches, one old and one new, is that both are designed to neatly finesse finesse the whole issue of accuracy. They do this by giving accuracy an operational definition which is not quite the same as its everyday meaning, but which has the advantage of being directly measurable. In the older research, assessment of a subject's accuracy at predicting what somebody else would say about himself or herself seemed to be a straightfoward matter. Similarly, the newer studies can measure quite directly the degree to which a subject correctly acquaintances. Both methodologies predicts the descriptions provided by his or her acquaintances. are appropriate, and indeed necessary, if one is interested in the ontogeny of the self from a sociological viewpoint (Shrauger & Schoeneman, 1979). Even from a psychological viewpoint, these methodologies are appropriate and unproblematic if an investigator wants to understand no more than how well people can predict generalizabileach others' questionnaire responses (and, more charitably, assuming generalizabil ity, how well they understand each others' perceptions). The use of operational definitions to finesse a conceptual issue never comes & Gallwey, 1958; 1958; Cronbach, without cost, however (cf. Bronfenbrenner, Harding, & 1955). In In this case, case, the the cost costisis that thatno no information informationisis gathered gatheredrelevant relevantto tothe theordinary ordinary meaning of accuracy. That is, there is no indication whatsoever as to whether the by the self or or the acquaintances actually correctly characterize descriptions provided by anybody. This is because, in both methodologies, one member member of of the self-other to describe describeaaperson. person.InInthe theolder olderresearch researchthe theacquaintance, acquaintance,and and dyad is not asked to judgments provided provided in the the newer newer research research the the target, target, is is merely merely asked asked to to describe describe judgments in by the the other other dyad dyad member. by member. The other way to ask the descriptive question is simpler and, for a personality psychologist, more informative. It is to ask the target simply, "what are you like?", and to ask his or her acquaintance the parallel question, "what is this person like?" & Dobroth, 1987; 1987; Park & & Judd, 1989). like?" (e.g., Funder & Self-other congruence in this context at least indicates how well two people Self-other agree about the target's actual personality, and assessing judgmental agreement agree about the target's actual personality, and assessing judgmental agreement

CHAPTER CHAPTER 24 24

SELF-OrnER SELF-OTHER CONGRUENCE CONGRUENCE

621 621

in this manner is the only procedure that takes the question of accuracy in seriously.^ personality judgment seriously.! D. D. Two Two Ways Ways to to Analyze Analyze Answers Answers to to the the Question Question However an investigator decides to ask the descriptive question, self-other agree agreement can be assessed in two fundamentally fundamentally different different ways. One involves the analysis of mean differences; the other involves the analysis of correlations. The first kind placement of each descriptive item between others' of analysis compares the mean placement and self-descriptions. For instance, do people give themselves higher ratings, on average, on desirable attributes than their acquaintances do? The second kind of of holistianalysis computes correlations between others' and self-judgments, either holisti cally or one item at a time. For instance, do people who give themselves relatively high ratings on sociability (compared to other subjects) tend to receive relatively high sociability sociability ratings ratings from from their their acquaintances? acquaintances? high was noted long ago ago by by Conrad (1932) and more recently by Funder (1980a), As was these two analyses and these two kinds of question are utterly orthogonal. It is entirely possible and plausible to find good agreement on ratings of a trait using one method and to find poor agreement on ratings of the same trait, among the difference could just as well same subjects, using the other method. A large mean difference difassociated with either a high or a low correlation-so correlation—so could a small mean dif be associated ference. When one fails to recognize that the two analyses are orthogonal, subsequent interpretations can be misleading. One such failure failure occurred in the often-cited often-cited interpretations self-other agreement by Shrauger and Schoeneman (1979). Their Their review, review, review of self-other self-other agreement, combined the results which reported finding no consistent self-other differences and correlational analyses (see their Table 2). of both mean differences 2). When the results they report are discriminated according to the method of analysis, a different different picture emerges. From among the studies we could discriminate on the basis of of of 12 12 samples self-other self-other mean mean differences differences seemed to to indicate indicate their Table 2, in 2 out of a lack of agreement (significant self-other differences), whereas in 26 out of 45 self-other correlations indicated the presence of agreement (significant (significant samples self-other correlations). Clearly, if this review had recognized the distinct nature of the two self-other agreement might well have analyses, its evaluation of research on self-other different. been different. It is important important to keep in mind, therefore, that the term "agreement" can different and independent independent meanings. In what follows, we will have two entirely different consider each meaning separately. Thereisisno noreason reasonwhy whythe theaccuracy accuracyquestion questionneeds needsto tobe bethe thecentral centralconcern concernof ofeverybody everybodywho who 1^There self-other agreement; as already noted, social psychologists and sociologists often have different different investigates self-other and perfectly legitimate questions in mind. Our assumption is, however, that the accuracy question is the of central concern to personality psychologists of the sort who might read a handboo handbook one of k like this one.

622 622

FuNDER AND COLVIN FUNDERANDCOLVIN

II. OF ABSOLUTE (MEAN) AGREEMENT IL ANALYSES OF Psychologists investigating absolute or mean differences between others' others' and self selfperceptions have focused on three issues: issues: self-enhancement biases, the actor actorobserver effect, and the difference between internal and external personality traits. A. A. Self-Enhancement Self-Enhancement Biases Biases Probably the most obvious place to look for differences between others' and self selfdescriptions of personality is in the area of self-enhancement or "self-serving" (e.g.. Brown, 1986; 1986; Kunda, 1987; 1987; Miller & & Ross, 1975). 1975). It certainly seems biases (e.g., reasonable and consistent with everyday experience to expect that people will describe themselves in more laudatory terms than they will be described by others. self-enhancement does not test this expectation However, most of the work on self-enhancement self-enhancedirectly. As Colvin and Block (1994) have argued, the assessment of self-enhance ment requires a comparison of one's self-perception of personality against a valid external criterion. Because very little research actually includes any such criterion, many studies that supposedly address "self-enhancement" are open to a variety of alternative explanations. A second area of research on self-enhancement focuses on causal attributions. by this thisresearch researchisisthat thatpeople peopletake takemore morecausal causalresponsi responsiThe usual hypothesis tested by bility for their successes than for their failures, relative to the attributions offered offered by outside observers. However, a review by Miller and Ross (1975) concluded that there is surprisingly little evidence in support of this hypothesis, at least in what these authors called its "most general form" (p. 213). While they concluded that the evidence shows people engage in self-enhancing attributions under conditions of success, they found little evidence of a self-enhancement bias under conditions of failure. Still, as Kunda (1987) has pointed out, research in a variety of domains seems to show with regularity, if not perfect consistency, that people engage in self-enhancement in perception (Erdelyi, 1974), memory (Greenwald & & Pratkanis, Pratkanis, 1984), attribution of responsibility (Lerner, 1980; 1980; Tetlock & & Levi, 1982), and social comparison (Taylor, 1983). The results of the studies just cited would seem to imply that, in the end, people should manifest more positive opinions of themselves than others will have of them, but none of these studies tests this prediction directly. The prediction was tested by Funder (1980a), who found no indication that self-descriptions were more favorable than descriptions provided by by acquaintances. However, However, we can can offer offer some some previously unreported data that reexamine this issue within a much larger sample (approximately four times as large). One hundred fifty-seven fifty-seven undergraduates dede scribed their own personalities using the California Q-sort (Block, 196111978), 1961/1978), as modified for nonprofessional use use by Bem and and Funder Funder (1978). (1978).These Thesesame sameundergrad undergraduates were also described by two close acquaintances, and the two descriptions & Dobroth, 1987). The were averaged (for more procedural details, see Funder &

CHAPTER 24 24 CiJAPTER

623 623

SELF-OTHER CONGRUENCE SELF-OrnER

TABLE II Self-Placement Self-Placement Higher Higher Than Than Acquaintance Acquaintance Placement Placement

Q-sort item item Q-sort 46. Engages Engages in in personal personal fantasy fantasy and and daydreams daydreams 46. 3. Wide Wide range range of of interests interests 3.

SelfSelf-

Acquaintance Acquaintance

ratings ratings

ratings ratings

5.90

5.06 5.06

6.70 6.70

6.01 6.01

t/ (156) (156)

p-level p-Ievel

5.70 5.70

.001 .001

5.08 5.08

6.62 6.62

5.82 5.82

4.79 4.79

6.41 6.41

5.62 5.62

4.76 4.76

66. Enjoys Enjoys aesthetic aesthetic impressions impressions 66.

6.57 6.57

5.93 5.93

4.12 4.12

Concerned with with philosophical philosophical problems problems 90. Concerned

6.08 6.08

16. Introspective 16. Introspective

60. Insight Insight into into own own motives motives and and behavior behavior 17. 17. Is Is sympathetic sympathetic or or considerate considerate

6.64 6.64

35. 35. Has Has warmth; warmth; compassionate compassionate

7.20 7.20

29. Is Is sought sought for for advice advice 29.

47. 47. Readiness Readiness to to feel feel gUilt guilt

6.04 6.04

4.34 4.34

6.57 6.57

4.05 4.05

5.75 5.75

5.19 5.19

4.93 4.93

4.43 4.43

3.73

58. Enjoys Enjoys sensuous experiences

39. Thinks and associates ideas in unusual unusual ways

5.45 5.45

79. Has persistent preoccupying thoughts

5.94 5.94

5.48 5.48

5.45 5.45

6.69 6.69

6.26 6.26

64. Perceptive to interpersonal interpersonal cues

5.90 5.90

5.45 5.45

19. 19. Seeks reassurance

5.65 5.65

5.28 5.28

71. High aspiration level

6.76 6.76

89. Compares self self to others

5.66 5.66

10. 10. Anxiety Anxiety and and tension tension produce produce bodily bodily symptoms symptoms

95. Gives advice

83. 83. Able Able to to see see to to the the heart heart of of important important problems problems

24. 24. "Objective," "Objective," rational rational

5.02 5.02 6.37 6.37

3.78 3.78

3.15 3.15

2.88 2.88

2.79 2.79 2.66 2.66

2.59 2.59

2.56 2.56

5.34 5.34

5.02 5.02

2.47 2.47

4.55 4.55

4.21 4.21

2.15 2.15

5.83 5.83

5.50 5.50

5.41 5.41

.05

2.49 2.49

5.28 5.28

5.68 5.68

.01 .01

2.26 2.26 2.10 2.10 2.05 2.05

Note, p-values p-values are are two-tailed. two-tailed. Note.

items placed higher in the self-ratings are shown in Table I; I; those items placed II.^*^ higher by the acquaintances are shown in Table lIP eviWhat would traditionally be labelled a self-enhancement bias does seem evi dent in these tables. Acquaintances' ratings are higher than self-ratings, on average, Selfon items such as "expresses hostility" and "dissatisfied" and "self-pitying." Self warmth" and "is perceptive." This self selfratings are higher on items such as "has warmth" be demonstrated demonstrated over over the theentire entireset setof of 100 100items: items:Ratings Ratings enhancement tendency can be of the favorability of each of the 100 items were correlated with each of the scores; the resultant rr was .37 (p < .(01), .001), a self-acquaintance mean difference scores; direct measure of the general tendency of targets to rate themselves higher on favorable traits than their acquaintances did. This rr might have been even higher but for a contrary effect. As will be discussed below, subjects seem to give themselves higher ratings than do acquain2^ All All probability probability levels levels reported reported in in this this chapter chapter are are two-tailed two-tailed values. values.

^ The The data data in in these these tables tables have have not not been been previously previously reported; reported; studies studies from from the the same same dataset dataset by by

3

Funder and and Dobroth Dobroth (1987) (1987) and and Funder Funder and and Colvin Colvin (1988) examine examine correlational correlational agreement agreement rather rather Funder than mean mean differences. differences. than

624 624

FuNDER AND COLVIN FUNDERANDCOLVIN

TABLE TABLED n Acquaintance Placement Higher Than Self-Placement Self-Placement Q-sort item 100. 36. 94. 45. 88. 52. 76. 7. 86. 9. 57. 99. 4. 59. 49. 97. 21. 14. 75. 61. 78. 74.

Does not vary roles Is subtly negativistic Expresses hostility Has brittle ego-defense system Personally charming Assertive Projects own feelings and motivations onto others Favors conservative values Denies unpleasant thoughts and conflicts Uncomfortable with uncertainty Interesting, arresting person Self-dramatizing Self-dramatizing Talkative individual Concerned with functioning of own body Distrustful of people Emotionally bland Arouses nurturant feelings Genuinely submissive Internally consistent personality Creates and exploits dependency in people Self-pitying Satisfied with self self

Selfratings

Acquaintance Acquaintance ratings

t (156)

p-level p-Ievel

3.53 2.87 2.87 3.34 3.32 5.45 5.35 4.10 4.11 3.52 4.36 5.92 3.66 5.52 5.24 3.32 2.99 4.04 2.75 5.02 2.78 2.70 4.55

4.40 3.43 3.92 3.81 5.91 5.82 4.60 4.62 3.96 4.76 6.25 4.06 5.89 5.59 3.66 3.38 4.42 3.11 5.39 3.09 3.00 4.86

-4.85 -3.68 -3.62 -3.38 -3.37 -3.04 -3.04 -3.01 -2.68 -2.67 -2.61 -2.60 -2.52 -2.36 -2.30 -2.29 -2.26 -2.13 -2.09 -2.04 -2.02 -2.00

.001

.01

.05

Note. Note. All p-values are two-tailed.

tances on traits that are relatively internal, internal, or not outwardly observable, whereas acquaintances give subjects higher ratings on observable traits than do the subjects themselves (the r between observability and self-acquaintance mean differences is .25, p < .05). -.25, .05). Within the 100 items of the Q-set, observability and favorability 32, p < .01); ways" or are positively related (r = = .32, .01); items like "thinks in unusual ways" "has persistent preoccupying thoughts" thoughts" are rated rather unfavorably and are also quite internal, whereas traits like Uke "talkative" "talkative" and and "assertive" "assertive"are are both both more more favor favorable and observable.(At (Atpresent, present,ititisisnot notcompletely completelyclear clearwhether whether and more moreexternal externalor orobservable. this relation holds generally across the trait domain or is specific to the 100 items in the Q-set we use.) To disentangle the conjoined but opposing effects of favorability and observobserv correlations. Across the 100 items, the correlation ability, we computed partial correlations. between favorability and self-acquaintance mean differences, with observability partialled, is r = = .49 (p < .001). (The correlation between observability and self self[p < .001].) = -.42 (p acquaintance mean differences, with favorability partialled, is r = seems to to provide provide convincing convincing evidence that that people will, will, indeed, indeed, provide provide This finding finding seems .

CHAPTER 24 24 SELF-OTHER SELF-OTHERCONGRUENCE CONGRUENCE CHAPTER

625 625

self-descriptions than other people will provide of them, separate more favorable self-descriptions from the influence influence of trait observability. The *'other "other people" in the results just cited were close acquaintances and, in most cases, close friends (the request we had made to our subjects was to help us recruit "the two people, among those on campus and available, who know you the best"). As was reported by Funder and Colvin (1988), each subject was also rated by judges who had seen them only on a 5-min videotape which showed him or her interacting spontaneously with another subject. The comparison between these "strangers' " judgments and self-judgments is dramatic. The self-descriptions were self-stranger much more favorable; across the 100 traits, the correlation between self-stranger differences and favorability was rr = = .66 (p < < .(01). .001). Interestingly, the compari comparimean differences son between the close acquaintances' ratings and the strangers' ratings yielded acquaintance-stranger mean differ differnearly the samefinding; finding; the correlation between acquaintance-stranger ences and favorability was rr = = .60 (p < < .(01); .001); acquaintances described the targets favorably than did strangers, to nearly the same extent as the self. much more favorably self-enhancement effect effect is These findings lead to two conclusions. First, the self-enhancement much stronger when comparing self-ratings to ratings by strangers than when the comparison is to ratings by close acquaintances. This is only to be expected; it is reasonable to suppose that people are generally viewed positively by their close friends and acquaintances. "selfThe second conclusion may be more interesting and important: The "self" enhancement bias may be poorly named, because the effect effect seems similarly strong whether the comparison is between self-ratings and strangers' ratings, or between acquaintances' ratings and strangers' ratings. A major controversy in this area has self-enhancement effects are motivational or cognitive been the question of whether self-enhancement & Ross, 1975). 1975). The The present present findings findings imply imply that thatifthe if the mechanism mechanism (Kunda, 1987; Miller & self-other mean differences differences is motivational, it is a motivational mecha mechathat produces self-other of an unusual sort. The motivation to enhance the self, if there is such a thing, nism of seems to be experienced to a nearly equal degree by one's close acquaintances. Are we motivated to think well of our acquaintances (cf. (Cialdini & Richardson, 1980; Tesser & Campbell, 1982), or is something else going on? The data currently available are not sufficient sufficient to answer this question. But we would like to offer an unusual speculation. Perhaps the self-enhancement self-enhancement "bias" is not a bias at all-not all—not motivational, and not cognitive, either. This mild heresy can yield a parsimonious explanation if one assumes, reasonably we think, that positive attributes generally are correct characterizations of the subjects in most research (and certainly in our own), all of whom have, after all, passed the prescreen prescreening of college admissions committees. The subjects are in the best position to know these favorable facts about themselves, close acquaintances are in nearly as good a position, and the poor strangers, reduced in our research to basing judgments on on 5-min observations of rather stilted, uncomfortable interactions, are simply not in an equally good position to see the truth. Indeed, if the strangers try to make appropriately regressive judgments, and guess the (unselected) population mean, their ratings will be both less positive and less accurate than the ratings provided

626 626

FUNDER AND COLVIN FUNDERANDCOLVIN

by the student targets and their acquaintances. (The same line of of reasoning can be extended extended to the interpretation interpretation of of many findings concerning interpretations for for success and failure. For the vast majority majority of college student subjects-who subjects—who are generally successful successful people or they would not be in the subject pool-personal pool—personal attributions for for success and situational situational attributions for for failure failure are probably correct correct situations.) in nearly all except artificial, experimental situations.) B. The Actor-Observer Actor-Observer ERect Effect Another Another mean difference difference between others' and self-judgments self-judgments of of personality, the the "actor-observer" **actor-observer*' effect, effect, was popularized popularized by Jones and Nisbett (1987/1971; (1987/1971; see Watson, 1982, for for a later review). The effect effect is that people ordinarily offer offer explana explanations for for their behavior that are relatively situational, whereas observers are more more likely to explain behavior in more dispositional terms. Although this effect effect is not uniformly found found (Monson & Snyder, 1977; Robins, Spronca, & Mendelsohn, 1996), uniformly found seems to involve point of of view. Several Several the basic mechanism when it is found from the point of of view of of the person person studies have shown that observing a behavior from performs it leads to situational attributions relative to observing observing it from from an who performs manipulated perspective via camera placement; placement; outside perspective. Storms (1973) manipulated manipulated perspective through through instructions to subjects subjects Regan and Totten (1975) manipulated to "empathize" with the target person; Funder (1980b) examined perspective as a function of of the observers' observers* levels of of dispositional dispositional empathy; and Krones and Funder Funder function manipulated observational perspective as a function function of time-viewing time—viewing one one(1989) manipulated self in the past or future future seems to cause one's attributions to become similar to self of an outside observer (and, hence, more dispositional). As Hirschberg Hirschberg (1978) those of noted, at the moment of of action "people are interested in deciding what to do" do" (p. 58), not analyzing how their traits affect affect what they do. Only later, if ever, can actors take an external, more objective view and see how other people might have behaved differently differently in the same situation, and therefore therefore how their own traits affected affected behaved acted themselves. how they acted discussed by Monson and Snyder (1977) and by Funder Funder (1982), the the As was discussed dispositional-situational dispositional-situational "dichotomy" is not a true dichotomy anyway. To say that that someone someone gave money money because because he is generous (a disposition) implies that the giving situation was voluntary voluntary and not at gunpoint; to say that someone gave money because she was asked by a Girl Scout (a situation) implies that the subject is the because of person who responds generously to such a request. These kinds of of attributions sort of contribution of both dispositional and situational influences influences implicitly recognize the contribution determination of of behavior. When the situational response is rare, one may to the determination legitimately suspect a dispositional inference, whereas if it is common, one might difference is in emphasis more more put more emphasis on the situation, but overall the difference for discussions of of this point). than content (see Funder, 1982, and Ross, 1977, 1977, for Nonetheless, the strong strong tradition within attribution theory theory is almost uniformly uniformly to assume dispositional attributions to be wrong, and situational attributions to be right (Funder, 1982; 1982; Nisbett, 1980; see als Miller & Porter, 1980; 1980; Moore, Sherrod, Sherrod,

CHAPTER 24 24

SELF-OntER SELF-OTHER CONGRUENCE CONGRUENCE

627 627

Liu. Liu, & Underwood. Underwood, 1979). Within this literature. literature, the tendency tendency to make dispositional dispositional attributions is often often referred referred to as the "fundamental "fundamental attribution attribution error" (Ross. (Ross, 1977). attributions Remarkably. Remarkably, this practice persists even though though attribution attribution research has identi identified many errors that produce attributions situational, not disposi attributions that are unduly situational, dispositional. One example is the false consensus bias (Ross, Greene, & House, 1977). effect leads people to see their their behavior as more more common common in the population This effect beliefs that one's than it really is. Because such a tendency tends to yield (false) beliefs behavior is simply "what "what anybody would do" d o " in the same same situation, it yields own behavior situational and less dispositional than attributions that are more situational than they normatively should be. correct nor thereSituational attributions are neither always correct nor immune to error, there fore, and in general the actor-observer not be regarded as a matter actor-observer effect effect should not matter of accuracy (Funder, 1982). However, there may be some exceptions even to this of consider a behavior that that is part of of a chronic, maladaptive maladaptive general rule. For example, consider pattern. If the individual who performs performs it sees such a behavior as situationally considered to be wrong. An An example was caused, he or she can reasonably be considered Funder (1989), who showed provided by McKay, O'Farrell, Maisto, Connors, and Funder tended to explain the causes of of that previously hospitalized, long-term alcoholics tended of immediate situational influences influences (a stressful stressful recent drinking relapses in terms of day at the office, a car breakdown). Their wives, however, were more likely to see Given that that most individuals do not the drinking as produced by the alcoholism. Given office with lengthy drinking bouts, and and that that these respond to bad days at the office diagnosed and treated treated for for alcoholism over a long particular individuals had been diagnosed situational period, it is hard not to conclude that in this example the alcoholics' situational attributions were correct. attributions were incorrect, and the wives' dispositional attributions On the other side of the actor-observer actor-observer divide, there is at least one aspect of behavior that actors are in an exceptionally good position to witness. Nobody of including personality psychologists would deny that people vary their behavior according to the person they are interacting with, and actors are in a better position to see this variation than is any single interaction interaction partner partner (Swann, 1984). This effect effect is reflected reflected in the Q-sort Q-sort data reported above. Self-ratings Self-ratings of of the item "Does not vary roles; relates to everyone in the same way" receives a lower rating in self selfdescriptions than in descriptions by either acquaintances acquaintances or strangers (the means, respectively, are 3.53, 4.40, and 5.15; p p for for both self-other self-other differences differences < < .001). C. Internal versus External Traits c.

The third category of of mean differences differences between between others' and self-judgments self-judgments of of personality has received considerably less attention attention than the first two. Funder Funder classified the 100 items of the California California Q-set of "out "out(1980a) classified Q-set along a dimension of or "charming" rated high, ward observability," on which items such as "talkative" or "fantasizes and daydreams" or "ruminates "ruminates and and worries" rated low. and items like "fantasizes of the latter kind are more observable observable by the people It seems obvious that traits of who possess them than by acquaintances, because fantasizing, worrying, and other

628 628

foNDER FUNDER AND COLVIN COLVIN

internal activities are only accessible to acquaintances to the extent that the persons who perform them let others in on the secret. It might be less obvious that the opposite comment could be made about more observable traits: traits: attributes such as talkativeness, attractiveness, and charm may be more observable to other people himself or herself. A particularly good example might be "is than to the actor himself charm;ififyou youbelieve believeyourself yourself We are are all allimmune immune to to our ourown own charm; personally charming." We to be charming, it is presumably not because you have charmed yourself, but rather because other people have responded to you in such a way that you conclude you must be charming. Our own access to this kind of external trait is almost as indirect as is the access of an acquaintance to our internal traits. This line of reasoning leads to the prediction that people should give higher ratings to to themselves themselves on oninternal internaltraits traitsthan thando dotheir theiracquaintances, acquaintances,whereas whereasacquain acquainwill tend to togive givehigher higherratings ratingsto topeople peopleon onexternal externaltraits traitsthan thanthe theindividuals individuals tances will award themselves. This prediction was confirmed in the study by Funder (1980a) replicated in in our ourmore morerecent recentdata. data.Funder Funder(1980a) (1980a)reported reportedthat thatthe thecorrelation, correlation, and replicated across 100 Q-items, between the difference between acquaintances' and self-ratings the outward outward observability observability of of each each item itemon onthe the other, other,was was.44. .44.InIn on the one hand, and the our more more recent recent data data this this correlation correlation was was .25 .25 (p but when when the the counteracting counteracting our (p < .05), .05), but influence of of favorability favorability (discussed (discussed above) above) was was partialled, partialled, this this correlation correlation rose rose to to influence r= = .42 .42 (p (p < < .001). .001). r Thesefindings findings seem to indicate a fundamental difference in the way we look at our own personality, relative to how it is viewed by others. From our own perspective, our internal, private experiences and mental activities seem a more important and salient part of what we are than they seem to acquaintances who must view these attributes from the outside. But conversely, certain attributes of our personality that we we project project outward outward in in our oursocial socialbehavior, behavior,such suchas astalkativeness, talkativeness, assertiveness, and charm, are more visible and perhaps even more important to our acquaintances than they are to ourselves. Like the dispositional-situational dispositional-situational of accuracy, "dichotomy," the difference difference here seems better characterized not as one of but of point of view. m. CORRELATIONAL ANALYSES OF AGREEMENT m.

The second way to assess self-other congruence in personality judgments is to calculate the correlation between the two sets of judgments. This correlation can can be computed on personality profiles or on individual variables or items. The first first method assesses the similarity between the complete set of personality judgments set of judgments made by an an acquaintance (e.g., Andersen, made by the self and the set & LaVoie, 1984). The X variables in the correlation are all the self self1984; Kenny & judgments of a particular individual, and the Y variables are all the judgments of that individual provided by a peer. This is "profile" "profile*' agreement, and is calculated for one target-judge pair at a time. The second method examines congruence one variable at a time. Instead of comparing whole profiles, this method correlates

CHAPTER 24 CHAPTER

SELF-OTHER CONGRUENCE SELF-OrnER

629 629

acquaintances' and self-judgments self-judgments on a single variable (e.g., Funder, 1980a; Funder & Colvin, 1988; 1988; Funder Funder & & Dobroth, Dobroth, 1987). 1987). The The X X variables variables in in this this analysis analysis are the self-judgments self-judgments of all the subjects in the sample on this one variable. The offered by the subjects' Y variables are the corresponding judgments on this variable offered acquaintances. The profile and variable or item methods each have advantages and disadvantages, as we shall see (see Bernieri, Zuckerman, Koestner, & Rosenthal, 1994, 1994, for for a comparative analysis). A. A. Methodological Methodological Issues: Issues: "Cronbach's ^^Cronbach's Complaint" Complaint"

of social psychology The study of self-other congruence was a thriving subarea of from the 1930s into the early 1950s (see Taft, 1955, for for a review). Work in this from area was brought up short, however, by the methodological critique published by (1955; Gage & Cronbach, 1955; foreshadowed by Hastorf Hastorf & Bender, 1955; foreshadowed Cronbach (1955; profile similarity scores that were calculated 1952). Cronbach demonstrated how the profile in nearly all studies might be contaminated to an unknown but probably large ^'stereotypic accuracy," "elevation," and degree by extraneous influences including "stereotypic "differential elevation." The dramatic effect effect of Cronbach's critique was to render "differential self-other congruence to that date seemingly uninterpretable. nearly all research on self-other The even more dramatic response of researchers on accuracy was to cease work on the topic, abruptly and nearly completely (Funder, 1987; Schneider, Hastorf, & Ellsworth, 1979). 1979). "Accuracy" had gotten a bad name. There was no real reason why this had to happen. The artifacts Cronbach identified applied only to measures of profile similarity, and were not particularly identified is hard to understand, difficult to obviate, in principle, in any case (Wiggins, 1973). It is difficult scientific community community to Cronbach's 35 years later, why the general reaction of the scientific article was one that could be described only as panic. One factor factor might have been the style in which this influential critique was written. The article was highly critical and even slightly sarcastic, but also rather murky and hard to follow in many spots. Moreover, many of its its analyses were presented in a highly (and perhaps needlessly) mathematized manner, including numerous equations utilizing unconventional nono tation. The result, apparently, was that many readers were as much intimidated as informed. informed. The substance of Cronbach's complaint was that profile similarity scores were typically influenced influenced by several factors aside from from the judge's ability to accurately eleva differential elevadiscriminate properties of the target. Two of these, elevation and differential tion, referred referred to the effect effect of shared response styles between judge and target; ifif the subject and his or her acquaintance happened to use the scale in a similar manner, their similarity scores would be artifactually artifactually increased; if they used the scale in a different different manner, their similarity would be artificially artificially lessened. Another Cronbachian confound confound was stereotype accuracy: The similarity of a judge's rating subject to the subject's self-rating self-rating can be influenced influenced to an important degree of a subject by the extent to which the judge's rating resembles the average self-rating of all fair amount of accuracy, in terms targets. In other words, a judge can usually earn a fair

630 630

FuNoER AND COLVIN FUNDERANDCOLVIN

of of profile profile similarity, just by guessing the mean for every rating, and ignoring the individual target of of judgment judgment altogether.4 altogether."* Given due care, these artifacts artifacts are not difficult difficult to eliminate. Those involving elevation can be removed by using forced-choice removed forced-choice rating techniques (e.g., a Q-sort) Q-sort) that constrain the ratings of all judges to have the same mean and variance across constrain of matter of of stereotypic accuracy is somewhat more complex, but can be items. The matter approached by at least three different different ways, depending on what the investigator investigator approached focus of of interest. regards as the focus 1. / . The The Social Social Relations Relations Model Model

An important important contribution contribution to the study of self-other self-other congruence is the "social '*social introduced by Kenny and LaVoie (1984; Kenny, 1994). The pur purrelations model*' model" introduced of this analytic model is to account explicitly for for every identifiable identifiable source of of pose of variance in others' and self-ratings. This purpose requires a complete, randomized randomized **round robin" design in which all targets are judged by all raters. The data are "round of variance, which yields then entered into a model based closely on the analysis of of variance accounted accounted for for by the judge, the target, and their inter interproportions of action. The main advantage of of this approach approach is its sophistication and thoroughness. Cronbachian components components Everything that can be pinned down is pinned down, and Cronbachian of variance are not eliminated but rather are separately estimated. The model has of for example, the degree to which people can yielded some interesting insights into, for rated by individual judges, as opposed to judges in predict the way they will be rated general (DePaulo, Kenny, Hoover, Webb, & Oliver, 1987), and the degree to which "target variance" can be found found even within ratings provided provided after after only minimal minimal valid "target acquaintance (Kenny, Albright, & Malloy, 1988). Still, this technique has several disadvantages. First, use of the model imposes serious procedural burdens. Because all targets must be evaluated by all judges, it is difficult difficult to do studies in which targets are judged by individuals who know them well (in fact, we are aware of of no such studies to date). Second, the results of of an analysis with this model do not not yield measures of of agreement, such as correlation correlation coefficients, that are easily communicated. Like item analyses, the method does not yield individual accuracy scores for for either targets or judges. The model seems to do a better job at comparing the relative proportions proportions of of variance accounted accounted for for by various sources under specific circumstances than at reflecting reflecting simply how much congruence there is between ratings. found may not be straightfor straightforFinally, even the relative proportions that are found sometimes wardly interpretable. Analyses employing the social relations model have sometimes four . . . studies there is at least twice as much yielded conclusions such as ''in "in four partner variance as actor variance" (Kenny & LaVoie, 1984, p. 154). While such partner ^ItItisispossible, possible,and andreasonable, reasonable,to toregard regardstereotype stereotypeaccuracy accuracyas asbeing beingnot notan anartifact artifactatatall, all,but but 4 an important important component component of of valid valid judgment judgment ((Jackson, 1982). Specifically, Specifically, it it can can be be viewed viewed as as composing composing an Jackson, 1982). the component of accuracy that stems from a judge's knowledge of people in general.

24 SELF-OrnER SELF-OTHERCONGRUENCE CONGRUENCE CHAPTER 24

631 631

statements are indeed accurate in the context of the numbers obtained by particular studies, one lesson we should have learned from the person-situation controversy is that it can be highly hazardous to move directly from the statistical statistical apportionment of of variance in a particular study to the conceptual conceptual apportionment of variance in general (Golding, 1975; similar similar problems problems bedevil bedevil behavioral behavioral genetics, genetics, Hirsch, Hirsch, 1986). 1986). For instance, the amount of of variance contributed by any source is critically influ influenced by its range. range.^5 A A social relations study that uses targets who are relatively similar to each other will find less target variance than a study that uses targets who are relatively different judges and "partner different from each other-the other—the same goes for forjudges ,, 6 variance."^ interaction And the most complex term in the analysis always includes variance. an error component that cannot be separated. 2. Pro Profile Partial Con-elations Correlations 2. file Partial

The second method for assessing self-other self-other congruence is relatively simple and yields an accuracy score for each target-acquaintance pair. It is simply to calculate a partial correlation between each set of acquaintance's and self-judgments, across across items, correcting for both the average self-description self-description and the average acquain acquaintance's description (if what is of interest is the (sheer phenomenon of congruence), calculating a semipartial correlation correcting for the average average self-judgment self-judgment (if (if of interest interest is the ability of the judge to discriminate among how different different what is of targets describe themselves), or a semipartial correlation correcting for average acquaintance's judgment (if (if what is of interest is the ability of the target to discrimi discriminate how he or she is viewed differently differently from other targets). In our experience, acquaintance judgment are highly correlated, the average self-judgment self-judgment and average acquaintance of partial correlation so as a practical matter it matters little exactly which kind of self-other one elects to compute. This procedure yields a congruence score for every self-other pair. The score can be correlated with properties of the target or acquaintance that are regarded as potential moderators of congruence, or the mean congruence scores can be compared between experimental conditions in studies that manipulate some someaffect congruence. thing believed to affect J. Item-Level Item-Level Analysis Analysis 3. This last technique is, on the one hand, a method for obviating the influences of stereotype accuracy, and, on the other hand, an approach for studying the differences differences ^To Totheir theircredit, credit,Kenny Kennyand andLaVoie LaVoie(1984) (1984)explicitly explicitlyacknowledged acknowledgedthis thislimitation limitation(p. (p.174); 174);see see 1994. 6 ^To Toexpress expressthis thispoint pointwith withmore moreprecision: precision:Even Evenininaafully fullycrossed crosseddesign design(every (everysubject subjectserves serves as as aa judge judge and and as as aa target targetof ofall allother othersubjects), subjects),actor actorvariance variancewill willbe berestricted restrictedto tothe theextent extentthe thesample sample of of subjects subjects is is homogeneous homogeneous with with respect respect to to the the properties properties that that are are judged: judged: partner partner variance variance will will be be restricted to the sample restricted to the the extent extent the sample of of subjects subjects is is homogeneous homogeneous with with respect respect to to properties properties that that affect affect how one judgments. Very Very little is known known about about how how these these two two kinds kinds of of properties properties might might differ differ how one makes makes judgments. little is from assume that from each each other, other, but but they they are are not the the same, same, and it it is is probably probably not safe safe to to assume that they are are equally equally variable variable in in aa given given subject subject sample. sample. Yet, Yet, this this assumption assumption isis fundamental fundamental to to some some interpretations interpretations of of results results from the the social social relations from relations model. model. S

also Kenny, Kenny, also

632 632

FUNDER AND COLVIN COLVIN FuNDER

between traits. The most simple method for dealing with the influence of stereotype accuracy is to correlate correlate item ratings ratings instead instead of individual profiles. When a correlation correlation is computed between acquaintances' and self-ratings self-ratings on a single item, as previously described, the result is a number completely immune to enhancement by stereotype agreement. In fact, if all subjects in the sample simply guess the mean "stereotype," then this correlation will approach 0 (technically, it would be undefined). In general, any tendency by the raters to give the same, stereotypic ratings to all targets will severely attenuate, not enhance, item correlations. The disadvantage of this method for each individual, nor for for each judge. is that it does not yield an accuracy score for fact makes cumbersome any attempt to investigate properties of the target That fact self-other agreement. or judge that might tend to enhance self-other However, disadvantage in one case is turned turned to advantage in another. This method of analysis has been used extensively by Funder (1980a), Funder and Colvin Colvin (1988), and Funder and Dobroth (1987) to investigate the differences differences between items that lead to better and worse self-other self-other agreement. In all three studies, for example, it was found found using this analysis that more observable traits yield higher self-other self-other agreement. In addition to being a useful useful method for studying differences differences between items, it also benefits from from simplicity of use and ease of communication. B. B. Substantive Substantive Issues Issues 1. Does Self -Other Congruence Self-Other CongruenceExist? Exist?

This basic question continues to arise occasionally. In an often-cited often-cited review, Shrauger and Schoeneman (1979) concluded that "there is no consistent agreement between people's self-perceptions self-perceptions and how they are actually viewed by others" (p. 549), a conclusion also reached by Bourne (1977) and by Kammann, Smith, Martin, and McQueen (1984). (1984). As mentioned earlier, Shraugher and Schoeneman failed failed to distinguish between the two fundamentally fundamentally different different kinds of agreement considered in this chapter. Moreover, their review failed to cite several studies that did find an important degree of congruence (e.g., Fiske & Cox, 1960; Hase & Norman, 1969; 1969;Norman Norman & &Goldberg, Goldberg, 1966; 1966;Scott Scott&&Johnson, Johnson,1972), 1972), Goldberg, 1967; Norman, 1984; Bem & Allen, 1974; and many others have appeared since (e.g., Andersen, 1984; Bemieri et al., 1994; Bledsoe Bledsoe & & Wiggins, Wiggins, 1973; 1973; Borkenau Borkenau & & Liebler, Liebler, 1993b; 1993b; Cheek, Cheek, Bernieri aI., 1994; 1982; Conley, 1985; Edwards & Klockars, 1981; Funder, 1980a; Funder & Colvin, 1985; 1988; 1988; Funder & Dobroth, 1987; 1987; Goldberg, Norman, & Schwartz, 1980; 1980; Gormly, 1984; Kenrick & Stringfield, 1980; 1980; Marsh, Marsh, Barnes, Barnes, & & Schwartz, Schwartz, 1980; 1980; Gormly, Gormly, 1984; 1984; Kenrick & Stringfield, 1980; 1980; Marsh, Marsh, Barnes, Barnes, & & Hocevar, Hocevar, 1985; 1985; McCrae, McCrae, 1982; 1982; Mon Monson, Tanke, & Lund, 1980; 1980; Moskowitz, 1990; 1990; Park & Judd, 1989; 1989; Paunonen, 1989; Paunonen & Jackson, 1985; Watson, 1989; Woodruffe, 1985). It appears appears that that self selfPaunonen Jackson, 1985; Watson, 1989; Woodruffe, 1985). It other congruence in personality ratings is in fact a fairly robust phenomenon. The only times it it is is not appear to to be be when overly strong strong self-presentational self-presentational only times not found found appear when overly pressures are present, unreliable rating scales are used, or the judges and and their their pressures are present, unreliable rating scales are used, or the judges targets have not had a chance to become acquainted with each other (Funder, targets have not had a chance to become acquainted with each other (Funder, 1980a, 1987). 1980a, 1987).

CHAPTER CHAPTER 24 24

SELF-OrnER SELF-OTHER CONGRUENCE CONGRUENCE

633

2. Congruence? 2. What Moderates Self-Other Self-Other Congruence?

self-other congruence in personality If we can accept that a substantial degree of self-other judgments is a fairly be fairly typical finding, the next and more constructive question becomes, what kinds of variables make the degree of congruence larger and smaller? agreement (and At the present time, we organize potential moderators of self-other self-other agreement more generally, potential moderators of judgmental accuracy) into four four categories, which we call good judge, good target, good trait, and good information. a. Good Judge. Judge. This is perhaps the most obvious moderator moderator one might want to examine, and historically it has received the most attention: Is there such a thing so, what are the properties of the good judge? as a good judge of personality, and, if so, Bronfenbrenner et a1. al. (1958) In an excellent review of the research to that time, Bronfenbrenner unequivocal answer to these concluded that the extant data did "not permit an unequivocal different today, in part because research of questions." The situation is not much different of the sort that could have addressed this question died out almost completely, ifif temporarily, from the late 1950s until quite recently. for investigators to Because the general reaction to Cronbach's article was for abandon the field rather than to improve their methodology, the question still has yet to receive the attention that it deserves, given its importance. More than three decades after after Bronfenbrenner's Bronfenbrenner's review, it still seems premature to close the book found reliable individual on the good judge of personality. In fact, a recent study found differences in judgmental ability that related to ratings of judges' concern about differences interpersonal relationships (Vogt & Colvin, 1996). This promising start suggests it of personality. may be a good time to take a second look at the good judge of for the generalized generalized good judge might Whatever the ultimate fate of the search for be, more recent research does suggest that certain judges might be particularly be, good at judging certain traits. A particularly interesting study is one by Park and for whom a certain trait term is "chronically Judd (1989), which found that judges for often and spontaneously, tend to yield judg judgaccessible," that is, who use the trait often ments of that trait that agreed highly with the target's own self-judgments. For instance, a judge who often often uses uses the the dimension dimension "intelligent/conscientious" "intelligent/conscientious" to tocharac characterize people also tends to rate this dimension in a way similar to the way targets rate themselves. It is interesting to combine these results with those by Lewicki (1983, 1984), (1983,1984), who has found found what he calls "self-image bias" which leads people to evaluate others using those trait terms they believe to be most desirable in themselves. The conclusion would seem to be that people will judge those traits they see as desirable in others, because, according according to to Lewicki, Lewicki, they theybecome become in themselves more accurately in Pratto, what Park and Judd would call chronically accessible traits (see also Bargh & Pratto, 1986; Higgins, King, & Mavin, 1982). 1986; 1982). Research in this area seems still to be in its sufficient to make plausible the idea that different different people people early stages, but does seem sufficient different traits more accurately. The possibilities have particular abilities to judge different further investigation seem promising, indeed. One possibility that deserves for further

634 634

FuNDER AND COLVIN FUNDERANDCOLVIN

of experience, investigation is that, if "chronicity" and resultant accuracy are a result of then some kind of of training might be possible to simulate this experience and improve the accuracy of of judgment for at least one trait at a time.

b. Good Good Target. Target A A second possible moderator of self-other self-other congruence is something that could be called "judgability": perhaps some individuals are easier A prominent investigation into this possibility was the to judge than are others. A study by Bern Bem and Allen (1974). These investigators simply asked their subjects, "how consistent are you?" on the traits of of friendliness friendhness and conscientiousness, and found that self-other self-other agreement on ratings of of these traits (as well as other correla correlations between judges, between behaviors, and between judgments and behaviors) was higher within the self-identified self-identified consistent group than in the inconsistent group. Does this finding replicate? Chaplin and Goldberg Goldberg (1985) reported an exhaus exhausself-rated consistency moderated tive attempt at replication that failed to find that self-rated self-other agreement. But Bem and Allen's basic result has been replicated in several self-other other studies, including Campbell (1985), Cheek (1982), Kenrick and Stringfield al. (1988) performed a (1980), and even Mischel and Peake (1982). Zuckerman et a1. meta-analysis of of several studies in this area and concluded concluded that the aggregate result was what they considered a small, but still significant, effect: there did seem to be a self-peer agreement. positive relationship between self-reported consistency and self-peer Later investigations by Zuckerman and his colleagues have tried to pin down of this effect. Zuckerman, Bernieri, Bemieri, Koestner, and Rosenthal (1989) investi investithe basis of gated three potential moderators of self-peer self-peer agreement: self-reported self-reported trait rele relevance, consistency of of behavior, and observability of behavior. All three of these variables (when they were rated through a ranking procedure) were found to have significant significant moderator effects. (And, the observability observability finding is consistent with several studies considered, further on, under the heading Good Trait. In futher research, Zuckerman, Miyake, Koestner, Baldwin, and Osborne (1991) found that individuals who see themselves themselves as particularly unusual on a given trait dimension tend to yield the best self-other self-other agreement agreement in ratings of that dimension. Cheek (1982) examined several personality personality variables as possible moderators of self-other congruence. The most promising results he reported were for the of self-other "acting ability" subscale of Self-Monitoring. Subjects earning high scores on this '^acting of agreement on each of four different scale were described with better self-other self-other different personality traits. Cheek interprets this finding as a reflection of of acting ability abiUty as a social skill: "If ''If social skill leads to successful communication of of one's self-image self-image . . .. then those who are socially skilled should have stronger agreements between between self-ratings self-ratings and peer ratings than those who are not so skilled" (p. 1265). Closely related to Cheek's notion of acting ability is the concept of "self "selfdisclosure." Jourard (1971) suggests that self-disclosure self-disclosure is the process by which a person reveals his or her "thoughts, feelings, hopes and reactions to the past" (p. 5) to another person. Jourard (1971) has conducted empirical studies that indicate that there are individual differences in the amount that people disclose and the type of of information information that is disclosed. A A more recent study by Koestner, Bernieri, Bemieri,

CHAPTER 24 24 SELF-OrnER SELF-OTHERCONGRUENCE CONGRUENCE CHAPTER

635 63S

and Zuckerman (1989) demonstrates that those traits which a person reports to be similar between his or her public versus private self self also tend to be judged with better agreement between the self and his or her acquaintances. self-other In a very unusual study, Davidson (1993) presented evidence that self-other disagreement disagreement may occur when individuals engage in repression or suppression. In the former case, people will avoid particular thoughts and feelings but still express them through observable behaviors. Consistent with this idea, Colvin (1993a) argued that individuals who manifest a discrepancy between their private inner self and public outer self will be relatively difficult to judge, as manifested in lower self-other self-other agreement. In the latter case, suppressors will avoid particular thoughts and feelings as well as their behavioral expression, although they may still respond to these "forbidden" ideas on a physiological physiological level. The resulting lack of self-knowledge self-knowledge can be expected to produce discrepancies between the views of personality by the self 1982; Colvin, 1993a). self and by others (Cheek, 1982; Colvin's research is an attempt to integrate the literature on "judgability" (1993a, 1993b). In his first study, Colvin employed a multioperational, person personcentered approach that demonstrated reliable reUable individual differences in self-other self-other agreement about entire personalities, rather than just specific traits. Individuals found to be most judgable, in this sense, were independently characterized as generrelatively extraverted, agreeable, emotionally stable, and conscientious. More gener ally, judgability was related to good psychological adjustment (Colvin, 1993a). A further study showed that individual differences in judgability were stable A from age 18 to age 23 (Colvin, 1993b). More important was the finding that adoles adolespsychological adjustment) predicted cent ego-resiliency (a cocept closely akin to psychological judgability during young adulthood ((rr = = .54 and .45 for men and women, respec respectively). This result is further evidence that individuals who are well adjusted are more likely to provide descriptions of their own personalities that agree well with the consensus of ratings by their friends and acquaintances.

c. Good Good Trait. Trait A A third potential moderator of of self-other self-other congruence is the nature of the trait being evaluated: Are some traits judged with better agreement than others? Here, at least, the answer appears to be relatively simple yes. Research about what might be a good trait in this sense goes back at least to Estes (1938). Subjects attempted to judge judge the personalities of stimulus stimulus a study by Estes persons viewed in a brief movie film, and their accuracy was evaluated through comparison with judgments rendered of these persons by a panel of of clinical judges. Estes found that, for example, inhibition-impulsion was judged more accurately than objectivity-projectivity. But no clear, overall pattern of results emerged, and it was also unclear whether the same traits were easiest to judge from the film would also be the easiest to judge in real life. More recent research has been more informative and somewhat more consis consistent. Borgatta (1964), John and Robins (1993, (1993,1994), 1994), Norman and Goldberg (1966), Watson Hase and Goldberg (1967), McCrae (1982), Funder and Dobroth (1987), and Watson (1989) all showed that traits relevant to extraversion tended to manifest higher

636 636

FuNDER AND COLVIN FUNDERANDCOLVIN

self-other self-other congruence. Funder and Dobroth calculated agreement correlations be between self-ratings and ratings by close acquaintances on each of of the 100 items of of the California Set (Block 196111978). 1961/1978). They then correlated these 100 correlations correlations with factor loadings of of the 100 traits (as determined by McCrae, Costa, & & Busch, 1986). The result is an indication of the kind of trait that yields the best agreement. self-acquaintance agreement agreement was found to correlate rr = = .29 with By this method, self-acquaintance = - . .53 5 3 with neuroticism (which is itself itself correlated rr = = -.32 -.32 extraversion, and rr = with extraversion). Heath, Neale, Kessler, Eaves, and Kendler (1992) also found self-other agreement than did neuroticism. that extraversion yielded better self-other The trait property of of subjective visibility has a quite general influence. For instance, the same traits that yield better self-acquaintance self-acquaintance agreement also tend & Colvin, to yield better agreement among acquaintances (according to Funder & = .57, p p < < .(01), .001), and so it is not surprising that agreement among acquain acquain1988, rr = = .43 (p < .(01). .001). Findings consistent tances correlates with subjective visibility with rr = (p < with these have been reported by John and Robins (1993) and by Watson and Clark (1991). Moreover, Funder and Colvin asked each acquaintance to view a brief (5 min) videotape of of a subject they did not know, and try to complete a Q Qbrief of that "stranger." A A surprising amount of agreement agreement was found between self selfsort of ratings strangers* ratings ratings even even in in such such impoverished impoverished circumstances circumstances (cf. (cf. ratings and and these these strangers' Albright et et aI., al., 1988; 1988; Watson, Watson, 1989; 1989; see see next next section). section). More germane to to the the present present Albright More germane point was was the the finding finding that that self-stranger self-stranger agreement agreement correlated correlated with with subjective subjective visibil visibilpoint ity with r = .40 (p < .001), and agreement between strangers correlated with ity with r = .40 < .(01), and agreement between strangers correlated with subjective visibility with r = .42 (p < .001). The effect of subjective visibility on subjective visibility with r = .42 ( p < .(01). The effect of subjective visibility on interjudge agreement seems powerful, robust, and general.^ interjudge agreement seems powerful, robust, and generaI.7 This effect has philosophical implications, as well. If personality traits were no more than arbitrary social constructs, as implied by the sociological perspective discussed at the beginning of of this chapter, then there would be no reason to expect some of them to be more observable than others. One can "construct" a trait like "fantasizes and daydreams" as readily as one can construct a trait like "is talkative." But, as Clark and Paivio (1989) point out, if if different different raters can agree better about more observable phenomena than they can about less observable phenomena, this finding implies that something is actually out there for these different different raters to observe! The findings of of higher interjudge agreement agreement on more observable observable traits can therefore therefore be added to the steadily accumulating evidence that personality traits are real properties of of people (Funder, 1991; 1991; Kenrick & & Funder, 1988). 7^A A possibility possibilitythat thatmust mustbe beborne borneininmind mindisisthat thatless lessvisible visibletraits traitsalso alsomanifest manifestless lessvariance, variance, and that that therefore therefore the the lesser lesser congruence congruence on on such such traits traits might might be be an an artifact artifact of of restricted restricted range. range. In and response to to aa suggestion suggestion by by Lewis Lewis Goldberg, Goldberg, we we computed computed the the correlation correlation between between item item variance variance and response "visibility," which which yielded yielded rr = = .52, and and the the correlation correlation between between item item variance variance and and seU-acquaintance self-acquaintance "visibility," agreement, which which yielded yielded r = = .49. .49. However, However, the the partial partial correlation correlation between between visibility visibility and and agreement, agreement, agreement, controlling for for variance, variance, was was still still significant significant (r (r = = .39). .39). These These results results yield yield two two conclusions. conclusions. First, First, less less controlling visible traits are are less less variable, variable, and and yield yield lower lower agreement agreement partially partially for for that that reason. reason. But But second, second, the the effect effect visible traits less when when the the items items of variance variance is is not not all all there there is is to to visibility. visibility. Judges Judges seem seem not not only only to to vary vary their their ratings ratings less of less accurately. accurately. are less less visible, visible, they they also also seem seem to to vary vary their their ratings ratings less are

CHAPTER CHAPTER 14 24 St:LF-OmER SELF-OTHERCONGRUENCE CONGRUENCE

637 637

A A different sort of "good **good trait" trait" has been proposed by Gangestad, Simpson, DiGeronimo, and Biek (1992). From a functionalist functionalist perspective, these investigators propose that traits that are important for survival and reproduction would become of evolutionary processes. They found that the trait of of more judgable as a result of "sociosexuality," the tendency to be willing wiUing to engage in sexual relations in the absence of a personal relationship, was judged with better agreement than (in order) social potency, social closeness, and stress reaction. Their finding that social potency yielded better agreement than stress reaction is completely consistent with the work summarized above showing that extraversion yields better agreement than neuroticism. The finding about sociosexuality adds a new and interesting wrinkle. d Good Good Information. Information, A A final possible moderator of self-other self-other congruence congruence d. information upon which the peers' judgment is based. is the amount or kind of information One conclusion that has emerged from recent research is that even a quite small amount of information can lead to judgments that appear to have a surprising amount of validity (see Ambady & Rosenthal, 1992, for a meta-analysis of of this finding). Research on this topic has frequently employed the unfortunate term "zero acquaintance" to describe the relationship between target and judge. The term is misleading because in none of these studies is acquaintance in fact "zero," which we presume could only mean that no information about the target of of judgment was available whatsoever. Each in fact provides the judge with minimal but real real— and apparently useful-information. useful—information. The judge may have observed the target only briefly in person, on a videotape, or in a photograph, and may have heard a voice recording or watched a brief behavioral episode. In any case, the conclusion that emerges from this research is that surprisingly valid ratings can emerge from minimal observation—but certainly not not "zero acquaintance." observation-but of particular interest. Each investigated self-other self-other agreement Four studies are of in personality judgments among subjects who had little or no acquaintance with each other. Albright et al. (1988) examined ratings by subjects who sat in a small group together, but had not been given a chance to talk. These subjects were asked to describe each other's personality personaHty on each of of five traits. Despite this minimal acquaintance, Albright et al. reported "a significant significant proportion of of the variance [in ratings] was due to the stimulus target" (p. 387). Funder and Colvin (1988) looked at agreement between self-ratings self-ratings on the Q-sort and judgments provided by observ observers who had viewed the subjects for only 5 min by watching a videotape. They found that 24 out of the 100 self-other agreement correlations were significant at the nominal .05 level, or nearly five times as many as would be expected by chance. In more recent research, Watson (1989) found that subjects in small groups who had heard each other speak their names, but nothing else, nonetheless mani manifested significant self-other self-other agreement correlations for the traits of of Extraversion, Conscientiousness, and (when the ratings of of several peers were averaged) Agree Agreesuffiableness. Borkenau and Liebler (1993a) report that minimal observation was suffi cient to yield significant self-other self-other agreement on the traits of of extraversion and conscientiousness, but not neuroticism, openness, or agreeableness.

638 638

FUNDER AND AND COLVIN COLVIN FuNDER

It might not surprise the reader to learn that we believe one of the betterbetter controlled studies of the effect of acquaintanceship on interjudge agreement may have been the one by Funder and Colvin (1988). In this study, as was mentioned earlier, targets were judged by two close acquaintances and by two strangers (ac(ac quaintances of other targets) who viewed them only on a brief videotape. The advantage of this design was that the acquaintanceship variable was manipulated targets and and informants informantswere werethe the same same under underboth boththe theacquainted acquainted experimentally; the targets and the unacquainted conditions, and all that varied was the pairing between judges and targets. As was already mentioned, the self and strangers' Q-sort ratings agreed more strongly than might have been expected. But Funder and Colvin also found that more acquaintanceship led to better agreement: across all 100 items, the mean self-acquaintance r = = .27 \p < .01), whereas the mean self-stranger rr = = .05 (ns). (p < These means are significantly different, but statistics are hardly necessary because These means are significantly different, but statistics are hardly necessary because self-acquaintance agreement was higher than self-stranger self-stranger agreement on each and every one one of of the the 100 100 Q-sort Q-sort items items (but (but see see Colvin Colvin & & Funder, 1991, for for aa boundary boundary Funder, 1991, every condition for this effect). condition for this effect). Another well-controlled study, by Stinson and Ickes (1992), generalized the conclusion by Funder and Colvin. Stinson and Ickes found that friends agreed better than did strangers when trying to judge a person's thoughts and feelings. Even when the similarity between the target and the judge was statistically concon Stinson and Ickes concluded that trolled, friends still demonstrated an advantage. Stinsoq friends are more accurate in their inferences about their partner's thoughts and feelings as a result of their accumulated base of behavioral knowledge, which they use for prediction. Watson and Clark (1991) extended the acquaintanceship effect in a different direction, finding finding that well-acquainted peers agreed better than did relative strangers about each others' specifically emotional traits. Funder et al. (1995) examined, and ruled out, a couple of possible artifactual explanations for the effect of acquaintanceship on self-other and interjudge agreement. These included interjudge communication and assumed similarity. They concluded the most parsiparsi acquaintancemonious explanation for the increase in judgmental agreement with acquaintance one: that as you know somebody longer, the person ship is also the most obvious one: becomes better known.

e. Moderators Comment. anan important Moderatorsof ofSelf-Other Self-OtherAgreement: Agreement:General General Comment InIn important and thought-provoking essay, Chaplin (1991) commented that the "blind" "blind" search for moderators of agreement is likely to produce only weak results that fail to replicate. However, moderator variables firmly firmly grounded in theory are another story. When there are good, theoretical reasons to believe that a given construct will moderate the relationship between two other constructs, the operationalization of the moderator will be more similar to the "construct validity" approach (Cron(Cron bach & & Meehl, 1955) than the currently more common "moderator of the week" approach. In the latter approach, researchers gather measurements of a moderator variable that might be interesting, and might be reliably measured, and give it a shot. In the former approach, much more attention will be given to the selection

24 CHAPTER 24

SELF-OTHERCONGRUENCE CONGRUENCE SELF-OrnER

639 639

and measurement measurement of of relevant, reliable, and valid measures. Such additional effort effort Theoretically derived moderators, as they is more likely to yield positive results. Theoretically hypothesized and then found, will be those that are most likely to begin to be hypothesized replicate across studies and prove to be important (Funder, 1995). IV. CONCEPTUAL ISSUES IV.

A. Self-Other Self-Other Agreement: Why Is It Important? The issue of of self-other self-other congruence congruence in personality judgments has generated a consid considerable amount amount of of research over the years, and the pace of of research research has accelerated accelerated Investigators seem to regard regard this issue as important important for for one of two reasons, recently. Investigators importantly different different kinds of of research as a result. and do subtly but importantly regard self-other self-other agreement as important important is as an end in The first reason to regard itself (e.g., DePaulo DePaulo et aI., al., 1987; Kenny Kenny & & LaVoie, LaVoie, 1984; 1984;Swann, Swann, 1984). 1984).The Theemphasis emphasis itself congruence is an intrinsically important important feature feature of of the here is on how judgmental congruence social world. For For instance, your ability to predict what others think of of you could have some obvious strategic value in social interaction and negotiation. This is a perfectly plausible reason to regard self-other self-other congruence as important, but re reperfectly search that that follows follows this approach approach sometimes seems unfortunately unfortunately reluctant to ac acknowledge personality personality as a real construct that could actually be judged accurately accurately or inaccurately, as opposed opposed to merely agreed or disagreed about (cf. Cook, 1984; 1984; Stryker & Gottlieb, 1981). This leads to a neglect of the possibility of gathering 1981). neglect of gathering Stryker other data, data, independent of others' others' and and self-judgments, self-judgments, that that might might help determine other independent of help determine judgmental vaHdity (Funder, 1995). judgmental validity (Funder, 1995). The other other reason to regard self-other self-other congruence as important is as a possible indicator of of judgmental accuracy. But one must be careful; two equally serious indicator often made about the connection connection between agreement agreement and kinds of mistake are often accuracy. The first is to regard regard the two terms of synonymous. More than a few few articles in the literature include accuracy accuracy in their title but only agreement agreement in their methods. Agreement Agreement is merely one possible and fallible indicator of of accuracy; it judgmental deserves a place in the array of converging methodologies for assessing judgmental accuracy, but agreement agreement is not accuracy accuracy itself. The second kind of of mistake is just bad—regarding agreement agreement as totally irrelevant to accuracy. Researchers who as bad-regarding enjoy enjoy investigating judgmental errors, for for instance, typically regard whatever judg judgagreement exists as evidence only that illusions can be shared (Ross, 1977). mental agreement B. The Relation between Agreement and Accuracy agreement as irrelevant irrelevant to accuracy is not a tenable position on either To regard agreement philosophical or empirical empirical grounds. Logically, the relationship between agreement agreement and accuracy is real although asymmetric: two judgments that agree may not be accurate, but two judgments that do not agree cannot both be accurate. In relation relation

640 640

FuNDER AND COLVIN FUNDERANDCOLVIN

to the accuracy issue, therefore, the investigation of of interjudge interjudge agreement agreement is techni technically a matter matter of of testing the null hypothesis. If agreement agreement is found, accuracy cannot cannot be implied. But if if agreement agreement is not found, then inaccuracy (of (of at least one judge) must be implied. Therefore, investigations of of interjudge interjudge agreement agreement serve to put at risk the hypothesis that judgments are accurate. Agreement Agreement is regularly found, as summarized summarized in this chapter, and so the accuracy hypothesis has in this sense repeat repeatedly survived potential disconfirmation. disconfirmation. Many other well-established well-established hypotheses in the psychological literature literature are supported, indirectly, in exactly the same way. Still, the logical connection connection between agreement agreement and accuracy remains less than airtight. Two judges can agree for for the wrong reasons. It is reassuring, therefore, therefore, that a survey of of the empirical data in this area also supports the existence of of a connection. A A fair fair amount of of research leads to the following conclusion: Everything Everything we can think of of that, it seems, should should improve judgmental accuracy, in fact fact does does improve interjudge interjudge agreement. Some evidence has already been reviewed. Traits Traits that seem, on common-sense grounds, to be more visible, in fact fact are judged with better agreement, no matter whose judgments are being compared (Funder & better matter compared (Funder Colvin, 1988). Knowing Knowing a person better, or at least longer, seems like something something that would improve improve accuracy. It is something something that improves agreement. agreement. C Next Steps: Beyond Agreement Agreement c. for including the study of of self-other self-other (and other otherThere seem to be good grounds for agreement as part part of of the study of of judgmental accuracy. But as accuracy other) agreement research begins to enjoy enjoy its recent renewal and second childhood, it will need to research branch out. Accuracy Accuracy research should begin to include other other criteria, including, branch prediction of of behavior (Funder, 1993, 1993,1995). 1995). most critically, the prediction endeavor will not be easy. Too much research already has consisted of of This endeavor some almost randomly randomly chosen but but convenient behavior in a sample of of measuring some personality subjects, failing to find correlations between this behavior and some personality judgment, and concluding the judgment to be faulty. To be done properly, research research careful establishing the reliability and construct construct in this area will have to be as careful validity of of the behavioral measures as psychometricians traditionally have been in selecting their questionnaire items ((Jackson harder to Jackson & Paunonen, 1985). It is harder necessary research will be measure behaviors than to ask questions though; the necessary difficult and is, to date, almost untried. difficult problem with using the ability to predict behavior as a criterion for for The basic problem from that entailed by using judgmental accuracy is, ironically, the opposite problem from agreement as a criterion. The problem problem with agreement agreement is that two judges that agree agreement still might not be accurate. The problem with behavioral predictability predictability is that a predict a particular behavior behavior still might be correct: maybe maybe judgment that fails to predict behavior was measured, or the right behavior behavior was measured, but with the wrong behavior prediction became impossible for for that reason alone. such low reliability that prediction of behavioral behavioral prediction prediction as a criterion for for accuracy, promising as it The use of (Cronbach & seems, will be no panacea. Accuracy is a lot like construct validity (Cronbach

CHAPTER 24 CHAPTER

SELF-OTHER CONGRUENCE SELF-OrnER

641 641

Meehl, 1955); it may be exactly the same thing. You You can never assess it directly, or prove it with one or even with several experiments. You You can only accumulate evidence, try to use judgment, and gradually become convinced convinced that it exists. The most convincing evidence consists of between data of very different of convergences convergences between data of different sorts, gathered through diverse and independent independent methods. Behavioral Behavioral prediction is an excellent excellent criterion for for accuracy, but it is imperfect. imperfect. In order to assess the accuracy of judgments of want to know, first, whether of personality, we will always want the judgments agree with each other. other.

ACKNOWLEDGMENTS of this chapter and the original research it reports were supported by NIMH Preparation of Grant Grant R01-MH42427 to D.C.F. D.C.F. We are grateful grateful to Oliver Oliver P. John and Robert R R. McCrae for for helpful helpful comments conunents on an earlier draft. draft.

REFERENCES (1988). Consensus in personality judgments at Albright, L., Kenny, D. A., & Malloy, T. E. (1988). of Personality Personalityand andSocial SocialPsychology, Psychology,55, 55,387-395. 387-395. zero acquaintance. Journal of of Ambady, N., & Rosenthal, R R. (1992). (1992). Thin slices slices of expressive behavior as predictors of interpersonal consequences: consequences: A A meta-analysis. meta-analysis. Psychological PsychologicalBulletin, Bulletin,111, 111, 256-274. 256-274. interpersonal Andersen, S. S. M. (1984). (1984). Self-knowledge Self-knowledge and social inference: II. The diagnosticity of cogni cognitive/affective tive/affective and behavioral data. Journal of of Personality Personality and and Social SocialPsychology, Psychology,46, 46, 294-307. 294-307. Bargh, J. J. A., A., & & Pratto, Pratto, F. F. (1986). (1986). Individual construct construct accessibility accessibility and and perceptual perceptual selection. selection. Bargh, of Experimental Experimental Social SocialPsychology, Psychology,22, 22,293-3 293-311. Journal of 11. Bem, D D.. J., J., & & Allen, Allen, A. A. (1974). (1974). O Onn predicting predicting some some ooff the the people people some some of of the the time: time: The The Bern, Psychological Review, search for for cross-situational consistencies in behavior. Psychological Review, 81, 506-520. 506-520. Bern, D. Bem, D. J., & Funder, D. D. C. (1978). (1978). Predicting Predicting more more of of the the people people more more of of the the time: time: Assessing Assessing

the personality personality of of situations. Psychological PsychologicalReview, Review,85, 85,485-501. 485-501. the Bender, I. I. E., E., & & Hastorf, Hastorf, A. A. H. H. (1950). (1950). The The perception perception of of persons: persons: Forecasting Forecasting another another Bender, person's responses on three personality scales. scales. Journal Journalof of Abnormal Abnormal and and Social SocialPsychol Psycholperson's ogy, 45, 45, 556-561. Bemieri, F. J., Zuckerman, M., Koestner, R, R., & Rosenthal, R R. (1994). Measuring person perception accuracy: Another look at self-other self-other agreement. Personality Personality and Social Psychology Bulletin, Bulletin, 4, 367-378. Psychology Bledsoe, J. J. c., C, & & Wiggins, Wiggins, R. R. G. G. (1973). (1973).Congruence Congruenceof ofadolescents' adolescents'self-concepts self-conceptsand andparents' parents'

perceptions of of adolescents' adolescents' self-concepts. self-concepts. Journal of of Psychology, Psychology, 83, 83,131-136. 131-136. perceptions Block, J. (1978). research. Palo Q-sort method methodin inpersonality personalityassessment assessmentand andpsychiatric psychiatric research, Palo (1978). The Q-sort Alto, CA: Consulting Psychologists Press. (Original work published 1961) Blumer, H. H. (1937). (1937). Social Social psychology. psychology. In In E. E. P. P. Schmidt Schmidt (Ed.), (Ed.), Blumer, Prentice-Hall. Prentice-Hall.

Man and society. NewYork: York: society. New

E. F. (1964). (1964). The Thestructure structure of of personality personality characteristics. characteristics.Behavioral BehavioralScience, Science,9,9,8-17. Borgatta, E. 8-17.

642 642

FuNoER AND COLVIN FUNDERANDCOLVIN

Borkenau, P., & & Liebler, Liebler, A. A. (1992). (1992). Trait Trait inferences: inferences: Sources Sources of of validity validity at at zero zero acquaintance. acquaintance. Borkenau, Journal 62,62, 645-657. Journalof of Personality Personalityand andSocial SocialPsychology, Psychology, 645-657. Borkenau, P., & Liebler, A. (1993a). (1993a). Consensus and self-other self-other agreement agreement for for trait inferences inferences

from minimal information. Journal of of Personality, Personality,61, 61,477-496. 477-496. Borkenau, P., & & Liebler, Liebler, A. A. (1993b). (1993b). Convergence Convergence of of stranger ratings ratings of of personality and and Borkenau, of intelligence with with self-ratings, self-ratings, partner partner ratings, ratings, and and measured measured intelligence. Journal of 546-553. Personality Personalityand andSocial SocialPsychology, Psychology,65,65, 546-553.

Bourne, E. E. (1977). (1977). Can Can we we describe describe an an individual's individual's personality? personality? Agreement Agreement on on stereotype stereotype Bourne, versus Social Psychology, 35, 35, 863-872. versus individual individual attributes. attributes. Journal Journalof ofPersonality Personalityand and Social Psychology, 863-872. Bronfenbrenner, Bronfenbrenner, V., U., Harding, J., & Gallwey, M. (1958). (1958). The measurement of of skill in social

perception. In In D. D. McClelland McClelland (Ed.), (Ed.), Talent Talentand andsociety: society:New New perspectives in the identifiperception. perspectives in the identifi cation cation of of talent. talent. New NewYork: York:Van VanNostrand. Nostrand. Brown, J. J. D. (1986). (1986). Evaluations Evaluations of of self self and and others: others: Self-enhancement Self-enhancement biases biases in in social social judg judgments. Social Social Cognition, Cognition,4,4,353-376. 353-376. ments. Cross-situationalconsistency consistency personality attributes. Paper Campbell, J. J. B. B. (1985, (1985, April). April). Cross-situational ofof personality attributes. Paper Campbell, presented at at the the annual annual meetings meetings of of the the Eastern Eastern Psychological Psychological Association, Association, Boston. Boston. Chaplin, W. W. F. F. (1991). (1991). The The next next generation generation of of moderator moderator research research in in personality personality psychology. psychology. Chaplin, Journal of of Personality, Personality,5959, 143-178. Journal , 143-178. Chaplin, W. W. F., F., & & Goldberg, Goldberg, L. L. R. R. (1985). (1985). A A failure failure to to replicate replicate the the Bem Bem and and Allen Allen study study Chaplin, of of individual individual differences differences in in cross-situational consistency. Journal of of Personality Personality and and Psychology, 47, 47,1074-1090. 1074-1090. Social Psychology, Social Chaplin, W. W. F., F., & & Panter, Panter, A. A. T. T. (1993). (1993). Shared Shared meaning meaning and and the the convergence convergence among among observ observChaplin, ers' personality personality descriptions. descriptions. Journal of of Personality, Personality,61, 61,553-585. 553-585. Cheek, Cheek, J. J. M. M. (1982). (1982). Aggregation, Aggregation, moderator moderator variables, variables, and and the the validity vaUdity of of personality personality tests: A peer peer rating rating study. study. Journal Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, 1254-1269. 43,43, 1254-1269. A Cialdini, R. R. B., & Richardson, Richardson, K. K. D. D. (1980). (1980). Two Two indirect indirect tactics tactics of of image image management: management: Cialdini, B., & Basking 39,39, 406-415. Basking and and blasting. blasting. Journal Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, 406-415. Clark, J. M., & Paivio, A. (1989). (1989). Observational and theoretical terms in psychology: A cognitive perspective on scientific language. American Psychologist, Psychologist,44, 44,500-512. 500-512. Colvin, C. R. (1993a). (1993a). Judgable people: people: Personality, behavior, and competing explanations. Journal 64,64, 861-873. Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, 861-873. Colvin, C. R. R. (1993b). (1993b).Childhood Childhoodantecedents antecedentsof ofyoung-adult young-adultjudgability. judgability.Journal Journalof ofPersonal Personality, ity, 61, 61, 611-635. 611-635. Colvin, C. R., & Block, J. (1994). (1994). Do positive illusions foster foster mental health? An examination of of the Taylor and Brown formulation. Psychological PsychologicalBulletin, Bulletin,116, 116,3-20. 3-20. Colvin, C. R., & Funder, D. C. C. (1991). (1991). Predicting personality personality and behavior: A boundary on Psychology, 60,60,884-894. 884-894. the the acquaintanceship acquaintanceship effect. effect. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, Conley, J. J. (1985). (1985). Longitudinal stability of personality traits: A multitrait-multimethod multitrait-multimethodmultioccasion Psychology, 49,49, 1266-1282. multioccasion analysis. analysis. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 1266-1282. Conrad, H. S. of S. (1932). (1932). The validity validity of personality personality ratings of of preschool preschool children. Journal of Educational EducationalPsychology, Psychology,23,23,671-680. 671-680. Cook, M. M. (1984). (1984). Issues in person perception. perception, London: London: Methuen. Methuen. Cook, Cronbach, L. L. J. J. (1955). (1955). Processes Processesaffecting affecting scores scoreson on"understanding "understandingof ofothers" others"and and"assumed "assumed Cronbach, similarity." similarity." Psychological PsychologicalBulletin, Bulletin,52,52,177-193. 177-193. L. J., J., & &Meehl, Meehl,P. P.E. E.(1955). (1955).Construct Constructvalidity validityininpsychological psychologicaltests. tests.Psychological Psychological Cronbach, L. Cronbach, Bulletin, Bulletin,52, 52,177-193. 177-193. Davidson, Davidson, K. K. W. W.(1993). (1993).Suppression Suppressionand andrepression repressioninindiscrepant discrepantself-other self-otherratings: ratings:Relations Relations with thought thought control control and and cardiovascular cardiovascular reactivity. reactivity. Journal of of Personality, Personality,61, 61,669-691 669-691. with .

24 SELf-OrnER SELF-OTHER CONGRUENCE CONGRUENCE CHAPTER 24

643 643

DePaulo, B. M., Kenny, D. A., Hoover, Hoover, C. C. W., W., Webb, Webb, W., W., & & Oliver, P. P. V. V. (1987). (1987). Accuracy Accuracy DePaulo, B. M.,

of person person perception: perception: Do Do people people know know what what kinds of of impressions they they convey? convey? Journal of of 52,52, 303-315. of Personality Personalityand andSocial SocialPsychology, Psychology, 303-315. Donahue, E. M., Robins, R. W., W,, Roberts, B. B. W., & John, O. P. (1993). (1993). The divided self: Concurrent and longitudinal effects effects of of psychological adjustment adjustment and social roles on Psychology, 64, 64, 834-846. self-concept Journalof ofPersonality Personalityand andSocial Social Psychology, 834-846. self-concept differentiation. differentiation. Journal Edwards, A. L., & &Klockars, Klockars,A. A.J.J.(1981). (1981).Significant Significantothers othersand andself-evaluation: self-evaluation:Relationships Relationships between perceived perceived and and actual actual evaluations. evaluations. Journal Journalof of Personality Personalityand andSocial Social PsycholPsychol between ogy, 7, 7, 244-251 244-251.. ogy, Erdelyi, M M.. H H.. (1974). (1974). A A new new look look aatt the the new new look: look: Perceptual Perceptual defense defense and and vigilance. vigilance. Erdelyi, PsychologicalReview, Review,81,81,1-25. 1-25. Psychological Estes, S. (1938). Judging S. G. (1938). Judging personality from from expressive behavior. Journal Journal of of Abnormal Abnormal and Social Psychology,33, 33,217-236. 217-236. Social Psychology, Fine, R. R. (1986). (1986). Narcissism, Narcissism, the theself selfand andsociety. society.New NewYork: York:Columbia ColumbiaUniversity UniversityPress. Press. Fine, (1960).The Theconsistency consistency of of ratings ratingsby bypeers. peers.Journal Journalof ofApplied Applied Fiske, D. W., & & Cox, Cox, J.J. A., A., Jr. Jr. (1960). Psychology, Psychology,44, 44,11-17. 11-17. Funder, Funder, D. D. C. C. (1980a). (1980a). On On seeing seeing ourselves ourselves as as others others see see us: Self-other Self-other agreement agreement and and discrepancy discrepancy in in personality personality ratings. ratings. Journal of of Personality, Personality,48, 48,473-493. 473-493. Funder, Funder, D. D. C. C. (1980b). (1980b). The The "trait" "trait" of of ascribing ascribing traits: traits: Individual Individual differences differences in in the the tendency tendency to trait trait ascription: ascription: Journal Journalof ofResearch ResearchininPersonality, Personality, 376-385. 14,14, 376-385. to Funder, Social Funder, D. D. C. C. (1982). (1982). On On the the accuracy accuracy of of dispositional dispositional vs. vs. situational situational attributions. attributions. Social Cognition,1,1,205-222. 205-222. Cognition, Funder, Funder, D. D. C. C. (1987). (1987). Errors Errors and and mistakes: mistakes: Evaluating Evaluating the the accuracy accuracy of of social social judgment. judgment. PsychologicalBulletin, Bulletin,101, 101, 75-90. Psychological 75-90. Funder, Funder, D. D. C. C. (1989). (1989). Accuracy Accuracy in in personality personality judgment judgment and and the the dancing dancing bear. bear. In In D. M. Buss Recent trends andand emerging directions Buss & & N. N.Cantor Cantor (Eds.), (Eds.),Personality Personalitypsychology: psychology: Recent trends emerging directions (pp. 210-223). New York: Springer-Verlag. Funder, Funder, D. C. C. (1991). (1991). Global Global traits: traits: A A Neo-Allportian Neo-Allportian approach approach to to personality. personality. Psychological Psychological Science, Science,2,2,31-39. 31-39. Funder, (1993). Judgments Funder, D. D. C. C. (1993). Judgments as as data data for for personality personality and and developmental developmental psychology. psychology. In In D. Funder, R. Parke, C. Tomlinson-Keasey, & K. Widaman (Eds.), Studying lives (pp.(pp. 121-146). Washington, through to to personality andand development throughtime: time:Approaches Approaches personality development 121-146). Washington, DC: DC: American Psychological Association. Funder, D. D. C. C. (1995). (1995). On On the the accuracy accuracy of of personality personality judgment: A realistic approach. Psycho PsychoFunder, logical Review, Review,102, 102,652-670. 652-670. logical Funder, D. c., C, & & Colvin, Colvin, C. C. R. R. (1988). (1988). Friends Friends and and strangers: strangers: Acquaintanceship, agreement, agreement, Funder, Social Psychology, and and the the accuracy accuracyof ofpersonality personalityjudgment. judgment.Journal Journalof ofPersonality Personalityand and Social Psychology, 55, 55, 149-158. Funder, D. c., C, & Dobroth, K. M. M. (1987). (1987). Differences Differences between traits: Properties associated with interjudge interjudge agreement. agreement. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 409-418. with Psychology, 52, 52, 409-418. Funder, D. D. C., C, Kolar, Kolar, D. D. c., C, & & Blackman, Blackman, M. M. C. C. (1995). (1995). Agreement Agreement among among judges judges of of Funder, personality: Interpersonal Interpersonal relations, relations, similarity, similarity, and and acquaintanceship. Journal of of Perpersonality: Per sonality and Social Psychology,69, 69,656-672. 656-672. sonality Social Psychology, Funder, Funder, D. D. C., C, & & West, West, S. S. G. G. (1993). (1993). Consensus, Consensus, self-other self-other agreement, agreement, and and accuracy accuracy in in personahty judgment: An introduction. Journal of of Personality, Personality, 61, 61,457-476. 457-476. personality Gage, Gage, N. N. L., & & Cronbach, Cronbach, L. L. J. J. (1955). (1955). Conceptual Conceptual and and methodological methodological problems problems in in interper interpersonal perception. perception. Psychological PsychologicalReview, Review,62,62, 411-422. sonal 411-422.

644 644

FUNnER FUNDER AND COLVIN COLVIN Gangestad, Gangestad, S. S. W., Simpson, Simpson, J. J. A., DiGeronimo, DiGeronimo, K., K., & & Biek, Biek, M. M. (1992). (1992). Differential Differential accuracy accuracy

in in person person perception perception across across traits: traits: Examination Examination of of aa functional functional hypothesis. hypothesis. Journal Journalof of Personality 688-698. Personalityand andSochll SocialPsychology, Psychology,62,62, 688-698. Goldberg, Goldberg, L. L. R, R., Norman, Norman, W. W.T., T.,&&Schwartz, Schwartz,E.E.(1980). (1980).The Thecomparative comparativevalidity validityofofquestion questionnaire naire data data (16PF scales) scales) and and objective objective test test data data (O-A (O-A battery) battery) in in predicting predicting five peer peerrating criteria. criteria. Applied Psychological PsychologicalMeasurement, Measurement, 183-194. 4, 4, 183-194. rating Golding, Golding, S. S. L. L. (1975). Flies Flies in in the the ointment: ointment: Methodological Methodological problems problems in in the the analysis analysis of of the the 82,82,278-288. 278-288. percentage percentage of of variance due due to to person personand andsituations. situations.Psychological PsychologicalBulletin, Bulletin, Gormly, (1984). Correspondence Gormly, J. J. (1984). Correspondence between between personality personality trait trait ratings ratings and and behaviorial behaviorial events. events. Journal Journal of of Personality, Personality,52,52,220-232. 220-232. Greenwald, self. In In R R. S. S. Wyer Wyer & & T. T. K. K. Srull SruU (Eds.), (Eds.), Greenwald, A. A. G., G., & & Pratkanis, Pratkanis, A. A. R R. (1984). The The self. Handbook of of social cognition cognition(pp. (pp.129-178). 129-178).Hillsdale, Hillsdale,NJ: NJ:Erlbaum. Erlbaum. Hase, Hase, H. H. D., & & Goldberg, Goldberg, L. L. R R. (1967). Comparative Comparative validity validity of of different different strategies strategies of constructing 231-248. constructing personality personality inventory inventory scales. scales. Psychological PsychologicalBulletin, Bulletin,67,67, 231-248. Hastorf, A. A. H., & Bender, I. E. E. (1952). A A caution caution respecting respecting the the measurement measurement of of empathic empathic Hastorf, H., & Bender, I. 574-576. ability. Journalof ofAbnormal Abnormaland andSocial SocialPsychology, Psychology,47,47, 574-576. ability. Journal Heath, (1992). Evidence Heath, A. A. C., C, Neale, Neale, M. M. c., C, Kessler, Kessler, R R. C., C, Eaves, Eaves, L. L. J., J., & & Kendler, Kendler, K. K. S. S. (1992). Evidence for genetic genetic influences influences on on personality personality from from self-reports self-reports and and informant informant ratings. ratings. Journal Journal for of Personality Personalityand andSocial SocialPsychology, Psychology,63,63, 85-96. of 85-96. Higgins, E. E. T., T., King, King, G. G. A., & Mavin, Mavin, G. G. H. H. (1982). Individual Individual construct construct accessibility accessibility and Higgins, A., & and subjective impressions impressions and and recall. recall. Journal Journalof of Personality Personalityand andSocial SocialPsychology, Psychology,43,43, subjective 35-47. Hirsch, Hirsch, J. J. (1986). Behavior-genetic Behavior-genetic analysis. analysis. In In J. J. Medioni Medioni & & G. G. Vaysse Vaysse (Eds.), (Eds.), Readings Readings from the the 19th 19thinternational international ethological conference: Genetic approaches to behavior ethological conference: Genetic approaches to behavior (pp. (pp. 129-138). Toulouse, Toulouse, France: France: Privat Privat Publisher. Publisher. traits. In In H. H. London London (Ed.), (Ed.), Personality: Personality:AAnew new Hirschberg, N. N. (1978). A A correct correct treatment treatment of of traits. Hirschberg, look at metatheories metatheories(pp. (pp.45-68). 45-68).Washington, Washington,DC: DC:Hemisphere. Hemisphere. Ickes, Ickes, W. W. (1993). Empathic Empathic accuracy. accuracy. Journal Journalof ofPersonality, Personality,61,61,587-610. 587-610. Jackson, D. N. (1982). Some preconditions for for valid person perception. In M. M. P. Zanna, Ontario E. E. T. T. Higgins, Higgins, & & c. C. P. P. Herman Herman (Eds.), (Eds.), Consistency Consistencyininsocial socialbehavior: behavior:The The Ontario Symposium Symposium (Vol. (Vol. 2). Hillsdale, Hillsdale, NJ: NJ: Erlbaum. Jackson, Jackson, D. N., Neill, Neill, J. J. A., & & Bevan, A. A. R R. (1973). An An evaluation evaluation of of forced-choice forced-choice and true-false Journalof ofResearch ResearchininPersonality, Personality, true-false item item formats formats in in personality personality assessment. assessment. Journal 7, 7, 21-30. Jackson, Jackson, D. D. N., & &,Paunonen, Paunonen,S.S.V. V.(1985). (1985).Construct Constructvalidity validityand andthe theprediction predictionof ofbehavior. behavior. Journal 49,49, 554-570. Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, 554-570. John, ofof altruism. InIn A.A. Angleitner, John, O. O. P. P.(1986). (1986).How Howshall shallaatrait traitbe becalled: called:AAfeature featureanalysis analysis altruism. Angleitner, trends A. Furnham, Furnham, & & G. G. van van Heck Heck (Eds.), Personality Personalitypsychology psychologyininEurope: Europe:Current Current trends and controversies controversies(pp. (pp.117-140). 117-140).Berwyn, Berwyn,PA: PA:Swets SwetsNorth NorthAmerica. America. John, John, O. P., & & Robins, Robins, R R. W. W. (1993). Determinants Determinants of of interjudge interjudge agreement agreement on on personality personality traits: The The Big Big Five Five domains, domains,observability, observability, evaluativeness, evaluativeness, and andthe the unique unique perspective perspective traits: of the the self. self. Journal Journalof ofPersonality, Personality,61,61, 521-531. of 521-531. John, O. O. P. P. & & Robins, Robins, R R. W. W. (1994). Accuracy Accuracy and and bias bias in in self-perception: self-perception: Individual Individual differ differJohn, ences in self-enhancement self-enhancement and and narcissisim. narcissisim. Journal Journalof ofPersonality Personalityand andSocial Social Psycholences in Psychol 66, 206-219. ogy, 66, Jones, Jones, E. E. E., & & Nisbett, Nisbett, R R. E. E. (1987). The The actor actor and and observer: observer: Divergent Divergent perceptions perceptions of of the the causes of of behavior. behavior. In In E. E. E. E. Jones, Jones,D. D.Kanouse, Kanouse,H. H.H. H.Kelley, Kelley,R R.E. E.Nisbett, Nisbett,S.S.Valins, Valins,&& causes

24 CHAPTER 24

SELF-OTHER CONGRUENCE CONGRUENCE SELF-OrnER

64S 645

causes of behavior (pp.(pp. 79-94). Hillsdale, B. Weiner (Eds.), (Eds.), Attribution: Attribution:Perceiving Perceivingthethe causes of behavior 79-94). Hillsdale, B. Weiner NJ: NJ: Erlbaum. Erlbaum. (Originally (Originally published published 1971). Jourard, S. Van transparentself self.New NewYork: York: VanNostrand. Nostrand. S. M. (1971). The transparent

Kamman, R, R., Smith, Smith, R, R., Martin, Martin, c., C, & & McQueen, McQueen, M. M. (1984). Low Low accuracy accuracy in in judgments judgments of Kamman, of others' psychological psychological well-being well-being as as seen seen from from aa phenomenological phenomenological perspective. perspective. Journal others' of of Personality, Personality,52, 52,107-123. 107-123.

Kenny, perception: A social relations analysis. NewNew York: Guilford. Kenny, D. D. A A. (1994). (1994). Interpersonal Interpersonal perception: A social relations analysis. York: Guilford. Kenny, Kenny, D. D. A., & & LaVoie, LaVoie, L. L. (1984). The The social social relations relations model. In In L. L. Berkowitz Berkowitz (Ed.), (Ed.),

Advances in experimental experimentalsocial socialpsychology psychology(Vol. (Vol.18,18,pp. pp.141-179). 141-179).New NewYork: York:Aca AcaAdvances demic Press. demic Press. Profiting from from controversy: controversy: Lessons Lessons from from the the person personKenrick, D. D. T., T., & & Funder, Funder, D. C. C. (1988). Profiting situation debate. debate. American American Psychologist, Psychologist,43,43, 23-34. 23-34. situation Kenrick, D. D. T., T., & & Stringfield, Stringfield, D. D. O. O. (1980). Personality Personality traits traits and and the the eye eye of of the the beholder: beholder: Kenrick, Crossing some some traditional traditional philosophical philosophical boundaries boundaries in in the the search search for for consistency consistency in in all all Crossing PsychologicalReview, Review,87,87,88-104. 88-104. of the the people. people. Psychological of Koestner, R, R., Bernieri, Bernieri, F., F., & & Zuckerman, Zuckerman, M. M. (1989). (1989). Trait-specific Trait-specific versus versus person-specific person-specific Koestner, moderators of of cross-situational cross-situational consistency. consistency. Journal of of Personality, Personality,57, 57,1-16. 1-16. moderators Krones, J. J. M., M., & & Funder, Funder, D. D. C. C. (1989). The The effect effect of ofperspective perspectiveon onattributions attributionsover overtime. time. Krones, Unpublished manuscript, manuscript, University University of of Illinois, Illinois, Urbana. Urbana. Unpublished Kruglarski, A A. W. W. (1989). TTie psychology psychology of of being being "right": "right": The The problem problem of of accuracy accuracy in in Kruglarski, (1989). The social perception perception and and cognition. cognition. Psychological PsychologicalBulletin, Bulletin,106, 106,395-409. 395-409. social Kunda, Z. Z. (1987). Motivated inference: inference: Self-serving Self-serving generation generation and and evaluation evaluation of of causal causal theories. Journal Journalof ofPersonality Personalityand and Social Psychology, 636-647. theories. Social Psychology, 53, 53, 636-647. Lepper, M. M. R., Greene, D., D., & & Nisbett, Nisbett, R R. E. E. (1973). Undermining children's children's intrinsic intrinsic interest interest Lepper, R, Greene, with extrinsic rewards: A A test test of of the the "overjustification" "oveijustification" hypothesis. hypothesis.Journal Journalof ofPersonal PersonalSocial Psychology, Psychology,28, 28,129-137. 129-137. ity and Social Lerner, M. M. J. J. (1980). The The belief in in a just world: world:AA fundamental fundamentaldelusion. delusion. New York: PleLerner, New York: Ple num Press. Press. num Lewicki, P. P. (1983). Self-image Self-image bias bias in in person person perception. perception. Journal of of Personality Personalityand andSocial Social Lewicki, Psychology,45, 45,384-393. 384-393. Psychology, Lewicki, P. P. (1984). Self-schema and and social social information information processing. processing. Journal of of Personality Lewicki, (1984). Self-schema Personality Social Psychology, and Social Psychology,47, 47,1177-1190. 1177-1190. Marsh, H. W., Barnes, J., & Hocevar, D. (1985). Self-other Self-other agreement on multidimensional multidimensional self-concept of self-concept ratings: ratings: Factor Factor analysis and multitrait-multimethod multitrait-multimethod analysis. Journal of Personality 1360-1377. Personalityand andSocial SocialPsychology, Psychology,49,49, 1360-1377. McCrae, R from self-reports self-reports R. R R. (1982). Consensual validation of personality traits: Evidence from Psychology, 43, 43, 293-303. and ratings. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 293-303. McCrae, R R. R, R., Costa, P. T., Jr., & Busch, C. M. (1986). (1986). Evaluating comprehensiveness in personality systems: The five factor The California California Q-set Q-set and and the thefive factor model. model.Journal Journalof ofPersonal Personality, 54, 54, 430-446. O'Farrell, T. J., Maisto, S. A., A, Connors, G. J., & Funder, D. McKay, J. R, R., OTarrell, D. C. (1989). Biases wives. Addictive Addictive Behaviors, Behaviors, Biases in in relapse relapse attributions attributions made made by by alchoholics alchoholics and and their their wives. 14, 14, 513-522. Mead, Mead, G. G. H. H. (1934). Mind, Mind, self selfand andsociety. society.Chicago: Chicago:University UniversityofofChicago ChicagoPress. Press. Miller, D. T., T., & & Porter, Porter, C. C. A. A. (1980). Effects Effects of of temporal temporal perspective perspective on on the the attribution attribution Miller, process. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 532-541. Psychology, 39,39, 532-541. process. Miller, Miller, D. D. T., T., & & Ross, Ross, M. M. (1975). Self-serving Self-serving biases biases in in the the attribution attribution of of causality: causality: Fact Fact or or fiction?Psychological PsychologicalBulletin, Bulletin, 213-225. fiction? 82, 82, 213-225.

646 646

FUNDER AND COLVIN FUNDERANDCOLVIN W., & Mischel, W., & Peake, P. P. K. K. (1982). (1982). Beyond Beyond deja d^ja vu vu in in the the search search for for cross-situational cross-situational , 730-755. consistency. consistency. Psychological PsychologicalReview, Review,8989, 730-755. Monson, Monson, T. T. c., C, & Snyder, Snyder, M. M. (1977). (1977). Actors, Actors, observers, observers, and and the the attribution attribution process: process: Toward Toward Psychology, 13, 13, 89-89-111. 111. a Journalof ofExperimental ExperimentalSocial Social Psychology, a reconceptualization. reconceptualization. Journal Monson, T. T. C., C, Tanke, Tanke, E E.. D D.,. , & & Lund, Lund, JJ.. (1980). (1980). Determinants Determinants ooff social perception perception in in a Monson, 14,14, 104-120. naturalistic Journalof ofResearch ResearchininPersonality, Personality, 104-120. naturalistic setting. setting. Journal Moore, B. B. S., S., Sherrod, Sherrod, D. D. R., Liu, T. T. J., J., & & Underwood, Underwood, B. (1979). The The dispositional dispositional shift shift in Moore, R., Liu, B. (1979). in Psychology, 15, 15, 553-569. attribution attribution over over time. time. Journal Journalof ofExperimental ExperimentalSocial Social Psychology, 553-569. Moskowitz, D. D. S. S. (1990). (1990). Convergence Convergence of of self-reports self-reports and and independent independent observers: observers:Dominance Dominance Moskowitz, Psychology, 58,58, 1096-1 106. and Journalof ofPersonality Personalityand andSocial Social Psychology, 1096-1106. and friendliness. friendliness. Journal Nisbett, R. R. E. E. (1980). (1980). The The trait trait construct construct in in lay lay and and professional professional psychology. psychology. In In L. L. Festinger Festinger Nisbett, (Ed.), (Ed.), Retrospections Retrospectionson onsocial socialpsychology psychology(pp. (pp.109-130). 109-130).New NewYork: York:Oxford OxfordUniver University Press. sity Press. C., Legant, Nisbett, E., Caputo, Caputo, C, Legant, P., & & Marecek, J. J. (1973). (1973). Behavior as as seen seen by by the the actor Nisbett, R. E., Social Psychology, 27, 27,154-165. 154-165. and and as as seen seen by by the the observer. observer. Journal Journalof ofPersonality Personalityand and Social Psychology, Nisbett, R. R. E., E., & & Ross, Ross, L. L. (1980). (1980). Human Human inference: inference:Strategies Strategies shortcomings of social Nisbett, andand shortcomings of social judgment. New NewYork: York:Prentice-Hall. Prentice-Hall. judgment. Norman, W. T. T. (1969). (1969). "To "To see see oursels oursels as as ithers ithers see see us,'': us!": Relations Relations among among self-perceptions, self-perceptions, Norman, W. peer-perceptions, and and expected expected peer-perceptions peer-perceptions of of personality personality attributes. Multivariate peer-perceptions, attributes. Multivariate Behavior Research, Research,4,4,417-433. 417-433. Behavior Norman, W. W. T., T., & & Goldberg, Goldberg, L. L. R. R. (1966). (1966). Raters, Raters, ratees, ratees, and and randomness randomness in in personality personality Norman, structure. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 4, 681-691. Psychology, 4, 681-691. structure. Park, B., & Judd, Judd, C. C. M. M. (1989). (1989). Agreement Agreement on on initial initial impressions: impressions: Differences Differences due due to to per perB., & Park, ceivers, trait trait dimensions, dimensions, and and target target behaviors. behaviors. Journal Journalof ofPersonality Personalityand andSocial Social PsyPsy ceivers, chology, 56, 56,493-505. 493-505. chology, Paunonen, S. S. V. V. (1989). (1989). Consensus Consensus in in personality personality judgments: judgments: Moderating Moderating effects effects of of targetPaunonen, target rater acquaintanceship acquaintanceship and and behavior behavior observability. observabiUty. Journal Journal of of Personality Personalityand andSocial Social rater Psychology,56, 56,823-833. 823-833. Psychology, Paunonen, Paunonen, S. S. V., V., & &Jackson, Jackson,D. D.N. N.(1985). (1985).Idiographic Idiographicmeasurement measurementstrategies strategiesfor forpersonality personality 92, and and prediction: prediction: Some Some unredeemed unredeemed promissory promissory notes. notes. Psychological Psychological Review, Review, 92, 486-51 1. 486-511. Regan, Regan, D. D. T., T., & & Totten, Totten, J. J. (1975). (1975). Empathy Empathy and and attribution: attribution: Turning Turning observers observers into into actors. Journal Psychology, 32,32, 850-856. Journalof ofPersonality Personalityand andSocial Social Psychology, 850-856. Robins, Robins, R. R. W., Spranca, Spranca, M. M. D., & & Mendelsohn, Mendelsohn, G. G. A A. (1996). (1996). The The actor-observer actor-observer effect effect revisited: revisited: Effects Effects of of individual differences differences and and repeated repeated social interactions interactions on on actor and Psychology, 71, 71, 375-389. and observer observer attributions. attributions. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 375-389. Ross, Ross, L. L. (1977). (1977). The The intuitive intuitive psychologist psychologist and and his his shortcomings. shortcomings. In In L. L. Berkowitz Berkowitz (Ed.), Advances in experimental experimentalsocial socialpsychology psychology(Vol. (Vol.10,10,pp. pp.174-214). 174-214).New NewYork: York:Aca Academic Press. demic Press. Ross, D., & & House, House,P. P.(1977). (1977).The Thefalse falseconsensus consensus phenomenon: phenomenon:An Anattributional attributional L., Greene, Greene, D., Ross, L., bias in in self-perception self-perception and and social social perception perception processes. processes. Journal Journal of of Personality Personality and and bias Social , 485-494. Social Psychology, Psychology,3434, 485-494. Schneider, Hastorf, A A. H., & & Ellsworth, Ellsworth, P. P. C. C. (1979). (1979). Person Person perception perception(2nd (2nd ed.). ed.). Schneider, D. J., Hastorf, Reading, Reading, MA: MA: Addison-Wesley. Addison-Wesley. direct and indirect personality Scott, Scott, W. W. A, A.,& &Johnson, Johnson,R. R.C.C.(1972). (1972).Comparative Comparativevalidities validitiesofof direct and indirect personahty tests. Psychology, 38, 38, 301-318. tests. Journal Journalof of Consulting Consultingand andClinical Clinical Psychology, 301-318. Shrauger, J. J. S., S., & & Schoeneman, Schoeneman, T. T. J. J. (1979). (1979). Symbolic Symbolic interactionist interactionist view view of of self-concept: self-concept: Shrauger, Through Through the the looking looking glass glass darkly. darkly. Psychological PsychologicalBulletin, Bulletin,86,86,549-573. 549-573.

CHAPTER 24 24

SELF-OrnER SELF-OTHER CONGRUENCE CONGRUENCE

647 647

Stinson, L., & Ickes, W. (1992). (1992). Empathic accuracy in the interactions of of male friends friends versus

male strangers. strangers. Journal Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, 787-797. male 62,62, 787-797. Storms, M. M. D. D. (1973). (1973).Videotape Videotapeand andthe theattribution attributionprocess: process:Reversing Reversingactors' actors'and andobservers' observers'

points 27,27, 165-175. points of of view. view. Journal Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, 165-175. Stryker, S., & &Gottlieb, Gottheb,A. A.(1981). (1981).Attribution Attributiontheory theoryand andsymbolic symbolicinteractionism: interactionism:AAcompari comparison. In J. H. Harvey, W. Ickes, & R. F. Kidd (Eds.), New directions directions inin attribution attribution research research (Vol. (Vol.3). 3).Hillsdale, Hillsdale,NJ: NJ:Erlbaum. Erlbaum. Swann, W. B., Jr. (1984). (1984). Quest for for accuracy in person perception: A matter of pragmatics. Psychological PsychologicalReview, Review,91,91,457-477. 457-477. Taft, Taft, R. (1955). (1955). The ability to judge people. Psychological PsychologicalBulletin, Bulletin,52, 52,1-23. 1-23. Taft, R. (1966). (1966). Accuracy of of empathic judgments of acquaintances and strangers. Journal of of

Personality 600-604. Personalityand andSocial SocialPsychology, Psychology,3, 3, 600-604. Taylor, S. S. E. (1983). (1983). Adjustment Adjustment to threatening events: events: A theory of cognitive adaptation. American 173. American Psychologist, Psychologist,38,38,1 161-1 1161-1173.

Tesser, A., A., & & Campbell, Campbell, J. J. (1982). (1982). Self-evaluation Self-evaluation maintenance maintenance and and the the perception perception of of friends friends Tesser, and strangers. strangers. Journal Journalof ofPersonality, Personality,50,50,261-279. 261-279. and

Tetlock, P. P. E., & & Levi, Levi, A. A. (1982). (1982). Attribution Attribution bias: bias: On On the the inconclusiveness inconclusiveness of of the the cognition cognitionTetlock, motivation debate. debate. Journal Journalof ofExperimental ExperimentalSocial Social Psychology, 68-88. motivation Psychology, 18, 18, 68-88. Watson, D. D. (1982). (1982). The The actor actor and and the the observer: observer: How How are are their their perceptions perceptions of of causality causahty Watson, divergent? Psychological PsychologicalBulletin, Bulletin,92,92, 682-700. 682-700. divergent? Watson, D. D. (1989). (1989). Strangers' Strangers' ratings ratings of of the the five robust robust personality personality factors: factors: Evidence Evidence of of aa Watson, surprising convergence convergence with with self-report. self-report. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, Psychology, surprising 57, 120-128. 57,

Watson, D., D., & & Clark, Clark, L. L. A. A. (1991). (1991). Self Self versus versus peer peer ratings ratings of of specific specific emotional emotional traits: traits: Watson, Evidence of of convergent convergent and and discriminant discriminant validity. validity. Journal of of Personality Personalityand andSocial Social Evidence Psychology,60, 60,927-940. 927-940. Psychology, Wiggins, J. J. S.S.(1973). (1973).Personality Personality and prediction: Principles of personality assessment. Reading, and prediction: Principles ofpersonality assessment. Reading, Wiggins, MA: Addison-Wesley. Addison-Wesley. MA: Woodruffe, C. C. (1985). (1985). Consensual Consensual validation validation of of personality personality traits: traits: Additional Additional evidence evidence and and Woodruffe, individual differences. differences. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 1240-1252. individual Psychology, 48,48, 1240-1252. Individualdiff differences judgmental ability. Paper Vogt, D., D., & & Colvin, Colvin, C. C. R. R. (1996, (1996, August). August). Individual Vogt, erences inin judgmental ability. Paper presented presented at the Annual Meetings of the American American Psychological Association, Toronto. Zuckerman, M., Bemieri, F., Koestner, R., & Rosenthal, R. (1989). (1989). To predict some of the people some of of the time: time: In search of moderators. Journal of of Personality Personalityand andSocial Social Psychology, Psychology, 57, 57,279-293. 279-293. Zuckerman, M., Koestner, R., DeBoy, T., Garcia, Garcia, T., Maresca, Maresca, B. C., C, & & Sartoris, Sartoris,J.J.M. M.(1988). (1988). To predict some of of the people some of the time: time: A reexamination of the moderator moderator Psychology, variable approach in personality theory. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 54, 54, 1006-1019. Zuckerman, M., Miyake, K., Koestner, R., Baldwin, C. C. H., & Osborne, J. W. (1991). (1991). Unique Uniqueness as aamoderator Psychology Bulletin, moderatorof ofself-peer self-peer agreement. agreement.Personality Personalityand andSocial Social Psychology Bulletin, 17, 17, 385-391.

This Page Intentionally Left Blank

CHAPTER 25 25

SELF-INSIGHT THE QUEST FOR SELF.INSIGHT THEORY AND RESEARCH ON ACCURACY AND BIAS SELF-PERCEPTION IN SELF-PERCEPTION

W . ROBINS RICHARD W. UNIVERSITY UNIVERSITY OF OF CALIFORNIA, CALIFORNIA, DAVIS DAVIS OLIVER P. GLIVER P. JOHN UNIVERSITY OF CALIFORNIA, UNIVERSITY OF CALIFORNIA, BERKELEY BERKELEY

I. INTRoDucnoN INTRODUCTION The quest for self-insight has been a pervasive concern for over 25 centuries. In the 7th century B.C., B.C., when the citizens of ancient Greece sought advice from the oracle at at Delphi they were weregreeted greeted with withthe the salutation, salutation,"know "knowthyself." thyself."The The influence influence of this maxim on Western thought is generally attributed to Socrates, who quesques tioned why people should pry into the heavens while they are still ignorant of their own selves. "The unexamined life is not worth living," living," Socrates told his disciples, for through self-knowledge self-knowledge lies the path to truth, virtue, and happiness. The ancient Greek mandate to "know thyselr' thyself foreshadowed a perennial problem: What is self-insight and how does one acquire it? Since the early days of psychology, researchers interested in self-insight have explored a number of intriintri questions: How well do people know themselves? Which psychological proguing questions: pro cesses promote accurate self-perception and which promote distortion? How can self-knowledge be distinguished from self-deception and hubris? What criteria true self-knowledge self-views? can be used to evaluate the veracity of a person's self-views? HANDBOOK OFPERSONALITY PSYCHOLOGY

649

HANDBOOK OF PERSONALITY PSYCHOLOGY

649

ComlGHT COPYRIGHT C© 1997 1997 n BY ACADEMIC ACADEMIC PlESS PRESS.. ALL liGHTS OF IEPRODUcnON IN ANY FOIM IESERVED.

ALL RIGHTS OF REPRooucnoN IN ANY FORM RESERVED.

650 650

ROBINS AND JOHN JOHN ROBINS AND

This chapter chapter provides a review and integration of of current theory and research research on self-perception personality and social psychology" self-perception accuracy accuracy in personaUty psychology.^ We address two broad conceptual issues in the study of of self-perception self-perception accuracy: (a) What is accuracy and how should it be measured? and (b) What psychological processes are involved self-perception? To organize our our review and analysis of of the literature, we discuss in self-perception? conceptual frameworks. frameworks. The first defines the various criteria researchers have two conceptual used to measure accuracy accuracy and classifies them into six broad categories: operational, operational, social consensus, consensus, functionaVpragmatic, functional/pragmatic, normative normative models, models, information information processing, processing, social and internal internal consistency consistency (Robins & John, 1996a). The second framework framework character characterand self-perception process from from four four different different theoretical perspectives and uses izes the self-perception metaphor to capture the essence of each perspective: the Scientist, the Consistency Consistency a metaphor Seeker, the Politician, Politician, and the Egoist. Egoist In the final section, we illustrate the heuristic heuristic Seeker, of these metaphors by applying them to our own research on self-enhancement self-enhancement value of bias. In In this this section, section, we we also also discuss discuss the the implications implications of of accurate accurate self-perception self-perception for for bias. mental health. health. Is Is self-insight self-insight worth worth pursuing, pursuing, as as Socrates suggested, or or are are people people mental Socrates suggested, better off maintaining positive illusions about themselves (cf. (cf. Taylor & Brown, Brown, better off 1988)? 1988)?

n. IS ACCURACY AND How HOW SHOULD IT BE MEASURED? n. WHAT Is A. The Criterion Problem How How do we know whether whether a person has self-insight? self-insight? At first, the answer seems deceptively simple: Individuals have self-insight self-insight if they perceive themselves accu accudeceptively rately. To study self-insight, self-insight, then, one need only compare a person's view of of him himor herself herself with what that person is truly like. Uke. And And therein therein lies the problem. We do do not know the true nature of of the person. That is, there are no absolute, perfectly perfectly objective measures of of a person's traits, capabilities, motives, and so on. Thus, although we can tell how tall people are by measuring their height with a ruler or how heavy they are by weighing them on a scale, we cannot tell how neurotic neurotic people are in an equally objective objective manner. For most attributes of of interest to psychol psychologists, we have only indirect indirect measures (e.g., a questionnaire scale of of neuroticism) neuroticism) from which the constructs constructs of of interest left with a from interest must be inferred. Thus, we are left study self-insight self-insight in the absence of of an absolute standard standard conundrum: How can we study for reality? reaUty? for

Inthis thischapter, chapter,we wedefine definethe theterm tennaccuracy accuracybroadly broadlytotoinclude includeboth bothvalidity validityand andbias. bias.Validity Validity 1^In is typically typically defined defined by by the the correspondence correspondence (e.g., (e.g., correlation) correlation) between between self-perceptions self-perceptions and and aa criterion, criterion, is whereas bias bias is is typically typically defined defined in in terms terms of of directional directional deviations deviations (e.g., (e.g., positive positive or or negative) negative) from from a whereas criterion. These These two two measures measures are are statistically statistically independent; independent; for for example, example, self-perceptions self-perceptions could could be be more more criterion. positive than than aa criterion criterion but but still still be highly highly correlated correlated with with that that criterion. criterion. positive

CHAPTER CHAPTER 2S 25

SELF-PERCEPTION ACCURACY AND BIAS IN SELF.PERCEmON B. B.

651 651

Defining Accuracy: Three Perspectives Perspectives

Unfortunately, Unfortunately, the criterion problem has no simple solution. Instead, researchers have conceptualized accuracy in ways that make it amenable to empirical inquiry. To provide an overview of these various conceptualizations, we summarize three perspectives below (see also Fiske, 1993; 1993; Hastie & & Rasinski, 1988; 1988; Judd & & Park, 1993; Swann, 1984). 1993; Jussim, 1993; Kruglanski (1989) differentiated consendifferentiated three notions ofjudgmental accuracy: consen sus, correspondence, and pragmatic utility. The first, consensus, consensus, implies impliesthat thataajudg judgment is accurate if it agrees with judgments by others. The problem, however, is correspondencenotion notion fallible and subject to bias. The correspondence that human judgments are fallible of accuracy refers to the relation between a subject's judgment and a criterion for reality. For example, self-ratings of personality can be said to be accurate if they major problem, accord accordcorrespond with ratings by knowledgeable informants. The major ing to Kruglanski (1989), is identifying identifying an appropriate appropriate criterion: criterion: "criteria "criteriafor for accurate accurate ing judgments are not invariably self-evident. Often Often they need to be justified by complex argument or indirect evidence . .. .. . . [T]he accuracy of any given criterion . .. . . is criticism.. . . [and] [and]accuracy accuracy standards standards are arethemselves themselvesjudgments judgments perennially open to criticism contingent on argument and evidence" (p. (p. 396). Kruglanski suggested that it is 396). especially important important for for subjects and experimenters to agree about the appropriate appropriateness of the criterion. Kruglanski's third notion of accuracy involves considerations pragmatic utility utility—the functional value of the judgment. From this of pragmatic the adaptive or functional perspective, a judgment is accurate if it is useful to the individual, that is, if it is related to successful successful task accomplishment, goal attainment, and other desirable outcomes. Accuracy, then, should be assessed on the basis of outcomes that have adaptive significance for for the individual. focused on the use of observer judgments in research Kenny (1991, 1994) focused & John, 1996b). on accuracy and consensus (i.e., interjudge agreement) (Robins & instrument' used by Noting that the human observer is "the most valued 'instrument' definition of accuracy formulated formulated psychologists" (p. 156), 156), Kenny provided a formal definition in terms of observer judgments: "the average judgment made by all possible 159). This definition, intended as observers of all possible target behaviors" (p. 159). carries three basic assumptions derived a theoretical ideal (like a true score), carries from psychometric theory. First, averaging judgments across observers helps from cancel out random errors in individual judgments. Second, the use of "all possible observers" eliminates systematic bias due to a particular sampling of observers. informaThird, the observation of all possible behaviors ensures that all relevant informa tion is available to the judges. To Kenny, then, the truth about a person can found at the intersection of judgments by all observers of all behaviors be found exhibited by the person. 1987, 1990, 1993) has written extensively on accuracy issues, Funder (e.g., 1987, particularly in the context of personality judgments. He has argued that "the study of accuracy in judgment is exactly the same thing as measurement validity, where -

652 652

ROBINS AND JOHN ROBINS AND JOHN

the measurements being validated are interpersonal judgments" (Funder, 1990, p. 208). From this perspective, a personality judgment is accurate if it agrees with judgments by others and predicts behaviors relevant to the trait being judged. Although Funder did not equate interjudge interjudge agreement agreement with accuracy, he pointed out that judgments that agree with each other are more likely to be accurate than judgments that do not agree. That is, consensus is a necessary but not sufficient sufficient condition for for accuracy. Funder also emphasized the importance of studying accuracy in real-world contexts, noting that a judgment deemed an error in the laboratory may be accurate, or at least adaptive, in a person's everyday life. In this context, paraFunder contrasted two broad categories of judgment research, the accuracy para real-world contexts by digm and the error paradigm. Accuracy is usually gauged in real-world comparing subjects' judgments to an external criterion. Errors, in contrast, are usually gauged in experimental contexts by comparing subjects' judgments to criteria derived from from a normative model that prescribes how the judgments should be made (e.g., Bayes' theorem). C. Measuring Measuring Accuracy: Accuracy: Six Six Categories Categories of of Accuracy Accuracy Criteria Criteria C. The writings of Kruglanski, Kenny, and Funder point to the myriad ways accuracy has been conceptualized and assessed. However, each of these accounts is incom incomframework that would organize plete and the field has needed a comprehensive framework the various criteria used in empirical research. To this end, we recently reviewed the research literature and proposed a framework categories framework consisting of six distinct categories normative of accuracy criteria: operational, operational,social socialconsensus, consensus,functionaVpragmatic, functional/pragmatic, normative models, information iriformationprocessing processing(cue (cueuse), use),and and internal internalconsistency consistency(see (see Robins Robins && John, 1996a, for for a more extensive discussion; 1996b). The The six six categories categories are are concep conceptually based, rather than being based on the specific method or data source used. That is, the categories reflect reflect different different definitions of accuracy and classify criteria according to the assumptions that justify their use as standards for accuracy. Al Although we have described the framework framework in terms of self-perception accuracy, the categories apply equally well to research on social perception. We next review each category and provide examples. Operational of Operational (or reality) reaUty) criteria are difficult difficult to find in the domain of personality. Nonetheless, some personality characteristics can be assessed in much the same way as height or weight. For example, talkativeness can be defined defined as the amount amount a person talks. Therefore, the number of words spoken in a conversation may be interpreted as a direct operational criterion for selffor self ratings of talkativeness. In this example, there is a direct correspondence between the definition interpretation definition of of the construct construct being rated (talkativeness) and the interpretation of the criterion measure (number of words spoken). Another example of a direct criterion for for reality comes from from research on self-perceptions self-perceptions of test performance where a subject's actual test performance provides an operational performance criterion (e.g., Beyer, 1990). 1990). A third example comes from from research on the of people's expectations about the future. We have used an actual accuracy of

CHAPTER CHAPTER 25 25

ACCURACY ACCURACYAND ANDBIAS BIASININSELF.PERCEmON SELF-PERCEPTION

653 653

outcome (course grades) as a criterion for students' expectations about what grade they will receive in the course (Robins, 1996). We found found that about half of the students expected to attain higher grades in the beginning of of the semester semester than they actually received, about one-quarter one-quarter of of the students expected lower grades than they received, and about one-quarter one-quarter were accurate, receiving exactly the grades they expected. Finally, in studies of self-ratings of personality, the absence of a clear operational criterion has led some researchers to ask questions "reflected for which a direct criterion is available. For example, research on "reflected beUefs about how they are seen by others (for (for appraisals" examines people's beliefs conceptually a review, see Kenny, 1994). In this context, there is an obvious and conceptually defensible operational criterion, namely, how the individual is actually seen by defensible of self selfothers. However, this research cannot replace research on the accuracy of perceptions, which addresses a different question: Do people's views of themselves perceptions, which addresses a different question: Do people's views of themselves correspond with with what what they they are are truely truely like? like? correspond The use of of social social consensus criteria reflects reflects the folk belief belief that self-insight self-insight means seeing oneself oneself as others see one. In fact, judgments by others (e.g., friends, spouses, psychologists) are widely widely used to evaluate the validity vaUdity of of self-reports of of personaUty (e.g.. Cheek, 1982; Funder & Colvin, this volume, Chap. 24; John & personality (e.g., & & Robins, 1993; McCrae & & Costa, 1990). McCrae (1982) referred to this use of social consensus criteria as the principle of consensual validation, which is based on the assumption that aggregating measurements (e.g., across judges) cancels out random error associated with any single measurement (or individual judge). In other words, in the consensus lies the truth (Hofstee, 1994). Like all criteria, consensual judgments properly aggregated should not be used unthinkingly. However, when judgments are properly and a case is made for their validity, social consensus criteria can play a central role in accuracy research. According to to functional functional (or pragmatic) pragmatic) criteria, a judgment can be considered According if it helps an individual adapt in the real world. From this perspective, the accurate if accuracy of belief should be evaluated according to how well it serves the goals of a belief of of the perceiver, rather than by its correspondence with some absolute reality (Baron, 1988; 1988; Gibson, 1979; 1979; Swann, 1984). For example, a functionalist might argue of self-perceptions self-perceptions of of height should be gauged by our success in that the accuracy of predicting whether or not we can walk through a doorway without bumping our of adaptiveness is not heads rather than by using a ruler. If a direct measure of available, researchers can use a proxy known to predict an adaptive outcome. For & Robins, 1994), we have argued that observer example, in our own research (John & functional assessments of performance in a managerial simulation task provide a functional self-assessments of of performance because they predict long-term ca cacriterion for self-assessments reer success. success. reer normative models models are commonly used to evaluate the Criteria derived from normative of human judgment (e.g., Einhorn Einhom & & Hogarth, 1981; 1981; Nisbett Nisbett & & Ross, Ross, 1980). 1980). quality of The basic assumption underlying this research is that statistical and probability models prescribe the optimal judgment against which human judgment can be Discrepancies between judgments and the prescriptions of the model compared. Discrepancies

6S4 654

ROBINS AND AND JOHN JOHN ROBINS

can therefore be interpreted as bias. Explicit normative models such as Bayes' theorem are rarely used in the self-perception literature (for an exception, see Krueger & 1993). Self-perception researchers typically rely on normative & Zeiger, 1993). models that are left implicit. For example, several studies have shown that people's self-ratings are more positive than their ratings of a hypothetical "average other" (e.g., Brown, 1986). This finding self(e.g.. jSnding has been widely interpreted as evidence of self enhancement bias because, according to these researchers (e.g.. (e.g., Brown, 1991; Tay Tay& Brown, Brown, 1988), 1988), itit isis logically logically impossible impossible for for the the majority majority of of people people to to be be better better lor & than average. The implicit model underlying this interpretation is the arithmetic of notion that the average of the individual self-ratings should equal the rating of the average other. There are a number of reasons why one might question the appropriateness of this model. Most important, it is not logically impossible for of aagroup groupof ofpeople peopleto tobe beabove aboveaverage: average:the themajority majorityof ofindividuals individuals more than 50% 50% of can be above the arithmetic mean when the characteristic being rated is negatively skewed (rather than normally distributed). Thus, almost everyone can be above average when a small percentage of individuals are substantially below average. As this example illustrates, it is often difficult to interpret a lack of correspondence with a normative model, and a case must be made for the appropriateness of the model as an accuracy criterion. information-processing(cue (cueuse) use)criteria criteriawas wasspurred spurredbybyBruns BrunsInterest in information-processing wik's (1956) lens model, which provides a powerful framework for studying the appropriateness or optimality of people's use of informational cues. The lens model describes the relations among the attributes of a stimulus object (e.g., a person), cues in the environment (e.g., information about the person), and the Selfjudgment made by a perceiver (the self in the case of self-perception). Self perception researchers have examined cue use by providing subjects with experi experimentally manipulated feedback about themselves. In a typical study, subjects are randomly raudomly provided with either positive or negative feedback about their performance in a task (e.g., Baumgardner, Kaufman, & Levy, 1989; 1989; Shrauger, 1975). In general, this research suggests that subjects operate in a self-serving manner when they process cues provided by the experimenter; that is, they tend to downplay the validity of negative feedback and inftate inflate the validity of positive feedback. Although such experimental research has provided insights into the processes involved in the formation of self-perception biases, we know little about the actual cues people use to form an evaluation of themselves in realistic settings. For example, when college students appraise their academic competence relative to other students, to what extent do they base their judgments on success and failure experiences in school, on performance on standardized tests (e.g., SAT), on feedback from teachers, or on information they have about how other students perceive them? In addition to studying how people use cues, self-perception future research should study the link between cue use and self-perception accuracy. Although it seems reasonable to assume that accurate judgments are more likely to result from the use of valid cues than from the use of invalid

CHAPTER CHAPTER 25 25

ACCURACY PERCEPTION ACCURACY AND AND BIAS BIAS IN IN SELF. SELF-PERCEPTION

655 655

cues, there is no necessary link between the process used to form form a judgment and the accuracy of the resulting judgment. Studies employing internal internal consistency consistencycriteria criteriaexamine examinewhether whetherindividuals' individuals' self-perceptions are consistent with their other beliefs about themselves. If a person self-perceptions sees him- or herself herself in an inconsistent manner (e.g., as both generally talkative and generally quiet), it seems unlikely that both beliefs are accurate. In a study of trait attributions, Borkenau and Ostendorf Ostendorf (1989) examined two types of consistency: descriptive (ascribing two two traits that have the same descriptive meaning) and evalua evaluative (ascribing two traits that are both evaluatively positive or evaluatively negative). found that individuals often described described themselves using two favorable traits They found even when the traits were descriptively inconsistent (e.g., generous and thrifty), but rarely described themselves using two descriptively consistent traits that were evaluatively inconsistent (e.g., thrifty thrifty and stingy). The internal consistency of the self-concept has also been conceptualized as the degree to which a person sees self-concept herself as having similar personality characteristics in different different social roles. him- or herself found that most people had self-views Donahue, Robins, Roberts, and John (1993) found that were fairly consistent across social roles; moreover, people who had consistent adjusted psychologically than were those who self-views across roles were better adjusted Note, however, however, that that although although an an inconsistent inconsistent "fragmented" self. Note, had a "divided" or "fragmented" self-concept may suggest a conflict in a person's self-views, these inconsistent views different roles. Thus, Thus, may accurately reflect variability in a person's behavior across different conclusions about accuracy based on internal consistency standards can be probprob lematic. D. Conclusions Conclusions and and Recommendations Recommendations D. Based on our review of the criteria used in accuracy research, we have made several recommendations (see Robins & John, 1996). First, accuracy researchers should be explicit about which of the six types of accuracy criteria they have chosen and thus clarify the conceptualization of accuracy entailed by that choice. Second, researchers need to make an argument for for why the criterion is appropriate. For example, social consensus is a better criterion criterion for for physical attractiveness than for subjective well-being. Much confusion confusion could be avoided if researchers would always present conceptual and empirical arguments for of for the cogency of their accuracy claims. Third, the criterion must be measured properly. For example, ratings of physical attractiveness made by blind people would not provide a good measure of the social consensus. Fourth, given the limitations of any one accuracy criterion, researchers should use multiple criteria to assess accuracy. For example, in our own research we have used several criteria to examine the accuracy of self-evaluations of performance performance in a group discussion 1994; Robins & John, 1997). The convergence of findings findings task (John & Robins, 1994; across all three criteria provides more powerful powerful evidence about the accuracy (or inaccuracy) of self-perceptions than any single criterion. Fifth, assertions found that self-perceptions self-perceptions about accuracy should be criterion specific (e.g., "We found

656 656

ROBINS ROBINS AND AND JOHN JOHN

were biased when compared to the prescriptions of of our normative model"). Divergence from one type of of accuracy criterion may not have the same psychologi psychological implications as divergence divergence from another type. Clearly, lack of correspondence correspondence with a social consensus criterion may not imply the same psychological psychological processes processes of correspondence with the prescriptions of of a normative model. Thus, as lack of self-perception accuracy is to examine an important next step in research on self-perception convergences across the types of of accuracy criteria described here and to identify identify convergences the psychological implications of of criterion. of deviating from each type of

m. in. SELF·PERCEmON SELF-PERCEPTION PROCESSES: FOUR METAPHORS OF mE THE SELF.PERCEIVER SELF-PERCEIVER Thus far we have discussed how accuracy has been conceptualized and assessed, and reviewed reviewed a conceptual framework framework that organizes the various accuracy criteria turn to the process of self-perception self-perception and used in self-perception research. We now tum ask: How do people form beliefs about themselves? In answering this question, of self-insight and attempt to understand how we explore the psychological roots of beliefs about the self self are formed and maintained. inaccurate beliefs Researchers have approached the question of of how people form beliefs about Researchers themselves from many perspectives. There are, in fact, hundreds of models and of the self-perception self-perception process. One reason why the literature on the self theories of is so complex is the lack of an overarching and integrative framework. We believe some order can be found by organizing the various models in terms of several broad rubrics, each representing a coherent theoretical perspective captured by a unique metaphor of of the person. Each metaphor embodies a set of of assumptions about the role of motivation, information processing, and affect affect in the self-percep self-perception process. We propose four metaphors as central to understanding the psychology psychology of the self-perceiver: the Scientist, the Consistency Seeker, the Politician, and the Egoist. In some ways people act like Scientists, seeking out information about themselves in a dispassionate search for truth; in some ways people are Consistency Seekers, striving to confirm their preexisting self-views with little regard for reality; in some ways people behave like Politicians, striving to present themselves in ways that create the most favorable favorable impressions on others; and in some ways people act like Egoists, narcissistically distorting information to enhance their self-worth. Our of these metaphors and their application to the self-perceiver self-perceiver was formulation of inspired by by existing models of of human judgment. The The Scientist Scientist and and the the Consistency Consistency inspired existing models human judgment. Seeker are are commonly commonly used used metaphors metaphors in in the the social social cognition cognition literature literature (Fiske (Fiske & Seeker Taylor, 1991, 1991, pp. pp. 9-12; Nisbett & & Ross, Ross, 1980, 1980, chap. chap. 1). 1). The The Politician Politician was was suggested suggested Taylor, 9-12; Nisbett by Tetlock (1992) as as a a general for research research on on judgment judgment and and choice. choice. The The by Tetlock (1992) general metaphor metaphor for Egoist metaphor derives from the large body of studies characterizing the selfEgoist metaphor derives from the large body of studies characterizing the self-

CHAFfER CHAPTER Z5 25

ACCURACY AND ANDBIAS BIAS IN IN SELF.PERCEmON SELF-PERCEPTION ACCURACY

657 657

perception process as ego driven (e.g., Greenwald, 1980; perception 1980; John & Robins, 1994; Taylor & Brown, 1988).^ Taylor 1988).2 The four four metaphors provide provide broad rubrics that summarize the basic processes governing self-perception. self-perception. They are not not intended as four four discrete syndromes or types of of individuals. Rather, Rather, each metaphor metaphor provides a different different lens through which the self-perceiver self-perceiver can be viewed, and each highlights a unique set of of processes. Together, the four four metaphors provide a framework framework for for organizing the various self selfprocesses that have been postulated postulated to underlie self-perception self-perception biases. Although Although this framework, framework, like most categorization systems, may blur distinctions among the theories and models categorized categorized within each metaphor, we believe that the heu heuristic value of of the framework framework outweighs the costs associated associated with broad broad catego categories. Figure 11 summarizes the metaphors and illustrates how they can be differenti differentiated in terms of: (a) the central motive motive driving self-perception; self-perception; (b) how information information about the self self-perception process; and self is processed; processed; (c) the role of of affect ajfect in the self-perception individual differences differences constructs relevant to each metaphor. metaphor. (d) the individual

A. The Scientist Every man is, in his own particular way, a scientist [whose] ultimate aim is to predict and control. (Kelly, (Kelly, 1995, 1995, p.p.5) 5)

assumption guiding the Scientist metaphor metaphor is that individuals are The basic assumption understand themselves and the world. This metaphor metaphor has its roots in driven to understand George Kelly's personal construct theory and in the early attribution theories of of Harold Kelley. George Kelly (1995) argued that just as the scientist Fritz Heider and Harold facts and observations to develops empirically based theories, the layperson uses facts develop "personal "personal constructs," or theories about about the self self and the world. Like develop scientific theories, these constructs allow people to interpret interpret and predict behavior. scientific Heider (1958) also noted noted the similarity between between the goals of of the scientist-to scientist—to Heider control—and the goals of of the layperson. Building on understand, predict, and control-and Obviously, other othermetaphors metaphors exist exist ininthe the literature. literature.The The Intuitive Intuitive Psychologist Psychologist (Heider, (Heider, 1958; 1958; 2^Obviously, & Ross, 1980) is essentially a variant of the Scientist metaphor. The Cognitive Miser metaphor Nisbett & is another variant of the Scientist metaphor which focuses on the shortcomings of the person as scientist & Taylor, 1991). The Computer metaphor (e.g., Greenwald & & Pratkanis, Pratkanis, 1984), 1984), which likens (Fiske & mental processes to the workings of a computer, also resembles the Scientist metaphor in that both focus on informational processes and deemphasize the importance of affect. The Lawyer metaphor was & Jaspers, Jaspers, 1980), 1980),but butmore morerecently recently used initially to to characterize characterize attributions attributions of of responsibility responsibility (Fincham & used has been used in a way that resembles the Consistency Seeker metaphor-to metaphor—to emphasize that people reason toward towardpredetermined, predetermined,specific specificconclusions conclusions(Baumeister (Baumeister&&Newman, Newman,1994). 1994).Finally, Finally,the the Economist reason Economist metaphor emphasizes emphasizes that that people people are are utility utility maximizers; maximizers; that that is, is, they act in in ways ways that that advance advance their metaphor they act their own self-interest. However, this this idea is not not unique unique to to the the Economist Economist metaphor metaphor as as virtually virtually all all the the other other own self-interest. However, idea is tomaximize maximizetheir theirpersonal personalutility, utility,but buteach eachmetaphor metaphorspecifies specifies metaphors assume assume that that individuals individuals attempt attempt to metaphors different type type of of utility utility or or motive. motive. aa different

The Consistency Seeker Seeker T he Consistency

The Scientist T he Scientist Motive:

Motive: Motive:

Accuracy Accuracy

-

Data driven - objective objective evaluation evaluation of of Informationproc~ing: D Information-processing: ata driven informationabout about the the self self information

Affect: Affect:

-

Irrelevant- dispassionate dispassionate search search for for Irrelevant

truth about about the the self self truth Individualdifferences: differences: Individual

-

Information-processing: Theory Theory driven driven - biased biased by by motivation motivation to to Information-processing: confirm internal internalrepresentation representationof of self self confirm

Affect: Affect:

Private self-consciousness. self-consciousnass, need need for for Private

cognition, attributional amibutional style style cognition.

Popularity Popularity

Individualdifferences: differences: Individual

Motive: Motive:

-

with impression impressionmade madeon on others; others; selfselfwith

Affective state state depends depends on on whether whether the the Affective impressionsothers others have haveof of oneself oneself are are impressions

clarity clarity

Needfor for approval. approval, self-monitoring. self-monitoring, Need

Self-enhancement Self-enhancement

-

protect and and enhance enhance self-worth self-worth through through protect

positively distorted distortedself-evaluation self-evaluation positively Affect: Affect:

Affective state state depends depends on on whether whether selfselfAffective

Individualdifferences: differences: Individual

Narcissism, self-esteem self-esteem NarCissism.

positive or or negative negative positive Machiavellianism Machiavellianism

Self-conceptstability. stability, self-concept self-concept Self-concept

biased by by motivation motivation to to Information-processing: Ego Ego driven driven - biased Information-processing:

perceptions as as self-presentations self-presentations perceptions

Individualdifferences: differences: Individual

perceptionsare are consistent consistent or or perceptions

The Egoist T he E goist

Information-procasing: Audience Audience driven driven- biased biasedby by concern concern Information-processing:

Affect: Affect:

Affective state state depends depends on on whether whether selfselfAffective inconsistent with with prior prior self-views self-views inconsistent

The Politician T he Pol itician Motive: Motive:

Consistency Consistency

Rcure11 Four Four metaphors metaphorsof of the the self-perceiver self-perceiver F'lGUIlE

perceptionsare are positive positive or or negative negative perceptions

CHAFfER CHAPTER 2S 25

SELF-PERCEPTION ACCURACY AND BIAS IN SELF.PERCEmON

659 659

Heider's theory, Harold Kelley (1967) drew a parallel between the methods used in science and those used in everyday life. Kelley proposed that laypeople make causal inferences about themselves and the world in the same way as scientists, formulated a model of the attribution process based on this assumption. and formulated Several theories of the self self also incorporate the notion that people act, at least in part, like scientists. In his social comparison theory, Festinger (1954) postulated a universal motive to acquire accurate self-knowledge: "There is a motivation in the correct opinions, beliefs, and ideas .. .. .. and to know human organism to hold correct precisely what his abilities enable him to do in this world" (p. 194); in the absence precisely (p. 194); of objective standards, people attempt to acquire an accurate appraisal of their abilities by comparing themselves to others. According to Bem Bern (1972), individuals self-knowledge by observing their own behavior in much the same way as gain self-knowledge would an an observer: observer: Just Just as as the the empirically empirically oriented oriented scientist scientist observes observes the the world world would and draws conclusions, we observe ourselves, note the behaviors we exhibit, and and draws conclusions, we observe ourselves, note the behaviors we exhibit, and form impressions accordingly. Thus, Bem's self-perception theory did not focus on form impressions accordingly. Thus, Bern's self-perception theory did not focus on the possibility that the process of perceiving oneself may be threatening and may the possibility that the process of perceiving oneself may be threatening and may therefore differ differ from from the the process process of of perceiving perceiving others (see John John & & Robins, Robins, 1993). therefore others (see 1993). Trope (1979) argued that people have a need for accuracy and found evidence that that Trope (1979) argued that people have a need for accuracy and found evidence people prefer to engage in tasks that provide them with valid information about people prefer to engage in tasks that provide them with valid information about abilities. Finally, Finally, Baumeister Baumeister and and Newman Newman (1994) (1994) explicitly explicitly draw draw on on ideas ideas from from their abilities. their the Scientist metaphor to understand the self-regulation of inferential processes. the Scientist metaphor to understand the self-regulation of inferential processes. In summary, the Scientist metaphor (a) focuses on the motivation to acquire informational processes accurate self-knowledge, (b) emphasizes perceptual and informational and highlights people's inferential inferential strategies and shortcomings, and (c) deempha deemphaaffective processes, assuming that people's beliefs about themselves are largely sizes affective immune to the influence of emotion. In short, the Scientist metaphor suggests that people are data driven and engage in a dispassionate search for accurate self-knowledge. selffor the accuracy of self What are the implications of the Scientist metaphor for perception? The Scientist metaphor implies that, like the idealized scientist, people form form opinions about themselves based on the available data, with no regard for how favorable favorable the information information is and with no regard for whether the information information is consistent with their previous beliefs about themselves. Thus, beliefs about the self self should be accurate, and their degree of correspondence with reality is con constrained only by the individual's perceptual and informational informational limitations. Nisbett and Ross (1980) epitomize this position, although they argue that people are not particularly good at being scientists: We from the We proceed proceedfrom the working working hypothesis hypothesis that that inferential inferential and and judgmental judgmental errors errors from nonmotivational-perceptual arise primarily primarilyfrom nonmotivational—perceptual and and cognitive-sources. cognitive—sources.Such Such errors, we we contend, contend,are are almost almostinevitable inevitableproducts productsofofhuman humaninformation-process information-processing ing strategies strategies .. .. .. .[M]any .[M]any phenomena generally generally regarded regarded as as motivational motivational (for (for example, self-serving self-serving perceptions perceptions and attributions, ethnocentric ethnocentric beliefs, beliefs, and example, and attributions, and many typesof of human humanconflict), conflict),can can be be understood understoodbetter betteras as products productsofofrelatively relatively many types

660 660

ROBINS ROBINS AND AND JOHN JOHN

information-processing errors than of deep-seated motivational passionless information-processing forces. (p. (p. 12) 12)

In the the personality literature, the Scientist perspective is is represented by McCrae as primarily and Costa's (e.g., 1990) 1990) interpretation of self-reports of personality as reflecting underlying personality traits, rather than motivated self-presentations self-presentations or response styles. What about individual differences? de differences? We We all know people who who are aptly described by the Scientist metaphor and other people who who are not. not. Although the refer to types of people, individuals may may vary in the degree to metaphors do not refer metaphor characterizes them. Which individual-difference individual-difference measures which each metaphor can be linked to the Scientist metaphor? Although there is no single personality construct that focuses thyself,'* measures of individual differ differfocuses on the need to "know thyself," ences in attention and information information processing (e.g., private self-consciousness, need for cognition, attributional style) should be relevant to the Scientist metaphor. For for self-consciousness (Buss, 1980) reflects the tendency to attend to example, private self-consciousness and become aware of internal thoughts, motives, and feelings; the scale includes items such as "I'm always trying tofigure figure myself out" and "I'm constantly examining selfmy motives." A number of studies have shown that individuals high in private self (e.g.. Cheek, 1982). 1982). Thus, the consciousness provide more accurate self-reports (e.g., metaphor should be particularly appropriate for characterizing the self selfScientist metaphor perception processes of privately self-conscious individuals, but less appropriate self-consciousness. for those low in private self-consciousness. Another individual-difference individual-difference variable that might seem relevant to the Scien Scientist metaphor is depression. Some researchers have argued that depressed individu individuals have more realistic self-views (e.g.. (e.g., Alloy & Mis & Abramson, 1988; 1988; Lewinsohn, Mischel, Chaplin, & of & Barton, 1980). We We do not believe the self-perception self-perception processes of depressed individuals conform conform to the Scientist metaphor for two two reasons. First, some research suggests that depressives have unrealistically negative, rather than accurate, self-perceptions self-perceptions (for critiques of the "depressive realism" literature, see Ackerman & DeRubeis, DeRubeis, 1991; 1991; Colvin Colvin & &Block, Block, 1994; 1994;John John&&Robins, Robins,1994). 1994).Second, Second, even if depressives do perceive themselves accurately, it is is unlikely that this occurs through the the dispassionate search for accurate information information captured by the Scientist metaphor. Thus, "depressive realism" illustrates the need to distinguish between the process of self-perception self-perception and the the accuracy accuracy of of the the outcome. outcome. The The Scientist Scientist metaphor metaphor describes a particular set of processes through which people form form self-perceptions; self-perceptions; it does not claim that the outcome of these processes will be necessarily accurate. B. The Consistency Seeker self-conceptions act like the rudder of a ship, bolstering people's confi confiStable self-conceptions dence in their ability to navigate through the sometimes murky seas of everyday social life. (Swann, (Swam, Pelham, Pelham, & & Krull, Kmll, 1989, 1989,p.p.783) 783)

CHAPTER CHAPTER 2S 25 ACCURACY ACCURACYAND ANDBIAS BIASIN INSELF.PERCEmON SELF-PERCEPTION

661 661

The Consistency Seeker metaphor assumes that individuals strive to see them themselves in a consistent manner. Its origins can be traced to self-consistency self-consistency theory (e.g., Lecky, 1945), balance theory (e.g., Heider, 1958), and cognitive dissonance theory (e.g., Festinger, 1957). In these theories, consistency has two meanings. First, people strive to maintain consistency consistency among their beliefs. Second, people strive to confirm their preexisting beliefs and maintain them across time and situations. All consistency theories share the view that the subjective experience of of inconsistency therefore motivated to reduce or produces an aversive state, and that people are therefore interpersonal function, ensuring prevent inconsistency. Consistency also serves an interpersonal that people will honor the identities they negotiated in previous social interactions and act similarly over time. Swann's (e.g., 1990) theory of self-verification self-verification builds on these earlier consis consistency theories and applies the notion of consistency directly to people's beliefs about themselves. He characterizes the person as an "architect of of social reality," extensive research program, striving to bring reality into harmony with the self. In an extensive Swann and his colleagues have amassed considerable evidence that people actively seek out and create contexts in which their self-views will be confirmed, even when their self-views self-views are are negative. negative. For For example, example, people people with negative self-views self-views prefer prefer their with negative interacting with individuals individuals who who have have aa negative negative impression to interacting interacting interacting with impression of of them them to with those those who who have have aa favorable favorable impression. impression. Another Another way way in in which which people people verify verify with their self-conceptions self-conceptions is is through through cognitive cognitive distortions distortions (e.g., selective attention) attention) that that their (e.g., selective allow people people to to see see more more self-confirmatory self-confirmatory evidence evidence than than actually Research allow actually exists. exists. Research in the the social social cognition cognition literature, literature, for for example, suggests that people process process in example, suggests that when when people information about about themselves themselves they they are are biased biased by by their their existing and information existing self-conceptions self-conceptions and expectations (Higgins (Higgins & & Bargh, Bargh, 1987; Markus & & Wurf, Wurf, 1987). 1987). In In summary, summary, the the expectations 1987; Markus Consistency-Seeker strives strives to to maintain maintain aa consistent set of of self-views self-views by by seeking seeking out out Consistency-Seeker consistent set self-congruent information information and and by by distorting distorting incongruent incongruent information to create create an an self-congruent information to illusion of of consistency; consistency; when when these these mechanisms mechanisms fail fail and and the the perceived perceived reality reality is is not not illusion in harmony harmony with with the the self-view, self-view, the the person person experiences experiences negative affect. in negative affect. self-perception, consistency theories assume With regard to the accuracy of self-perception, of self. Thus, informa informathat preexisting self-conceptions anchor and drive perceptions of tion that is inconsistent with current self-views, including valid information, may be distorted or simply avoided. In a sense, prior beUefs beliefs about the self self provide a framework for processing new information and thus form form a Procrustean bed for the acquisition of new beliefs about the self. Thus, the Consistency Seeker metaphor is particularly appropriate for generating hypotheses in in research evaluating the accuracy of self-perceptions against information-processing information-processing criteria. For example, research on schematic processing suggests that people may fail to attend to valid autoinformation if it is incongruent with their self-schemas. Similarly, research on auto biographical memories suggests that people selectively remember life events that are consistent with how they see themselves currently; that is, they reconstruct their past to fit their current self-views (Ross, 1989). Are there individual differences in the tendency to act like the Consistency Seeker? In other words, do some people have a stronger need to confirm and

662 662

ROBINS AND AND JOHN JOHN ROBINS

maintain their preexisting self-conceptions self-conceptions than others? Rosenberg's (1979) Stability of of Self Self Scale was designed to measure individual differences differences in the tendency to see oneself oneself as the same person over time and across situations, and includes items such D o you ever find that on one day you have one opinion of yourself yourself and as ""Do another day you have a different different opinion?" More recently, Campbell (1990; on another self-concept clarity, which Campbell & Lavallee, 1993) 1993) developed the construct of self-concept **self-beliefs are clearly and confidently confidently defined, defined, she defines as the extent to which "self-beliefs temporally stable, and internally internally consistent" (Campbell, Chew, & Scratchley, 1991, temporally 475). Campbell (1990) found found that individuals high in self-clarity self-clarity showed more p. 475). between their self-concepts self-concepts and their subsequent perceptions of their congruence between found a sex behavior in specific situations. Finally, Beyer and Bowden (in press) found difference in consistency consistency seeking; compared compared to men, women based their self-evalua self-evaluadifference tions of task performance performance more on their initial (low) (low) expectations than on their actual performance, performance, contributing contributing to to aa self-diminishment self-diminishment bias. bias. Thus, Thus, the the Consistency Consistency actual Seeker metaphor seems seems particularly particularly appropriate appropriate for for characterizing characterizing some some self selfSeeker metaphor perception processes of women women and and those those of of individuals individuals high high in in self-concept self-concept stability stability perception processes of and clarity. clarity. and C. The Politician c. myself which II try to create in my own mind .. . .. is different different from from The image of myself the minds of others in order that they may the image which II try to create in the love me. me. (W. (W, H. K Auden Auden in Snyder, Snyder, 1987) 1987)

The view of the person as a political animal traces its historical roots to Aristotle and Machiavelli. In its contemporary contemporary form, the Politician metaphor metaphor repre represents a fusion fusion of ideas from from symbolic interactionism, role theory, and impression impression management theory, as well as more recent elaborations of these theories (e.g., Hogan, Hogan, 1983; Schlenker, 1980; Snyder, 1987; Tetlock, 1992). These theories empha emphasize the reciprocal reciprocal nature of social interaction; the person is both a product product of the social context and a creator creator of social reality. Social reality reality is constructed and negoti negotiated through through interactions with others. In these interactions, people's behaviors behaviors represent represent public performances performances that "present images of the self self for the social world to see and evaluate" (Schlenker, 1985. p. 21). 1985. p. 21). The primary goals of these self selfpresentations are to get along along (i.e., gain approval from from others) and to get ahead (i.e., attain social status and power) (Hogan, Jones, & Cheek, 1985). Thus, in the social as well as the political arena. arena, people people attempt attempt to influence and manipulate manipulate how they are seen by others. Just as politicians strive to gain the approval of their constituencies constituencies to become elected elected or remain remain in office, people attempt attempt to make make favorable favorable impressions on others in order order to gain approval approval and status. Building on these ideas, Tetlock Tetlock (1992) argued argued that the demands people face face in their their everyday life are similar to those faced faced by politicians. Like politicians, people people accountable to different different audiences (or constituencies), which place multiple and are accountable often conflicting conflicting demands and expectations expectations on them. The primary primary goal of the often

CHAPTER CHAPTER 25 25

ACCURACY ACCURACYAND ANDBIAS BIASIN INSELF·PERCEmON SELF-PERCEPTION

663 663

person-as-politician is to "maintain the positive regard of important constituencies to whom he or she feels accountable" (Tetlock, 1992, p. 332). How are these ideas related to the self-perception A core assumption self-perception process? A of symbolic interactionism, as well as of other theories subsumed by the Politician metaphor, is that the self-concept is defined, constructed, and negotiated through self-concept in two funda fundainteractions with others. Self-presentations influence the self-concept mental ways. First, people are influenced by how others perceive and respond to their social behavior; that is, the impressions of others serve as a "looking glass," reflecting back an image of ourselves. Second, to some extent people "believe" the reflecting beliefs roles they play during their self-presentations and thus incorporate these beliefs into their self-concepts (e.g., Schlenker, Dlugolecki, & Doherty, 1994). More gener generally, there is a reciprocal relation between self-presentations self-presentations and self-conceptions: self-concept is shaped by attempts to gain acceptance and status through social the self-concept interactions, and in turn, people's self-presentations self-presentations are guided by their self-concep self-conceptions. In summary, the central assumptions driving the Politician metaphor are that individuals (a) are motivated by a concern with the impression impression they make on others, self-presentations to achieve status and approval from others, and (b) alter their self-presentations (c) experience negative affect when they fail to gain approval. selfaffect the accuracy of self How does a concern with others' impressions affect perception? At the most general level, many theories adopting the perspective of the Politician metaphor assume that people have no core self self or personality; like an onion, when the layers of social roles and public presentations are peeled away, nothing remains. Thus, the person is viewed solely as a product of the social context, as nothing but a constantly shifting self-presentation. From this perspective, the accuracy question is misguided. self-reports Another perspective is to assume that personality exists, but that self-reports should not be taken as a direct indication of an individual's personality because self-presentational goals (e.g., Johnson, 1981; 1981; Mills & they may be distorted by self-presentational Hogan, 1978). That is, when people respond to questions on a personality question questionnaire they do not report how they see themselves but how they would like to be seen, just as the consummate politician poUtician does not present personally held opinions to others but rather adopts whatever opinions are most popular. According to this view, people engage in impression management when they complete a questionnaire just as they engage in impression management when they interact with others. If self-reports are analogous to self-presentations, then their validity should be influenced by whether they are obtained in a public or private context. Self Selfreports should be more indicative of a person's true self-views in a private context, where the expectations and demands of others are less salient and the impetus to engage in impression managment is consequently weaker. This is akin to asking politicians about their views during an interview on national television as compared "off the record" in the privacy of their homes; the answers one would get in to "off these two contexts are likely to be quite different. Of Of course, one could argue that a politician is never truly off the record, just as psychologists have argued that

664 664

ROBINS AND AND JOHN JOHN ROBINS

even private self-reports reflect reflect self-presentations self-presentations to real or imagined audiences audiences (including presentations to oneself). The Politician metaphor also suggests that manipulating a person's account accountability abihty to others will influence the accuracy of their self-reports. Although the effects effects of of accountability accountabiUty on self-perception self-perception have not been examined directly, we do know that individuals are less prone to certain judgmental errors (e.g., overconfidence) overconfidence) when they are accountable to others, in part because accountability promotes a self-critical process that improves judgment in some contexts (Tetlock, 1992). For self-critical Tetlock and Kim (1987) found that subjects who thought that they would example, Tetlock have to justify their judgments to others were more accurate at predicting how others would respond to personality questionnaires and also had more realistic of confidence confidence in their judgments. In addition, Lerner, Lemer, Spranca, and Tetlock levels of self-reported (1994) found that people's people's behaviors were more consistent with their self-reported attitudes when they they were were accountable accountable to to others. others. attitudes when Similar research designs could be employed to examine the effect of account accountof self-perception. For example, it would be interesting to ability on the accuracy of study whether people evaluate themselves more accurately when they think they justify their self-evaluations self-evaluations to others. In this case (i.e., when publicly publicly will have to justify accountable to others), the person must weigh the potential benefits making a favorable impression on others against the potential costs of public exposure as a Schlenker and Leary (1982) point out, "self-presentation "self-presentation fraud or imposter. As Schlenker involves maintaining a delicate balance among self-enhancement, accuracy, and humility" (p. 89). humility" (p. 89). of the interpersonal processes that may hinder accurate self-percep self-percepIn terms of tion, the Politician metaphor suggests that people will seek out "friendly" audiences who see them as they would like to be seen, regardless of reality. Conversely, Conversely, people will avoid "hostile" audiences that provide feedback feedback counter to their self selfpresentational goals, or suppress or distort the feedback if they cannot avoid such audiences. More generally, people attempt to negotiate with others an identity that suits their interpersonal and personal goals. For example, consider an individual who thinks that her boss values competitiveness but her spouse does not: she may attempt to get her boss to see her as competitive competitive and her spouse to see her as cooperative. There are a number of of personality scales that should be relevant to the processes described by the Politician metaphor. Some of of these scales directly mea measure the tendency to manipulate others to achieve desired goals (e.g., Machiavellian Machiavellianism), whereas others measure individual differences in concern with and sensitivity to the impressions of of others (e.g., impression management, need for approval, public self-consciousness, self-monitoring; for reviews, see Paulhus, 1990; Snyder, 1987). Probably the best known measure is Snyder's (e.g., 1987) Self-Monitoring Self-Monitoring Scale, which includes items such as "In order to get along and be liked, I tend to be what people expect me to be rather than anything else." else." The high self-monitor, according to Snyder (1987), is a person who "is particularly sensitive to cues to the situational ness of appropriateness of his or her social behavior, and who uses these cues as guidelines appropriate

CHAPTER CHAPTER 25 25

ACCURACY ACCURACY AND AND BIAS BIAS IN IN SELF.PERCEmON SELF-PERCEPTION

665 665

for for monitoring monitoring . .. . . his or her expressive behavior and self-presentations" self-presentations" (p. 14). Thus, with with regard regard to their their motivation, information-processing information-processing orientation, and af affective responses, high self-monitors self-monitors should be particularly particularly well characterized characterized by the Politician metaphor. metaphor.

D. The Egoist self-enhancNormal human thought is marked not by accuracy but by positive self-enhanc ing illusions.

(Taylor, 1989, 1989, p.p. 7)7) (Taylor,

metaphor is that people want to like themselves themselves The basic premise of of the Egoist metaphor and will adopt cognitive and interpersonal interpersonal strategies to create and maintain a reflected in many personality and self-concept self-concept positive self-image. This perspective is reflected AUport, 1937; Epstein, 1990; Greenwald, 1980; James, 1890; Kohut, theories (e.g., Allport, virtually every self-theory self-theory posits 1971; Rogers, 1959; Rosenberg, 1979). In fact, virtually variant of of the motive to protect protect and enhance self-worth self-worth (Wells & Marwell, some variant 1976). Allport AUport (1937) referred referred to the defense defense of of the ego as "nature's eldest law," enhancement is a fundamental fundamental human human motive tied to the and believed that ego enhancement need for for survival. Kohut (1971) argued that three motivational motivational systems drive the need and the need need for for self-esteem. Greenwald Greenwald (1980) referred referred person: ambitions, ideals, and "totalitarian ego" to convey the idea that self-perceptions self-perceptions are driven by to the "totalitarian egocentrism; that is, the self reconstructs personal history history to to fit needs. egocentrism; that is, the self reconstructs personal fit self-worth self-worth needs. Finally, the the self self theories theories of of Epstein Epstein (1990), (1990), Rosenberg Rosenberg (1979), (1979), and and many many others others Finally, include self-enhancement self-enhancement as as aa basic basic motive. motive. include of self-perception contemporary models of self-perception have been infused infused with ideas Many contemporary from Egoist metaphor. Researchers building on Festinger's social comparison from the Egoist found that people use social comparisons to bolster their self-views; theory have found for example, people often often compare themselves to less fortunate fortunate others so that they for will seem better (e.g., Wills, 1991). Similarly, the self-worth better in comparison comparison (e.g.. self-worth motive is central to Tesser's (1988) self-evaluation self-evaluation maintenance model, which is based on social comparison comparison theory. Cognitive dissonance research has also been reconceptual reconceptualized in terms of of self-esteem self-esteem motivation; Greenwald and Ronis (1978) noted that dissonance theory theory has become "focused "focused on cognitive changes occurring in the service of of ego defense, or self-esteem self-esteem maintenance, rather than in the interest of of preserving psychological consistency" (pp. 54-55). Steele (1993) has reinterpreted and extended findings findings in the dissonance literature on the basis of of his self-affirmation self-affirmation theory. In the attribution attribution literature, researchers have moved moved beyond perceptual perceptual and informational informational accounts accounts of of self-other self-other differences differences to explore motivational ac accounts (cf. Tetlock Tetlock & Levi, 1982). More generally, there has been a shift shift away from from conceptualizing conceptualizing self-perception self-perception in purely cognitive-informational cognitive-informational terms toward the of self-perception self-perception as driven driven by multiple motivations, including including the need for for view of self-worth. of self-based self-based accounts of of social phenomena, a large body Riding on the crest of of research emerged emerged in the 19808 1980s aimed specifically specifically at documenting documenting positivity biases of

666 666

ROBINS ROBINS AND AND JOHN JOHN

self-perception in a wide range of contexts, including self-conceptions of personal personalin self-perception self-attributions of of success and failure, perceptions of of control, re reity attributes, self-attributions sponses to feedback, and beliefs beliefs about the likelihood of future events. In a review of this literature, Taylor and Brown (1988) argued that self-perceptions self-perceptions exhibit pervasive and enduring positive distortions, presumably stemming from the basic motive toward self-enhancement self-enhancement (see also Lockard & Paulhus, 1988). Recent re research in the Egoist tradition has explored the psychological bases of of these biases, including the role of of cognitive and affective processes and situational factors such as ego involvement. In summary, the Egoist (a) is motivated toward self-enhance self-enhancement, (b) distorts information information about the self to protect and enhance self-worth, affect by protecting self-worth; that is negative self-views are and (c) regulates affect avoided because they produce negative affect. affect. In the personality personality assessment assessment literature, the Egoist metaphor is most notice noticeable in research on socially desirable responding. Paulhus (1984) distinguished responding-impression management and between two kinds of of socially desirable responding—impression self-deceptive manipulapositivity. Whereas impression management (the conscious manipula self-deceptive tion of of one's self-reports self-reports to make a favorable impression on others) falls within the domain of of the Politician metaphor, self-deceptive self-deceptive positivity reflects reflects honestly held of the but unrealistically positive self-views, and therefore falls within the domain of Egoist metaphor. Egoists truly believe that they are exceptional people who are superior to others. With regard to individual differences, the most theoretically theoretically relevant construct is narcissism. According to the DSM-IV criteria, the defining characteristics of the Narcissistic Personality Personality include a grandiose sense of of self-importance, a tendency to Narcissistic ''speexaggerate accomplishments and talents, and an expectation to be noticed as "spe cial" even without appropriate achievement. All clinical accounts of of narcissism (e.g., Freud, 1914/1953; Kohut, 1971; 1971; Millon, Millon, 1990) 1990) concur concurthat that narcissistic narcissistic individuals individuals hold hold unrealistically exaggerated beliefs about their abilities and achievements. In support of of these accounts, research suggests that narcissistic individuals respond to threats to their self-worth by perceiving themselves more positively than is justified (Gabriel, Critelli, & Ee, 1994; John & Robins, 1994; Robins & John, 1997) and by denigrating others (Morf (Mod & Rhodewalt, 1993). The effects effects of of another individual-difference individual-difference variable, self-esteem, have been of self-enhancement self-enhancement biases. Compared with low examined in numerous studies of self-esteem individuals, high self-esteem individuals are more likely to describe themselves more positively than they describe the "average other" (Brown, 1986), in compensatory self-enhancement self-enhancement following negative feed feedmore likely to engage in back (e.g., Baumeister, 1982), more likely to believe their abilities are unusual and their failings are common (e.g., Campbell, 1986), and more likely to derogate sources of negative feedback feedback (e.g., Baumgardner et at, al., 1989). Thus, the Egoist metaphor of seems self-processes of selfto capture the self-processes of individuals high in narcissism and in self seems esteem. esteem.

CHAPTER CHAPTER 25 25

ACCURACY ACCURACY AND AND BIAS BIAS IN IN SELF.PERCEmON SELF-PERCEPTION

667 667

IV, AN A N APPLICATION THE FOUR METAPHORS TO TO IV. ApPLICATION OF mE

SELF.ENHANCEMENT SELF-ENHANCEMENT BIAS At this point, one of one may wonder which metaphor offers offers the best characterization of the psychology of the self-perceiver. We feel that this question is misguided because each metaphor emphasizes a different different aspect of the self-perception self-perception process. Like men and the elephant, each metaphor by itself itself reveals only the fable of the blind men the picture, picture, but but together they provide a more complete account. Thus, a part of the may be gained when findings in the self-perception self-perception literature are new insights may from each of the four perspectives. Below we we use use the four metaphors interpreted from one kind of self-perception to generate complementary accounts of one self-perception bias that we own research, namely, unrealistically positive self-perceptions. have studied in our own In several studies, we have measured self-enhancement self-enhancement bias in a group discusdiscus sion task in which subjects compete for a fixed amount of money that must be allocated consensually by the group (see (see John & & Robins, 1994; 1994; Robins Robins &&John, John, 1997). After After the group discussion is completed, subjects evaluate their performance performance assess the accuracy of these self-evaluations, self-evaluations, relative to the other group members. To assess members we compare them with performance evaluations made by the other group members who observe the task but do not enter into the and by a group of psychologists who discussion (Le., (i.e., two two social consensus criteria). criteria). Thus, Thus,self-insight self-insight isis defined defined as as knowing knowing as in ancient Greece selfone's place relative to the other group members (just Gust self insight implied knowing one's place in life-above life—above the animals but beneath the Gods). We find that, on average, subjects evaluate themselves more positively than selfthey are evaluated by either their peers or the psychologists. This general self enhancement bias is one one of three positive illusions discussed by Taylor and Brown (1988). Brown (1988). How can the metaphors help us understand the mechanisms and motives that underlie self-enhancement bias? Below we discuss our research within the theoretical framework framework provided by the four four metaphors and describe hypotheses generated by each metaphor. A. The The Scientist Metaphor: Metaphon Perceptual and Infonnational Informational Limitations

self-knowlThe Scientist metaphor assumes that people strive to acquire accurate self-knowl informational processes in self selfedge and emphasizes the role of perceptual and informational self-perception biases are apparent, perceptual and perception. Therefore, when self-perception informational mechanisms are are invoked to explain the findings. For example, the informational self-enhancement bias in in our research could be due to the different different visual perspective self-enhancement may be important because it can influence of self and others. Visual perspective may When people people information that is available and salient to a judge (Storms, 1973). the information 1973). When from the same perspective as are given the opportunity to observe their behavior from

668 668

ROBINS ROBINS AND AND IOHN JOHN

others see them, they may be better able to see their their own shortcomings and thus be less inclined to self-enhance. In effect, effect, reversing visual perspective could change change the information influence what is perceptually perceptually salient, and as information available to the self, influence a result make reality reaUty less ambiguous and more constraining. constraining. To test this hypothesis, we examined examined the effects effects of visual perspective perspective (manipu (manipulated via videotape) on self-enhancement self-enhancement bias (Robins & John, 1997). Subjects Subjects from two visual perspectives: evaluated their performance performance in a group discussion task from (a) the "normal" perspective experienced by the self self and (b) a "self-focused" **self-focused" perspective in which subjects watched watched themselves on videotape videotape from from the perspective perspective of an external observer. Visual perspective perspective did not affect affect the degree of self-enhance self-enhanceof overestimated their their performance ment bias; on average, subjects overestimated performance to the same extent view ourselves regardless of their visual perspective. Apparently, even when when we view from the visual perspective of others, we still see ourselves from from the emotional emotional from perspective of the self. B* The Consistency Seeker: Seeken Confinning Confirming Prior Beliefs about the tlie Self Self B. metaphor raises the possibility possibility that in our experiment experiment The Consistency Seeker metaphor confirm their prior beliefs beliefs about posindividuals were striving to confirm about themselves. It is pos dissible, for example, that when subjects evaluate their performance performance in a group dis beliefs about about their effectiveness effectiveness cussion task they are biased by their more general beliefs currently under under way, in group discussions. In a study currently way, we tested this hypothesis by perform in group discusmeasuring subjects' beliefs about how well they typically perform discus sions several weeks prior to the experiment. Subjects who reported reported that they are typically very effective effective tended to rate their performance performance in the group discussion subjects reporting more positively than was justified. Conversely, subjects reporting relatively poor performance rated their performance performance more negatively typical performance negatively than justified. Thus, subjects based their self-evaluations, in part, on how they expected expected they would subjects perform rather rather than on how they actually performed. performed. This finding finding is consistent consistent with perform consistency contributes contributes to self-perception self-perception bibi the idea that the motive to maintain consistency ases.^ ases.3

C. The Politician: Managing Public Impressions The Politician metaphor metaphor raises the possibility that unrealistically unrealistically positive self-evalua self-evaluations reflect reflect self-presentations self-presentations aimed at gaining the approval approval of others, rather rather than privately believed believed distortions about about the self. However, in most most research on self^Although Althoughsubjects' subjects*self-evaluations self-evaluationsfailed failedtotoconform conformwith withthe thesocial socialconsensus consensuscriteria criteriawe weused used 3 to assess accuracy, one (i.e., one might argue argue that subjects subjects were appropriately weighting prior probabilities (i-e.,

their previous performance) when evaluating their current performance, and were therefore acting in

Bayes* theorem. Thus, this might be be aa case case in which subjects were inaccurate relative accordance with Bayes' one kind of criterion (social consensus) but but accurate relative to another (the (the prescriptions of a to one normative model).

CHAPTER25 CHAPTER 2S

ACCURACY AND AND BIAS BIASININSUF.PERCEmoN SELFPERCEPTION ACCURACY

669 669

enhancement biases, including our own, self-evaluations are obtained obtained in an anon anonenhancement own, self-evaluations ymous context with no explicit accountability demands. Nonetheless, we can about the effects effects of manipulating private versus versus public context and acspeculate about ac would not need justify their countability. In our research, subjects knew they would need to justify self-evaluations self-evaluations to others. Consequently, the subjects subjects may have felt felt free free to self selfenhance without without fear fear of appearing conceited conceited and boastful. boastful. If we had manipulated evaluations accountability (e.g., by telling subjects they would have to discuss their evaluations metaphor would predict predict a decrease decrease with the other group members), the Politician metaphor in self-enhancement self-enhancement bias. In some public contexts, however, individuals individuals may be more inclined to self-enhance; for example, professors professors discussing discussing their their research research in the departmental hallways may exaggerate the quality of their their research when the department chair chair joins the conversation. Another Another way to link the Politician meta metadepartment phor to self-enhancement self-enhancement bias is to examine whether whether individual-difference measures individual-difference measures related to impression management management predict predict individual individual differences differences in the degree of related of self-enhancement bias. To test this possibility, we correlated several social desirabil desirabilself-enhancement ity scales with self-enhancement self-enhancement bias and found found weak weak and nonsignificant nonsignificant relations (John & Robins, 1992). Apparently, individuals who are more inclined to manage ( John 1992). their impressions are not more likely to show self-enhancement self-enhancement bias in the private context of of our our research. research. context

D. The Egoist: Narcissistic Self-Aggrandizement Self-Aggrandizement The Egoist metaphor provides the most direct and compelling account of the self selfenhancement effect-individuals effect—individuals want to like themselves and will attempt to do so performing poorly is threatening to the sub subregardless of reality. In our research, performing (MBA students) self-image as successful successful future future managers, which may lead jects' (MBA would them to distort reality to convince themselves that they are as capable as they would be. This account explains the general self-enhancement self-enhancement effect. However, it like to be. differences in self-enhancement self-enhancement that we does not explain the substantial individual differences found in several samples. We find that only about about 35% of the subjects show have found self-enhancement bias whereas about 50% are relatively accurate and about about a clear self-enhancement 15% actually show self-diminishment self-diminishment bias ((John John & Robins, 1994).^ 1994).4 fact that some individuals individuals self-enhance How can we account for the fact self-enhance whereas individual-differences framework framework others do not? Narcissism theories provide an individual-differences for research on self-enhancement self-enhancement bias. According to these theories, narcissistic for individuals hold unrealistically positive beliefs beUefs about their abilities and achieve4 ^This This finding findingcontrasts contrastswith withTaylor Taylor and andBrown's Brown's (1994a) (1994a) claim claim that that the thepercentage percentage of of subjects subjects (p. 973). 973). This apparent inconsistency is a 95%" (p. who show self-enhancement bias is typically "above 95%" good example of the confusion that arises when researchers do not specify the criterion they use to justify their accuracy claim. The Taylor and and Brown claim is based on a normative-model criterion justify (i.e., the the discrepancy other" should should discrepancy between expectations expectations for the self and expectations for the "average other" our accuracy claim is based on an an expert-based social consensus criterion. criterion. Apequal zero) whereas our Ap parently, by their their criterion criterion virtually virtually everyone everyone self-enhances self-enhances whereas by our our criterion criterion less less than than half parently, by whereas by half do so. do so.

670 670

ROBINS AND JOHN JOlIN ROBINS

ments. Thus, we predicted that individual differences in self-enhancement self-enhancement bias would be a function of narcissism. In support of this hypothesis, we found that both self-report and observer measures of narcissism predicted who would show the most pronounced self-enhancement bias in our experimental task (John & Robins, are summarized in Figure 2. Note that narcissists and and nonnarcis nonnarcisfindings are 1994). Our findings sists did not differ differ in actual performance, as defined by our accuracy criterion (judgments by 11 11 psychologists trained to evaluate performance in this task). HowUudgments task). How ever, they differed considerably in their self-evaluations: narcissists overestimated their performance relative to the criterion whereas nonnarcissists underestimated nonnarcis slightly. Thus, although narcissists did not actually perform better than nonnarcissists, they believed believedthey theyperformed performed substantially substantially better better (almost (almost aafull full standard standarddevia deviation). This finding finding replicated when we assessed accuracy using a second social consensus criterion (peer evaluations) and an operational criterion (an objectively assessed task outcome) (Robins & & John, John, 1997). What processes and mechanisms underlie the link between narcissism and self-enhancement bias? Our findings findings are are consistent with the assumption that narcis narcisthan other otherindividuals individualsbecause becausetheir theirsense senseof ofself-importance self-importance sists are more defensive than and superior competence is inflated and thus more easily threatened (e.g., Westen, 1990). Thus, when they do not perform well in an ego-involving context, narcissistic

4 .5 h 4.5

................. ........... .

.

. . .... ..... . . .

.

.

.

.

. .

...

.... .. .

.

High N arci ss i s m Narcissism Q) (D CJ O C C CtI

4.0 4.0

E E .... o

't:

& tf

3.5 3.5

." ..................

� . . . . . .

.....'.............. ................ . low arci ss i s m Low N Narcissism

---r---r---� 3.0 3.0 �T" T Psychologist Self-evaluation Psychologist criterion criterion Self-evaluation

FIGUREl2 FiGURl

Psychologist evaluations evaluations (criterion) (criterion) and andself-evaluations self-evaluations of of performance performance as as aafunction function of of narcis narcisPsychologist sism. (These analyses are are based based on on data data from from John John & &Robins, Robins, 1994.) 1994.)

CHAPTER CHAPTER 25 25 ACCURACY ACCURACY AND AND BIAS BIAS IN IN SELF·PERCEPTION SELF-PERCEPTION

671 671

individuals are particularly particularly motivated to bolster their self-image by positively distort distorting their self-perceptions. Indeed, a reanalysis of our data ((John John & Robins, 1994) 1994) shows that the narcissism narcissism effect effect is even stronger for those subjects who did not perform well (as defined defined by the psychologist criterion); the correlation between perform narcissism and self-enhancement self-enhancement was .62 .62 among the worst performers, compared compared performance (and to performers, respectively. This pattern indicates that poor performance therefore therefore greater greater ego threat) accentuates the difference difference in self-enhancement self-enhancement bias between subjects high and low in narcissism. A found when we manipulated self-focused self-focused A similar accentuation effect effect was found attention via videotape (Robins & John, 1997). Watching their performance performance on videotape further further increased the disparity between the self-evaluations self-evaluations of narcissistic narcissistic and nonnarcissistic individuals. Specifically, narcissists evaluated themselves even Selfmore positively whereas nonnarcissists saw themselves in a less positive light. Self focused attention, like poor performance, is particularly threatening because self selffocused evaluative processes are triggered, heightening awareness of the discrepancy bebe inflated internal standards of the narcissists. Apparently, the tween reality and the inflated self-focused condition condition further further activated narcissists' self-esteem self-esteem protective mecha mechaself-focused inflated self-evaluations. self-evaluations. nisms, leading to even more inflated Self-Enhancement Bias for Mental Health E. Implications of Self-Enhancement self-enhancement bias is relevant to an important important The Unk link between narcissism and self-enhancement debated in the literature: Hterature: What are the implications of self-enhance self-enhanceissue currently debated ment bias for mental health (Colvin & Block, 1994; Shedler, Mayman, Mayman, & Manis, 1993; 1993; Taylor & Brown, 1988, 1988,1994)? generally considered to to Taylor 1994)1 Narcissistic characteristics are generally personalbe indicative of psychological problems, and, at extreme levels, constitute a personal selfity disorder. Thus, our narcissism findings suggest that unrealistically positive self compared against against both social consensus and operational criteria) reflect reflect views (as compared maladjustment. This view is consistent consistent with those of Freud, Jahoda, and many many maladjustment. clinical psychologists, who assumed assumed that self-insight self-insight (i.e., the absence of either either positively positively or negatively biased self-perceptions) self-perceptions) is an essential aspect aspect of mental mental health. As Allport AUport (1937) noted, "Not infrequently infrequently insight is exalted to the highest place among the virtues, or therapeutically is regarded regarded as a panacea panacea for all mental ills" (p. (p. 221). 221). This traditional view has recently been challenged by Taylor Taylor and Brown (1988, 1994 Brown, 1993), who have argued (1988,1994 argued that "positive illusions" about about the self self promote and maintain mental health. Our narcissism findings findings appear to contradict the Taylor and Brown thesis. Still, they do not provide direct counterevi counterevidence because one could argue that narcissism theories are wrong and narcissism narcissism is in fact fact adaptive. To provide provide a more direct direct test, we examined whether self-enhancement self-enhancement bias is associated associated with maladjustment, maladjustment, as our narcissism findings suggest, or with better better adjustment, adjustment, as Taylor Taylor and Brown (1988) have suggested. Using data available for the subjects in the John and Robins (1994) study, we examined whether subjects subjects who showed self-enhancement self-enhancement bias were better or worse adjusted adjusted than subjects subjects

ROBINS ROBINS AND AND JOHN JOHN

672 672 0. 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Q) 0) ... o o U o ell 0) I I N

N .... C C Q) 0)

SB ell

0.3 h

0.2 . . . . . . . . . . . . . . . . . . . . . .

Accurate. Accurat^s

0.2 h

0. '

......................

0.1 h

0O rr----,-�--�._--rl

(0 = . ... "a xT
o O

-0 . 3

lf.Dlmlnl sh e rs . Se $4itt*DiminUhers

.......................

............................................ SiilfliiiHMbils e l f·Enhancers

-0.3

FIGURE 33 Optimal Optimal adjustment adjustment as as aa function function of of self-perception self-perception bias: bias: Self-enhancers Self-enhancers are are not not well well adjusted adjusted f1cuu according to psychologists. (These analyses are are based on data from John & & Robins, 1994.)

who were relatively accurate or showed a self-diminishment self-diminishment bias. The 102 subjects were divided into three groups: self-enhancers, who substantially overestimated their performance in the group discussion; accurates, who were relatively accurate in their self-evaluations; and self-diminishers, who substantially underestimated their performance. To assess adjustment, we used Block's (196111978) (1961/1978) index of of "optimal psychological adjustment," which is computed from Q-sort personality personaUty descriptions. In our study, five psychologists provided Q-sort personality descripdescrip tions of each subject; these psychologists observed the subjects in a wide range of activities over the course of a weekend-long assessment program and had ample adjustment.^ opportunity to evaluate adjustment.s Figure 3 shows the level of psychological adjustments for the self-enhancers, ANOVA accurates, and self-diminishers. A one-way ANOV A showed that the three groups p <.05. Contrary to Taylor and = 3.1, P differed significantly in adjustment: F(2, 99) = least well-adjusted of the Brown's (1988) prediction, the self-enhancers were the least t{99) = = 2.4, P p <.05. This three groups and differed significantly from the accurates, 1(99) findingbolsters bolstersour ourinterpretation interpretationofofthe thenarcissim narcissimeffect: effect:positive positiveillusions illusions(relative (relative finding Weused usedthe thepeer peercriterion criteriontotoindex indexself-enhancement self-enhancementbias, bias,thus thusensuring ensuringthat thatthe thebias biasmeasure measure 5^We was independent of the psychologists' ratings of adjustment. adjustment.

CHAFfER CHAPTER 25 25 ACCURACY ACCURACYAND ANDBIAS BIASININSELF·PERCEPnON SELF-PERCEPTION

673 673

to a social consensus criterion) are not necessarily associated with adjustment, and they may even be maladaptive. The two findings together (self-enhancement (self-enhancement bias maladjustis associated with both maladjustment and narcissism) suggest that the maladjust ment of self-enhancers is narcissistic in character, rather than taking the form of overt anxiety or low self-esteem. This interpretation receives further support from self-enhancement bias was associated associated a recent longitudinal study which found that self-enhancement Q-sort correlates (Colvin, Block, & & with a generally narcissistic pattern of Q-sort Funder, 1995). Our findings, although replicated repUcated in an independent independent study, are nonetheless insufficient to warrant the conclusion that self-enhancement self-enhancement bias is necessarily insufficient maladaptive. This issue is complicated and more complex hypotheses need to be considered. Rather than pitting the two opposing claims against each other, future of when self-enhancement research should turn toward the question of self-enhancement is adaptive, longand when it is maladaptive. For example, does it have short-term benefits but long term negative consequences? Is it adaptive in some contexts (e.g., when one has term optimal margin no control over the outcome), but maladaptive in others? Is there an optimal 1989)—is a small bias beneficial beneficial but a large one harmful? of illusion (Baumeister, 1989)-is Finally, so so far far we we have have studied studied self-enhancement self-enhancement bias bias relative relative to to two of criteria criteria Finally, two types types of (social consensus consensus and and operational); operational); whether whether or or not not self-enhancement self-enhancement relative relative to to (social other types types of of criteria criteria is is maladaptive maladaptive awaits awaits empirical empirical test. test. In In conclusion, conclusion, the the impor imporother tant question question of of how how self-insight self-insight and and adjustment adjustment are are related related requires requires further further exami examitant nation to to evaluate evaluate these these various various possibilities. possibilities. nation V. CONCLUSION V. thyself" inspired an enduring interest in the The ancient Greek mandate to "know thyself" issue of self-insight. Our review and analysis of current theory and research reveals many complex issues with which researchers must grapple. The scientific study of self-perception accuracy requires an understanding of of the person at three levels: as object (e.g., what is the true nature of the person?), as perceiver (e.g., how do people form self-perceptions?), and as researcher (e.g., how can we determine whether a self-perception self-perception is accurate?). At each level, philosophical, conceptual, and methodological issues must be considered. In this chapter, we have discussed these issues in the context of of two conceptual frameworks, one that organizes the various criteria researchers use to assess the accuracy of of self-perception, and another different theoretical accounts of of the self-perception self-perception process. We that organizes the different self-insight hope that these frameworks will stimulate and guide future research on self-insight and its causes and consequences.

ACKNOWLEDGMENTS of R.W.R. This chapter is based, in part, on the doctoral dissertation of RW.R Preparation of this was supported supportedby byaaNational NationalScience ScienceFoundation FoundationGraduate GraduateFellowship Fellowshipand andaaChancelChancelchapter was

ROBINS ROBINS AND AND JOHN JOHN

674 674

lor's Dissertation Fellowship from from the University of California California to RW.R R.W.R. and National to O.P.J. O.P.J.We Wethank thankSylvia SylviaBeyer, Beyer,William WilhamChaplin, Chaplin, Institute of Mental Health Grant MH49255 to Jonathan Cheek, Kenneth Craik, David Funder, Samuel Gosling, Ravenna Helson, Robert Jennifer Pals, Hogan, William Ickes, John Johnson, Joachim Krueger, Gerald Mendelsohn, Jennifer Saucier, and William Swann for their comments on an earlier version. Delroy Paulhus, Gerard Saucier,

REFERENCES Ackerman, R, R., & DeRubeis, R R. J. (1991). (1991). Is depressive realism real? Clinical Clinical Psychology Psychology

Review, Review, 11, 11, 565-584.

Four theoretical perspectives. Alloy, L. B., B., & Abramson, L. Y. (1988). Depressive realism: Four L. B. B. Alloy (Ed.), Cognitive Cognitive processes processes in in depression depression (pp. (pp. 223-265). 223-265). New New York: York: In L. Guilford Guilford Press. Personality:AApsychological psychologicalinterpretation. interpretation. New York: Holt. AUport, New York: Holt. Allport, G. W. (1937). Personality: R. M. M. (1988). (1988). An ecological framework framework for for establishing a dual-mode theory of social Baron, R & A. A. W. W.Kruglanski Kruglanski (Eds.), (Eds.), The Thesocial socialpsychology psychologyof ofknowledge knowledge knowing. In D. Bar-Tal & York: Cambridge University Press. (pp. 48-82). New York: R. F. (1982). (1982). A self-presentational self-presentational view of of social phenomena. Psychological Psychological Baumeister, R Bulletin, 91, 91,3-26. 3-26. Bulletin,

R. F. (1989). The optimal margin of illusion. Journal of of Social and Clinical Clinical Baumeister, R Psychology,8,8,176-189. 176-189. Psychology,

& Newman, Newman,L. L.S.S.(1994). (1994).Self-regulation Self-regulationofofcognitive cognitiveinference inferenceand anddecision decision R. F., & Baumeister, R processes. Personality 20,20, 3-19. Personalityand andSocial SocialPsychology PsychologyBulletin, Bulletin, 3-19. affect interpersonally: Baumgardner, A. H., Kaufman, C. M., & Levy, P. (1989). Regulating affect Journal of of Personality Personalityand andSocial Social When low self-esteem leads to greater enhancement. Journal Psychology, Psychology, 56, 56,907-921. 907-921.

Bern, experimental Bem, D. J. (1972). Self-perception Self-perception theory. In L. Berkowitz (Ed.), Advances in experimental -62). New social social psychology psychology (Vol. (Vol. 6, 6, pp. pp. 11-62). New York: York: Academic Academic Press. Press. Beyer, S. S. (1990). (1990). Gender differences differences in the accuracy of self-evaluations self-evaluations of performance. performance.

Journal Psychology, 59,59, 960-970. Journal of ofPersonality Personalityand andSocial Social Psychology, 960-970. S., & Bowden, E. M. (in press). Gender differences Beyer, S., differences in self-perceptions: Convergent evidence from Personalityand andSocial SocialPsychol Psycholfrom three measures of accuracy and bias. Personality ogy Bulletin. Bulletin, Block, J. J. (1978). The assessment andand psychiatric research. Palo Palo The Q-sort Q-sortmethod methodininpersonality personality assessment psychiatric research, Alto, CA: Consulting Psychologists Press. (Original work published 1961). desirabihty in self self& Ostendorf, F. (1989). Descriptive consistency and social desirability Borkenau, P., & EuropeanJournal JournalofofPersonality, Personality, 3, 31-45. 3, 31-45. and peer reports. European self and others: Self-enhancement Self-enhancement biases in social judg judgBrown, J. D. (1986). Evaluations of self Cognition,4,4,353-376. 353-376. Social Cognition, ments. Social R. Snyder & & D. F. Forsyth Brown, J. D. (1991). Accuracy and bias in self-knowledge. In C. R perspective (pp.(pp. 158(Eds.), Handbook of of social social and and clinical clinicalpsychology: psychology:The Thehealth health perspective 158178). New York: Pergamon Press. Brown, J. D. (1993). (1993). Coping with stress: The beneficial beneficial role of positive illusions. In A. P. Cognitive coping, coping,families, families,and anddisability disability(pp. (pp.123-137). 123-137). BaltiTurnbull et al. (Eds.). Cognitive Balti more: Paul H. Brookes.

CHAPTER 2S 25

SELF-PERCEPTION ACCURACY AND BIAS IN SELF.PERCEmON

675 675

Brunswik, E. E. (1956). (1956). Perception Perception and andthe therepresentative representativedesign design psychological experiments Brunswik, ofof psychological experiments (2nd ed.). Berkeley: University of California Press. Self-consciousnessand and social anxiety. Francisco: Freeman. Buss, A. H. (1980). Self-consciousness social anxiety. SanSan Francisco: Freeman. Campbell, J. J. D. D. (1986). (1986). Similarity Similarity and and uniqueness: uniqueness: The The effects effects of of attribute attribute type, type, relevance, relevance, Campbell, Journal of and individual differences in self-esteem and depression. Journal ofPersonality Personalityand and Social Social Psychology, Psychology,50, 50,281-294. 281-294. Campbell, J. D. (1990). Self-esteem and clarity of the self-concept. Journal Journal of ofPersonality Personality and Social Social Psychology, Psychology,59, 59,538-549. 538-549. Campbell, J. D., & J. D., & Lavallee, L. F. (1993). Who am I? The role of self-concept confusion in understanding the the behavior behavior of of people people with with low low self-esteem. self-esteem. In In R R. F. F. Baumeister Baumeister (Ed.), understanding Self-esteem:The Thepuzzle puzzleofoflow lowself-regard self-regard (pp. 3-20). New York: Plenum Press. Self-esteem: (pp. 3-20). New York: Plenum Press. Campbell, J. J. D., Chew, Chew, B., B., & & Scratchley, Scratchley, L. L. S. S. (1991). (1991). Cognitive Cognitive and and emotional emotional reactions reactions to to Campbell, daily events: events: The The effects effects of of self-esteem self-esteem and and self-complexity. self-complexity. Journal Journal of of Personality, Personality, daily 59, 473-505. 473-505. 59, Cheek, J. M. (1982). Aggregation, moderator variables, and the validity of personality tests: A peer peer rating rating study. study. Journal Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, 1254-1269. 43, 43, 1254-1269. A & Block, Block, J. J. (1994). (1994). Do Do positive positive illusions illusions foster foster mental mental health? health? An An examination examination Colvin, C. C. R, R., & Colvin, of the the Taylor Taylor and and Brown Brown formulation. formulation. Psychological PsychologicalBulletin, Bulletin,116, 116,3-20. 3-20. of C. R, R., Block, Block, J., J., & & Funder, Funder,D. D. C. C. (1995). (1995).Overly Overlypositive positiveself-evaluations self-evaluationsand and personal personalColvin, C. Colvin, ity: Negative Negative implications implications for for mental mental health. health. Journal Journalof ofPersonality Personality andSocial Social Psycholity: and Psychol ogy, 68, 68, 1152-1162. ogy, 1152-1162. Donahue, E. E. M., M., Robins, Robins, R R. W., W., Roberts, Roberts, B. B. W., W., & & John, John, O. O. P. P. (1993). (1993). The The divided divided self: Donahue, self: Concurrent and and longitudinal longitudinal effects effects of of psychological psychological adjustment adjustment and and social social roles roles on on Concurrent -concept differentiation. Journal self-concept Journalof ofPersonality Personalityand andSocial Social Psychology, Psychology, 64, 64,834-846. 834-846. self Einhorn, H. J., & & Hogarth, Hogarth, R R. M. M. (1981). (1981). Behavioral Behavioral decision decision theory: theory: Processes Processes of of judgment judgment Einhorn, H. Psychology,32, 32,53-88. 53-88. and choice. choice. Annual Annual Review Reviewof of Psychology, and Epstein, S. S. (1990). (1990). Cognitive Cognitive experiential experiential self-theory. self-theory. In In L. L. A. A. Pervin Pervin (Ed.), (Ed.), Handbook Handbook of of Epstein, personality:Theory Theoryand andresearch research (pp. 165-192). New York: Guilford Press. (pp. 165-192). New York: Guilford Press. personality: Nebraska Festinger, L. L. (1954). (1954). Motivation Motivation leading leading to to social social behavior. behavior. In In M. M. R R. Jones Jones (Ed.), (Ed.), Nebraska Festinger, Symposiumon onMotivation Motivation(Vol. (Vol.2,2,pp. pp.191-218). 191-218).Lincoln: Lincoln:University University Nebraska Press. Symposium ofof Nebraska Press. Festinger, L. (1957). A theory Stanford theory of of cognitive cognitive dissonance. dissonance.Palo PaloAlto, Alto,CA: CA: StanfordUniversity University Press. Fincham, F. D., & Jaspers, J. J. M. M. (1980). (1980). Attribution Attribution of of responsibility: responsibihty:From Fromman man the thescientist scientist Advances in in experimental experimentalsocial social psychology to man man the the lawyer. lawyer. In In L. L. Berkowitz Berkowitz (Ed.), (Ed.), Advances to psychology (Vol. 13, pp. 81-137). New York: 13, pp. York: Academic Press. Annual Review Fiske, S. (1993). Social cognition and social perception. Annual Review of of Psychology, Psychology,44, 44, 155-194. Fiske, S. MA: S. T., T., & & Taylor, Taylor,S. S. E. E. (1991). (1991).Social Socialcognition cognition(2nd (2nded.). ed.).Reading, Reading, MA:Addison-Wesley. Addison-Wesley. Freud, S. S. (1953). (1953). On On narcissism: narcissism: An An introduction. introduction. In In J. J. Strachey Strachey (Ed.), (Ed.), The standardedition edition Freud, The standard of the the complete completepsychological psychologicalworks works(Vol. (Vol.14, 14,pp. pp.69-102). 69-102).London: London:Hogarth HogarthPress, Press, of (Original work work published published 1914). (Original Funder, D. D. C. C. (1987). (1987). Errors Errors and and mistakes: mistakes: Evaluating Evaluating the the accuracy accuracy of of social social judgment. judgment. Funder, PsychologicalBulletin, Bulletin,101, 101,75-90. 75-90. Psychological Funder, D. D. C. C. (1990). (1990). Process Processversus versuscontent contentininthe thestudy studyof ofjudgmental judgmentalaccuracy. accuracy.Psychologi PsychologiFunder, cal Inquiry, Inquiry,1,1,207-209. 207-209. cal Funder, D. D. C. C. (1993). (1993). Judgments Judgments as as data data for for personality personality and and developmental developmental psychology: psychology: Funder, Error versus versus accuracy. accuracy. In In D. C. C Funder, R R. D. Parke, C. Tomlinson-Keasey, Tomlinson-Keasey, & & K. K. Error

676 676

ROBINS AND AND IOHN JOHN

Personality andand development (pp.(pp. 121Widaman (Eds.), Studying Studyinglives livesthrough throughtime: time: Personality development 121146). 146). Washington, DC: DC: American Psychological Association. of Gabriel, M. T., Critelli, J. J. W., & & Ee, J. S. (1994). Narcissistic illusions in self-evaluations of intelligence and attractiveness. Journal Journal of ofPersonality, Personality,62,62,143-155. 143-155. Gibson, J. J. (1979). The ecological perception. Boston: Houghton-Mifflin. ecologicalapproach approachtotovisual visual perception, Boston: Houghton-Mifliin. Greenwald, A. G. (1980). The totalitarian ego: ego: Fabrication and revision of personal history. American Psychologist, Psychologist,35,35, 603-618. American 603-618. Greenwald, A. G., & R. S. Wyer & T. K. Srull (Eds.), & Pratkanis, A. R R. (1984). The self. In R Handbook of of social social cognition cognition(Vol. (Vol.3,3,pp. pp.129-178). 129-178).Hillsdale. Hillsdale,NJ: NJ:Erlbaum. Erlbaum. Greenwald, A. G., & & Ronis, D. L. (1978). Twenty years of cognitive dissonance: Case study of the evolution of a theory. Psychological PsychologicalReview, Review,85, 85,53-57. 53-57. Hastie, R, Bar R., & Rasinski, K. A. (1988). The concept of accuracy in social judgment. In D. BarTal & & A. W. Kruglanski (Eds.), The social psychology psychology of of knowledge knowledge (pp. (pp.193-208). 193-208). New York: York: Cambridge Cambridge University University Press. New Press. Heider, F. F. (1958). The The psychology psychology of of interpersonal interpersonalrelations. relations.New New York: Wiley. York: Wiley. Heider, Higgins, E. E. T., T., & & Bargh, Bargh, J. J. A. A. (1987). Social Social cognition cognition and and social social perception. perception. Annual Annual Review Review Higgins, of Psychology, Psychology,338, 369-425. of 8, 369-425. Hofstee, W. W. K. K. B. B. (1994). Who Who should should own own the the definition definition of personality? European EuropeanJournal Journal Hofstee, of Personality, Personality,8,8,149-162. 149-162. of NebraskaSympo SympoHogan, R R. (1983). A A socioanalytic socioanalytic theory theory of of personality. personality. In In M. M. Page Page (Ed.), (Ed.), Nebraska Hogan, sium on Motivation Motivation (pp. (pp.55-89). 55-89).Lincoln: Lincoln:University Universityof ofNebraska NebraskaPress. Press. sium Hogan, R, R., Jones, Jones, W. W. H., H., & & Cheek, J. J. M. M. (1985). Socioanalytic Socioanalytic theory: theory: An An alternative alternative to to Hogan, and social life armadillo psychology. In B. R R. Schlenker (Ed.), The The self and life(pp. (pp.175-198). 175-198). New York: York: McGraw-Hill. McGraw-Hill. New James, W. (1890). The principles principlesof ofpsychology. psychology,Cambridge, Cambridge, MA: MA:Harvard Harvard University. University. variance ininself-perception: self-perception:Self SelfJohn, O. P., & & Robins, Robins, R R. W. W. (1992, August). August). Method variance John, O. enhancement,selself-deception, social desirability. Paper presented inSymposium the Symposium enhancement, f-deception, andand social desirability. Paper presented in the "New Directions in Social Desirability Research" (M. Botwin, Chair), annual convenconven tion of the American Psychological Association, Washington, DC. John, O. P., & & Robins, R R. W. W. (1993). Determinants of interjudge agreement on personality traits: traits: The The Big BigFive Five domains, domains,observability, observability, evaluativeness, evaluativeness, and and the the unique unique perspective perspective of the self. Journal 521-551. Journalof ofPersonality, Personality,61,61, 521-551. John, O. P., & & Robins, R R. W. (1994). Accuracy and bias in self-perception: Individual differences Journal of differences in self-enhancement and the role of narcissism. Journal ofPersonality Personalityand and Social Psychology, Psychology,66, 66,206-219. 206-219. Johnson, J. A. (1981). The "self-disclosure" and "self-presentation" views of item response Psychology, dynamics and personality scale validity. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 40, 40, 761-769. Judd, C. C. M., & & Park, Park, B. B. (1993). Definition and and assessment assessment of of accuracy in social social stereotypes. Judd, (1993). Definition PsychologicalReview, Review,100, 100,109-128. 109-128. Psychological Jussim, L. (1993). Accuracy in interpersonal expectations: A reflection-construction reflection-construction analysis of current current and and classic classic research. research. Journal Journalof ofPersonality, Personality,61,61,637-668. 637-668. of Kelley, H. H. (1967). Attribution theory in social psychology. In D. Levine (Ed.), Nebraska Nebraska Symposium on Motivation Motivation (Vol. (Vol. 15, 15,pp. pp.192-240). 192-240). Lincoln: Lincoln:University University of of Nebraska Nebraska Symposium Press. Press. Kelly, G. G. A. A. (1955). The psychology psychology of of personal constructs. constructs.New NewYork: York:Norton. Norton. Kelly, Kenny, D. D. A. A. (1991). A A general general model model of of consensus consensus and and accuracy accuracy in in interpersonal interpersonal perception. perception. Kenny, PsychologicalReview, Review,98,98,155-163. 155-163. Psychological

CHAPTER 2S 25 CHAFfER

IN SELF.PERCEmON SELF-PERCEPTION ACCURACY AND BIAS IN

677 677

Kenny, D. A. (1994). Interpersonal A Asocial relations analysis. NewNew York: Guil Interpersonalperception: perception: social relations analysis. York: Guilford Press. ford Kohut, H. (1971). The International The analysis analysisof ofself. self.New NewYork: York: InternationalUniversities UniversitiesPress. Press. Krueger, J., & & Zeiger, J. S. (1993). Social categorization and the truly false consensus effect.

65, 65, 670-680. Journal ofPersonality Personalityand andSocial SocialPsychology, Psychologyy 670-680. Journal of

Kruglanski, A. W. (1989). The psychology of being 'right': *right': The Theproblem problem of of accuracy accuracy in insocial social perception and cognition. Psychological PsychologicalBulletin, Bulletin,106, 106,395-409. 395-409. Lecky, P. (1945). Self-consistency: ofof personality. New York: Island Press. Self-consistency:AAtheory theory personality. New York: Island Press. Lerner, J. S., Spranca, M. D., & & Tetlock, P. E. (1994, June). Accountability Accountability and andallocating allocating resources: another. Poster presented at the annual confer resources:Doing Doingone onething thingbut butsaying saying another. Poster presented at the annual conference of the American Psychological Society, Washington, DC. Lewinsohn, P. P. M., M., Mischel, Mischel, W., W., Chaplin, Chaplin, W., & & Barton, Barton, R. R. (1980) Social Social competence competence and Lewinsohn, depression: The role of illusory self-perceptions. Journal of Abnormal Psychology, Journal of Psychology,

89, 203-212. 89,

Lockard, J. S., & Paulhus, D. L. Englewood L. (1988). Self-deception: Self-deception:An Anadaptive adaptivemechanism? mechanism? Englewood NJ: Prentice-Hall. Cliffs, NJ: Markus, H., H., & & Wurt, Wurf, E. E. (1987). (1987).The Thedynamic dynamicself-concept: self-concept:AAsocial socialpsychological psychologicalperspective. perspective. Markus,

Review of of Psychology, Psychology,38, 38,299-337. 299-337. Annual Review

McCrae, R. R. (1982). Consensual validation of personality traits: Evidence from self-reports and ratings. ratings. Journal Journal of ofPersonality Personalityand andSocial SocialPsychology, Psychology, 293-303. 43,43, 293-303. and McCrae, R. R. R. R. & & Costa, Costa, P. P. T. T. (1990). Personality Personalityininadulthood. adulthood.New NewYork: York: GuilfordPress. Press. McCrae, Guilford Millon, T. T. (1990). (1990).The The disorders disordersof ofpersonality. personality.In In L. L. A. A.Pervin Pervin(Ed.), (Ed.),Handbook Handbookof ofpersonality: personality: Millon, Theory and andresearch research(pp. (pp.339-370). 339-370).New NewYork: York:Guilford GuilfordPress. Press. Theory c., & Mills, C, & Hogan, R. (1978). A role theoretical interpretation of personality scale item responses. Journal Journalof ofPersonality, Personality,46,46,778-785. 11^-1%S. responses. Morf, C. C. c., C, & & Rhodewalt, Rhodewalt, F. F. (1993). Narcissism Narcissism and and self-evaluation self-evaluation maintenance: maintenance: ExploraMort, Explora Personalityand andSocial SocialPsychology PsychologyBulletin, Bulletin, 668-676. tions in in object object relations relations Personality 19, 19, 668-676. tions Nisbett, R. R. E., & Ross, Ross, L. L. (1980). Human Human inference: inference:Strategies Strategies and shortcomings of social Nisbett, E., & and shortcomings of social judgment.Englewood EnglewoodCliffs, Cliffs,NJ: NJ:Prentice-Hall. Prentice-Hall. judgment. of Paulhus, D. D. L. L. (1984). Two-component Two-component models models of of socially socially desirable desirable responding. responding. Journal Journal of Paulhus,

Personalityand andSocial SocialPsychology, Psychology,46,46, 598-609. Personality 598-609.

Paulhus, D. D. L. L. (1990). Measurement Measurement and and control control of of response response bias. J. P. Robinson, P. P. R. R. Paulhus, bias. In In J. P. Robinson, Shaver, & Measures of ofpersonality personalityand andsocial-psychological social-psychological & L. L. Wrightsman (Eds.), Measures attitudes (pp. (pp.17-59). 17-59).San SanDiego, Diego,CA: CA:Academic AcademicPress. Press. attitudes Robins, R. R. W. W. (1996). Unpublished Unpublished data, data. University University of of California California at at Berkeley, Berkeley, Department Department Robins, of Psychology. Psychology. of Robins, R. R. W., W., & & John, John, O. O. P. P. (l996a) (1996a) A conceptual conceptualframework frameworkfor forthe thestudy studyof ofaccuracy accuracyand and bias in in self-perception: self-perception:Issues Issues definition measurement. Unpublished manuscript. Unpublished manuscript, bias ofof definition andand measurement. University of of California California at at Berkeley. Berkeley. University Robins, R. R. W., W., & & John, John, O. O. P. P. (l996b). (1996b). Toward Toward aa broader broader agenda agenda for for research research on on self self and Robins, and PsychologicalInquiry, Inquiry,7, 7,279-287. 279-287. other perception. perception. Psychological other Robins, R. R. W., W., & & John, John, O. O. P. P. (1997). Self-perception, Self-perception, visual visual perspective, narcissism: Is Robins, perspective, and and narcissism: Is seeing believing? believing? Psychological PsychologicalScience Science(in(inpress). press). seeing Rogers, C. C. R. R. (1959). (1959). A Atheory theoryof oftherapy, therapy,personality, personality,and and interpersonal interpersonalrelations, relations,developed developed Rogers, in the client-centeredframework. framework. In S. Koch (Ed.), Psychology: Psychology:AAstudy studyof ofaascience science (Vol. 3, pp. pp. 185-256). New New York: McGraw-Hill. (Vol. York: McGraw-Hill. Rosenberg, M. M. (1979). Conceiving Conceivingthe theself. selfNew NewYork: York:Basic BasicBooks. Books. Rosenberg,

678 678

ROBINS AND AND JOHN JOHN ROBINS

Ross, M. (1989). Relation of implicit theories to the construction of personal histories. Psychological PsychologicalReview, Review,96,96,341-357. 341-357. Schlenker, B. R The self-concept, social identity, and inter R. (1980). Impression Impressionmanagement: management: The self-concept, social identity, and interpersonal relations. relations, Monterey, Monterey,CA: CA:Brooks/Cole. Brooks/Cole. R. (1985). Introduction: Introduction: Foundations Foundations of of the the self self in in social social life. life. In In B. B. R R. Schlenker Schlenker Schlenker, B. R self and social life. life. New York: McGraw-Hill. (Ed.), The self R., Dlugolecki, D. W., W., & & Doherty, K. (1994). The impact of self-presentations self-presentations Schlenker, B. R, on self-appraisals self-appraisals and behavior: The power of public commitment. Personality Personality and and Psychology Bulletin, Bulletin,20, 20,20-33. 20-33. Social Psychology R., & Leary, Leary, M. M.R R.(1982). (1982).Audiences' Audiences'reactions reactionsof ofself-enhancing, self-enhancing,self-denigrat self-denigratSchlenker, B. R, and accurate accurate self-presentations. self-presentations. Journal Journal of of Experimental ExperimentalSocial Social Pscyhology, ing, and Pscyhology, 18, 18, 89-104. illusion of mental health. American Shedler, J., Mayman, M., & Manis, M. (1993). The illusion Psychologist,48, 48,11117-1131. Psychologist, 117-1131. function of of initial self-perceptions. self-perceptions. Shrauger, J. S. (1975). Responses to evaluation as a function PsychologicalBulletin, Bulletin,82,82,581-596. 581-596. Psychological Snyder, M. appearances/privaterealities: realities: psychology of self-monitoring. Snyder, M. (1987). Public appearances/private TheThe psychology of self-monitoring. New York: Freeman. Steele, C. C. (1993). Self-image Self-image resilience resilience and and dissonance: dissonance: The The role role of of affirmational affirmational resources. resources. Steele, Journal of of Personality Personalityand andSocial SocialResearch, Research, 885-896. Journal 64, 64, 885-896. D. (1973). (1973).Videotape Videotape and andthe theattribution attribution process: process:Reversing Reversingactors' actors'and andobservers' observers' Storms, M. D. points of of view. view. Journal Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, 165-175. points 27, 27, 165-175. for accuracy in person perception: A matter of pragmatics. Swann, W. B., Jr. (1984). Quest for PsychologicalReview, Review,91,91,457-477. 457-477. Psychological Swann, W. W. B., Jr. (1990). To be adored or to be known: The interplay of self-enhancement self-enhancement self-verification. In R R. M. Sorrentino & E. T. Higgins (Eds.), Motivation Motivation and and self-verification. cognition cognition (Vol. 2. pp. 33-66). New York: Guilford Guilford Press. (1989). Agreeable fancy or disagreeable B. W., & KruU, Krull, D. S. (1989). Swann, W. B., Jr., Pelham, B. truth? Reconciling self-enhancement self-enhancement and self-verification. self-verification. Journal of of Personality Personalityand and Social Social Psychology, Psychology,57, 57,782-791. 782-791. Taylor, S. E. (1989). Positive andand thethe healthy mind. New Positive illusions: illusions:Creative Creativeself-deception self-deception healthy mind. New York: Basic Books. Taylor, S. E., & Brown, J. (1988). Illusion and well-being: A social psychological perspective PsychologicalBulletin, Bulletin,103, 103, 193-210. 193-210. on mental health. Psychological Taylor, S. E., & Brown, J. (1994). Positive illusions and well-being revisited: Separating fact fromfiction. fiction. Psychological 21-27. PsychologicalBulletin, Bulletin,116, 116, 21-27. self-evaluation maintenance model of social behavior. In L. Tesser, A. (1988). Toward a self-evaluation Berkowitz (Ed.), Advances experimental social socialpsychology psychology(Vol. (Vol.21, 21,pppp181-227). 181-227). Advances in experimental Orlando, FL: FL: Academic Press. Toward aa social social Tetlock, P. E. (1992). The impact of accountability on judgment and choice: Toward contingency model. model. In In M. P. P. Zanna Zanna (Ed.), (Ed.), Advances Advancesin inexperimental experimental social psychology contingency social psychology (Vol. 25, 25, pp. 331-376). San Diego, CA: Academic Press. Accountability and judgment processes in a personality Tetlock, P. E., & Kim, J. I. (1987). Accountability prediction task. Journal ofPersonality Personalityand andSocial Social Psychology, 700-709. prediction Journal of Psychology, 52,52, 700-709. & Levi, A. (1982). Attribution bias: On the inconclusiveness of the cognition cognitionTetlock, P. E., & motivation debate. debate. Journal Journalof ofExperimental ExperimentalSocial Social Psychology, 68-88. motivation Psychology, 18, 18, 68-88. Y. (1979). (1979). Uncertainty-reducing Uncertainty-reducing properties properties of of achievement achievement tasks. tasks.Journal Journalof ofPersonal PersonalTrope, Y. Social Psychology, Psychology,37, 37,1505-1518. 1505-1518. ity and Social

CHAPTER CHAPTER 25 25

IN SELF.PERCEmON SELF-PERCEPTION ACCURACY AND BIAS IN

679 679

L. E., & Marwell, G. (1976). Self-esteem: Self-esteem: Its Its conceptualization conceptualizationand andmeasurement. measurement, Wells, L. CA: Sage. Beverly Hills, CA: Hand(1990). Psychoanalytic approaches to personality. personaUty. In L. L. A. A. Pervin (Ed.), Hand Westen, D. (1990). book of of personality: personality: Theory Theoryand and research research(pp. (pp.21-65). 21-65).New NewYork: York:Guilford GuilfordPress. Press. T. A. A. (1991). Similarity and self-esteem in downward comparison. In J. Suls & Wills, T. T. and research (pp.(pp. 51-78). T. A. A. Wills (Eds.), Social Social comparison: comparison:Contemporary Contemporarytheory theory and research 51-78). Hillsdale, NJ: Erlbaum.

This Page Intentionally Left Blank

CHAPTER CHAPTER 26

SELF-CONCEPT, IDENTITY, SELF.CONCEPT, AND SELF.EsTEEM SELF-ESTEEM AND THE SELF SELF LOST LOST AND AND FOUND FOUND THE

Roy ROY F. F . BAUMEISTER BAUMEISTER CASE UNIVERSITY CASE WESTERN WESTERN RESERVE RESERVE UNIVERSITY

I. I. DEFINITIONS DEFINITIONS The most obvious and common things are sometimes the most difficult difficult to to define. This certainly applies to to the the self. People use use the the word "self," especially with its many prefixes and suffixes, dozens of times each day, and yet it is difficult to pause and say what is is meant by self. For purposes of the present discussion, it it seems best to to define define certain certain terms. terms. The term self corresponds to to its its everyday usage in in colloquial speech. As As such, it encompasses the the direct feeling feeling each person has of privileged access to to his his or or her her own thoughts and feelings and sensations. It It begins with the the awareness of one's own body and and and is is augmented by by the the sense of being able to to make choices and initiate action. It also also encompasses encompasses the the more morecomplex complexand and abstract abstractconstructions constructionsthat that embellish the self. In everyday speech, the familiar expressions "to find find yourself' yourself or "to know yourself" do do not not ordinarily mean to to locate one's body and be be able to recognize it; rather, those expressions refer to some difficult difficult act regarding complex, abstract knowledge. The term self-concept self-conceptrefers refers totothe thetotality totality of ofinferences inferences that that aaperson personhas has or herself. These refer centrally to to one's personality traits made about himself or may also involve an an understanding of one's social roles and schemas, but but they may and relationships. IIANDBOOK HANDBOOK OF OF PERSONALITY PERSONALITY PSYCHOLOGY PSYCHOLOGY

681 681

CorYKiGHT C 0 1991 1997 BY BY ACADEMIC PlEss. PRESS. ComroHT ALL liGHTS OF IEPlODUcnON IN ANY FOD RESERVED.

ALL RIGHTS OF REPRODUOION IN ANY FORM RESERVED.

682 682

Roy F. BAUMEISTER ROYF.BAUMEISTER

Lastly, the term identity identity refers to the definitions definitions that are created created for and superimposed superimposed on the self. These definitions refer refer to concepts about who the the person is and what the person person is like. Identity Identity can be analyzed as consisting of an interpersonal interpersonal aspect (a set of roles and relationships), a potentiality aspect (a concept of who the person might become), and a values aspect (a set of values and priorities) priorities) (Baumeister, 1986). 1986). Identity differs differs from from self-concept self-concept in that it is socially defined. defined. That That is, is, the self-concept self-concept is wholly contained in the person's own mind, whereas whereas identity is often often created created by the larger larger society, although individuals typically have have some opportunity opportunity to refine refine or negotiate the identities that society gives them. Identity must be regarded regarded as an aggregate definition definition of self. That is, is, a self can be defined defined in many ways-with ways—with many traits and many different different roles. Identity is thus the product product of the many definitions definitions of self self that exist. Likewise, the self selfinferences that the person person concept is a loose combination of the many ideas and inferences him- or herself. The term self-esteem self-esteem refers refers to the evaluative dimension dimension has about himof the self-concept. self-concept. of II. FINDING TIlE THE SELF: SELR CREATION OF IDENTITY D. self-esteem) are constructed. It will This section will examine how identity (and self-esteem) begin with the role of culture, by considering how different different historical periods have treated the single self. Then it will turn to how knowledge about the self self is actually actually treated Then tum formed and organized. Next, it will take a closer look at the formation and mainteformed formation mainte nance of self-esteem. Lastly, it will consider consider briefly briefly some of the ways people try to control the information information about about the self, usually in order to fit it into established established or control preconceived patterns. preconceived Self in Historical Historical Perspective A. The Self Although the concept of self self tends to seem so natural and normal as to be inevitable, Although it is in fact fact far from from universal. Ideas about about the nature of the self self have varied widely across cultural and historical boundaries. Although more detailed accounts of these 1986, 1987; Markus & Kitayama, variations are available elsewhere (see Baumeister, 1986,1987; Kitayama, 1991; Triandis, 1989), it is necessary necessary to summarize summarize briefly briefly some of the factors factors that distinguish the sense of self self in the modem modern Western world. world. appropriate starting point. Modern Western Western The sense of uniqueness is an appropriate society treats each person as a special, unique individual and encourages encourages people to other) that way. way. This is a fairly recent development, development, regard themselves (and each other) people were not identical, but however. To be sure, the ancients recognized that people placed relatively little importance on these differences. What mattered were were they placed form and function. Similarly, in the Middle Ages people were were the similarities in form regarded as having functions functions according according to their place in society. The person's identity regarded was intimately bound up with his or her social rank, family family ties, and occupation, occupation, and people were not supposed to want to change any of these. More important, and supposed important,

CHAPTER . SELF.CONCEPT. CHAPTER 26 26 IDEN11TY IDENTITY, SELF-CONCEPT,AND ANDSELF.EsTEEM SELF-ESTEEM

683 683

the models of human potential were essentially the same for the vast Christian majority: salvation in heaven, as a result of living a moral and pious life, was the goal for everyone. At most, people differed according to how well they served their functions, that is, whether they fulfilled fulfilled their duties and obeyed the rules (see Macintyre, Maclntyre, 1981). It was not until the early modem modern period (roughly 1500-18(0) 1500-1800) that people became fascinated with all the small characteristics that made one person different different from another (e.g., Weintraub, 1978). One sign of this new interest was a great biographical and autobiographical writing, including a greatly increased increase in biographical emphasis on accurate description of factual details about the person's life (Altick, 1965; Weintraub, 1978). This new interest was associated with a great social change philosophitoward an emphasis on individuality. Politically, economically, socially, philosophi self-contained cally, and in other ways, society came to treat each person as a unique, self-contained unit. People began to think of themselves as capable of changing roles, to search for their own unique traits and destiny, to campaign for individual rights and social equality, and to do other things that reflected this new sense of the individual. self expanded greatly during this same period. The The notion of an inner self self is a metaphor for one's private access to, or privileged possession of, inner self one's thoughts and feelings and intentions. There is some evidence that the notion of an inner self self began to gain in importance in our culture around the 16th century of (e.g.. Trilling, 1971). It may have been derived in some way from the Christian (e.g., self may have been simply a way of Gnostic notion of soul. At first, the inner self of thinking about about hypocrisy, hypocrisy, deception, deception, and and insincerity: insincerity: people people were were not not always always really really thinking the way way they they appeared the surface surface to to be. the appeared on on the be. self expanded. People began to Over time, however, the notion of an inner self believe that their inner selves contained their true personality traits, the basis for creativity, and even their most strongly held values and opinions. Poets and other artists attracted attracted great public attention because they were believed to lead rich inner (Altick, 1965). For example, instead of just enjoying Shakespeare's Shakespeare's plays, lives (Altick, people began to wonder seriously about what kind of person Shakespeare must have been (Altick, 1965). By the 19th century, the culture had come to regard each person as containing a vast inner realm of hidden material. The culture had also fulfillment depended substantially on come to think that the path to personal fulfillment discovering this inner self self and developing it (Baumeister, 1986; 1986; Sennett, 1974; 1974; Weintraub, 1978). shift toward an expanded concept Two developments are associated with this shift of self-knowledge had come to seem increasingly difficult. difficult. of the inner self. First, self-knowledge Confidence self-knowledge eroded over the subsequent centuries through a series Confidence in self-knowledge of developments that included the Puritan discovery of the pervasiveness of self selfof deception, the Victorian fascination Freudfascination with involuntary disclosure, and later the Freud ian exploration of the unconscious. The second development is the evolution of the idea of identity crisis. Erik identity crisis in the 1940s and it Erikson (1968) claimed that he coined the term identity immediately gained a wide usage. The instant popularity of the term suggests that

684

ROY F. F. BAUMEISTER BAUMEISTER Roy

there was already a broadly familiar phenomenon that it defined, so one must assume that identity crises were occurring before then. But the weight of the evidence suggests that the modern modem form of identity crisis only became generally apparent late in the 19th century. In the Middle Ages, for example, people had no term or concept for an identity crisis, nor did they apparently undergo the sort of experience to which it now refers. the religious conver converThe historical predecessors of identity crises were, first, first, the sion experience, experience, and, and, second, second, the the belief beUef that that one one did did not not belong belong in in the the role role or or life life sion oneself. These These experiences experiences are are important important because because they they imply imply in which which one one found found oneself. in a separation of the person from his or her beliefs and and actions (which are are changed in religious conversion) and from his or her place in society. Thus, the self is conceived of as something that exists prior to and apart from its beliefs and roles (see MacIntyre, Maclntyre, 1981, 1981, for elaboration). Undoubtedly, this new view of the self was encouraged by the rise in social, occupational, and geographical mobility, which showed how the same person could switch to a very different place in society. The notion of identity crisis is based on the belief that a person person is is conceptually conceptually separate from his or her place in society, and on the belief that a person can find find inside him him or or herself herself the the basis basis for for choosing choosing an an identity. identity. These These beliefs, beliefs, as as we we have have inside modern beliefs, and they would have been inconceivable to earlier eras. seen, are modem Westem phenomenon, reflecting the new ideas Identity crisis is thus a modem, Western self. The The modern modem notions notions of of self self are are more more complex complex and and sophisticated sophisticated about the the self. about than other notions, but they are not necessarily more accurate, and they also carry variety of of burdens burdens and and potential potential problems. problems. The The identity identity crisis crisis is is one one symptom symptom of of aa variety the modem burden of selfhood. Most cultures cultures in in the the history history of of the the world world have have not not required required people people to to create create Most definitions of themselves that could serve as the basis for their adult lives, and so most cultures have not produced large numbers of identity crises. Indeed, even our own culture did not make such a requirement until recently. One's occupation was arranged by one's parents, who also took aa leading leading role role in inarranging arrangingone's one'smarriage, marriage, and so the individual did not have very many choices to make in determining his or her own adult identity. Now, however, parents at most provide advice, and the young person can can (and must) choose from a bewildering variety of possible career opportunities and potential marriage partners (e.g., Kett, 1977). Thus, although the modem modern self self is associated with a great deal of freedom and opportunity and flexibility, flexibility, it is also a problem. The 20th century has seen an expanding fascination with the problems of the self, as reflected in everything from the popular culture (e.g., books and movies in which people try to understand or find find themselves) to scientific research (e.g., social science research on the self). What was wasonce onceaasimple, simple,straightforward straightforwardmatter matterhas hasnow nowbecome becomesomething somethingdifficult, difficult, uncertain, and problematic. Complicating the matter further is the fact that modern modem society has turned tumed to selfhood to solve some of its more pressing problems. In particular, the main meaning in in life life for for modern modern Western Westem individuals individuals is is that that of of finding finding difficulty in infinding difficulty finding meaning basis for for values. values. In In response response to to this this value value gap, gap, modern modem society society has has placed placed an an aafirm firm basis

CHAPTER CHAPTER 26 26

IDEN11TY IDENTITY,, SELF.CONCEPT, SELF-CONCEPT, AND AND SELF.EsTEEM SELF-ESTEEM

685 685

increasing emphasis on the self as a fundamental fundamental good and source of value. The rise of the self as a basic value is one of the most important cultural changes of of the 20th century and has transformed transformed the way people understand understand their identities and modem modern attitudes about work, family, love, and death (Baumeister, 1991b). Partly as a result of this burden of selfhood, self (along with its selfhood, the modem modem self concepts and definitions) carries a substantial ambivalence. On the one hand, the modem self self is a unique and special entity, a fascinating fascinating puzzle, and a presumed source of much that is valuable in life. On the other hand, the self self can be a difficult difficult and uncertain puzzle, a focus of fears and anxieties, and a burdensome source of of demands and obligations. Whereas our ancestors had one word, "selfishness," to refer refer to orientation toward oneself oneself (and to condemn condemn it as morally undesirable), modem modem society has added a wide assortment of words that express many nuances of that orientation over a wide evaluative spectrum: individualism, individuality, egoism and egotism, self-confidence, so forth. self-confidence, self-centeredness, self-expression, and so The modem psychology of self self must be understood against the context of this ambiv ambivalence.

B. B.

Self-Knowledge Self-Knowledge

self may be understood as a unity, the self-concept self-concept is not really a Although the self single, unified concept so much as a loosely connected set of ideas, inferences, and self-knowledge has abandoned illusions. Indeed, one recent thrust of research on self-knowledge self-concept per se and focused focused instead on self-schemas, selfschemas, that the notion of a single self-concept is, specific concepts of various features of the self self (Markus, 1977). The The self-concept self-concept is, can be regarded as a collection of these particular schemas about its traits. Developmental psychologists have addressed the question of how children accumulate knowledge about themselves (see Damon & & Hart, 1982, for for review). It appears that children begin early in life to form form notions of of who they are, but the formation formation of self-concept self-concept is dependent on a wide variety of of factors, including the sources of feedback feedback available from peers and the child's own developing ability abiUty to understand itself itself in abstract or complex ways. Children are born bom with identities (insofar (insofar as they belong to particular families) families) but must develop self-concepts. The child's sense of of self self apparently begins with the discovery that some events are contingent on its own acts-perhaps acts—perhaps most vividly, seeing one's image in the mirror and noticing that it moves whenever oneself oneself moves (Lewis & Brooks-Gunn, 1979). The first contents of the self-concept 1979). self-concept appear to be as a member of a particular family, as being young and small, and as being male or female. Among older preschool children the self self is understood understood especially in terms of capabilities and competencies (e.g., Keller, Ford, & Meacham, 1978). 1978). This This emphasis continues through the early school years, although it becomes increasingly is, the child evaluates competence relative to others' levels and comparative; that is, relative to measured standards of competence (e.g., Erikson, 1950, 1968; 1968; Ruble, 1983). In adolescence, people increasingly come to think of themselves in terms of of

686

Rov F. BAUMEISTER ROYF.BAUMEISTER

abstract and interpersonal traits, issues of choice, and values (e.g., Erikson, 1968; Montemayor & & Eisen, 1977). The development of self-knowledge is hardly hardly aasmooth smoothor oreasy easyprocess. process.Indeed, Indeed, Rosenberg (1979) found that over half the preadolescents in his sample a study by Rosenberg felt that their parents or other adults knew them better than they knew themselves. The notion that a person has privileged access to his or her inner states is one that children come to accept only after a long period of development. Throughout life, an important source of self-knowledge is the social feedback forpeople receive from each other. An extreme statement of this view was put for selfward by the symbolic interactionist George Herbert Mead, who proposed that self knowledge is essentially distilled from feedback received from other people. But an extensive literature review by Shrauger and Schoeneman (1979) found the symbolic interactionist view of the self-concept to be inadequate. People's self-concepts do not correspond very closely to how they are regarded by others. others. On the other hand, there is considerable evidence that people are less than fully accurate in their 1980; Taylor, 1989; perceptions of how others evaluate them (e.g., Greenwald, 1980; & Brown, Brown, 1988). 1988).Self-deceptions Self-deceptionsenable enablepeople peopleto tosustain sustainviews viewsofofthemselves themselves Taylor & that are more favorable than their actual feedback would tend to warrant. Thus, the feedback people receive from others may be subject to substantial distortions. Shrauger and Schoeneman (1979) found that people's self-concepts were highly correlated with how they believed others regarded them, even though the self-concepts were not correlated with how others others actually actually regarded them. them. Thus, Thus, it may be most accurate to suggest that the self-concept is the product of some negotiation between one's interpersonal feedback and one's preferred beliefs beUefs about oneself. Swann and Hill Hill (1982) (1982) demonstrated that that the the effects effects of of social social feedback feedback depend depend on how people are able to respond to it. When When subjects subjectswere weregiven givenbogus boguspersonality personality feedback with no chance to respond to it, they tended to accept it and shift their private views of themselves to agree more with it. In contrast, when other subjects were given bogus feedback plus a chance to dispute it (which they did), they were not swayed by it. The implication is that the passive self may be shaped directly by external feedback, but the active self tends to take an aggressive and critical respondresponse to feedback so as to measure it against what it already knows. By respond ing actively to feedback, people can maintain their views of themselves despite contrary evidence. By adolescence and certainly throughout adulthood, people have a collection of concepts about themselves. They have fairly detailed (although not necessarily coherent) concepts about who and what they are. They may also have fairly elaboelabo rate concepts about who and what they might become. Identity begins with an awareness of one's body, but in an adult human being identity is generally oriented toward goals (Baumeister, 1986). These goals include becoming a certain kind of of person and not becoming another kind of person. These concepts of what oneself might become have been termed possible selves by Markus and Nurius (1986, 1987). These researchers began with the selves

CHAPTER CHAPTER 26 26

IDEN11TY, ESTEEM IDENTITY, SELF.CONCEPT, SELF-CONCEPT, AND AND SELF. SELF-ESTEEM

687 687

older notion of an ideal self (e.g., Rogers & 1954). People have some & Dymond, 1954). concept of themselves, not as they currently are, but rather as they would ideally be, and their efforts efforts are often often directed at becoming more like their ideal like to be, self. Markus and Nurius added, however, the important notion that people have becoming—such as overweight, unloved, or a failure. concepts of what they fear becoming-such Often Often these undesirable possible selves are quite detailed and elaborate concepts, Uke these feared efforts to avoid becoming like and people exert efforts feared selves. Indeed, in self becomes a major major motivating factor many respects the undesired self factor (Ogilvie, 1987). These conceptions of possible selves are not only important in shaping behav behaveffect on emotions. Higgins (1987; also ioral motivations, they also have a strong effect & Strauman, 1987) 1987) has hasproposed proposed that that emotional emotional patterns patterns are are strongly strongly Higgins, Klein, & influenced by two sets of concepts about the self. In his view, people compare how how self The The ideal self, they perceive themselves with an ideal self and with an ought self. how one one would like to be, again, is how be, and the ought self self is the way one feels some obligation or duty to be. Agitated emotions, such as anxiety and guilt, arise when oneself and one's ought self. Dejected Dejected emotions, one sees a discrepancy between oneself as sadness, depressed mood, and disappointment, arise when one sees a discrep discrepsuch as oneself and one's ideal self. ancy between oneself efforts to become more like their ideal selves have been studied by People's efforts Wicklund and Gollwitzer (1982). These researchers examined how how people try to claim desired identities. When people's sense of being able to reach their goals is threatened, they try harder to achieve some success or even some symbolic gesture that will help them feel that they are reaching these goals. symboUc gestures for claiming identity involves An important category of symbolic oneself as having that identity. Thus, it is not enough convincing others to see oneself simply to believe privately that one one is reaching one's ideal self; identity demands 1982b; Schlenker, Schlenker, 1980, 1980,1985,1986). Wicklund public recognition (e.g., Baumeister, 1982b; 1985, 1986). Wicklund and Gollwitzer found repeatedly that people will respond to an identity threat by trying to prove themselves to somebody. Interestingly, it seemed not to matter who this other person was. was. If you want to be an artist, and somehow that creative ability is questioned, you will tend to try to persuade someone of your artistic talent and accomplishments. Although one one might think that other artists or art critics would be the most relevant audiences, empirically people seem to settle for whomever they can find 1986). The important factor find and persuade (Gollwitzer, 1986). factor is thus the social validation of one's identity, almost regardless of who provides it.

c. C. The The Basics Basics of of Self·Esteem Self-Esteem

may well well begin begin early early in in life. life. Kagan Kagan (1981) (1981) reviews reviews The importance of self-esteem may evidence that the words "good" and "bad" "bad" are among the most common ones spoken to young children, across many cultures. Moreover, by the second year of of of goodness and badness, including life, children compare their behavior to standards of so the the habit habit of of self-evaluation self-evaluation is is acquired acquired standards of competent performance, and so early and is pervasive.

688

Roy F. BAUMEISTER ROYF.BAUMEISTER

Self-esteem is a central trait, in the sense that it is one of the most important elements of the self-concept and and that it affects many other elements (Greenwald, 1979) reviewed the research Bellezza, & & Banaji, 1988). Indeed, when Wylie (1974, (1974,1979) literature on self-concept, she found that the vast majority of it focused on self selfesteem. To be be sure, sure, there there are are other other aspects aspects to to how how people think think of of themselves themselves than than the evaluative dimension, but the evaluative aspect is extremely important and and has has captured the primary interest of most researchers. Despite the appeal of the symbolic interactionist arguments, most researchers have come to believe that there are two main sources of self-esteem. One is indeed the person receives from others (however distorted it may may the evaluative feedback the be). and success (or failure). There is be). The other is direct experiences of efficacy and some evidence that these two two aspects of self-esteem self-esteem are not strongly related to Franks & Marolla, MaroUa, 1976). 1976). People People may maybe beinsecure insecure about abouthow how others others each other (e.g., Franks abiUty to do things right, or the reverse. regard them but quite confident about their ability self-esteem typically features a questionThe study of individual differences in self-esteem question naire measure, which is used to sort people according to how favorably they regard 1974,1979). themselves. There is aa wide variety of such measures (see Wylie, 1974, 1979). to regard regard self-esteem. self-esteem. One One There have been several controversies about how to day to to day day controversy concerns the stability of self-esteem: does itfluctuate t1.uctuate from day or remain stable? Most studies have found it to be quite stable across time. Some researchers are currently undertaking to revive the notion of self-esteem states that but this this work work has has to to overcome overcome the the stable stable tendencies tendencies of of self selffluctuate rapidly, rapidly, but t1.uctuate esteem (e.g., Heatherton & & Polivy, 1991; 1991; Kernis, Kernis, 1993). 1993). Baumeister Baumeister (1991c) (1991c) found found .904 across 22 weeks on on aa self-esteem scale, indicating a test-retest reliability of .904 has found that self-esteem self-esteem can indeed change, very high stability. Harter (1993) has particularly at major transition points in life (e.g., graduation), but still it tends to remain quite stable most of the time. Another issue concerns whether self-esteem should be considered as a sinsin gle quantity as as opposed to aa collection of independent (and (and uncorrelated) self selfevaluations. That is, is it appropriate to think of people as having high or low self selfesteem overall, or are are people more likely to think well of themselves in some spheres (such as as socially) while thinking poorly of themselves in others (such as has evolved toward a compromise. A athletically)? Current thinking on this issue has Flemhierarchical facet model has been proposed by Fleming and Courtney (1984; Flem & Watts, 1980). 1980). This model says that there is indeed a global level of self selfing & esteem that ret1.ects reflects a person's overall evaluation of self, but there are are also specific levels of self-esteem with respect to various specific spheres. Researchers should therefore consider carefully whether they want to study global self-esteem or some particular dimension of self-esteem. self-knowledge Self-esteem levels are centrally linked to differences in self-knowledge J. D. D. Campbell, 1990; J. D. Campbell & & Lavallee, 1993). (Baumgardner, 1990; 1990; J. 1990; J. People with high self-esteem appear to have clear, consistent, and stable views about themselves. People with low self-esteem, in contrast, do not seem to know

CHAPTER 26 CHAPTER26

IOEN11TY , SELF.CONCEPT, AND SELF·EsTEEM IDENTITY,SELF-CONCEPT, ANDSELFESTEEM

689 689

themselves well. Their self-concepts appear to be confused, contradictory, unstable, uncertain, and full of gaps. Self-esteem BaumeisSelf-esteem has been shown to influence a variety of behaviors (see Baumeis ter, 1993, self-esteem appear appear to be more 1993, for such a compilation). People with low self-esteem susceptible to influence than people with high self-esteem self-esteem (Brockner, 1983; 1983; Cohen, 1959; 1959; Janis, 1954; Janis & & Field, 1959). 1959). Initial failure is apparently quite aversive to people with high self-esteem, and they become determined to avoid repeating the experience. They respond either by trying harder on the second trial (e.g., Shrauger & Sorman, 1977; 1977; Silverman, 1964) 1964) or by avoiding the task if they can (e.g., Baumeister & & Tice, 1985). 1985). People with low low self-esteem self-esteem respond to initial failure by remaining at about the same level of effort, effort, although some researchers have interpreted effort in comparison with the responses interpreted this as a withdrawal of effort self-esteem (cf. Maracek & Mettee, 1972). of people with high self-esteem is important to realize, however, that although self-esteem predicts responses responses It is to such feedback, there is very little evidence of any general tendency for people low self-esteem self-esteem to perform perform worse than people with high self-esteem; indeed, with low perform about the same found the two groups to perform most laboratory studies have found 1978; J. D. D. Campbell & & Fairey, 1985). 1985). Thus, despite the (e.g., Brockner & Hulton, 1978; self-esteem is not associated with large occasional benefits of confidence, high self-esteem low self selfadvantages in success or achievement. It is unwarranted to assume that low esteem is an accurate assessment of one's general lack of competence (see (see also A. Campbell, 1981). may be differences differences in success in life that do not depend on However, there may competence. Managing oneself oneself effectively, effectively, such as by choosing appropriate tasks and making commitments that one can keep, can be just as important as overall competence in bringing success, and recent evidence indicates that people with high effective at setting appropriate goals and living up self-esteem are generally more effective self-esteem & Tice, 1993), partly due to their to their commitments (Baumeister, Heatherton, & superior (J. D. D. Campbell, 1990). On the other hand, ego ego threats superior self-knowledge self-knowledge (J. produce extreme and irrational reactions from from people with high self-esteem, and they become prone to respond in nonoptimal and even self-defeating self-defeating ways to such threats. Their effective effective self-management self-management seems to vanish in such circumstances (Baumeister et aI., al., 1993). People with low self-esteem self-esteem are more likely than highs to say that their behavior varies across situations (e.g., Goldberg, 1981; 1981; Paulhus & Martin, 1988; 1988; see 1993). This J. D. Campbell & & Lavallee, 1993). This may may reflect reflect the the greater greater plasticity plasticity or or flexibility flexibility of people with low self-esteem and itit may may also also reflect reflect their general general self-esteem (Brockner, 1983), and lack of firm self-knowledge D. Campbell, 1990). 1990). Self-esteem also influences the self-knowledge (J. (J. D. way people respond to public situations. Many behavioral differences differences are found 1972; Wilson & only in in public situations (e.g., Archibald & & Cohen, 1971; 1971; Shrauger, 1972; Benner, 1971), which suggests that self-esteem is associated with important differ differences in self-presentational self-presentational patterns (e.g., Arkin, 1981; 1981; Baumeister, 1982a). For a long time, the evidence suggested that people with low self-esteem self-esteem were more likely to hold negative stereotypes and prejudices than people with high self-

690 690

ROY F. F.BAUMEISTER BAUMEISTER Roy

esteem, but recent work has revealed a serious flaw in this evidence. People with self-esteem are more critical of of others and of of themselves than are people with low self-esteem discrepancy between evaluation of self self and evaluation of of high self-esteem. The discrepancy about the same for for people of of all levels of of self-esteem self-esteem (Crocker (Crocker & outgroups is about Schwartz, 1985). Self-Concept D. Motivations Regarding the Self.Concept self-concept typically includes reference reference to certain We have already seen that the self-concept goals, including trying to reach one's ideals and avoid certain undesirable possible selves. Beyond these broad goals, however, it appears that people spend a consider considerable amount amount of of effort effort on their self-concepts. However, researchers have been sharply of these efforts. efforts. divided as to the nature and goal of The two main motivations regarding the self-concept self-concept are consistency consistency and favorability. It is clear that once a person has formed formed a certain certain concept or evaluation evaluation of of self, and if it is acceptable, people seek to maintain it, and they resist external influences designed to change it. But is this because the strive for for consistency or because they desire favorability? favorability? The evidence is divided. The view that people desire to hold positive positive views of of themselves has a long majority of of people strive to sustain favorable favorable views history. It seems clear that the majority of themselves (Taylor, 1989; Taylor & Brown, 1988). They blame their failures failures of for successes, they convince themselves that others like externally but take credit for them, and they exaggerate their degree of of control and efficacy efficacy (e.g., Greenwald, Greenwald, persuade themselves themselves that their abilities are unique but that their their opin opin1980). They persuade validated by most other other people (J. D. Campbell, 1986). ions are validated On the other other hand, there is evidence that people people seek seek consistency. They They strive to confirm confirm their views of of themselves, they dispute feedback feedback that is discrepant discrepant from from their self-concepts, and they will even avoid someone whose opinion of of them them differs differs from self-concept (Swann, 1987). from their self-concept The main test test case, of of course, is what what happens happens when people have formed formed unfavorable unfavorable opinions of of themselves. In this case, if they desire consistency, consistency, they should prefer prefer to receive unfavorable unfavorable evaluations that confirm confirm their low self-esteem. self-esteem. On On the other other hand, if they mainly desire favorable favorable views of themselves, then they should prefer prefer favorable favorable evaluations. An extensive review of of the early research research literature literature on this topic found found some support support for for both predictions, but the preponderance preponderance of of studies supported supported the favor favorability hypothesis (S. (S. C. Jones, 1973). A A later and more careful careful review by Shrauger Shrauger (1975) found, found, however, that one could could explain the discrepant findings findings by sorting the work work according to how the response response to evaluations evaluations had been measured. Shrauger Shrauger found found that when the measures were primarily primarily affective, affective, people showed showed a clear preference preference for for favorable favorable feedback, feedback, regardless of of their level of of self-esteem. However, However, when the measures were primarily cognitive, people seemed more inclined to believe and accept feedback feedback that was consistent with their views of of themselves. In short, self-esteem are more likely likely to enjoy enjoy receiving favorable favorable feedback feedback people with low self-esteem

CHAPTER 26 CHAPTER26

IDENTITY,SELF-CONCEPT,AND SELFESTEEM IDENlTfY, SELF.CONCEPT, AND SELF.EsTEEM

691 691

but they are more likely to believe unfavorable unfavorable feedback. Subsequent work has borne out Shrauger's conclusion (McFarlin & & Blascovich, 1981; 1981; Swann, Griffin, Griffin, & Gaines, 1987). Predmore, & favorably of oneself oneself can be placed in a broader context, The desire to think favorably especially if one accepts the view that the motivation is linked to emotional patterns. Becker (1973) proposed that that self-esteem is is aa vital means means of protecting protecting oneself oneself against against anxiety. This hypothesis has has recently been revived and elaborated by Greenberg, Pyszczynski, and Solomon (1986). According to this view, human beings are unique among animal species in that they know that they are going to die. of die. This fear of death gives rise to an existential terror that is the main cause of anxiety in life. People therefore therefore desperately need some defenses against this threat of death, or Self-esteem furnishes a vital protec protecelse they would be in a constant state of terror. Self-esteem tion against anxiety, because it casts the individual as a valued participant in a after the individual dies. By regarding them themcultural drama that will continue even after selves as important, worthy individuals, people can begin to overcome the feelings insignificance and ephemerality that are caused by the realization that they will of insignificance die (Greenberg et al, aI., 1986). has generated some controversy, mainly The terror management hypothesis has fear of death is is really because there is some question as to whether the existential fear has proposed that the central cause of all human anxiety. An alternative view has anxiety is a natural response to exclusion from social groups (Baumeister, 1990a; & Leary, 1995; 1995; Baumeister Baumeister & & Tice, 1990). 1990). In other words, people feel Baumeister & Baumeister and so so forth, or even anxiety when others reject them, dislike them, avoid them, and nonethewhen there is merely some threat of rejection. The social exclusion view is nonethe selfless quite compatible with Becker's and Greenberg et al.'s al.'s hypothesis that self conesteem is an important defense against anxiety. People with high self-esteem con sider themselves to be competent, virtuous, and attractive by definition, so they are less worried than others that they will be rejected or excluded. As a result, they are less troubled by anxiety. People with high self-esteem expect others to like them and to want to be associated with them, and they confidently pursue these outcomes. i n . LOSING mE THE SELF: SELR PROBLEMS OF SELF·CONCEPT SELF-CONCEPT AND IDENTITY m.

modem interest in self must At the beginning of this chapter, I proposed that the modern be understood in the context of the tension between the great desire to know and self and the concomitant burdens and difficulties difficulties associated with the express the self difficulties that self. This section will examine some of the specific problems and difficulties are associated with the self. A. The The Puzzle Puzzle of of Low Low Self-Esteem Self-Esteem A. Psychology has generally been sympathetic to people with high self-esteem. Indeed, adjustment have often often treated treated self-esteem self-esteem as one one measure measure of of adjustment, adjustment. studies of adjustment

692 692

Roy F. BAUMEISTER ROYF.BAUMEISTER

such that the higher a person's self-esteem score, the better adjusted the person is assumed to be. be. If that were all there were to it, society should perhaps simply encourage everyone to be as conceited as possible! It is not difficult difficult to understand the goals and motives of the people with high self-esteem. They want to succeed, to be loved and admired, and to enjoy their lives and accomplishments. In contrast, the motives of people with low self-esteem self-esteem have been a mystery. Their goals have been relatively uncertain. Indeed, as we saw in the precious section of of this chapter, many psychologists have proposed that people with low self-esteem self-esteem opindesire failure and rejection, because this feedback will confirm their negative opin 1962; Aronson & Mettee, 1968; ions of themselves (e.g., Aronson & Carlsmith, 1962; 1968; Maracek & Mettee, 1972). The accumulated evidence has shown, however, that self-esteem desire success just as much as anyone else (e.g., McFar McFarpeople with low self-esteem lin & Blascovich, 1981). The solution to the puzzle of low self-esteem self-esteem may be somewhat complex. To perceive it, it isfirst classified first necessary to realize that most research subjects who are classified self-esteem are not low in an absolute sense. Baumeister, Tice, and Hutton as low in self-esteem self-esteem scores for many different different scales in (1989) reviewed the distributions of self-esteem different studies, and they found that invariably there were only a few people many different whose scores were genuinely low. Many people score at the high end of the scale, and most of the rest score in the middle. Thus, in an absolute sense, most people moderate in self-esteem. Low scores are only should be labeled as either high or moderate low; in an absolute sense, they are moderate. relatively low; Next,· Next, it is vital to recognize that people with low self-esteem do not seem to are, as already mentioned (Baumgardner, have a firm sense of who and what they are, 1990; J. D. Campbell, 1990; J. D. Campbell & Lavallee, 1993). This pervasive "self "selfconcept confusion" (J. D. Campbell & Lavallee, 1993) 1993) may underlie a broad range of their thoughts, feelings, and actions. Furthermore, one must take into account the evidence that these "low" "low" self selfesteem individuals have mixed reactions to success, as already noted. They would like to succeed, but they do not expect to do so (McFarlin & Blascovich, 1981; 1981; Shrauger, 1975; Swann Swann et et at, al., 1987). 1987). Thus, Thus, they they are are somewhat somewhat insecure insecure about about achiev achieving the outcomes they desire, in contrast to the people with high self-esteem who are confident confident that they can achieve whatever they try. There is also some evidence suggesting that different different levels of aspiration are associated with different different levels of self-esteem. In a study by Baumeister and Tice (1985), (198S), people received initial success or failure and then had an opportunity to persist at the task or to devote their time to something else. Not surprisingly, people with high self-esteem self-esteem showed great interest in the task when they initially succeeded, but they tended to avoid the task if they had initially failed. People with low self selfesteem, however, showed the opposite pattern. One way of interpreting these results is to suggest that people with high self selfesteem are interested in achieving exceptional successes, whereas people with low self-esteem mainly want to avoid failures. When the person with low self-esteem self-esteem receives failure feedback, it is discouraging, but the person will tend to work on

26 IDEN'I1TY, IDENTTTY,SELF.CONCEPT, SELF-CONCEPT,AND ANDSELF.EsTEEM SELF-ESTEEM CHAPTER 26

693 693

this problem to try to remedy the deficit. That way, the likelihood likelihood of future future failure failure is reduced. In contrast, initial success signifies to someone with low self-esteem self-esteem that he or she is already performing performing at an adequate, passable level, so there is no need need to work on it. Indeed, it may be prudent work prudent to avoid the task, so that there is no danger danger of ruining one's initial success by failing at it on a second try. Thus, people people with low self-esteem self-esteem may be oriented toward remedying remedying their their deficits and and overcom overcoming their faults, so as to reach reach an adequate adequate or passable level. A forward by Baumeister Baumeister A more general formulation formulation has recently been put forward al. (1989). This formulation formulation distinguishes between between the motive to protect protect one's one's et al. self-concept and the motive to enhance enhance it (i.e., make it more favorable). Self Selfself-concept enhancement requires seeking out opportunities to achieve, succeed, and stand out; enhancement self-protection involves avoiding chances of failure, rejection, or humiliation. Tice self-protection showed that even when similar behaviors are involved, they appear to be (1990) showed different motivations: People with low self-esteem self-esteem are mainly concerned concerned driven by different self-protection, whereas people with high self-esteem self-esteem are mainly concerned concerned with self-protection, self-enhancement (see also Arkin, 1981; 1981; Baumeister Baumeister et al., a l , 1989; Wolfe, Len Lenwith self-enhancement nox, & Cutler, 1986). A final final and important important piece of the puzzle has been suggested by Steele (1988) A self-esteem is a resource, and Spencer, Josephs, and Steele (1993). In their view, self-esteem self-esteem simply do not have as much of it as people with and people with low self-esteem stressed or threatened, they have less to draw high self-esteem. Accordingly, when stressed approach life in general-in general—in a more protective protective respond-and indeed approach upon, and they respond—and defensive way. and defensive self-esteem. People Thus, a solution is slowly emerging to the puzzle of low self-esteem. self-esteem measures typically lack a clear and definite stock of who score low on self-esteem of particular they suffer suffer from from a lack of helpful, positive views self-knowledge, and in particular about themselves. They desire and enjoy enjoy success, but their actions are influenced influenced about regular or frequent frequent by their doubts that they will be able to achieve success on a regular basis. They They focus on protecting themselves against failures failures and rejections, such as by presenting presenting themselves themselves in a cautious or modest fashion. They orient themselves toward toward finding finding out their shortcomings and inadequacies so as to remedy these. Unlike people with high self-esteem, who focus on their strengths and try to cultivate cultivate these so as to become outstanding, people with low self-esteem self-esteem strive to be adequate adequate by focusing focusing on their weaknesses and overcoming them. These are of course only broad, general general patterns. People with high self-esteem self-esteem dislike failure failure and will work work hard to avoid avoid humiliating experiences. But when they have a choice, their their primary primary goal is to achieve great success rather rather than to avoid failure. People with low self-esteem self-esteem will tend toward the opposite choice.

B. Self.Defeating Self-Defeating Behavior One of the greatest paradoxes of human behavior, and certainly in the study of of self-defeating behavior. It is clear that people sometimes do things that the self, is self-defeating Self-defeating behavior spans cause themselves pain, harm, loss, and even death. Self-defeating

694 694

ROYF.BAUMEISTER Roy F. BAUMEISTER

a wide spectrum, ranging from getting into debt or making poor investments to suicide. Self-preservation Self-preservation and the pursuit of self-interest self-interest are widely widely regarded as the self-defeating behaviors seem quintessen quintessenessence of rational behavior, and so these self-defeating tially irrational. Considerable information is available about the ways people harm themselves and sabotage their projects (see Baumeister & Scher, 1988). To make sense of this selfinformation, it is first necessary to distinguish several possible categories of self defeating behaviors. The purest form would be cases in which people engage in suffering that it will bring them. In these some action for the sake of the loss or suffering self-destructive behavior, the person both foresees and desires cases of deliberately self-destructive the harm to self. At the other extreme, people may harm themselves almost by accident; in these cases, people neither desire nor foresee the harm to self, but their efforts toward positive goals are undermined by counterproductive means or strategies. Lastly, an intermediate category includes cases in which the harm to self is perhaps foreseen but is not desired. In this category, typically, people are engaging in trade-offs, trade-offs, so so they they engage engage in in the the behavior behavior for for the the sake sake of of positive positive benefits benefits and in and accept the the risks risks and and costs costs that that accompany accompany it. accept it. There is very little evidence that normal adult human beings engage in the self-destructive behavior (i.e., deliberate self-destruction). However, first kind of self-destructive there is considerable evidence of counterproductive strategies. People use various 1981) or ingratiation strategies (E. E. Jones & Wort Wortbargaining strategies (Pruitt, 1981) man, 1973) 1973) that backfire and produce undesired results. They persist in failing endeavors far past the point at which they should rationally cut their losses and 1976; Teger, 1980). They start over elsewhere (Rubin & Brockner, 1975; 1975; Staw, 1976; respond to pressure situations by focusing on themselves, which tends to impair skilled performance (Baumeister, 1984). Even learned helplessness can be consid considered a maladaptive withdrawal of effort (Seligman, 1975; 1975; also Roth & Kubal, 1975). Lastly, there is considerable evidence of self-defeating self-defeating behavior that occurs as a result of trade-offs between competing, incompatible goals. People handicap their performances so as to give themselves an excuse for failure (e.g., E. E. Jones & Berglas, 1978). They use drugs and alcohol, which can cause considerable damage to one's health and relationships, in order to avoid realizing unpleasant things about themselves (Hull, 1981). They disregard and disobey medical advice from their physicians, even skipping important appointments and failing to take their medicines (e.g., Dunbar & Stunkard, 1979; 1979; Sackett & Snow, 1979). They sacrifice tangible rewards to avoid temporary embarrassment or to take revenge against others 1968; Brown & Garland, 1971). Despite their desire to have friends, shy (Brown, 1968; people avoid others and avoid social interactions, so they remain lonely and isolated (e.g., Cheek & Busch, 1981; 1981; W. W. H. Jones, Freemon, & Goswick, 1981; 1981; Maroldo, 1982; Schlenker & Leary, 1982). One somewhat surprising conclusion that has emerged emerged regarding self-defeating self-defeating behavior is that it often appears to be motivated motivated by states of high self-awareness self-awareness (Baumeister & Scher, 1988). When attention is focused focused on the self, especially in an aversive fashion, people are more likely to do things that will produce harmful

CHAFfER 26 CHAPTER26

IDENTITY, SELF·CONCEPT, AND SELF.EsTEEM IDENTITY,SELF-CONCEPT,AND SELFESTEEM

695 695

outcomes. In many cases, this appears to occur because people are eager to escape from from an awareness of the self's shortcomings. The willingness to accept costs and risks for the sake of immediate relief relief is increased when the current state involves an aversive awareness of self. It seems unwarranted, then, to infer infer that people have self-destructive self-destructive urges or motivations. Self-defeating Self-defeating behavior occurs among normal people either as an unwanted by-product of some desirable outcome or as an unwanted result of poor judgment and ill-advised strategies.

C. C. Identity Identity Crisis

1968). identity crisis crisis apparently originated in the 1940s (Erikson, 1950, The term identity 1950, 1968). was that an identity crisis is a normal, possibly universal stage of Erikson's view was of human development, typically associated with adolescence. The universality of of identity crises has become an increasingly untenable hypothesis. For one thing, identity crises appear to be historically and culturally relative to some extent; as far as we can tell, people did not commonly have identity crises before the 19th although itit isis plausible plausible that that exceptional exceptional individuals individuals century (e.g., Baumeister, 1986), although 1958). Furthermore, there is substantial evioccasionally had them (see Erikson, 1958). evi dence that many people today do not report anything resembling an identity crisis. crises could could be be unconscious, unconscious, so sopeople people might mightnot notbe beaware aware In Erikson's view, identity crises difficult to evaluate and perhaps impossible to of having them. Such a hypothesis is difficult found people who reveal no sign of disprove, but researchers have consistently found of 1966,1967). identity crises even in response to in-depth interviews (e.g., Marcia, 1966, 1967). The discovery that many people show no signs of identity crises has led 1967; Orlof researchers to formulate a taxonomy of identity identity statuses statuses(Marcia, (Marcia,1966, 1966,1967; Orlofsky, Marcia, & & Lesser, 1973). 1973). People are sorted according to whether they have formed a secure had identity crises or not, not, and according to whether they have formed identity with roles and commitments or not. not. The four four statuses deserve some explana explana1981; Bourne, 1978). tion and comment (for reviews, see Bernard, 1981; People who have had identity crises and resolved them successfully successfully are classi classified as identity identity achieved. achieved. These people are typically mature, well adjusted, and flexible. flexible. Indeed, Indeed, they they score score highest highest on onmost most adjustment adjustment measures, measures,suggesting suggesting that that identity crises are generally beneficial beneficial in the long run. People who have had identity crises but have not resolved them are classified moratoriums. as as moratoriums. Typically, Typically,they theyare arecurrently currentlyinvolved involved in inthe thecrisis. crisis.The Theterm term"crisis" **crisis" carries a connotation of disaster and suffering, which is only partly accurate. Morato suffering, is Moratorium subjects often often appear to be open to new experiences, actively exploring a wide range of ideas and lifestyles, and often often exhilarated by some of what they find, find, although of course there are periods of confusion, confusion, depression, and dismay. Some researchers have recently come to prefer the term term "exploration" rather than "crisis," "crisis," simply to avoid the melodramatic implications of the latter term. A third category, foreclosures, foreclosures, involves people who have commitments to adult identity patterns without having gone through through a substantial period of crisis

696 696

Roy F. BAUMEISTER ROYF.BAUMEISTER

or exploration. Most children have foreclosed identities, for for they tend to accept the beliefs, values, and goals that their parents instill in them. The adolescent identity crisis often often begins with a rejection of these parental lessons. Hence, people who do not experience the adolescent adolescent crisis typically retain their allegiance allegiance to what their parents taught them. Foreclosures tend to seem mature earlier than their often conform to a pattern of of stable, continuous progress peers, and their lives often inflexible and they do not adapt toward long-term goals. However, they tend to be inflexible stressful circumstances. There is some evidence suggesting that well to changing or stressful & the foreclosure pattern is maladaptive for males but not for females (Marcia & Scheidel, 1983), although it would be premature to draw a broad conclusion. The fourth category, identity identity diffusion, refers to people who have not formed the commitments to adult identity but are not engaged in any active search or effort effort to do so. This category is generally regarded as the most maladaptive and even pathological of of the four. At a minimum, these individuals tend to resemble the adolescent*' who postpones the responsibilities, decisions, and commit commit"perpetual adolescent" ments of adult life as long as possible. The nature and processes of of identity crisis have remained shrouded in mystery. metaThe vagueness of the concept, combined with its multiple usages (including meta difficult to study the process closely. phorical and colloquial ones), has made it very difficult One review of the available evidence concluded that there are actually two major & Tice, 1985; also Baumeister, 1986). types of identity crisis (Baumeister, Shapiro, & The first type of of identity crisis can be called an identity identity deficit. deficit This is the state created when the person's identity is inadequate to make the choices facing it. It especially when the is commonly associated with adolescence and midlife, arising especially of thinking and acting that have person questions and then rejects the patterns of guided the person over the preceding years. In the adolescent, it is often often associated independently (e.g., with breaking away from parents and learning to think and act independently Bios, 1962). At midlife, it may often be prompted by the sense that one's life is BIos, passing by and so one must reassess where best to devote one's time and efforts efforts (see Levinson, 1978). The identity deficit is often accompanied by radical shifts in feelings and behaviors. The other type of identity crisis can be called an identity identity conflict. It typically herself in terms of of multiple commitments, arises when the person has defined him or herself conflicting demands on the person. Examples of of this type of of identity and these make conflicting conflicts between family ties and religious beliefs, and conflicts between crisis include conflicts occupational advancement and personal or home life. Unlike the wide mood swings of the identity deficit, the identity conflict is often characterized by a pervasive, of of being trapped, guilty, or traitorous. These crises also do not oppressive sense of show the exploratory openness to experience that characterizes the deficit crises. The person person suffering suffering from from an an identity identity conflict conflict does does not not want want new information or or The new information alternatives, for he or she already has too many commitments. Instead, there may alternatives, for he or she already has too many commitments. Instead, there may be a tendency for the identity conflict conflict to breed a passive attitude, as the person postpones making making any any irrevocable irrevocable decision decision and and hopes hopes for for aa solution solution to to emerge. emerge. postpones

CHAPTER 26 26 CHAPTER

IDENTITY, SELF.CONCEPT, SELF-CONCEPT, AND SELF.EsTEEM SELF.ESTEEM IOENlTIY,

697

Thus identity crises are not universal but rather are associated with particular individuals, circumstances, and cultural or historical patterns. There are two broad types of identity crises, and and research should distinguish between them. Identity deficits appear to be linked to particular stages in life, associated associated with the desire to reject and replace some definitions of the self, and associated with beneficial definitions beneficial outcomes. Identity conflicts can occur at any age, they arise when the age, the situation conflicts can forces the person to choose between different different definitions definitions of self, and and these crises do not apparently benefit the individual.

D. D. Escape from Self If the self can be a burden or problem, then sometimes people may want to avoid self-awareness. Escapist motivations may may be strongest when the self is linked to aversive emotional states. As Higgins (1987) has has proposed, such states arise when people fall short of their standards, including ideals and moral obligations. Self Selfawareness is centrally concerned with comparing oneself oneself with standards, so when the self painful or unpleasant to focus attention on self falls short, it may be especially painful oneself (e.g., Duval & & Wicklund, 1972; 1972; Wicklund, 1975). 1975). Research has has shown a oneself and motivate variety of circumstances that make self-awareness especially aversive and shift their attention inward. These people to try to avoid anything that would shift & Wicklund, 1972), hearing circumstances include receiving a bad evaluation (Duval & difficult to correct (Steenbarger & that one has personality problems that would be difficult 1979), receiving a rejection and putdown by an attractive member of the Aderman, 1979), & Wicklund, 1976), or performing performing actions that run counter opposite sex (Gibbons & & Musham, 1981). 1981). Most of these studies belief and values (Greenberg & to one's belief confronting the subject with a mirror and from self-awareness by confronting measured escape from assessing the subject's attempts to avoid it, such as by choosing a seat facing away quickly and and leaving leaving the the room. room. from the mirror or by finishing from finishing quickly In everyday life, of course, escape from from aversive self-awareness is not always as easy as walking away from a mirror. When the self is cast in an unfavorable unfavorable may find find themselves locked into undesirable emotional states and light, people may unable to distract themselves from from the unpleasant thoughts about their failures failures and inadequacies. self is generally a source of burdensome demands and Also, if the modern self constraining definitions, people may may find find it exhilarating to escape from from self selfawareness even when nothing bad has happened. States of ecstasy appear to depend powerful centrally on loss of ordinary awareness of self. Religious mystics speak of powerful the ego ego is dissolved (see (see Goleman, 1988). "Peak" or "flow" "flow" experiences in which the often characterized by absorption absorption in some experiences of ordinary individuals are often may involve a suspension of one's normal awareness of self self (e.g., activity, which may Thus,although although escapist escapist motivations motivations may mayarise arisefrom from specific, specific, Csikszentmihalyi, 1982). Thus, unhappy thoughts and feelings connected with the self, they may be attractive in their own own right as appealing experiences.

698 698

Roy F. BAUMEISTER ROYF.BAUMEISTER

How do people go about escaping from from self-awareness? self-awareness? It is not easy to stop difficult to prevent unwanted being aware of of oneself. People generally find it difficult thoughts (Wegner, Schneider, Carter, & White, 1987), and and the self self may be especially difficult to suppress. After After all, one cannot monitor one's success at not thinking difficult about oneself, for for in order to monitor oneself oneself one must attend to oneself. A A cognitive about effort to avoid self-awareness self-awareness may therefore therefore be a paradoxical, impossible task. effort What people appear appear to do instead, therefore, is to deconstruct deconstruct the self. Identity construct, that is, an entity consisting of of meaning and involving connections is a construct, and relationships among many events, stimuli, and contexts. Deconstruction Deconstruction is a matter matter of of breaking those connections and dissolving those relationships, thereby reducing the sense of of self self back to its bare minimum: a mere body. By focusing focusing broadly meaningful narrowly on physical movement movement and sensation, people can avoid broadly meaningful awareness, including awareness of of implications impUcations about about the self self (Baumeister, 1989, Vallacher & Wegner, 1985, 1985,1987). 1990a, 1990b, 1991a; Vallacher 1987). A A variety of escapist behaviors behaviors can be understood understood on the basis of of this process of shifting shifting attention down to minimal levels. Cognitive deconstruction deconstruction creates a of state characterized characterized by a narrow narrow time frame frame (focused (focused on the immediate present), concrete and rigid thinking, a rejection rejection of of meaningful meaningful thought, a focus on means concrete rather than ends, a passive or impulsive style of of behavior, and and techniques rather reduced reduced or suppressed suppressed emotion. The deconstructed deconstructed state may make the person's person's behavior more inconsistent, because it takes meaningful meaningful integration integration to recognize behavior inconsistencies. It may also remove inhibitions, because inhibitions typically require high-level evaluations of of the meanings of possible acts (see Baumeister, 1990a, 1990b, 1991a). 1. L Alcohol Alcohol Use

Hull (1981) proposed that alcohol use is often of escaping from often a means of from self selfawareness. Alcohol use impairs high-level cognitive processes and meaningful meaningful thought, focusing focusing attention instead on sensations and movements. Even small doses have this effect, of having just a drink effect, and so this view helps explain the appeal of or two. of alcohol in escaping Experimental work has established the effectiveness effectiveness of from from unpleasant awareness of of self. Alcohol makes people less likely to refer refer to themselves in speech and reduces the number number of of first-person first-person pronouns they use (Hull, Levenson, Young, & Sher, 1983). People consume more more alcohol after after experi experiencing failure failure than after after success (Hull & Young, 1983). Research Research on stress has failed to find that all all forms forms of of stress increase alcohol alcohol consumption, consumption, but people do failed consumption when the stress reflects unfavorably unfavorably on the self self (Hull, 1981). increase consumption Indeed, one study examined examined the relapse rates for for alcoholics who completed completed a detoxification program. People who experienced aversive life events tended to detoxification experienced life tended they were inclined to reflect reflect on relapse more quickly than others, but only if they When life stress was not accompanied accompanied themselves (Hull, Young, & Jouriles, 1986). When self-awareness, there was presumably no drive to escape it by getting drunk. by high self-awareness,

CHAPTER , SELF·CONCEPT, CHAPTER 26 26 IDEN11TY IDENTITY, SELF-CONCEPT,AND ANDSELF·EsTEEM SELF-ESTEEM

699 699

Alcohol is a good illustration of both the the positive and the the negative aspects of deconstructing the self. As Hull's work has shown, people tend to consume alcohol to forget unpleasant implications about themselves. Undoubtedly, however, alcoholic intoxication is often often an appealing state even in the absence of the need to escape escape from unpleasant emotions. By disconnecting certain aspects of the self and focusing narrowly on the immediate present, people are able to enjoy themselves more. Alcohol does appear to reduce inhibitions and make people more able to act in ways that are inconsistent with some of their abstract beliefs and values (e.g., & Southwick, 1985). The uninhibited behaviors associated with wild parties Steele & are a familiar illustration of these effects of alcohol. These behaviors may involve self-awareness and meaningful thought, for intoxicated intoxicated people people the same escape from self-awareness do things that are inconsistent with the way they normally regard themselves and want to to be be regarded regarded by by others. others. But But in in this this case case the the impetus impetus for for consuming consuming alcohol alcohol want is not not so so much much to to end end unpleasant unpleasant feelings feelings as as the the positive positive attractions attractions of of the the intoxi intoxiis cated state. state. cated Masochism 2. Masochism

Masochism means obtaining sexual pleasure and and arousal in connection with pain, definition to bondage, and/or humiliation, and some theorists have extended the definition nonsexual enjoyment of pain, helplessness, and humiliation as well. Masochism is one of psychology's long-standing puzzles. Most theorists who have written about have regarded regarded it it as as aa variety variety masochism have worked from clinical observations and have of self-destructive self-destructive behavior. Recent research has shown, however, that the majority of of masochists appear to be normal, healthy, well-adjusted individuals who show of no signs of mental illness apart from their deviant sexuality (e.g., Scott, 1983; 1983; T. Weinberg & & Kamel, 1983). We saw earlier that normal people do not apparently self-destructive behavior except in connection with positive, desirable engage in self-destructive goals (Baumeister & Scher, 1988). It is necessary, therefore, to furnish a new theory of of masochism. Based on current evidence, it seems reasonable to conclude that masochism is a set of techniques for removing one's ordinary identity from awareness (see Baumeister, 1988a, 1988b, 1989). Masochism deconstructs the self in mUltiple multiple ways. To appreciate this, one must consider the common features of masochistic activity. These include humiliation and embarrassment, loss of control (especially through bondage), and pain. Humiliation and embarrassment are a direct attack on the dignity of the self. Earlier in this chapter, it was noted that people are generally motivated to maintain EarUer favorable views of themselves; indeed, this is almost an axiom of the psychology of of self. Masochists, however, seek out degrading experiences such as being dressed in embarrassing costumes, being kept on a leash like an animal, having to kiss pracanother person's feet, being displayed naked, and so forth. These masochistic prac tices thus contradict one of the most pervasive functions of the self, and they make it impossible for the person to maintain his or her normal sense of identity.

700 700

Roy F. BAUMEISTER ROYF.BAUMEISTER

A A second pervasive motivation of the self self is to gain and maintain control over the environment. People are relentlessly motivated to maintain maintain control, and where where real control is lacking they cultivate the illusion of of control (e.g., Brehm, 1966; Langer, 1975; Rothbaum, Weisz, & Snyder, 1982; White, 1959). Masochists, however, seek the illusion of having lost all control. They desire to be tied up, blindfolded, blindfolded, gagged, and otherwise restrained. They seek a partner partner who will give them them arbitrary com commands and take over all initiative. Thus, the self self as an active agent ceases to exist masochism. in masochism. The desire for pain is perhaps the most puzzling feature feature of masochism, for it hard to understand understand how pain could become pleasure. Evidence suggests that that is hard masochists do not actually come to enjoy enjoy the pain (e.g., M. S. Weinberg, Williams, & Moser, 1984; also Scott, 1983). Also, the masochistic masochistic desire for pain is not accompa accompanied by any desire for injury, and in fact masochists appear appear to be very concerned concerned safety (e.g., Baumeister, 1988a; Scott, 1983; with safety 1983; M. S. Weinberg et al., aI., 1984). shifting attention attention to the immediate immediate present. It Pain is quite effective, however, at shifting deconstructs the world, preventing meaningful meaningful thought thought and focusing focusing attention attention on immediate sensations (Scarry, 1985). Pain is thus a tool for manipulating awareness, from ordinary ordinary self-awareness. self-awareness. to help bring about the escape from 3. 3.

Binge Eating Eating Binge

Another modern life is binge Another odd behavior pattern pattern that appears to be on the rise in modern eating. Patterns of binge eating range from from the temporary temporary indulgences of dieters who, having broken their diets, feel that all rules are off and so eat large quantities of bulimia nervosa. of fattening fattening foods, to the pathological patterns of bulimia There is some evidence that binge eating is associated associated with escape from from self selfawareness (Heatherton (Heatherton & Baumeister, 1991). First, eating binges are linked to negative views of self self and awareness of the self's deficiencies deficiencies (e.g., (e.g.. Gamer, Garner, Olm Olm1988; Katzman 1984; Katzman & Wolchik, 1984; sted, Polivy, & Garfinkel, 1984; Gross & Rosen, 1988; Schlesier-Stropp, 1984). Second, manipulations threat or aversive manipulations that that involve involve ego threat moods do increase the eating by obese or dieting subjects subjects (i.e., those most prone prone to engage in eating binges), unlike control subjects Baucom & Aiken, 1981; 1981; subjects (e.g., Baucom Frost, Goolkasian, Ely, Ely, & Blanchard, 1982; Ruderman, 1985; Slochower Slochower & Kaplan, 1980). Third, binges do not occur occur if people are kept kept in a state of high self-awareness, self-awareness, whereas the binge is associated with the loss of attention attention to self. In particular, particular, people cease to monitor their eating during a binge (e.g., Polivy, 1976). This fits the view that inhibitions and restraints involve meaningful meaningful awareness of self, and so deconstruction deconstruction removes them. Eating binges may be pleasant pleasant and desirable in themselves, but the available powerful role of unpleasant emotions and aversive aware awareevidence does suggest a powerful Self-awareness and aversive emotions are minimized, thinking becomes ness of self. Self-awareness focuses on immediate sensations (especially the concrete and rigid, and the person focuses food) rather than long-range considerations or goals. Such binges therefore food) therefore appear from an unpleasant unpleasant to be more commonly motivated by the desire to get away from

CHAPTER CHAPTER 26 26

IDENTITY,, SELF.CONCEPT, SELF-CONCEPT, AND SELF.EsTEEM SELF-ESTEEM IoEN11TY

701 701

state than an attraction to the state of eating, although especially among dieters the eating may be intensely enjoyable. enjoyable. 4. Suicide

self is suicide. There There are are multiple multiple patterns patterns and and causes causes A last example of escaping the self of suicide, but escape appears to be the centrally important one (e.g., (e.g., Baechler, Baumeister, 1990b, 1990b, 1991a). 1991a). Indeed, Indeed, some some researchers researchers have have found found escape escape 1975/1979; Baumeister, for suicide combined (e.g., Smith & to be more common than all other motives for Bloom, 1985). The pattern pattern of events preceding a suicide attempt appears to conform to the same process of escaping the self self that was discussed in the preceding examples (see 1990b). Suicide is associated with a sense of falling short of one's Baumeister, 1990b). goals and standards (including the expectations other people have for one), which self as incompetent, blameworthy, undesirable, and so produces an acute sense of self emoforth. This awareness is initially accompanied by strong patterns of negative emo tion, including depression and anxiety. from this aversive state, the person attempts to avoid meaningful meaningful To escape from thought. The mental state of the suicidal individual conforms very closely to the focused narrowly features of the deconstructed state. The person's sense of time is focused (e.g.. Greaves, 1971; on the present (e.g., 1971; Neuringer & Harris, 1974; 1974; Yufit & Benzies, 1976; Perrah & & Wichman, (1973). Thinking is rigid and concrete (e.g., Henken, 1976; stifled amid a general passivity (e.g., Henken, 1976; 1976; Ringel, 1976) 1987). Initiative is stifled or channeled into impulsive acts. Emotion is broadly stifled, so that even positive emotions are suppressed (Williams & Broadbent, 1986). The suicide attempt itself itself may be a result of the person's inability to maintain the escape using less drastic methods. The person's mental state oscillates between periods of numbness (which are felt as boring and empty) and brief, intense doses affect that arise whenever the person happens to resume meaningful meaningful of negative affect thought. As the latter are felt as intolerable, the person is attracted to the presumed oblivion of death. The deconstructed state removes the person's normal inhibitions against taking his or her own life, and so a suicide attempt results (Baumeister, 1990a, 1990b, 1991a). Suicide represents the most negative and maladaptive aspect of of escapist moti motivations. It was suggested earlier that many common forms of self-defeating self-defeating behavior are motivated by a desire to escape from from an aversive state of high self-awareness self-awareness (Baumeister & Scher, 1988). Suicide may often often be an unfortunately unfortunately extreme case of that principle. People attempt to take their own lives as a desperate strategy to bring an end to the emotional misery associated with an awareness of the self's self's failures and shortcomings. IV. IV. SUMMARY AND AND CONCLUSION The self self begins with simple and universal psychological experiences, such as having beginning. a body and being a distinct member of a social unit. From this crude beginning,

702 702

Roy F. BAUMEISTER ROYF.BAUMEISTER

self can be defined and understood in a wide variety of ways, and however, the self different different cultures and historical periods have indeed taken very different different approaches to selfhood. society constructs the self in a complex and elaborate Our modern Western society expresfashion. The great cultural emphasis on cultivating a well-developed, unique, expres sive, and successful successful self self links the self to a variety of powerful powerful motivations. These fulfilling the are both positive and negative. The opportunities for developing and fulfilling self are greater in our modern modem culture than in nearly any other. At the same time, self these patterns create demands, obligations, and threats that make the self especially self in some ways resembles the prize fish in problematic and burdensome. The self Ernest Hemingway's novel The Old Old Man and and the Sea: It is a great treasure and of dangers and difficulties. difficulties. opportunity, and at the same time it is a source of On the positive side, people are very interested in self-knowledge. They desire to learn about themselves, although they have strong preferences regarding what they might find out. People seek to manage and control the information about themthemselves. Typically, people want to confirm their favorable opinions of them self, including possible future selves, selves. People hold multiple conceptions of mUltiple images of of how they ideally would like Uke to be and how they ought to be, detailed (if inaccurate) concepts of of how they really are and how they appear to others, and more. People People with with high high self-esteem self-esteem are are guided guided by by aa desire desire to to stand stand out, out, to to excel, excel, more. and to make strongly favorable impressions on others. People with low self-esteem and to make strongly favorable impressions on others. People with low self-esteem are torn torn between between aa desire desire for for favorable favorable feedback feedback and and aa tendency tendency to to distrust distrust and and are disbelieve it. They appear to be guided by a desire to avoid failure, rejection, and disbelieve it. They appear to be guided by a desire to avoid failure, rejection, and humiliation, such as by remedying weaknesses and avoiding risks. humiliation, such as by remedying weaknesses and avoiding risks. self is associated with a variety of threats and On the negative side, the self problems. When people discover a discrepancy between how they are and how they want or ought to be, they suffer suffer a variety of of unpleasant emotions. Self-esteem Self-esteem may play an important role in defending the individual against anxiety, and so threats to self-esteem self-esteem may trigger acutely aversive emotional states. It does not appear that people are generally motivated to suffer, but they do engage in a wide variety of of self-destructive self-destructive or self-defeating self-defeating behaviors as a result of of poor judgment or conflicting conflicting goals. Identity crises appear to be one symptom of the modern emphasis on requiring each person to create and define his or her own identity. There are at least two of identity crisis. Identity deficits begin when the person rejects the major types of values and behavior patterns that have shaped his or her life up to that point, and longtypically a period of exploration and experimentation follows, usually with long term beneficial results. Identity conflicts arise in conflict situations that require the person to betray some personal commitments or self-definitions. self-definitions. modem burden of selfhood selfhood has fostered a great increase in the variety The modern of means people use to escape from self-awareness. The most common process of appears to involve deconstructing the identity by focusing narrowly on movements and sensations in the immediate situation. Alcohol use, sexual masochism, binge eating, suicide, and other patterns reflect reflect this pattern of of escape.

CHAPTER 26 26 CHAnER

SELF-CONCEPT, AND SELF.ESTEEM IDENTITY, SELF.CONCEPT,

703 703

The modern fascination fascination with self self seems likely to endure, for it is deeply rooted rooted The modern probably going to continue. For the near future future in current social patterns that are probably self is likely to continue to be a great source of of challenge and at least, defining the self source of of threat and difficulty. difficulty. The construction of satisfaction, as well as a great source self is one of of the major life tasks to confront confront the modern individual. self

REFERENCES Lives and and letters: A history history of of literary biography biography in England England and America. America. Altick, R. (1965). Lives New York: Knopf. Archibald, W. P., & Cohen, R. L. (1971). Self-presentation, embarrassment, and facework of self-evaluation, conditions of of self-presentation, and feedback from as a function of Personality and others. Journal Journal of of Personality and Social Social Psychology, Psychology, 20, 287-297. Arkin, R. M. (1981). Self-presentational Self-presentational styles. In J. T. Tedeschi (Ed.), Impression Impression manage management ment theory theory and and social social psychological psychological research (pp. 311-333). New York: Academic Press. Aronson, E., & Carlsmith, J. M. (1962). Performance Performance expectancy as a determinant of of actual performance. Journal Journal of of Abnormal Abnormal and and Social Social Psychology, Psychology, 65, 178-182. of differential differential levels of Aronson, E., & Mettee, D. (1968). Dishonest behavior as a function of Journal of of Personality Personality and and Social Social Psychology, Psychology, 9, 121-127. induced self-esteem. Journal Suicides, New York: Basic Books. (Original work published 1975) Baechler, J. (1979). Suicides. Effect of of depressed mood on eating among obese and Baucom, D. H., & Aiken, P. A. (1981). Effect nonobese dieting persons. Journal of of Personality Personality and and Social Social Psychology, Psychology, 41, 577-585. Baumeister, R. F. (1982a). Self-esteem, self-presentation, and future interaction. A A dilemma of of reputation. Journal Journal of of Personality, 50, 29-45. A self-presentational self-presentational view of of social phenomena. Psychological Psychological Baumeister, R. F. (1982b). A Bulletin, 91, 3-26. Self-consciousness and paradoxical effects effects Baumeister, R. F. ((1984). 1984). Choking under pressure: Self-consciousness of incentives on skillful performance. Journal of Journal of of Personality Personality and Social Psychology, Psychology, 46, 610-620. Baumeister, R. F. (1986). Identity: Cultural change and and the struggle for for self. self New York: Oxford Oxford University Press. Baumeister, R. F. (1987). How the self self became a problem. A A psychological review of historical research. Journal Journal of of Personality Personality and and Social Social Psychology, Psychology, 52, 163-176. 28-59. Baumeister, R. F. (1988a). Masochism as escape from self. Journal Journal of of Sex Research, 25, 25,28-59. Baumeister, R. F. (1988b). Gender differences differences in masochistic masochistic scripts. Journal Jourrml of of Sex Sex Research, Research, 25, 478-499. Masochism and the self. self Hillsdale, NJ:: NJ: Erlbaum. Baumeister, R. F. (1989). Masochism (1990a). Anxiety and deconstruction: On escaping the self. In J. M. M. Olson Olson & Baumeister, R. F. (199Oa). M. P. Zanna (Eds.). Self-inference processes: processes: The Ontario Ontario Symposium Symposium (Vol. 6, pp. 259-291). Hillsdale, NJ: Erlbaum. (1990b). Suicide as escape from self. Psychological Psychological Review, Review, 97, 90-113. Baumeister, R. F. (199Ob). Escaping the sel self: Alcoholism, spirituality, spirituality, masochism, masochism, and and other other f.' Alcoholism, Baumeister, R. F. (1991a). Escaping flights from from the burden burden of of selfhood. selfhood. New York: Basic Books. flights Baumeister, R. F. (1991b). Meanings of of life. life. New York: Guilford Press.

704 704

Roy F. BAUMEISTER ROYF.BAUMEISTER

Baumeister, R R. F. (1991c). On the stability of variability: Retest reliability of metatraits. Personality 17, 17, 633-639. Personalityand andSocial SocialPsychology PsychologyBulletin, Bulletin, 633-639. York: Baumeister, R R. F. (Ed.). (1993). Self-esteem: Self-esteem:The Thepuzzle puzzleofoflow lowself-regard. self-regard.New New York: Plenum Press. Baumeister, R R. F., Heatherton, T. F., & Tice, D. M. (1993). (1993). When ego threats lead to self selfPersonality regulation failure: Negative consequences of high self-esteem. Journal of of Personality and Social Social Psychology, Psychology,64, 64,141-156. 141-156. Baumeister, R R. F., & Leary, M. R. (1995). The need to belong: Desire for for interpersonal attachments as a fundamental 7, 497-529. fundamental human motivation. Psychological PsychologicalBulletin, Bulletin,11117,497-529. Baumeister, R F., & Scher, S. J. (1988). R. F., (1988). Self-defeating Self-defeating behavior patterns among normal individuals: individuals: Review and analysis of common self-destructive tendencies. Psychological Psychological Bulletin, Bulletin,104, 104,3-22. 3-22. of Baumeister, R R. F., & Senders, P. S. (1989). Identity development development and the role structure of 150, 19-37. children's games. Journal Journalof of Genetic GeneticPsychology, Psychology, 150, 19-37. R. F., Shapiro, J. J., & Tice, D. M. (1985). (1985). Two kinds of identity crisis. Journal Journal Baumeister, R of Personality, Personality,53, 53,407-424. 407-424. of R. F., & Tice, D. M. (1985). Self-esteem and responses to success and failure: Baumeister, R Subsequent performance and intrinsic motivation. Journal of of Personality, Personality,53, 53,450-467. 450-467. Baumeister, R R. F., & Tice, D. M. (1990). Anxiety and social exclusion. Journal of of Social and and Psychology,9,9,165-195. 165-195. ClinicalPsychology, Clinical R. F., Tice, D. M., & Hutton, D. G. (1989). Self-presentational Self-presentational motivations and Baumeister, R differences in self-esteem. Journal of of Personality, Personality,57, 57,547-579. 547-579. personality differences To know know oneself oneself isis to to like like oneself: oneself: Self-certainty Self-certainty and and self-affect. self-affect. Baumgardner, A. H. (1990). (1990). To Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, 1062-1072. Journal 58,58, 1062-1072. death. New York: Free Press. of death. Becker, E. (1973). The denial of (1981).Identity Identityformation formationduring duringlate lateadolescence: adolescence:AAreview reviewof ofsome someempirical empirical Bernard, H. S. (1981). findings. findings. Adolescence, Adolescence,16, 16,349-357. 349-357. Bios, P. (1962). On adolescence. adolescence.New NewYork: York:Free FreePress. Press. Bourne, E. (1978). The state of research on ego identity: A review and appraisal. Part II. Journal 371-392. Journal of of Youth Youthand andAdolescence, Adolescence,7, 7, 371-392. Brehm, J. W. (1966). A theory of of psychological psychological reactance. reactance.New NewYork: York:Academic AcademicPress. Press. Brockner, J. (1983). Low self-esteem self-esteem and behavioral plasticity: Some implications. In L. Wheeler & & P. Shaver (Eds.), Review of ofpersonality personalityand andsocUlI socialpsychology psychology(Vol. (Vol.4,4,pp. pp. 237-271). Beverly Hills, CA: Sage. Brockner, J., & Hulton, A. (1978). (1978). How to reverse the vicious cycle of low self-esteem: The importance of attentional focus. Journal Journal of of Experimental ExperimentalSocUlI SocialPsychology, Psychology,15, 15, 564-578. Brown, B. R R. (1968). The effects effects of need to maintain face on interpersonal bargaining. Journal Psychology, 4, 107-122. Journal of of Experimental ExperimentalSocial Social Psychology, 4, 107-122. R., & Garland, Garland, H. H. (1971). (1971).The Theeffects effects of ofincompetency, incompetency,audience audienceacquaintanceship, acquaintanceship, Brown, B. R, feedback on face-saving face-saving behavior. Journal of of Experimental and anticipated evaluative feedback Experimental SocUlI Social Psychology, Psychology,7,7,490-502. 490-502. Campbell, A. (1981). (1981). The sense of of well-being well-beingininAmerica. America.New NewYork: York:McGraw-HIll. McGraw-Hill. Campbell, J. D. (1986). Similarity and uniqueness: The effects effects of attribute type, relevance, and individual differences differences in self-esteem and depression. Journal of of Personality Personality and and Social Social Psychology, Psychology,50, 50,281-294. 281-294. Self-esteem and and the the clarity clarity of of the the self-concept. self-concept. Journal Journalof ofPersonality Personality Campbell, J. D. (1990). (1990). Self-esteem , 538-549. and Social Social Psychology, Psychology,5959, 538-549.

CHAPTER CHAPTER 26 26

IDENlTIY, IDENTITY, SELF.CONCEPT, SELF-CONCEPT, AND AND SELF·EsTEEM SELF-ESTEEM

705 70S

Campbell, J. D., & Fairey, P. J. (1985). Effects Effects of self-esteem, hypothetical explanations, and verbalization of expectancies on future future performance. Journal of of Personality Personality and and Social Social Psychology, Psychology,48, 48,1097-1111. 1097-1111. Campbell, J. D., & Lavallee, L. F. (1993). Who am I? The role of of self-concept self-concept confusion confusion in understanding the behavior of people with low self-esteem. In R. Baumeister (Ed.), Self-esteem: (pp. 3-20). New York: Plenum Press. Self-esteem:The Thepuzzle puzzleofoflow lowself-regard self-regard (pp. 3-20). New York: Plenum Press. M., & Busch, C. M. (1981). The The influence influence of of shyness shyness on on loneliness in in aa new new situation. situation. Cheek, J. M., Personalityand andSocial SocialPsychology Psychology Bulletin, 7, 572-577. Personality Bulletin, 7, 572-577. (1959). Some Some implications implications of of self-esteem self-esteem for for social social influence. influence. In In C. C. 1.I. Hovland Hovland & & Cohen, A. R. (1959). Personalityand and persuasibility persuasibility(pp. (pp.102-120). 102-120).New NewHaven, Haven,Cf: CT:Yale Yale I. L. Janis (Eds.), Personality University Press. Crocker, J., & Schwartz, I. (1985). (1985). Prejudice and ingroup favoritism favoritism in a minimal intergroup 11, situation: Effects of self-esteem. Personality Personality and Social Psychology Psychology Bulletin, Bulletin, 11, 379-386. Crowne, D. P., & Marlowe, D. (1964). The approval motive. New York: Wiley. Csikszentmihalyi, M. (1982). Toward a psychology of optimal experience. In L. Wheeler (Ed.), Review Review of of personality personalityand andsocial socialpsychology psychology(Vol. (Vol.2,2,pp. pp.13-36) 13-36)Beverly BeverlyHills, Hills, CA: Sage. Hart, D. D. (1982). (1982).The Thedevelopment development of ofself-understanding self-understanding from from infancy infancy through through Damon, W., & Hart, Development,53, 53,841-864. 841-864. adolescence. Child Development, Dunbar, J. M., & Stunkart, A. J. (1979). Adherence to diet and drug regimen. In R. Levy, heart disease B. Rifkind, B. Dennis, & & N. Ernst (Eds.), Nutrition, Nutrition,lipids, lipids,and andcoronary coronary heart disease (pp. (pp. 391-423). New York: Raven Press. Duval, S., & Wicklund, R. A. (1972). (1972). A theory of of objective objective self-awareness. self-awareness.New NewYork: York:AcaAcademic Press. Erikson, E. H. (1950). (1950). Childhood and society. society. New NewYork: York: Norton. Norton. Erikson, E. H. (1958). Young man Luther. Luther. New York: Norton. Erikson, E .H. (1968). Identity: crisis. New NewYork: York:Norton. Norton. Identity: Youth and crisis. (1984). The dimensionality dimensionality of self-esteem: II. Hierarchical Fleming, J. S., & Courtney, B. E. (1984). Journalof ofPersonality Personalityand andSocial Social Psycholfacet model for revised measurement scales. Journal Psychol 404-421.. ogy, 46, 46, 404-421 ogy, (1980). The dimensionality of of self-esteem: Some results for Fleming, J. S., & Watts, W. A. (1980). Psychology, 39,39, 921-929. a college sample. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 921-929. MaroUa, J. (1976). (1976). Efficacious Efficacious action and social approval as interacting Franks, D. D., & Marolla, reformulation through construct validation. dimensions of self-esteem: A tentative reformulation Sociometry,39, 39,324-341 324-341. Sociometry, . Freud, S. (1961). 1. Strachey, (1961). Beyond the pleasure pleasure principle principle ((J. Strachey, Trans.). Trans.). New New York: York: Norton. Norton. (Original work published 1920). Frost, R. 0., O., Goolkasian, G. A., Ely, R. J., & Blanchard, F. A. (1982). Depression, restraint Behavior Research Researchand andTherapy, Therapy, 113-121. and eating behavior. Behavior 20,20, 1 13-121. Garner, D. M., Olmsted, M. P., Polivy, J., & Garfinkel, P. E. (1984). Comparison between weight preoccupied women and anorexia nervosa. Psychosomatic 46, Psychosomatic Medicine, Medicine, 46, 255-266. X., & Wicklund, R. A. (1976). (1976). Selective exposure to self. Journal of of Research Gibbons, F. X., Research in Personality, Personality,10, 10,98-106. 98-106. in from uncertain, neutral, and Goldberg, L. R. (1981). Unconfounding situational attributions from ambiguous ones: A psychometric analysis of descriptions of oneself oneself and others. Journal Journal of Personality Personalityand andSocial SocialPsychology, Psychology, 517-552. of 41,41, 517-552.

706 706

Roy F. BAUMEISTER ROYF.BAUMEISTER

Goleman, D. (1988). The meditative ofof meditative experience. NewNew York: meditativemind: mind:The Thevarieties varieties meditative experience. York: St. Martin's Press. Gollwitzer, GoUwitzer, P. M. M, (1986). Striving for specific identities: The The social social reality reality of of self-symbolizing. self-symbolizing. In R. Baumeister (Ed.), Public Public self and private self (pp. 143-159). New York: Springer-Verlag. Springer-Veriag. 1020. Perceptualand and Motor MotorSkills, Skills,33, 33,1020. Greaves, G. (1971). Temporal orientation in suicidals. Perceptual Greenberg, J., & & Musham, C. (1981). Avoiding and seeking self-focused self-focused attention. Journal Journal of 191-200. of Research ResearchininPersonality, Personality,IS,15, 191-200. Greenberg, J., Pyszczynski, T., & & Solomon, S. (1986). The causes and consequences of self selfand private self self esteem: A terror management theory. In R. Baumeister (Ed.), Public and (pp. 189-212). New York: Springer-Verlag. Springer-Veriag. (pp. Greenwald, A. G. (1980). The totalitarian ego: ego: Fabrication and revision of personal history. American American Psychologist, Psychologist,35, 35,603-613. 603-613. & Banaji, M. R. (1988). Is self-esteem a central ingredient Greenwald, A. G., Bellezza, F. S., & Personalityand andSocial SocialPsychology PsychologyBulletin, Bulletin, 34-45. 14,14, 34-45. of the self-concept? Personality & Rosen, J. C. (1988). Bulimia in adolescents: Prevalence and psychosocial corre correGross, J., & lates. International IntemationalJournal ofEating EatingDisorders, Disorders,7, 7, 51-61. Journal of 51-61. Harter, S. (1993). Causes and consequences of low self-esteem in children and adolescents. Self-esteem:The Thepuzzle puzzleof oflow lowself-regard self-regard (pp. 87-116). New In R. Baumeister (Ed.), (Ed.), Self-esteem: (pp. 87-116). New York: Plenum Press. & Baumeister, R. F. (1991). Binge eating as escape from self-awareness. Heatherton, T. F., & PsychologicalBulletin, Bulletin,110, 110,86-108. 86-108. Psychological & Polivy, 1. J. (1991). Development and validation of a scale for measuring Heatherton, T. F., & Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, 895-910. state self-esteem. Journal 60,60, 895-910. Henken, V. J. (1976). Banality reinvestigated: A computer-based content analysis of suicidal Suicideand andLife-Threatening Life-Threatening Behavior, 6, 36-43. Behavior, 6, 36-43. and forced-death documents. Suicide Higgins, E. T. (1987). Self-discrepancy: A theory relating self and affect. Psychological Psychological Review, 94, 94, 319-340. Review, Higgins, E. T., Klein, R. L., & Strauman, T. J. (1987). Self-discrepancies: Distinguishing among self-states, self-state conflicts, and emotional vulnerabilities. In K. Yardley & T. Honess (Eds.), Selfand perspectives (pp. Chichester, Self and identity: identity:Psychosocial Psychosocial perspectives (pp.173-186). 173-186). Chichester, England: Wiley. England: Hull, J. G. (1981). A self-awareness model of the causes and effects effects of alcohol consumption. Journal of Journal of Abnormal AbnormalPsychology, Psychology,90, 90,586-600. 586-600. Hull, J. G., Levenson, R. W., Young, R. D., & & Sher, K. J. (1983). Self-awareness-reducing effects of alcohol consumption. Journal Journal of of Personality Personality and and Social SocialPsychology, Psychology,44,44, 461-473. & Young, R. D. (1983). Self-consciousness, self-esteem, and success-failure as Hull, J. G., & determinants of alcohol consumption in male social drinkers. Journal Journalof of Personality Personality and Social Social Psychology, Psychology,44, 44,1097-1109. 1097-1109. Hull, J. G., Young, R. D., & Jouriles, E. (1986). Applications of the self-awareness model Journal of ofPersonality Personality of alcohol consumption: Predicting patterns of use and abuse. Journal and Social Social Psychology, Psychology,51, 51,790-796. 790-796. (1954). Personality Personalitycorrelates correlatesof ofsusceptibility susceptibilityto topersuasion. persuasion.Journal JournalofofPersonality, Personality, Janis, I. L. (1954). 22, 504-518. & Field, P. (1959). (1959). Sex Sex differences and and personality factors related to persuasibility. Janis, I. L., & & I. Janis (Eds.), Personality Personalityand andpersuasibility persuasibility(pp. (pp.55-68, 55-68,300-302). 300-302). In C. Hovland & New Haven, CT: CT: Yale Yale University University Press. Press.

CHAPTER 26 26

IDEN11TY, IDENHTY, SELF.CONCEPT, SELF.CONCEPT» AND AND SELF.ESTEEM SELF-ESTEEM

707 707

selfJones, E. E., & Berglas, S. C. (1978). Control of attributions about the self through self handicapping strategies: The appeal of alcohol and the role of underachievement. Personality Personality and Social Social Psychology Psychology Bulletin, 4, 200-206. Jones, E. E., & Wortman, C. (1973). Ingratiation: An An attributional attributional approach. approach. Morristown, NJ: General Learning Press. Jones, S. C. (1973). Self- and interpersonal evaluations: Esteem theories vs. consistency theories. Psychological Psychological Bulletin, 79, 185-199. Jones, W. H., Freemon, J. E. & Goswick, R R. A. (1981). The persistence of of loneliness: Self and other determinants. Journal Journal of of Personality, Personality, 49, 27-48. second year: The emergence of of self-awareness. Cambridge, MA: Harvard Kagan, J. (1981). The second University Press. A Katzman, M. A., & Wolchik, S. A. (1984). Bulimia and binge eating in college women: A comparison of personality and behavioral behavioral characteristics. Journal Journal of of Consulting Consulting and Psychology, 52, 423-428. Clinical Psychology, of self-concept self-concept in preschool Keller, A., Ford, L. H., & Meacham, J. A. (1978). Dimensions of Developmental Psychology, Psychology, 14, 483-489. children. Developmental stabihty and level of self-esteem self-esteem in psychological function functionKernis, M. H. (1993). The roles of stability R. Baumeister (Ed.), Self-esteem: The puzzle puzzle of of low low self-regard self-regard (pp. 167-182). ing. In R New York: Plenum Press. Rites of of passage: Adolescence Adolescence in America America 1790 1790 to the present. New York: Kett, J. (1977). Rites Basic Books. conXxoX.-Journal Personality and andSocial SocialPsychology, Psychology, Journal ofofPersonality Langer, E. (1975). The illusion of control. 29, 253-264. of a man's man's lif life. Levinson, D. J. (1978). The seasons of e. New York: Ballantine. Social cognition cognition and and the acquisition self New York: acquisition of of self. Lewis, M., & Brooks-Gunn, J. (1979). Social Plenum Press. Maclntyre, A. (1981). After After virtue. Notre Dame, IN: University of of Notre Dame Press. MacIntyre, selfMaracek, J., & Mettee, D. (1972). Avoidance of continued success as a function of self of esteem certainty, and responsibility for success. Journal Journal of of Personality Personality esteem, level of Social Psychology, Psychology, 22, 98-107. and Social validation of of ego-identity status. Journal Journal of of Personality Personality Marcia, J. E. (1966). Development and validation and Social Social Psychology, Psychology, 3, 551-558. and Marcia, J. E. (1967). Ego identity status: Relationship to change in self-esteem, "general maladjustment,'* and authoritarianism. Journal Journal of of Personality, Personality, 35, 118-133. maladjustment," and gender. Marcia, J. E., & Scheidel, D. G. (1983). Ego identity, intimacy, sex role orientation, and Presented at the annual meeting of the Eastern Psychological Association, Philadelphia. Markus, H. (1977). Self-schemata and processing information about the self. Journal Journal of of Personality Personality and and Social Social Psychology, Psychology, 35, 63-78. Markus, H., & Nurius, P. S. (1986). Possible selves. American American Psychologist, Psychologist, 41, 954-969. Markus, H., & Nurius, P. S. (1987). Possible selves: The interface between motivation and the self-concept. In K. Yardley & T. Honess (Eds.), Self Self and and identity: Psychosocial Psychosocial perspectives perspectives (pp. 157-172). Chichester, England: Wiley. Mar�us, Markus, H. R, R., & Kitayama, S. (1991). Culture and the self: Implications for cognition, emotion, and motivation. Psychological Psychological Review, Review, 98, 224-253. Maroldo, G. K. (1982). Shyness and love on the college campus. Perceptual Perceptual and and Motor Motor Skills, 55, 55, 819-824. self-esteem and performance feedback McFarlin, D. B., & Blascovich, J. (1981). Effects of self-esteem Journal of on future affective affective preferences and cognitive expectations. Journal of Personality Personality and Social Psychology, Psychology, 40, 521-531. Social .

708 708

Roy F. BAUMEISTER ROYF.BAUMEISTER

Menninger, K. (1966). Man against himself. New York: Harcourt, Brace, & World. (Original work published 1938). Montemayor, R, R., & & Eisen, M. (1977). The development of self-conceptions from childhood to adolescence. Developmental DevelopmentalPsychology, Psychology,13, 13,314-319. 314-319. Neuringer, c., C, & & Harris, R R. M. M. (1974). The perception of the passage of time among death deathinvolved hospital patients. Life-Threatening Life-Threatening Behavior, Behavior,4, 4,240-254. 240-254. Ogilvie, D. D. M. M. (1987). The undesired self: A neglected variable in personality research. Journal 52,52, 379-385. Journal of ofPersonality Personalityand andSocial SocialPsychology, Psychology, 379-385. Orlofsky, J. L., E., & L., Marcia, J. E., & Lesser, I. M. M. (1973). Ego Ego identity status and the intimacy versus isolation crisis of Social Psychology, of young young adulthood. adulthood.Journal Journalof ofPersonality Personalityand and Social Psychology, 27, 27, 211-219. flexibility: A new conception of interper Paulhus, D. L., & & Martin, C. L. L. (1988). Functionalflexibility: interpersonalflexibility. flexibility. Journal 55,55, 88-101. Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, 88-101. Suicideand and Life LifePerrah, M., & Wichman, H. (1987). Cognitive rigidity in suicide attempters. Suicide ThreateningBehavior, Behavior,17,17, 251-262. Threatening 251-262. J. (1976). (1976).Perception Perceptionof ofcalories caloriesand andregulation regulationof ofintake intakein inrestrained restrainedand andunrestrained unrestrained Polivy, J. Behaviors,1,1,237-243. 237-243. subjects. Addictive Behaviors, D. G. (1981). Negotiation Negotiation behavior. behavior.New NewYork: York:Academic AcademicPress. Press. Pruitt, D. Suicide and Life-Threatening Life-Threatening Behavior, Behavior, 6, Ringel, E. (1976). The presuicidal syndrome. Suicide 131-149. C , & Dymond, R R. (1954). Psychotherapy Psychotherapyand andpersonality personalitychange. change, Chicago: University Chicago: University Rogers, c., of Chicago Press. M. (1979). Conceiving Conceiving the theself. self New NewYork: York: Basic Basic Books. Books. Rosenberg, M. Roth, S., & Kubal, L. L. (1975). Effects Effects of noncontingent reinforcement reinforcement on tasks of differing differing S., & of Personality Personality and andSocial Social importance: Facilitation and learned helplessness. Journal of Psychology, 332, 680-691. Psychology, 2, 680-691. Rothbaum, F., Weisz, J. R, R., & Synder, S. (1982). Changing and world and changing the self: self: Social Psychology, A two process processmodel modelof ofperceived perceivedcontrol. control.Journal Journalof ofPersonality Personalityand and Social Psychology, 42, 42, 5-37. 5-37. Rubin, J. Z., Z., & & Brockner, J. (1975). Factors affecting affecting entrapment in waiting situations: The The Rosencrantz and and Guildenstern effect. Journal Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, 31, 1054-1063. Ruble, D. (1983). The The development development of of social social comparison comparison processes processes and and their their role role in inachieve achievement-related self-socialization. self-socialization. In E. T. Higgins, D. Ruble, & W. Hartup (Eds.), Social Social (pp.(pp. 134-157). New York: cognition perspectives cognition and andsocial socialbehavior: behavior:Developmental Developmental perspectives 134-157). New York: Cambridge University Press. Ruderman, A. J. J. (1985). (1985). Dysphoric Dysphoricmood moodand andovereating: overeating:A Atest testof ofrestraint restrainttheory's theory'sdisinhibi disinhibiJournalof ofAbnormal AbnormalPsychology, Psychology,94,94,78-85. 78-85. tion hypothesis. Journal Sackett, D. D. L., L., & & Snow, J. C. C. (1979). The The magnitude of compliance and noncompliance. In R. B. B. Haynes, D. D. W. W. Taylor, & & D. D. L. L. Sackett (Eds.), Complaince Complaince in inhealth healthcare care(pp. (pp. R 11-22). Baltimore: Johns Hopkins University Press. Scarry, E. (1985). The The body in in pain: pain: The The making making and and unmaking unmakingofofthe theworld. world.New NewYork: York: Oxford University Press. Oxford Schlenker, B. B. R R. (1980). Impression Impression management. management Monterey, CA: BrookS/Cole. Brooks/Cole. Schlenker, B. B. R R. (1985). The self and social life. life. New New York: McGraw-Hill. R. (1986). Self-identification: Toward an integration of the private and public Schlenker, B. R R. Baumeister (Ed.), Public self and and private self self (pp. 21-62). New New York: self. In R Springer-Verlag.

CHAPTER CHAPTER 26 26

IDENTITY, SELF·CONCEPT, SELF-CONCEPT, AND SELF.EsTEEM SELF-ESTEEM IDENTIn',

709 709

(1982). Social Socialanxiety anxiety and andself-presentation. self-presentation. A A conceptualiza conceptualizaSchlenker, B. R., & Leary, M. R. (1982). 641-669. tion and and model. Psychological PsychologicalBulletin, Bulletin,92,92, 641-669. Schlesier-Stropp, B. B. (1984). Bulimia: A review of the literature. Psychological Psychological Bulletin, 95, 247-257. Schneider, D. J., & Turkat, D. (1975). Self-presentation Self-presentation following following success successor orfailure: failure:Defensive Defensive self-esteem of Personality Personality43, 43, 127-135. 127-135. self-esteem models. Journal of Scott, G. G. (1983). Erotic and submission. Secaucus, Erotic power: power:An Anexploration explorationofofdominance dominance and submission. Secaucus, NJ: Citadel Press. development, andand death. San San Fran-FranM. E. P. P. (1975). Helplessness: Helplessness:On Ondepression, depression, development, death. Seligman, M. cisco: Freeman. The fall of of public man. man. New New York: Random House. Sennett, R. (1974). The Shainess, N. N. (1984). Sweet suffering: suffering:Woman Womanasasvictim. victim.New NewYork: York:Simon Simon&&Schuster. Schuster. Shrauger, J. S. (1972). Self-esteem and reactions to being observed by others. Journal of of Personality 23,23, 192-200. Personalityand andSocial SocialPsychology, Psychology, 192-200. Shrauger, J. S. S. (1975). Responses to evaluation as a function function of initial self-perceptions. self-perceptions. Psychological , 581-596. PsychologicalBulletin, Bulletin,8282, 581-596. Symbolic interactionist interactionist view view of of the the self-concept: self-concept: Shrauger, J. S., & Schoeneman, T. S. (1979). Symbolic Through the looking glass darkly. Psychological PsychologicalBulletin, Bulletin,86, 86,549-573. 549-573. Shrauger, J. S., S., & Sorman, P. B. B. (1977). (1977). Self-evaluations, initial success and failure, and Journalof ofConsulting Consultingand andClinical Clinical Psyimprovement as determinants of persistence. Journal Psy chology, 445, 784-795. chology, 5, 784-795. Silverman, I. (1964). Self-esteem and differential differential responsiveness to success and failure. Jour Journal of of Abnormal and Social Social Psychology, Psychology,69, 69,115-119. 115-119. Slochower, J., & & Kaplan, S. S. P. (1980). Anxiety, perceived control, and eating in obese and normal weight persons. Appetite, Appetite, 1, 1, 75-83. & Bloom, Bloom, I.I. (1985). (1985). A A study study in in the the personal personal meaning meaning of of suicide suicide in in the the context context Smith, G. W., & Behavior, 15, 15, 3-13. of Baechler's typology. Suicide Suicideand andLife-Threatening Life-Threatening Behavior, 3-13. S. J., Josephs, R. A., & & Steele, C. C. M. M. (1993). Low self-esteem: The uphill struggle Spencer, S. for for self-integrity. In R. Baumeister (Ed.), Self-esteem: Self-esteem: The Thepuzzle puzzle of oflow low self-regard self-regard (pp. 21-36). New (pp. New York: Plenum Press. Staw, B. M. (1976). Knee-deep in the big muddy: A study of escalating commitment to a B. M. chosen course of action. Organizational Performance, 16, 27-44. OrganizationalBehavior Behaviorand andHuman Human Performance, 16, 27-44. M. (1988). The psychology of self-affirmation: Steele, C. C. M. self-affirmation: Sustaining the integrity of the L. Berkowitz (Ed.), Advances Advances ininexperimental experimental social psychology (Vol. social psychology (Vol. 21,21, pp.pp. self. In L. 261-302). Orlando, FL: Academic Press. & Southwick, L. L. (1985). (1985). Alcohol and social behavior I: The psychology of Steele, C. M., M., & of drunken excess. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 18-34. Psychology, 48, 48, 18-34. Steenbarger, B. N., & Aderman, D. (1979). (1979). Objective self-awareness as a nonaversive state: Effect of anticipating discrepancy reduction. Journal of of Personality, Personality,47, 47,330-339. 330-339. Effect W. B. two roads meet. Journal of of Personality Swann, W. B. (1987). Identity negotiation: Where two Personality Social Psychology, Psychology,53, 53,1038-1051. 1038-1051. and Social Swann, W. B., Griffin, W. B., Griffin, J. J., Predmore, S. S. c., C, & & Gaines, B. B. (1987). The cognitive-affective cognitive-affective self-consistency confronts self-enhancement. Journal of of Personality crossfire: When self-consistency Personality and Social Psychology,52, 52,881-889. 881-889. Social Psychology, W. B., B., & Hill, c. C. A. (1982). When our identities are mistaken: Reaffirming Reaffirming self selfSwann, W. ofPersonality Personalityand andSocial Social Psychology, conceptions through social interaction. Journal Journal of Psychology, 43, 59-66.

710 710

RoY ROY F. F.BAUMEISTER BAUMEISTER

Positive illusions: illusions:Creative Creativeself-deception self-deception healthy mind. Taylor, S. E. (1989). Positive andand thethe healthy mind. NewNew York: Basic Books. Taylor, S. E., E., & & Brown, Brown,J.J.(1988). (1988).Illusion Illusionand andwell-being: well-being:Some Somesocial socialpsychological psychologicalcontribu contributions to a theory of mental health. Psychological PsychologicalBulletin, Bulletin,103, 103,193-210. 193-210. Teger, A A. I. (1980). Too much invested invested to to quit. quit.New New York: York: Pergamon Pergamon Press. Press. M. (1990). Esteem protection or enhancement? Self-handicapping Self-handicapping motives and Tice, D. M. Journalof ofPersonality Personalityand andSocial Social Psychology, attributions differ by trait self-esteem. Journal Psychology, 60, 711-725. 60, Triandis, H. C. C. (1989). (1989).The Theself selfand andsocial socialbehavior behaviorinindiffering differingcultural culturalcontexts. contexts.Psychological Psychological . Review, 96, 506-520. 506-520. Review, 96, Trilling, L. (1971). Sincerity and and authenticity. authenticity.Cambridge, Cambridge,MA: MA:Harvard HarvardUniversity UniversityPress. Press. (1971). Sincerity Vallacher, R R. R, R., & Wegner, D. M. (1985). A theory of of action action identification. identification. Hillsdale, Hillsdale, NJ: Erlbaum. R. R R. & Wegner, D. M. (1987). (1987). What do people think they're doing: Action Vallacher, R PsychologicalReview, Review,94, 94,3-15. 3-15. identification and human behavior. Psychological identification R., & White, T. L. (1987). Paradoxical effects effects Wegner, D. M., Schneider, D. J., Carter, S. R, Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, 5-13. 53,53, 5-13. of thought suppression. Journal (1984). The The social social constituents constituents of of sadomasoch sadomasochWeinberg, M. S., Williams, C. J., & Moser, C. (1984). Social Problems, Problems,31, 31,379-389. 379-389. ism. Social 5 and M: Studies in sadomasochism. Buffalo, Weinberg, T., & Kamel, W. L. (Eds.). (1983). S sadomasochism. Buffalo, NY: Prometheus. NY: The value valueof ofthe theindividual: individual: Self circumstance in autobiography. Self andand circumstance in autobiography. Weintraub, K. J. (1978). The Chicago: University of Chicago Press. R. W. (1959). Motivation reconsidered: The concept of competence. Psychological Psychological White, R Review, 66, 66, 297-333. 297-333. Review, Wicklund, R (1975). Objective self-awareness. In L. Berkowitz (Ed.), Advances in experi R. A A. (1975). experipsychology (Vol. (Vol. 8, 8, pp. pp. 233-275). 233-275). New New York: York: Academic Academic Press. Press. mental social social psychology Wicklund, R R. A., (1982). Symbolic self-completion. self-completion. Hillsdale, Hillsdale,NJ: NJ:Erl ErlA, & Gollwitzer, P. M. (1982). baum. Williams, J. M., & Broadbent, K. (1986). (1986). Autobiographical memory in suicide attempters. Journal Journal of ofAbnormal AbnormalPsychology, Psychology,95, 95,144-149. 144-149. Wilson, S. R, R., & Benner, Benner, L. L. A A. (1971). (1971). The Theeffects effects of ofself-esteem self-esteem and and situation situation upon upon compari comparison choices during ability evaluation. Sociometry, Sociometry, 34, 34,381-397. 381-397. Wolfe, R N., Lennox, R R. N., R. D., & Cutler, B. B. L. (1986). Getting along and getting ahead: Empirical support for for a theory of protective and acquisitive self-presentation. Journal of Personality Personalityand andSocial SocialPsychology, Psychology, 356-361. of 50,50, 356-361. The self-concept: self-concept:Vol. Vol. 1. A review of methodological considerations R. C. (1974). The Wylie, R 1. A review of methodological considerations and and measuring instruments. instruments.Lincoln: Lincoln:University UniversityofofNebraska NebraskaPress. Press. measuring R. C. C. (1979). The Theself-concept: self-concept: Vol2. 2.Theory Theory and research on selected topics. Lincoln: Wylie, R Vol. and research on selected topics. Lincoln: University of Nebraska Press. Yufit, R R. I., & Benzies, B. B. (1973). Assessing suicidal potential by time perspective. Life LifeThreateningBehavior, Behavior, 270-282. Threatening 3, 3, 270-282.

CHAPTER 27 27

MENTAL REPRESENTATIONS OF OF PERSONS AND PERSONALITY JOHN F F.. KIHLSTROM YALE UNIVERSITY YALE UNIVERSITY REID HASTIE UNIVERSITY OF COLORADO UNIVERSITY OF COLORADO

The science of personality has long been troubled by the conflict between biosocial (AUport, 1937). and biophysical conceptualizations of its subject matter (Allport, 1937). Empirical view. Thus, Thus, AUport's time has has emphasized the biophysical view. work on personality since Allport's the traditional psychometric approach to personality has expressed its concern with in what might be be called the the Doctrine of Traits (Kihlstrom, individual differences in 1988), in which persons are are viewed as as collections of intrapsychic dispositions, analo analogous to physical characteristics, which give surface behavior a high degree of coher coherand predictability. predictabihty. Still, ence, stability across time, consistency across situations, and the Doctrine of Traits has has been under attack, off and and on, on, for more than half a century (Hartshorne & & May, 1928; 1928; Mischel, 1968; 1968; Nisbett, 1980; 1980; Peterson, 1968). According to its social-psychological critics, the correlations among topographically different behaviors, and and among among semantically semantically related related traits, traits, are aretoo toolow lowto toprovide provide more than the most abstract coherence: both observed behavior and inferred traits at best only modest stability over even short periods of time and substantial show at variability from one situation to another; another; and the extent to which specific behavior can be be predicted from generalized traits isis very verylow, low, even evenwhen whenbehavior behaviorisis measured measured can the aggregate. Thus, coherence, stability, consistency, and and predictability-the predictabihty—the in the raison d'etre d'etrefor forthe theDoctrine DoctrineofofTraits-appear Traits—appeartotobebemore moreininthe theeye eyeofofthe the very raison in the the person beheld. beholder than in lIANDBOOK OFOF PERSONALITY PSYCHOLOGY HANDBOOK PERSONAUTY PSYCHOLOGY 711

711

CoPYII GHT C COPYRIGHT 0 1997 BY ACADEMIC PRESS PRESS.. ALL liGHTS IEPltODUCTION IN ANY FORM RESERVED. A L LOFRIGHTS OF REPRODUCTION IN ANY FORM RESERVED.

KnusntOM AND HASTIE KIHLSTROMANDHASIIE

712 712

social-psychological critique of of Traits leads In this way, the social-psychological of the Doctrine of naturally to an emphasis on the biosocial rather than biophysical aspects of of personal personalimpressions of personality, rather than personality per se. But, of ity-on ity—on impressions of course, one need not adopt the social-psychological of traditional psychometric social-psychological critique of approaches to personality in order to be interested in how people perceive, remem personahty remember, and categorize themselves and others. In this chapter, we introduce contempo contempodirectly on the mental representation rary research in social cognition which bears directly of other persons and their personalities (for (for a parallel review review of of people's mental of representations of themselves, see Kihlstrom & Klein, 1994).

I.I.

THE DOMAIN OF SOCIAL COGNITION Cognitive psychology is concerned with mental representations of of the world and of acquiring the mental processes that operate on these representations in the course of manipulating, and utilizing knowledge stored in the mind. Cognitive psychology casts a very broad net; indeed, insofar as psychology is defined at "the science of mental life" (James, 1890/1981, p. 15), the idea of of a cognitive psychology is almost redundant. Of course, there was a time when psychology psychology was defined as the science of behavior and psychologists limited their work to tracing the functional of functional relations rebetween environmental stimuli and the organism's muscular and glandular re sponses to them. Cognitive psychology does not abjure an interest in behavior, but it does assume that a complete understanding of of behavior requires an explication of of the mental structures and processes that mediate between stimulus and response. Thus behavior is a window on the mind, and the visible expression of of mind is intelligent action. Behavior that is not under cognitive control, what might be called reflexive or instinctual, is more properly the province of of disciplines such as neuroscience and ethology. Cognitive approaches to personality, social, and clinical psychology psychology share the assumption that behavior is cognitively mediated-guided by our our perceptions mediated—guided by perceptions of of the the current current situation, situation, memories memories of of similar similar situations situations encountered encountered in in the the past, past, impres impressions of ourselves and other people in in the current environment, attributions concern concerning their (and our own) experiences and actions, and other sorts of of judgments and inferences inferences that go beyond the information given in the stimulus situation. When we speak of of social social cognition, then, we speak of cognition in the social domain, both elements broadly construed. At one level, the study of of social social cognition is simply the study of of our knowledge of social entities-of entities—of ourselves, other people, the situations in which we encounter them, and the interpersonal behaviors which are exchanged in those situations. At another level, it is concerned with the structure of of mental representations of these social entities, as they are currently perceived or retrieved from memory, and the ways in which these representations are constructed, doreconstructed, and used to guide our experience, thought, and action in social do mains.

CHAPTER 27 27

MENTALREPRESENTATIONS REPRESENTATIONSOF OFPERSONS PERSONS MENTAL

713 713

The formal formal study of social cognition is only about 50 years old. It had its beginnings beginnings in the work work of Heider (1944) on phenomenal phenomenal causality, Asch Asch (1946) on impression formation, and Bruner and Tagiuri (1954) on implicit personality theory. Cognitive concerns are also represented represented in much of the classic work on cognitive cognitive Cognitive consistency, balance, and dissonance, clinical judgment, and attribution theory pro produced during during the "Golden Years" of of experimental experimental social social psychology. But the cogni cognisocial psychology during this period were relatively relatively informal, per pertive concerns of social because the field field of modem modern cognitive psychology was also in its infancy. infancy. haps because This situation changed radically in the late 1960s, as social psychologists pro proposed posed formal models models concerning the mental representation representation of persons persons and the involved in causal attribution and impression impression formation. Of particular judgments involved importance were Rosenberg's spatial models for the representation of persons (e.g., & Vivekananthan, 1968) and N. H. Anderson's (1965) algebraic Rosenberg, Nelson, & models for social judgment processes. Through them, social cognition began to make closer contact with the emerging cognitive psychology. The connection has tightened since since then, with many many psychologists from both sides crossing the border tightened frequently and effortlessly. effortlessly. The result has been that the study of social cognition cognition frequently is characterized characterized by by sophisticated sophisticated concepts, concepts, theories, theories, and and methodologies, methodologies, many many of of is which are are ripe ripe for for application application in in the the fields fields of of personality personality and and clinical clinical psychology. psychology. which n. ALTERNATE VIEWS OF SOCIAL COGNITION n.

At A t this point in time, several general approaches to social cognition have emerged that are coherent and distinctive enough to stand as identifiable identifiable theoretical para paradigms (for a fuller fuller discussion, see Hastie, 1983). Each of these differs differs from from the of the completeness of of its account of social cognition, and each each others in terms of offers a different different budget budget of of theoretical assets and liabilities. offers Role Role theory theory (Sarbin, 1954) is based on a "dramaturgical metaphor" in which which actors play out scripts before before audiences. From this point of view, personal conduct conduct is governed by social roles imposed by the context in which behavior takes place; socialization socialization involves acquiring a repertoire repertoire of roles and understanding understanding the roles roles ultimately a cognitive cognitive required by various situations. In principle, role theory is ultimately theory because roles are abstract ideas that an actor must learn in order to behave behave in conformity conformity with social demands, expectations, and norms. In practice, however, however, Sarbin and his associates associates have not explored the sorts of of mental structures and processes that are relevant to mainstream cognitive psychology. By focusing focusing its analysis on the situational context in which social roles are enacted (Sarbin, 1982), role theory theory identifies itself itself as a version version of situationism, and is more more sociological sociological than cognitive in its orientation to its subject matter. The cognitive associated with the research of cognitive algebra algebra approach is primarily associated Anderson (1974, (1974,1978,1981) whereby Norman Anderson 1978, 1981) and his colleagues on the processes whereby information from from several sources is integrated into a unitary impression of a single information person. Unlike role theory, cognitive algebra provides a complete description of a

AND IlAsm Kun.sTROM KIHLSTROMANDHASIIE

714 714

cognitive cognitive system, including including psychophysical relations governing governing perception, psycho psychomotor laws governing response generation, and a set of cognitive processes that mediate between stimulus and response. These mediating processes consist of of alge algefor transforming transforming stimulus values (usually according to a weighted averag averagbraic rules for ing rule). A A closely related related program of of research, labeled social social judgment judgment theory theory (Arkes & Hammond, 1986; Brunswik, 1956; Hammond, 1955), has also utilized regression equations to capture the processes employed employed algebraic models based on regression judgment tasks such as psychiatric diagnosis (Brehmer & Joyce, 1988; in social judgment Slovic & Lichtenstein, 1971). An alternative judgment judgment heuristics heuristics approach is represented represented in the research research An and theory of of Kahneman and Tversky Tversky (1974; Kahneman, Slovic, & Tversky, 1982), Nisbett and Ross (1980; Ross, 1977), and a collection of of researchers who have Nisbett studied causal attributions for for behavior and predictions of of social events (e.g., Heider & Simmel, 1944; Jones, 1979; Jones & Davis, 1965; Kelley, 1967, 1973). Heider of cognitive algebra appear to require considerable considerable While the mental operations of effort, judgment judgment heuristics invoke shortcuts that permit efficient efficient judgments under effort, conditions of uncertainty, but increase the likelihood of of error. Thus, judgment conditions of account easily for for the departures from from normative rationality that are so heuristics account frequently observed observed in in social social cognition. cognition. frequently From a bird's eye view, role theory has emphasized the contents of of roles and From for action, but has neglected the manner in which these might their implications for mentally represented and processed processed in the mind. The cognitive algebra and be mentally judgment heuristics approaches approaches have both emphasized emphasized the manner manner in which informa information is processed, but have also neglected the manner in which it is represented. A A fourth approach, symbolic symbolic information information processing processing theory, provides a more thorough thorough fourth account of of both the representation representation and the processing components of of a complete complete account information processing approach dominates the study of of cognitive theory. The information nonsocial cognition and provides the framework framework for for the remainder of of the present (for a fuller fuller description, see Hastie, 1986). review (for

A. The Computer Metaphor The information information processing approach is derived from from a computer computer metaphor of of the mind and employs computer computer program program simulations which are designed to provide "languages" in which to write cognitive theories. These simulation models are then tested to determine whether whether they mimic the actual actual behavior of of people. tested of the Mind Mhid B. The Architecture of Most current versions of of the "architecture of cognition" (e.g., J. R. Anderson, 1983) provide for for the movement movement of of information information "inward" from from the sensory registers, "through" short-term short-term memory, "toward" long-term memory, and "back" to short shortterm memory again. Recently there has been a shift shift toward a unitary conception conception of of the memory store in which sensory registers are peripheral and a general long-

CHAPTER CHAPTER 1.7 27

MENTALREPRESENTATIONS REPRESENTATIONSOF OFPERSONS PERSONS MENTAL

715 715

term store is the locus of thinking in an activated activated portion portion described as short shortterm term memory. A variant variant on this point of of view distinguishes distinguishes between between merely merely active portions of of memory and portions of of memory memory that that are in dynamic dynamic transformation, transformation, labeled the working memory. memory. labeled A Taxonomy of of Knowledge Structures C. A

A A useful useful conceptual distinction has been borrowed from from computer science: declara declaraof general beliefs concerning concerning the nature of world and of the world tive knowledge consists of specific memories of personal experience; proce of events that have occurred in one's personal procedural knowledge consists of the skills, strategies, and rules with which we manipulate manipulate dural transform declarative declarative knowledge knowledge as well as take action world (J. R. action in the world and transform domain of of declarative knowledge it Anderson, 1983; 1983; Winograd, 1975). 1975). Within the domain common to distinguish distinguish between between semantic semantic knowledge, which comprises abstract, abstract, is common information, perhaps best conceptualized conceptualized as a combination categorical information, combination dictionary episodic knowledge, which consists of of autobiographical autobiographical mem memand encyclopedia, and episodic of events encoded encoded with reference reference to the self self and and experienced experienced within a particular ories of context (Tulving, 1983). temporal, spatial, emotional, and motivational context With respect to social cognition, declarative-semantic declarative-semantic knowledge consists of of the categories which we use to classify classify social stimuli-other stimuli—other people, ourselves, interpersonal actions, and the situations in which social interaction interaction takes place. interpersonal Declarative-episodic social knowledge has sometimes sometimes been form of of been studied studied in the form other persons, but it is more more generally represented represented by an individual's memory for other the person's autobiographical autobiographical memory. Procedural Procedural social knowledge consists of of the form impressions of of others, social competencies, strategies, and rules by which we form make causal attributions and other judgments, encode and retrieve social memories, plan and execute social behaviors, and manage other other people's impressions of of us. At the boundary between declarative and procedural social knowledge are scripts At the boundary between declarative and procedural social knowledge are scripts for social social interactions interactions (Abelson, (Abelson, 1981; 1981; Schank Schank & & Abelson, Abelson, 1977). 1977). As As semantic semantic for knowledge structures, scripts are used to help categorize the situations that people knowledge structures, scripts are used to help categorize the situations that people find themselves themselves in in and and to to make make inferences inferences about about what happened in in the the past past find what has has happened and what will happen in the future; as procedural knowledge structures, they guide and what will happen in the future; as procedural knowledge structures, they guide the actor's actor's behavior behavior in in the the situation situation from from start start to to finish. finish. the

D. The Activation of of Ideas A critical issue for research and theory has been been to characterize the manner in A which each of these structures is located and activated activated in its appropriate memory. The reigning principle, called "spreading activation," postulates that one activation of one concept in long-term memory, whether by perception of perception or or thought, activates closely related concepts according to the degree to which the new concepts share features or associative links Unks to the initial concept. At At some point, the idea becomes features accessed for for utilization by the information information processing processing system. active enough to be accessed

716 716

Knns TROM AND HAsnE KIHLSTROMANDHASHE

important corollary to the spreading activation principle is that activation activation An important of of any concept takes time to decay. If a concept retains an amount of of activation, it can be more easily reactivated reactivated on a subsequent subsequent occasion. This "priming **priming effect" effect" diminishes quickly, but some systems of of social nodes (such as personality concepts) concepts) have been hypothesized hypothesized to be permanently permanently activated (Bargh, Bond, Lombardi, & Tota, 1986; Higgins, King, & Mavin, 1982; Markus & Sentis, 1984). Chronic activa activation explains certain individual differences differences in sensitivity to types of of information information or habits in social comparison comparison processes. For example, some perceivers seem to be "tuned "tuned to" t o " channels of of information information about other other people people that that have implications for for their intelligence, while others are chronically concerned concerned with attractiveness or with their athletic ability. Elementary Information Processes E. Elementary The information method information processing approach is based on a reductionist theoretical method that assumes complex complex performances performances can be decomposed decomposed into a collection collection of of elemen elemeninformation processes. Thus, a complex complex achievement achievement like Uke the judgment judgment of of an tary information appUcant's suitability for for a job or the response to a request for for help from from an applicant's acquaintance can be described ultimately as a chain of of elementary elementary processes processes that acquaintance transform information. This basic cognitive level is operate to activate, store, and transform hypothesized to describe a level of of organization organization just above the neural substrate. hypothesized

F. Control of of Thought of a person's goals, including global Working memory contains representations of goals and current current subgoals that have to be achieved on the way to the ultimate goal, which join perceptual inputs as sources of of activation of of ideas in memory. Presumably there is an executive control structure that allocates priorities among Presumably competition multiple goals, coordinates goals when possible, and attempts to resolve competition conflicting goals. From the point of of view of of social cognition, this executive among conflicting important component of of the self. control structure is an important

G. Linking Mind and Body Cognitive neuroscientists are beginning to fulfill fulfill some of of the promises to link Unk mind and body that were the subject subject of of the earliest eariiest philosophical speculations about human psychology. This has led some psychologists to jump to the conclusion that theory at the cognitive level will soon be antiquated and that another level is more suitable closer to if not identical to the physical level of of brain modeling. The most popular solution is to propose "neurally inspired" "connectionist" models that are popular of intermediate between the cognitive level (frequently (frequently glimpsed through the lens of consciousness) and the neural level studied by anatomists.

CHAnER CHAPTER 27 27

MENTAL MENTAL REPRESENTATIONS REPRESENTATIONS OF OF PERSONS PERSONS

717 717

m. III PERSON PERCEPTION For much of its early history, social cognition was defined as the study of person & Tagiuri, 1954). The study of person perception begins with perception (Bruner & an analogy to the perception of nonsocial objects. The person exists as an object independent of the mind of the social perceiver, and the perceiver's task is to form an internal, mental representation of the person. In the same way that nonsocial perception extracts information from the array of physical energies impinging on sensory surfaces in order to address questions concerning the form, location, and activity of some object, social perception extracts information from the stream of behavior in order to address questions of the thoughts, moods, motives, and traits of other people. The study of person perception begins with the work of Asch (1946, p. 207), who defined the general problem as follows:

How do do we we organize organize the the various various data data of of observation observation into into aa single, single, relatively How relatively unified, impression? impression? How How do do our our impressions impressions change change with time and and further further unified, with time experiences with withthe the person? person?What Whateffects effectsononimpressions impressionsdodoother other psychological experiences psychological processes, such such as as needs, needs, expectations, expectations, and and established processes, established interpersonal interpersonal relations, relations, have? have? In order to study this problem, Asch invented the impression-formation paradigm traits describing another person which aa subject subject is is presented with with an an ensemble of of traits in which (the target) and is asked to report an impression of the target by completing a free findings (e.g., description, adjective checklist, or rating scales. scales. He interpreted his findings the discovery of central centraltraits) traits)asassupporting supportingthe theview, view,explicitly explicitlyderived derivedfrom fromGestalt Gestalt approaches to nonsocial perception, that the unified impression is greater than the sum of its individual elements. Asch's experiments largely set the agenda for the next 20 years of research on person perception. For example, N. H. Anderson's (1965,1974,1978) (1965, 1974, 1978) work on cognitive algebra analyzed the mathematical rules (e.g., adding vs. averaging) that govern how trait information is combined. Wishner (1960) and Rosenberg et al. (1968) showed that central traits (e.g., Asch's Asch's "warm-cold" "warm-cold" pair) pair) have havehigh highloadings loadings on the superordinate factors that summarize the trait lexicon. A major product of of this line Une of research was was the concept of implicit implicitpersonality personalitytheory theory(IPT; (IPT; for fora areview, review, see Schneider, 1970). 1970). Bruner and Tagiuri (1954) argued that perception of all kinds goes "beyond the information given" given" in the stimulus array, and depends on the perceiver's expectations and goals as well as general and specific world knowledge retrieved from memory. Thus, in the case of person perception, it is necessary to understand the "naive, implicit theories of personality" that people reason with, in order to understand how they form impressions of others. For Bruner and Tagiuri, IPT comprises the learned relations among various (biosocial) aspects of of personality—relations that might be quite different different from those present in actual, personality-relations IPT to empirical (biophysical) database. Cronbach (1955) expanded the concept of IPT include a list of the important dimensions of personality, estimates of population

AND IlAsm KniLsTROM KIHLSTROMANDHASTIE

718 718

means and variances on each of of these dimensions, and estimates of of the covariances covariances among them. Later, Rosenberg Rosenberg and his colleagues (Kim & Rosenberg, 1980; Rosenberg & Sedlak, 1972) evaluated evaluated a three-factor three-factor semantic diffential diffential model derived from from Os Osgood's (Osgood, Suci, & Tannenbaum, 1957) connotative theory theory of of meaning, and found found that evaluation evaluation (social and intellectual) was the only perceptual perceptual dimension common to all subjects. More recently, Goldberg (1981) and others have proposed proposed that Norman's (1963) "Big Five" model for for the structure of of personality (extraver (extraversion, agreeableness, conscientiousness, emotional stability, and culturedness) is a universally universally applicable structure of of perceived personality-at personaUty—at least so far far as Western, Western, industrialized industrialized cultures cuUures are concerned. A continuing debate concerns the origins of of IPT. The question was initially A extracted the same raised by an early study by Passini and Norman (1966), who extracted factors from from personality ratings of of strangers and of of friends. Since the judges five factors could not have known the actual covariation of of features features in the personaUties personalities of the target people, it seemed seemed likely that the perceived correlations, as reflected reflected in the factor structures, were contaminated contaminated by the judges' expectations and beliefs. This factor argument has been made most forcefully forcefully by Shweder and D'Andrade D'Andrade (1979; D'An D'Anargument drade & «& Shweder, Shweder, 1987; 1987;Shweder, Shweder, 1982), 1982), and and has has been been opposed opposed by by Weiss Weiss and and his his drade Thorne, 1979; Weiss & Mendelsohn, Mendelsohn, colleagues among others (Block, Weiss, & Thome, 1986). Social Perception A. The Ecological Approach to Sodal

Asch (1946) described his stimulus persons in terms of of lists of of traits. This has the advantage of of experimental convenience and mimics the ways in which people de deinteraction (Fiske & Cox, 1979; Peevers & Secord, 1973; scribe each other in social interaction Shweder Shweder & Bourne, 1981). At At the same time, it is clear clear that people people do not really perceive each other as lists of of traits (although much "secondhand" ''secondhand" information information is conveyed conveyed indirectly in third-person third-person descriptions of of a person; Gilovich, 1987). In the case of of direct acquaintances, we perceive the physical characteristics and behaviors behaviors of of other other people. Trait lists may come close to the mental representation representation of of personal personalfar from from the faces, voices, and ity stored in memory, but these abstractions are far of stimulation encountered encountered in the social gestures that make up the actual array of analyses environment. Accordingly, Ittelson and Slack (1958) raised the concern that analyses of person perception perception based on the Asch paradigm are incomplete, if not fundamen fundamenof McArthur and Baron (1983; Baron, 1981; 1981; Baron Baron & & tally misleading. Subsequently, McArthur argued for for an ecological approach to social Boudreau, 1987; Zebrowitz, 1990) have argued perception inspired by the work of of J. J. Gibson (1966, 1979), 1979), who asserted that perception phenomenal experience experience is is the the unmediated, direct perception perception of of stimulation, quaUtaphenomenal unmediated, direct stimulation, qualita tively the the same same as as perceiving perceiving the the pitch pitch of of sound sound or or the the color color of of light. light. Ultimately, Ultimately, tively Gibson proposed proposed that that what what we we really really perceive perceive are are affordances: affordances: the the functional functional utilities utilities Gibson of objects for organisms with certain action capabilities. In other words, we perceive of objects for organisms with certain action capabilities. In other words, we perceive

CHAPTER 17 27 CHAPTER

REPRESENTATIONS OF PERSONS MENTAL REPRESENTATIONS

719

the world in terms of the actions that we can take with respect to the objects in the world. The Gibsonian approach has been applied to various topics in social percep perception. For example, one interpretation of attribution theory is that the environment information needed to render a judgment of causal responsibility responsibility— supplies all the information that when the relevant information information concerning consensus, consistency, and distinc distinctiveness is available, causality "jumps out" at the perceiver (McArthur & Baron, 1983). While it might seem unlikely that specific neuronal structures have evolved to produce attributions of causality to actors, targets, and contexts, human beings do seem to possess some capacity for for picking up the sorts of covariation information information that lies Ues at the core of Kelley's (1967, (1967,1973) 1973) ANOVA model of causal attribution. That this apparently cognitive ability is part of our innate biological endowment fact that the ability to process covariations (or conditional is suggested by the fact probabilities) among environmental events is essential for classical conditioning to occur (Rescorla & Wagner, 1972)-a 1972)—a form form of learning that is accessible to all vertebrate, and many invertebrate, organisms (Razran, 1971). in which perceivers respond to human faces seems to be especially The manner in conducive to analysis in terms of Gibsonian direct realism. For example, Ekman following Darwin, have argued for an innate mechanism for and Friesen (1971), following information from from his or perceiving another person's emotional states by extracting information her face. Similarly, Rosenthal, Hall, DiMatteo, Rogers, and Archer (1979) have summarized the evidence for cross-cultural invariances in the perception of emotion features from extralinguistic verbal cues. Age can be accurately perceived from such features from as the ratio of head to body length, position of eyes with respect to the top of the head, size of eyes and length of nose and ears, and round versus pointed head shape (Shaw & Pittenger, 1977). 1977). Possession of "babyish" features lead adults to be perceived as low in strength and dominance (McArthur, 1982). A broad face or receding hairline increase perceptions of dominance (Keating, Mazur, & Segall, 1981). Global perceptual properties of of the body also seem to support some remark remarkably subtle conclusions about the person perceived. A number of such studies involve adaptations of the point-light technique of Johannsen (1973), in which which target target persons are clothed in black leotards to which point-light stimuli have been attached. When targets are photographed photographed against a black background the resulting stimulus gives no clues to body morphology; yet Kozlowski and Cutting (1977) found found that subjects were able to reliably discriminate between males and females on the basis of gait. It is not yet clear how well the direct or ecological approach to perception will succeed, in either social or nonsocial domains. Still the Gibsonian concern with from sterile ecological validity is pushing researchers in social cognition to move from stimulus materials such as still photographs, trait lists, and verbal descriptions of of behavior to more life-like materials. Even if the direct/realist approach to person perception should prove to be misguided in its rejection of mental structures and

KnnmtOM AND HAsm: KIHLSTROMANDHASIIE

720 720

processes, research within this tradition will lead a better description of the stimulus that is represented by internal mental structures and processes.

IV. IV. PERSON CONCEPTS Forming mental representations of persons and other social stimuli is fundamentally fundamentally a problem of perception, and, as Bruner (1957) noted, every act of perception is an act of of categorization. We naturally sort stimuli into equivalence equivalence classes based of features, attributes, or properties, forming concepts-mental concepts—mental repre repreon similarity of sentations of of categories of of objects. The concepts that we use to guide social percep perception are basic components in the repertoire of social intelligence: they form the of our social world (Cantor & background against which we organize and make sense of 1987,1989; Kihlstrom, 1987, 1989; Kihlstrom & Cantor, 1989; Lingle, Altom, & Medin, 1984). They are also the cognitive basis for consistency in social behavior: people respond in a roughly equivalent fashion to persons, situations, and behaviors which belong to the same category. Thus, we cannot understand a person's social behavior unless we understand the person's repertoire of social concepts and how they are formed, organized, and used. As Kelly (1955, 1963) noted, individual differences differences in social behavior may arise from individual differences in the categories used to construe social objects and events. of social categories. (1) Categories There appear to be at least four major types of of persons persons are labeled by nouns that designate types of people-extraverts people—extraverts and of neurotics, jocks and nerds, preppies and yuppies; there are also categories of social roles, such as parent, lover, teacher, and doctor. (2) Categories of actions are labeled by adjectives that designate qualities of behavior, for example, extraverted, agreeable, conscientious, emotionally stable, intelligent, cultured, or open. (3) Cate Categories of situations situations are labeled by nouns designating the types of situations in which social behavior is displayed, for example, weddings, funerals, seminars, cocktail parties, interviews, and bar-mitzvahs. (4) Viewed as declarative knowledge struc structures, scripts scripts are also concept-like, in that they contain bundles of features that various specific instances of a class of interactions have in common. Although the content of of social categories may differ from one individual or culture to another, the structure of these categories is probably pretty much the same for everyone (for comprehensive coverage, see E. E. Smith & Medin, 1981). Over Over the past two decades, the classical view of category structure, dominant from the time of of Aristotle, has been replaced by a probabilistic or prototype view. This exemplar view of of categorization. As an tum, has been challenged by a new exemplar view, in turn, example of of the difference difference among these views, consider Kant's (1798/1978) fourfold fourfold of temperamental types: phlegtaxonomy of types: melancholic, choleric, sanguine, and phleg matic. Kant described the melancholic individual as anxious, worried, unhappy, suspicious, serious, and thoughtful, and the phlegmatic individual as reasonable, high-principled, controlled, persistent, steadfast, and calm. Thus, under the classical view, all melancholies possess each of certain features in common, and any individual

CHAPTER CHAPTER 27 27

MENIAL MENTAL REPRESENIAll0NS REPRESENTAHONS OF OF PERSONS PERSONS

721 721

who possesses the entire set of features was thereby classified classified as a melancholic as opposed to a phlegmatic type of person. While such a scheme might fairly represent ideal personality types, they might not prove useful useful in the actual business of of classify classifying people. For example, what would we do with Bob, who is anxious and worried but not serious and thoughtful? thoughtful? Or Tom, who is anxious and worried, serious and thoughtful, but also controlled and persistent? Proper-set Proper-set definitions of of categories seem to leave no room for the partial and combined expression of personality for a related, empirically based critique of traditional types (see Achenbach, 1980, for psychodiagnostic categories). Probabilistic approaches classi approaches solve this problem by classifying individuals in terms of the central tendencies of their traits. Bob would be labeled as melancholic if he displayed some critical number of features of of central features of melancholia, even if he did not possess all of them and even if he possessed a few features normally associated with phlegmatics; alternatively, Tom would be labeled as melancholic if the average value of his melancholic traits were higher than that exemplar view, by contrast. of his phlegmatic traits. Under the exemplar contrast, Bob and Tom would exemplify melancholia melancholia or phlegmaticity, be compared to specific individuals who exemplify as opposed opposed to to summary summary prototypes. prototypes. If If Bob Bob resembles Dave, and and Dave Dave has has been been as labeled as as aa melancholic, melancholic, then then Bob Bob will will also also be be classified classified as as aa melancholic; if if Tom Tom labeled does not not resemble resemble any any known known melancholic, melancholic, then then he he will will escape escape this this particular label. label. does appHcation of probabilistic, fuzzy-set An extremely interesting application fuzzy-set approaches to categorization has been in the area of psychiatric diagnosis (Cantor & Genero, 1986; Cantor, Smith, French, & Mezzich, 1980). 1980). Psychiatric diagnoses traditionally 1986; have been construed in terms of the classical view of categorization; a diagnostic defined by the singly necessary and jointly sufficient category must be defined sufficient features that set In contrast, Cantor and her colleagues have argued cogently define a proper set. fuzzy sets sets of features features that are correlated with, that the diagnostic categories are fuzzy sufficient for, category membership. The principal but not singly necessary or jointly sufficient category members, such result of this situation is considerable heterogeneity among category that they are related by family resemblance more than any set of common defining features. This probabilistic point of view was implicitly adopted in the 1987 revision of the Diagnostic and Statistical Manual of Mental Disorders, and is maintained in the fourth edition (American Psychiatric Association, 1994). Certainly the probabilistic view has dominated studies of of social categorization (Lingle et al., aI., 1984). 1984). For example, example. Cantor and her colleagues (Cantor & Genero, 1986; 1982; Genero & 1986; Cantor & & Mischel, 1979; 1979; Cantor, Mischel, & Schwartz, 1982; & Cantor, 1987; 1987; Niedenthal & & Cantor, 1984) performed performed seminal research on the role of feature feature list prototypes in the categorization of persons into types (see also Brown, 1980). Buss and and Craik Craik (1983), (1983), and and John, Hampson, Hampson, and Goldberg (1989) (1989) Hampson (1982), Buss offered similar analyses of the classification classification of have offered of specific behaviors by traits. By of persons and their behaviors and large, this research has shown that our concepts of imperfectly nested, heterogeneous, and probabilistically as fuzzy sets, imperfectly are organized probabilistically summarized by category prototypes. However, there has been little research testing and even less attention given to comparing comparing alternatives within the probabilistic view, and the prototype view with the exemplar view.

KnlLSTROM AND HASm: KIHLSTROMANDHASHE

722 722

Another outstanding issue concerns the nature of of natural categories of of persons. The question arises of whether there are natural categories of of persons that have some degree of universality, regardless of the sociocultural affiliations affiliations of of perceiver or target. Brown (1980) has proposed that social stereotypes based on race, sex, or nationality may serve the function of basic person categories. Consulting word frequency norms to determine which person categories are most frequently used in everyday discourse, and which might therefore represent the most psychologically salient and general categories, he turned up ethnic stereotypes such as Oriental and Jew, kinship terms such as mother and husband, and terms pertaining to professions and social roles (teacher, lieutenant, poet, cook, nurse, culturally specific professions etc.). Furthermore, Brown noted that college campuses, with their rich repertoire of labels relating to socioeconomic status (preppie, yuppie), political stance (hippie, of of residence or voluntary association (Tri-Delt, Skull and eco-freak), and place of stereotype-based category labels. However, most Bones) provide another rich set of stereotype-based of these categories of persons are specific to a particular culture or subculture. of (e.g.. Cantor & & Mischel, 1979) took a more theoretical tack and derived Cantor (e.g., type labels, expressed as noun phrases, from Norman's (1963) "Big Five" structure of personality traits. Thus, extraversion was translated into "PR *TR type" and "comic of joker," conscientiousness into "religious devotee" and "social activist," and so forth. While Cantor's system has the advantage of of a basis in current personality theories, it is not clear that either system captures universal distinctions in disposi dispositions, temperament, emotion, and motivation that are inherent in our categorizations of personality. of Recently, researchers have noted a number of of cracks in the empirical facade of of the probabilistic prototype and exemplar models and have concluded that people carry more than feature lists around in their heads, and that classification classification processes involve more than similarity judgments (Medin, Goldstone, & & Gentner, 1993). Just what this "What more?" more?" comprises is a very open question, but theoreticians believe believe that important concepts are associated with common sense premises which serve as an explanatory theory to account for differential differential weights on features, feature intercorrelations, and the nature of the core features that seem to be the essence of of concepts.

V. PERSON MEMORY v.

Perceptual activity leaves its traces in memory, to be retrieved at a later time and of characteristic features, used to guide action. Concepts, with their prototypes, lists of of abstract and sets of exemplars, are one way to think about the long-term storage of information about persons and personalities. But much of of the social information in our memories is much more concrete. For example, we hold a voluminous store of memories of of particular people and their behaviors—a person of behaviors-a domain known as person memory (for reviews, see Hastie & & Carlston, 1980; Hastie, Park, & & Weber, 1984). memory

27 CHAPTER 27

MENTALREPRESENTATIONS REPRESENTATIONSOF OFPERSONS PERSONS MENTAL

723 723

The manner in which a person remembers and thinks about other people is of central importance in theories of of personality. Many of of the seminal schools of of development and dynamics of of personality put special empha emphathought concerning the development significant others and the manner manner in which others sis on the person's relations with significant remembered (Munroe, 1955). Even "modern" "modem" social social learn learnare conceptualized and remembered theories of of personality personality place place a special special emphasis on role models and vicarious ing theories reinforcement processes processes that depend on the person's memories of learning and reinforcement other people people (Bandura, 1977, 1977,1986; analyses other 1986; Mischel, 1973). In addition, scientific analyses of clinical assessment assessment methods and most types of of psychotherapy psychotherapy require a theoretical of understanding of of the interviewer's, therapist's, and client's understanding client's social perception and memory processes. processes. memory eariiest information information processing theories tended to construe construe knowledge knowledge The earliest represented by a verbal code-lists code—lists of of features or instances associated associated with as represented of objects and events. In concepts or sentence-like propositional descriptions of current descendants of of these early models, concepts are the basic units the current which knowledge structures are built. In the typical representation of from which nodes an experienced event, nodes representing concepts are linked to other nodes of those concepts. Propositions consist consist representing the characteristic features of of higher higher order order networks networks built built of of concept concept nodes nodes that that represent represent the the event event in in of memory. For For example, example, in in J. J. R. Anderson's Anderson's (1983) (1983) HAM H A M and and ACf A C T structure structure memory. system, aa proposition proposition consists consists of of links links between nodes representing representing subject subject and and system, between nodes predicate; the predicate, in turn, turn, consists consists of of links links representing representing relation relation and and predicate; the predicate, in argument. These These terms terms correspond, correspond, roughly, roughly, to to the the subject, subject, verb, verb, and and object object of of argument. seventh grade grade grammar grammar class class diagrams. diagrams. seventh Other theorists have favored favored a dual-coding hypothesis, arguing that knowledge knowledge can be represented in an analogue as well as a propositional format. For some time, lively debate among among psychologists concerning the comparative merits there was a lively of single-code and dual-code general agreement agreement of dual-code theories. However, there is now general that information may be stored in the form of of meaning-based abstract propositions, or as perception-based spatial images (preserving configural configural information), or as temporal strings (preserving order and contiguity information). In what follows, however, we focus focus on verbal representations of of persons and personality. There are various formal formal models of of memory available in the information information processing tradition, all variations on the theme of of associative networks, and all implemented as computer simulations. Similarly, there are several several information information processing processing models of of social social memory memory (Hastie & Carlston, 1980; Hastie et aI., al., 1984). Explicit Explicit proposals, within the information processing framework, framework, have been made made by Hamilton (Hamilton, Katz, & Leirer, 1980), Hastie (1980, 1981, 1988; Hastie & (1980,1981,1988; & Kumar, 1979), Klein and Loftus (1993), Ostrom (Ostrom, Lingle, Pryor, & Geva, 1980; Ostrom, Pryor, & Simpson, 1981), and Wyer and Srull SruU (1989; Wyer, 1974, 1989; Wyer & Carlston, 1979; Wyer & Gordon, 1984), among others. The simplest representational representational structures that have been proposed for for individu individuassociative or semantic networks with unlabeled links, which can store varied als are associative of information, including propositions, images, and emotional responses. An An types of

724 724

Knn.s TROM AND KlHLSTROM AND HAsnE HASTIE

example is given in Figure lA, where a particular person is represented by node P1, PI, two of his or her traits by nodes T1 Tl and 1'2, T2, and six of his or her behaviors by (1980,1981,1988), R. Anderson Anderson & & nodes B1-B6. Hastie (1980, 1981, 1988), J. R. Anderson (1983; (1983; J.J. R. SruU (1981) have been the primary advocates of these simple Hastie, 1974), and Srull network structures. They assume that social memory is organized primarily by are represented represented as as single single nodes nodes in in the the network. network. Events Events are arerepresented represented persons, who are by propositions which describe both the event and the context in which the event Unked to the nodes representing the suboccurred. Nodes representing events are linked sub jects of the propositions describing those events. There may also be episodic and semantic links to other events. In addition to information about specific episodes, nodes representing abstract descriptive information can be linked to the person node. This would occur, for B1 _iB1

A

_»B2 B2 _^B3 B3 P1

_•T1 T1 T2 _^T2 B4 _^B4 B5 _^B5 _^B6 B6

B1

B B

B2 B3 P1

B4 _... . .....�

B5 B6

FIGURE !1 Two Twonetwork network representations representations of ofmemory memory for foraaperson's person's traits traits and and behaviors. behaviors. (A) (A)Trait Trait and and FIGvu behavioral behavioral information information encoded encoded independently. independently. (B) (B) Behavioral Behavioral information information organized organized by by its its trait trait implica implications.

CHAPTER 27 27 CHAmR

MENTALREPRESENTATIONS REPRESENTAHONSOF OFPERSONS PERSONS MENTAL

725 725

example, when the subject forms forms an impression of of a person as talkative or intelligent intelligent based based on observations of of behavior. An A n inferred abstraction abstraction serves serves as a background background for for processing new information information about the person by providing providing category-based category-based expec expectations that are compared compared with the implications implications of of new new experiences. However, However, research has shown that these abstract inferences inferences are stored independent independent of of the events on which they are based, at least for for impressions impressions of of casual acquaintances acquaintances (N. H. Anderson & Hubert, 1963; Hastie Hastie & Park, 1986). Furthermore, abstractions abstractions about about a person like Uke traits affect affect the encoding of of new information information because because compari comparisons are made between new information information and current current salient salient abstractions. If If the information is surprising in the context of of a salient salient trait (or other abstraction), new information Hastie & & Kumar, 1979), Srull SruU (1981), and others have Hastie (1980, 1981, 1984; Hastie receive extra attention and processing, and is likely to be hypothesized that it will receive associatively linked linked to to other other information information about about the the person person (Hastie, (Hastie, 1984). 1984). However, However, associatively this on-line trait effect at encoding does not guarantee that behavioral information this on-line trait effect at encoding does not guarantee that behavioral information will be be clustered clustered in in the the memory memory representation representation "under" "under" the the trait trait attributed attributed to to the the will person (Hastie (Hastie & & Kumar, Kumar, 1979). 1979). person of person memory begins by activating activating the node node corresponding corresponding to Retrieval of along associative associative pathways emanating emanating the person, and then activation spreads out along reaches a node representing representing an event, that event event is from that node. If activation reaches of working working memory. The continues retrieved, that is, designated as part of The process continues fact has been found) process is until task goals are met (e.g., the desired fact found) or the process exhausted (e.g., several attempts to retrieve new information have failed). exhausted several retrieve new information empiriThis model makes some subtle predictions that have been been confirmed confirmed empiri remembered than expected expected cally. For example, surprising acts tend to be better remembered expected and surprising surprising acts will be better acts, and a person who performs both expected expected acts. But, the current model model is remembered than one who performs only expected all-or-none links links between limited in several respects: it utilizes unlabeled, all-or-none between nodes; judgment is characterized anchorsearch is random and undirected; and judgment characterized as a simple anchor averaging process. Hastie's model might might be fruitfully compleand-adjust weighted averaging fruitfully comple mented by another model, currently currently being developed developed by Smith, that also uses the of ACf* ACT* production systems to generate generate inferences inferences based on stored architecture of knowledge implemented as a knowledge (E. R. Smith, 1984). The two models, combined combined and implemented computer computer simulation, would would constitute a significant signiJScant theoretical theoretical advance in social cog cognition. A A closely related structure in the form form of of a hierarchy hierarchy with higher level level "control "control of the remembered remembered person's character has been been elements" associated with traits of proposed by Cantor and Mischel (1979), Hamilton (1989; Hamilton et al., 1980), proposed Hamilton Hamilton aI., Wyer and Gordon Gordon (1984), and others. The guiding precept precept is that the trait terms Wyer prominent in people's spontaneous spontaneous descriptions descriptions of of one one another another (Allport, (AUport, that are so prominent 1937; Fiske & & Cox, 1979; Peevers & & Secord, 1973) have a privileged status in social organize the other types of of information information that memory representations, serving to organize we have stored about a person (Hamilton, Driscoll, & & Worth, 1989). 1989). Such a structure we is illustrated in Figure Figure 1B, IB, where where the the trait trait nodes T l and and 1'2 T2 fan fan out out from from the the person person is illustrated in nodes T1

726 726

KottsnOM AND HAm! KIHLSTROMANDHASIIE

node P1, PI, and representations of events exemplifying these traits (B1-B3, B4-B6) B4-B6) fan out from their respective trait nodes. However, the strong assumption that traits play a central role in memory representations of individuals has met with some opposition. Studies of of the sponta spontaneous self-concept self-concept find that personality traits play a much smaller role in self selfdescriptions than they do in descriptions of other people (McGuire & McGuire, 1988). And, from the beginning (Hastie & Kumar, 1979), many person memory studies have not found that recall memory is "clustered" **clustered" by trait categories. Ostrom and his colleagues (1980, 1981) have proposed a model that relaxes the strong of organization by traits. They hypothesize that memory for events is requirement of usually organized by persons when the participants are familiar to the subject, but that other "themes" such as temporal sequence, situational context, self-reference, self-reference, or group membership dominate memory organization when the focal person is unfamiliar or the social goals of of the perceiver perceiver do not foreground the other person unfamiliar as an individual. SruU (1989) have proposed a "bin" model, aimed at Finally, Wyer and Srull impression-formation experiments, that also does accounting for the results of impression-formation not require that person memories be inevitably organized by traits. They simply combined the two basic models, scrambled associative network and trait organized hierarchy, and claimed that both representations are created, often with duplica duplication of the specific information nodes in a dual representational structure. One of a individual's individuars representation in memory is hypothesized to be a scrambled part of collection of behaviors behaviors associated associated with with aa summary evaluative node, and another another collection of summary evaluative node, and separate part is a hierarchical network of behaviors clustered under trait node separate part is a hierarchical network of behaviors clustered under trait node elements. Long-term person memory is described as a warehouse of contentelements. Long-term person memory is described as a warehouse of content addressable storage storage bins, bins, each each tagged tagged with with the the name object described described by by addressable name of of the the object its contents. Within each bin, the contents are organized according to the its contents. Within each bin, the contents are organized according to the temporal order in which they were experienced; otherwise, there is no inherent of bin contents. organization of Wyer and Srull SruU (1989) have noted that a major weakness of of all of of the person theoretical focus on information memory models derives from the empirical and theoretical formation stimulus ensembles, rather presented in the unnatural form of impression formation than more representative social contexts. In everyday everyday life, our experiences seem Memory represen represento be organized into temporally and causally structured episodes. Memory tations for social information organized into autobiographical episodes have been proposed by by Kolodner Kolodner (1984), (1984), Pennington Pennington and and Hastie Hastie (1986), (1986), Wyer (Fuhrman & proposed Wyer (Fuhrman Wyer, 1988; Wyer, Shoben, Fuhrman, & Bodenhausen, 1985), and others. These Wyer, 1988; Wyer, Shoben, Fuhrman, Bodenhausen, 1985), and others. These structures are hypothesized to be more complex than simple networks in two structures are hypothesized to be more complex than simple networks in two regards: the embedded within-episode components organized to reflect our culturregards: the embedded within-episode components organized to reflect our cultur ally shared expectations about about the the components components of of a a well-formed well-formed episode episode and and the the ally shared expectations links between episodes are labeled, with access to a link only permitted to memory links between episodes are labeled, with access to a link only permitted to memory probes include the the correct correct "key" features to to "open" "open" the the link. link. probes that that include "key" features

27 CHAPTER 27

REPRESENTATIONS OF PERSONS MENTAL REPRESENTATIONS

727 727

VL PROSPECTS FOR THE INFORMATION PROCESSING PROCESSING ApPROACH TO VI. PROSPECTS APPROACH TO

SOCIAL COGNITION five elements common to all A minimal model for social cognition should include five cognitive theories (Hastie & simu & Carlson, 1980): 1980): (a) a vocabulary to describe the simuIus; mem lus; (b) a characterization of the processes by which stimuli are encoded in memory; (c) a description of the encoded mental representation of the stimulus event; (d ma (d)) a characterization of the processes by which encoded representations are manipulated and transformed in the course of memory and judgment tasks; and (e) a the four general approaches vocabulary to describe the response to the stimulus. Of the to social cognition discussed at the outset of this paper, information processing satisfying these requirements. The information processing theory comes closest to satisfying approach is is general general and and itit should shouldapply applyto to the thecognition cognitionof ofalmost almostanything, anything,including including approach oneself. However, However, information information entities in in the the social social world world such such as as other other people people and and oneself. entities processing theory, developed developed in in aa nonsocial laboratory laboratory domain, domain, should should not not be be processing applied uncritically, without modification modification to to the the social social realm realm (Holyoak (Holyoak & & Gordon, Gordon, applied uncritically, without 1984; Ostrom, Ostrom, 1984). 1984; One limit on information processing theory derives from its failure to thor thoroughly address emotional and motivational phenomena in everyday life. There have been preliminary discussions of the manner in which information processing 1981; Clark & & Fiske, 1982; models could handle these "hot" "hot" phenomena (Bower, 1981; 1984; Mandler, 1984; & Collins, 1988; 1988; Simon, 1967), Leventhal, 1984; 1984; Ortony, Clore, & 1967), but there has been relatively little laboratory research to adequately test theoretical models. We believe that the information processing approach will provide the best medium to develop hypotheses about interpersonal goals and other purposeful & Abelson, 1977; 1977; Showers & & Cantor, 1985). 1985). But, again^ again, social behavior (Schank & more remains to be done than has been accomplished. conceptualThe information processing approach has not yet provided a full conceptual ization of the conscious versus unconscious process distinction nor has it done jus justice to phenomena associated with unconscious processes that have been rere vealed through the history of research on personality and psychopathology (but see Kihlstrom, 1987, 1990). There is, of course, a long tradition of research on implicit (subliminal, preconscious) perception and and perceptual defense and and vigilance, but this activity has made little contact with the concerns of social cognition per se. se. Moreover, the analysis of unconscious phenomena in everyday social interaction is still virtually untouched (Bargh, 1994; 1994; Greenwald & & Banaji, 1995; 1995; Kihlstrom, 1994). Another challenge that exceeds current information processing treatments arises from the complexity of the social world. The properties of objects studied in experiments on nonsocial cognition tend to be stable over time and across even contexts. However, people change considerably from situation to situation and even from moment to moment within a single situation (Mischel, 1968, 1973). This

728 728

KniLsTROM AND HAsTIE KIHLSTROMANDHASTIE

means that models that have been developed characterize the formation developed to characterize formation and maintenance of mental representations of of stable (laboratory) entities entities are bound bound to complexity is introduced provide inadequate accounts of of social cognition. A A further further complexity by the fact fact that many of of the entities thought about about in the social world are also sentient, independent, and likely to react to the belief belief that that they are being thought of inferences inferences occurs when when people wonder about. Thus, a complex recursive sequence of what other people are thinking about them and anticipate that other people that the other wondering whether they are thinking about what what the other are wondering other person person is thinking, of social social cognition cognition must take into and so forth. What this means is that theories of representation of of representations representations of of representations representations within a single account the representation mind, where each mind is sensitive to other minds around around it. We do not believe that these difficulties difficulties imply that we should should abandon abandon the information processing approach approach to social cognition. To the contrary, we believe information information processing processing approach approach provides the best for a theory that is the information best hope for theory that complex enough to handle emotion, motivation, the unconscious, and recursive self-conscious inferences inferences about persons and personalities. self-conscious

ACKNOWLEDGMENTS The point of view represented herein is based on research supported in part by Grants MH35856 and MH-44739 from from the National Institute of Mental Health, Grants SES-9122154 and SBR-9308369 from the National Science Foundation, and Subcontract 1122SC from the Program on Conscious and Unconscious Mental Processes of the John D. and Catherine T. MacArthur Foundation, through the University of California, San Francisco. We thank Bamhardt, Nancy Cantor, Lawrence Couture, Marilyn Dabady, Mahzarin Banaji, Terrence Barnhardt, DougMartha Glisky, Stanley Klein, Sheila Mulvaney, Paula Niedenthal, Nancy Pennington, Doug Titcomb, and Betsy Tobias for their comments during the preparation las Tataryn, Allison Titcomb; of this paper. Particular appreciation goes to James Cutting, William Epstein, and Mary Peterson for their help in understanding the implications of Gibson's work in perception for social cognition.

REFERENCES R. P. (1981). (1981). Psychological status of the script concept. American Psychologist, Psychologist, Abelson, R 36, 715-729. 36, Achenbach, T. M. (1980). (1980). DSM-III DSM-III in in light light of of empirical empirical research research on on the the classification classification of of child child American Academy Academy of of Child ChildPsychiatry, Psychiatry,19,19,395-412. 395-412. of the American psychopathology. Journal of Personality:AApsychological psychologicalinterpretation. interpretation. New York: Holt. AUport, G. W. (1937). (1937). Personality: Allport, New York: Holt. Diagnosticand andstatistical statisticalmanual manualofofmental mentaldisor disorAmerican Psychiatric Association. (1994). Diagnostic DC: Author. ders (4th ed.). Washington, DC: R. (1983). The architecture architectureof ofcognition. cognition,Cambridge, Cambridge,MA: MA:Harvard HarvardUniver UniverAnderson, J. R sity Press. R, & Hastie, R Anderson, J. R., R. (1974). (1974). Individuation and reference reference in memory: Proper names and definite descriptions. Cognitive CognitivePsychology, Psychology,6,6,495-514. 495-514.

CHAPTER 17 27

MENTAL MENTALREPRESENTATIONS REPRESENTAHONSOF OFPERSONS PERSONS

729 729

Anderson, N. H. (1965). Averaging versus adding as a stimulus-combination rule in impres impresJournalof of Experimental ExperimentalPsychology, Psychology, 394-400. sion formation. Journal 70,70, 394-400. Anderson, N. H. (1974). Cognitive algebra: Integration theory applied to social attribution. Advancesin inexperimental experimental social psychology (Vol. 1-101). In L. Berkowitz (Ed.), Advances social psychology (Vol. 7, 7, pp.pp. 1-101). New York: Academic Press. Anderson, N. H. (1978). Progress in cognitive algebra. In L. Berkowitz (Ed.), Cognitive theories ininsocial socialpsychology psychology(pp. (pp.103-126). 103-126).New NewYork: York:Academic AcademicPress. Press. theories (19S1). Foundations Foundationsof ofinformation informationintegration integration theory. New York: AcaAnderson, N. H. (1981). theory. New York: Aca demic Press. Effects of concomitant verbal recall on order effects effects Anderson, N. H., & Hubert, S. (1963). Effects Journalof ofVerbal VerbalLearning Learning and Verbal Behavior, in personality impression formation. Journal and Verbal Behavior, 2, 379-391. H. R., R., & &Hammond, Hammond,K. K.RR.(Eds.). (Eds.).(1986). (1986).Judgment Judgment and decision making: An interdisciArkes, H. and decision making: An interdisci reader.New NewYork: York:Cambridge CambridgeUniversity UniversityPress. Press. plinary reader. Asch, S. E. (1946). Forming impressions of personality. Journal of of Abnormal Abnormal and Social Psychology, Psychology,41, 41,258-290. 258-290. Bandura, A. (1977). Social learning learning theory. theory,Englewood Englewood Cliffs, Cliffs, NJ: NJ:Prentice-Hall. Prentice-Hall. Bandura, A. (1986). Social ffoundations oundations of AA social cognitive theory. ofthought thoughtand andaction: action: social cognitive theory, Englewood Cliffs, NJ: Prentice-Hall. Bargh, J. A. (1994). The four horsemen of automaticity: Varieties of automatic influence in social perception and cognition. In J. S. Uleman & J. A. Bargh (Eds.), Unintended Unintended thought (pp. 3-51). New York: Guilford Press. Bargh, J. J. A., A., Bond, Bond, R R. N., N., Lombardi, Lombardi, W., W.,& &Tota, Tota,M. M.E. E.(1986). (1986).The Theadditive additivenature natureofofchronic chronic and temporary and Social Psychology, 50, 50, temporary sources sourcesof ofaccessibility. accessibility.Journal Journalof ofPersonality Personality and Social Psychology, 869-878. of Baron, R R. (1981). Social knowing from from an ecological event perspective: A consideration of the relative domains of power for cognitive and perceptual modes of knowing. In J. Cognition,social socialbehavior, behavior, and environment 61-92). Hillsdale, Harvey (Ed.), Cognition, and thethe environment (pp.(pp. 61-92). Hillsdale, NJ: Erlbaum. R, & Boudreau, L. A. (1987). An ecological perspective on integrating personality Baron, R., and social psychology. Journal Psychology, 53,53, 1222-1228. Journalof ofPersonality Personalityand andSocial Social Psychology, 1222-1228. Block, J., Weiss, D. D. S., S., & & Thome, Thome,A. A. (1979). (1979).How Howrelevant relevantisisaasemantic semanticsimilarity similarityinterpreta interpretation of personality ratings? Journal Psychology, 37,37,1055-1074. 1055-1074. Journalof ofPersonality Personalityand andSocial Social Psychology, Bower, G. H. (1981). Mood and memory. American Psychologist, Psychologist,36, 36,129-148. 129-148. Brehmer, R, judgment: The SJT SJT view. view, Amsterdam: B., & Joyce, C. R R. B. (1988). Human judgment: North-Holland. Brown, R R. (1980). Natural Natural categories categoriesand andbasic basicobjects objectsininthe thedomain domainof ofpersons persons(Katz (KatzNewcomb Lecture). Ann Arbor: University of Michigan, Institute for Social Research. Bruner, J. S. (1957). On perceptual readiness. Psychological PsychologicalReview, Review,64, 64,123-152. 123-152. R. (1954). Person perception. In G. Lindzey (Ed.), Handbook of Bruner, J. S., & Tagiuri, R of psychology (Vol. (Vol. 2, 2, pp. pp. 634-654). 634-654). Reading, Reading, Ma.: Ma.: Addison-Wesley. Addison-Wesley. social psychology Brunswik, E. (1956). Perception design ofpsychological experiments. Perceptionand andthe therepresentative representative design of psychological experiments, Berkeley: University of California Press. Buss, D. M. (1989). Sex differences differences in human mate preferences: Evolutionary hypotheses Behavioraland andBrain BrainSciences, Sciences, 1-49. tested in 37 cultures. Behavioral 12,12, 1-49. The act act frequency frequency approach approach to to personality. personality. Psychological Psychological Buss, D. M., & Craik, K. H. (1983). The Review, 90, 90,105-126. 105-126. Review,

730 730

KnruiTROM AND HAm! KIHLSTROMANDHASIIE

N., & Genero, N. (1986). Psychiatric diagnosis and natural categorization: A close Cantor, N., analogy. In T. Millon & G. Klerman (Eds.), Contemporary Contemporaryissues issuesininpsychopathology psychopathology (pp. 223-256). New York: Guilford Press. Englewood Cliffs, Cantor, N., & & Kihlstrom, J. F. (1987). Personality Personalityand andsocial socialintelligence. intelligence, Englewood Cliffs, NJ: Prentice-Hall. Cantor, N., & & Kihlstrom, J. F. (1989). Social Social intelligence intelligence and and cognitive cognitive assessments assessments of of person person& T. K. Srull SruU (Eds.), Advances in social cognition cognition (Vol. 2, pp. ality. In R. S. Wyer & 1-59). Hillsdale, NJ: Erlbaum. Cantor, N., & Mischel, W. (1979). (1979). Prototypes in person perception. In L. Berkowitz (Ed.), Advances in experimental experimentalsocial socialpsychology psychology(Vol. (Vol. 12, 12,pp. pp.3-52). 3-52).New NewYork: York:Aca Academic Press. Cantor, N., N., Mischel, W., & Schwartz, J. (1982). (1982). A prototype analysis of psychological situa situations. Cognitive CognitivePsychology, Psychology,14,14,45-77. 45-77. & Mezzich, 1. J. (1980). Psychiatric diagnosis as N., Smith, E. E., French, R. DeS., & Cantor, N., of Abnormal Abnormal Psychology, Psychology, 89, 89,181-193. 181-193. prototype categorization. Journal of & Fiske, S. T. (1982). Affect Affect and cognition. cognition. Hillsdale, NJ: Erlbaum. Clark, M. S., & J. (1955). (1955).Processes Processesaffecting affecting scores scoreson on "understanding "understanding of ofothers" others"and and"assumed "assumed Cronbach, L. J. PsychologicalBulletin, Bulletin,52,52,177-193. 177-193. similarity." Psychological D'Andrade, & Shweder, Shweder, R. R. A. A. (1987). (1987). The Thesystematic systematicdistortion distortion hypothesis: A clarificaD' Andrade, R. G., & hypothesis: A clarifica tion tion and update. update. Unpublished manuscript, University of California, San Diego. Ekman, P., & & Friesen, W. V. (1971). Constants across culture in the face and emotion. Journal 17,17, 124-129. Journal of of Personality Personalityand andSocial SocialPsychology, Psychology, 124-129. effect of target familiarity and descriptive purpose Fiske, S. M., & Cox, M. G. (1979). The effect of Personality, Personality,47, 47,136-161. 136-161. on the process of describing others. Journal of &Wyer, Wyer,R. R.S.S.(1988). (1988).Event Eventmemory: memory:Temporal-order Temporal-orderjudgments judgmentsofofpersonal personal Fuhrman, R. W., & Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, 365-384. 54,54, 365-384. life experiences. Journal Genero, N., & Cantor, N. (1987). Exemplar prototypes and clinical diagnosis: Toward a 5, 59-78. cognitive economy. Journal Journalof ofSocial Socialand andClinical ClinicalPsychology, Psychology, 5, 59-78. systems. Boston: Houghton-Mifflin. Gibson, J. J. (1966). The senses sensesconsidered consideredasasperceptual perceptual systems, Boston: Houghton-Mifflin. perception. Boston: Houghton-Mifflin. Gibson, J. J. (1979). The ecological ecologicalapproach approachtotovisual visual perception, Boston: Houghton-Mifflin. Experimental Gilovich, T. (1987). Secondhand information information and social judgment. Journal of of Experimental Social Social Psychology, Psychology,23, 23,59-74. 59-74. Goldberg, L. R. (1981). Language and individual differences: The search for for universals in personality lexicons. In L. Wheeler (Ed.), Review Review of ofpersonality personalityand andsocial socialpsychology psychology (Vol. 2, pp. 141-165). Beverly Hills, CA: CA: Sage. Greenwald, A. G., & Banaji, M. R. (1995). Implicit social cognition: AttitUdes, Attitudes, self-esteem, self-esteem, and stereotypes. Psychological PsychologicalReview, Review,102, 102,4-27. 4-27. Hamilton, D. L. (1989). Understanding impression formation: What has memory research contributed? In P. R. Solomon, G. R. Goethals, C. M. M. Kelleyy, & B. B. R. Stephens Interdisciplinary approaches approaches (pp. 221-242). New York: Springer Springer(Eds.), Memory: Interdisciplinary Verlag. Hamilton, D. L., L., Driscoll, DriscoU,D. D.M., M.,&&Worth, Worth,L.L.T.T.(1989). (1989).Cognitive Cognitiveorganization organizationofofimpressions: impressions: Effects Effects of incongruency in complex representations. Journal of of Personality Personalityand andSocial Social Psychology, 57, 57,925-939. 925-939. Psychology, & Leirer, V. O. (1980). Organizational processes in impression Hamilton, D. L., Katz, L. B., & & D. Carlston formation. In R. Hastie, T. Ostrom, E. Ebbesen, R. Wyer, D. Hamilton, & Personmemory: memory:The The cognitive basis ofsocial perception 121-153). Hillsdale, cognitive basis ofsocial perception (pp.(pp. 121-153). Hillsdale, (Eds.), Person NJ: Erlbaum.

CHAPTER CHAPTER 27 27 MENTAL MENTALREPRESENTATIONS REPRESENTAHONSOF OFPERSONS PERSONS

731 731

R. (1955). Probabilistic functioning and the clinical interview. Psychological Psychological Hammond, K. R Review, 62, Review, 62, 255-262. 255-262. Hampson, S. (1982). The AnAn introduction. London: Routledge & & The construction constructionofofpersonality: personality: introduction. London: Routledge Kegan Paul. Hartshorne, H., & I. Studies in deceit. & May, M. M. A A. (1928). Studies Studies ininthe thenature natureofofcharacter: character: I. Studies in deceit. New York: Macmillan. R. (1980). Memory Memory for for behavioral behavioral information information that that confirms confirms or contradicts contradicts aa personal personalHastie, R R. Hastie, T. M. Ostrom, E. B. B. Ebbesen, R R. S. Wyer, D. L. Hamil Hamility impression. In R Person memory: memory:The Thecognitive cognitive basis of social perception & D. E. Carlston (Eds.), Person ton, & basis of social perception (pp. 155-177). Hillsdale, NJ: Erlbaum. Hastie, R R. (1981). Schematic Schematic principles principles in in human human memory. memory. In In E. E. T. T. Higgins, Higgins, C. C. P. P. Herman, Herman, & & M. P. Zanna (Eds.), Social cognition: The TheOntario OntarioSymposium Symposium(Vol. (Vol. 1,1,pp. pp.39-88). 39-88). Social cognition: Hillsdale, NJ: Erlbaum. R. (1983). Social inference. Annual Annual Review of of Psychology, Psychology, 34, 34,511-542. 511-542. Hastie, R Personalityand andSocial Social R. (1984). of Personality Hastie, R (1984). Causes and effects of causal attribution. Journal of Psychology, Psychology, 46, 46,44-56. 44-56. R. (1986). A primer of information information processing theory for for the political scientist. In Hastie, R R. Lau & D. O. Sears (Eds.), Political Politicalpsychology psychology (pp. (pp. 11-39). 11-39).Hillsdale, Hillsdale,NJ: NJ:Erlbaum. Erlbaum. R R. (1988). A computer simulation model of person memory. Journal Journalof of Experimental Hastie, R Social Psychology, Psychology,24, 24,423-447. 423-447. Social R., & & Carlston, D. (1980). (1980). Theoretical principles in person person memory. In R R. Hastie, Hastie, R, R. S. Wyer, D. L. Hamilton, & D. E. Carlston (Eds.), T. M. Ostrom, E. B. Ebbesen, R Person memory: memory: The The cognitive cognitive basis basis of of social social perception perception (pp. (pp. 1-53). 1-53). Hillsdale, Hillsdale, NJ: NJ: Person Erlbaum. R., & Kumar, P. A A. (1979). Person Person memory: memory: Personality Personality traits traits as as organizing organizing principles principles Hastie, R, Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, 25-38. in memory for behaviors. Journal 37,37, 25-38. R., & Park, B. B. (1986). The relationship between memory and judgment depends Hastie, R, on-Une. Psychological Psychological Review, Review, on whether the judgment task is memory-based or on-line. 93, 258-268. 93, R., Park, B., & Weber, R R. (1984). Social memory. In R. & T. K. Srull SruU (Eds.), Hastie, R, B., & R S. Wyer & Handbook of of social cognition cognition (Vol. 2, pp. 151-212). Hillsdale, NJ: Erlbaum. Handbook PsychologicalReview, Review,51, 51, Heider, F. (1944). Social perception and phenomenal causality. Psychological 358-374. R, & & Simmel, M. (1944). An experimental study of of apparent behavior. American Heider, F., Journalof ofPsychology, Psychology,57, 57,243-259. 243-259. Journal A., & & Mavin, G. H. (1982). Individual Individual construct accessibility and Higgins, E. T., King, G. A, Journal of of Personality Personalityand andSocial SocialPsychology, Psychology, subjective impressions and recall. Journal 43,43, 35-47. Information processing and social cognition. In Holyoak, K. J., & Gordon, P. C. (1984). Information R. S. Wyer & T. K. Srull (Eds.), Handbook Handbook of of social cognition cognition (Vol. 1, pp. 39-70). R Hillsdale, NJ: Erlbaum. & Slack, C. W. (1958). The perception of of persons as visual objects. In R R. Ittelson, W. H., & & L. Petrullo PetruUo (Eds.), Person Personperception perceptionand andinterpersonal interpersonal behavior 210Tagiuri & behavior (pp.(pp. 210228). Stanford, CA: Stanford University Press. (1981). Principles Principles of of psychology psychology (2 (2 vols.). vols.). In In F. F. Burkhardt Burkhardt (Ed.), (Ed.), The The works works of of James, W. (1981). William pub William James. James, Cambridge, MA: Harvard University Press. (Origianl work published 1890)

KniLsTROM AND HArnE KIHLSTROMANDHASTIE

732 732

Johannsen, Johannsen, G. G. (1973). (1973). Visual Visual perception perception of of biological biological motion motion and and aa model model for for its its analysis. analysis. Perception Psychophysics, 14, 14, 201-211 . Perception &<& Psychophysics, 201-211.

John, John, O O.. P., P., Hampson, Hampson, SS.. E., E., & & Goldberg, Goldberg, L L.. R R.. (1989). (1989). Is there there aa basic basiclevel levelof of personality personality description? description?Unpublished Unpublishedmanuscript, manuscript,University Universityof ofCalifornia, California,Berkeley. Berkeley. Jones, E. E. (1979). The rocky road from from acts to dispositions. American Psychologist, Psychologist, 34, 34,

107-117. 107-117. Jones, E. E. E., E., & & Davis, Davis, K. K. E. E. (1965). (1965). From From acts acts to to dispositions: dispositions: The The attribution attribution process process in in Jones, person perception. In L. Berkowitz (Ed.), Advances in experimental experimentalsocial socialpsychology psychology (Vol. 2, pp. 219-266). New York: Academic Press. Kahneman, Kahneman, D., Slovic, Slovic, P., P., & Tversky, Tversky, A. A. (1982). (1982). Judgment Judgment under under uncertainty: uncertainty:Heuristics Heuristicsand and biases. biases. New NewYork: York:Cambridge Cambridge University University Press. Press.

Kant, I. I. (1978). (1978). Anthropology Anthropology from from aa pragmatic pragmaticpoint pointof ofview. view.Carbondale: Carbondale:Southern SouthernIllinois Illinois University Press. Press. (Original work published published 1798) Keating, C. C. F., F., Mazur, Mazur, A., A., & & Segall, Segall, M. M. H. H. (1981). (1981). A A cross-cultural cross-cultural exploration exploration of of physiog physiogKeating, nomic traits traits of of dominance dominance and and happiness. happiness. Ethology and Sociobiology, Sociobiology, 2,2,41-48. 41-48. nomic Kelley, H. H. H. H. (1967). Attribution theory theory in in social social psychology. psychology. In In D. D. Levine Levine (Ed.), (Ed.), Nebraska Kelley, (1967). Attribution Symposium on Motivation (pp. 192-238). Lincoln: University of of Nebraska Press. Kelley, H. H. (1973). The 107-128. The processes processes of of causal causal attribution. attribution. American AmericanPsychologist, Psychologist,28, 28,107-128. Kelly, G. G. (1955). (1955). The psychology of of personal constructs constructs (Vol. (Vol. 1). 1).New New York: York: Norton. Norton. Kelly, Kelly, G. G. (1963). (1963). The psychology psychology of of personal constructs constructs (Vol. (Vol.2). 2).New NewYork: York:Norton. Norton. Kelly, Science,237, 237,1445-1452. 1445-1452. Kihlstrom, J. J. F. F. (1987). (1987). The The cognitive cognitive unconscious. unconscious. Science, Kihlstrom, yearsof ofpsychology: psychology:Essays Essays Kihlstrom, J. J. F. F. (1988). (1988). Personality. Personahty. In In E. E. R. R. Hilgard Hilgard (Ed.), (Ed.), Fifty years Kihlstrom, of Floyd Ruch (pp. (pp. 139-152). 139-152). Glenview, Glenview, IL: IL: Scott, Scott, Foresman. Foresman. in honor of Kihlstrom, J. J. F. F. (1990). (1990). The The psychological psychological unconscious. unconscious. In In L. L. Pervin Pervin (Ed.), (Ed.), Handbook Handbook of of Kihlstrom, personality personality theory and research research(pp. (pp.445-464). 445-464).New NewYork: York:Guilford Guilford Press. Press. Kihlstrom, J. J. F. F. (1994). (1994). Unconscious Unconscious processes processes in in social social interaction. In In S. S. Hameroof, Hameroof, A. A. W. W. for consciousness (pp. 93-104). Kaszniak, & A. A. C. C. Scott Scott (Eds.), (Eds.), Toward Towardaascientific scientificbasis basis for consciousness (pp. 93-104). Cambridge, MA: MIT Press Kihlstrom, J. F., F., & Cantor, N. N. (1989). (1989). Social intelligence and personality: There's room for for growth. In R. S. Wyer & T. K. Srull SruU (Eds.), Advances in social social cognition cognition (Vol. (Vol.2,2,pp. pp. 197-214). Hillsdale, NJ: Erlbaum. Kihlstrom, Kihlstrom, J. F., & Sc Klein, Klein, S.S.B. B. (1994). (1994).The The self self as as aa knowledge knowledge structure. structure.In In R. R. S.S.Wyer Wyer & & T. K. pp. 153-208). Hillsdale, K. Srull SruU(Eds.), (Eds.),Handbook Handbookof ofsocial socialcognition cognition(2nd (2nded., ed..Vol. Vol.1, 1, pp. 153-208). Hillsdale, NJ: Erlbaum. Erlbaum. Kim, M. M. P., P.,& &Rosenberg, Rosenberg,S.S.(1980). (1980).Comparison Comparisonofoftwo twostructural structuralmodels modelsofofimplicit implicitpersonal personality theory. Journal of Personality and andSocial SocialPsychology, Psychology,38, 38,375-389. 375-389. of Personality Klein, S. B., B., & Loftus, J. (1993). (1993). The mental representation representation of of trait and autobiographical autobiographical knowledge knowledge about about the the self. self. In In T. T. K. K. Srull SruU & & R. R. S. S. Wyer, Wyer, Jr. Jr. (Eds.), (Eds.), Advances in social social -50). Hillsdale, NJ: cognition cognition (Vol. 5, pp. 11-50). NJ: Erlbaum. Kolodner, J. J. L. L. (1984). (1984). Retrieval Retrieval and organizational organizational strategies strategies ininconceptual conceptualmemory: memory:AA Kolodner, computer model. model Hillsdale, NJ: Erlbaum. Kozlowski, L. L. T., T., & & Cutting, Cutting, J. J. E. E. (1977). (1977). Recognizing Recognizing the the sex sex of of aa walker walker from from aa dynamic dynamic Kozlowski, point-light display. Perception Perception &&Psychophysics, Psychophysics,21, 21,575-580. 575-580. Leventhal, H. H. (1984). (1984). A A perceptual-motor perceptual-motor theory theory of of emotion. emotion.In In L. L.Berkowitz Berkowitz(Ed.), (^±), Advances Advances Leventhal, experimentalsocial socialpsychology psychology(Vol. (Vol.17, 17,pp. pp.1117-182). NewYork: York:Academic AcademicPress. Press. in experimental 17-182). New Lingle, J. J. H., Altom, Altom, M. M. W., & & Medin, Medin, D. D. L. L. (1984). (1984). Of Of cabbages cabbages and and kings: kings: Assessing Assessing the the extendibility extendibiUty of of natural natural object object concept concept models models to to social social things. things. In In R. S. S. Wyer Wyer & & T. T. K. K. Srull ) Handbook of SruU (Eds. (Eds.), of social socialcognition cognition(Vol. (Vol.1,1,pp. pp.71-117). 71-117).Hillsdale, HiUsdale,NJ: NJ:Erlbaum. Erlbaum. ,

27 CHAPTER 27

MENTALREPRESENTATIONS REPRESENTATIONSOF OFPERSONS PERSONS MENTAL

733 733

Mind and body. body. New York: Norton. Mandler, G. (1984). Mind Markus, H., & Sentis, K. (1984). The self in social information processing. In J. Suls (Ed.), Psychological I, pp. 41-70). Hillsdale, NJ: Erlbaum. Psychological perspectives perspectives on the self self (Vol. {Vol 1, McArthur, L. Z. (1982). Judging a book by its cover: A cognitive analysis of the relationship between physical appearance and stereotyping. In A. Hastorf Hastorf & A. Isen (Eds.), Cogni Cognisocial psychology psychology (pp. 149-211). New York: Elsevier. tive social McArthur, L. Z., & Baron, R. (1983). Toward an ecological theory of of social perception. Psychological Review, Review, 80, 252-283. Psychological McGuire, W. J., & McGuire, C. V. (1988). Content and process in the experience of the Advances in experimental experimental social social psychology psychology (Vol. 21, pp. self. In L. Berkowitz (Ed.), Advances 97-144). Orlando, FL: Academic Press. Medin, D. L., Goldstone, R. L., & Gentner, D. (1993). Respects for similarity. Psychological Psychological 100, 254-278. Review, JOO, (1968). Personality Personality and assessment. New York: Wiley. Mischel, W. (1%8). Mischel, W. (1973). Toward a cognitive social leaming learning reconceptualization reconceptualization of of personality. Psychological Psychological Review, Review, 80, 252-283. Schools of of psychoanalytic psychoanalytic thought. New York: Hole, Rinehart & Munroe, R. L. (1955). Schools Winston. Niedenthal, P. M., & Cantor, N. (1984). Making use of social prototypes: From fuzzy concepts to firm decisions. Fuzzy Fuzzy Sets and Systems, 14, 5-27. 5-27. trait construct in lay and professional psychology. In L. Festinger (Ed.), Nisbett, R. (1980). The trait Retrospections on social social psychology psychology (pp. 109-130). New New York: Oxford University Press. Retrospections and shortcomings social judg judgshortcomings in social Nisbett, R., & Ross, L. (1980). Human inference: Strategies and ment. Englewood Cliffs, NJ: Prentice-Hall. of person attributes: Replicated factor Norman, W. (1963). Toward an adequate taxonomy of of Abnormal and Social Abnormal and structures in peer nomination personality ratings. Journal of Psychology, Psychology, 66, 66, 574-583. Ortony, A., Clore, G. L., & Collins, A. (1988). The cognitive cognitive structure of of emotions. emotions. New York: Cambridge University Press. measurement of of meaning. meaning. Osgood, C. E., Suci, G. J., & Tannenbaum, P. H. (1957). The measurement Urbana: University of Illinois Press. Ostrom, T. M. (1984). The sovereignty of social cognition. In R. S. Wyer & T. K. SruU Srull Handbook of of social social cognition cognition (Vol. I1,, pp. 1-38). (Eds.), Handbook 1-38). Hillsdale, NJ: Erlbaum. Ostrom, T. M., Lingle, J. H., Pryor, J. B., & Geva, N. (1980). Cognitive organization of person impressions. In R. Hastie, T. M. Ostrom, E. B. Ebbesen, R. S. Wyer, D. L. Hamilton, & D. E. Carlston (Eds.), Person memory: memory: The cognitive cognitive basis of of social social perception perception (pp. 55-88). Hillsdale, NJ: Erlbaum. Ostrom, T. M., Pryor, J. B., & Simpson, D. D. (1981). The organization of of social information. In E. T. Higgins, C. P. Herman, & M. P. Zanna (Eds.), Social Social cognition: The Ontario Ontario Symposium Symposium (pp. 3-38). Hillsdale, NJ: Erlbaum. A universal conception of of personality structure? Passini, F. T., & Norman, W. T. (1966). A Journal Journal of of Personality Personality and Social Social Psychology, Psychology, 4, 44-49. of descriptive concepts Peevers, B., & Secord, P. (1973). Developmental changes in attribution of Journal of of Personality Personality and Social Social Psychology, Psychology, 27, 120-128. to persons. Journal Pennington, N., & Hastie, R. (1986). Evidence evaluation in complex decision making. Journal of of Personality Personality and Social Social Psychology, Psychology, 51, 242-258. Journal clinical study study of of social social behavior. AppletonPeterson, D. R. (1968). The clinical behavior. New York: Appleton Century-Crofts.

KmLSTROMANOHASIlE Knn.snt OM AND HAsTIE

734 734

Razran, G. (1971). Mind in evolution. evolution, Boston: Boston:Houghton-Mifflin. Houghton-Mifflin. Razran, Mind in Rescorla, R R. A., A., & & Wagner, Wagner, A. A. R R. (1972). A A theory theory of of Pavlovian Pavlovian conditioning: conditioning: Variations Variations Rescorla, in the the effectiveness effectiveness of of reinforcement reinforcement and and non-reinforcement. non-reinforcement. In In A. A. H. H. Black Black & W. W. F. F. in Prokasy (Eds.), Classical Classicalconditioning conditioningIIII(pp. (pp.64-99). 64-99).New NewYork: York:Appleton-Century Appleton-CenturyCrofts. Crofts. Rosenberg, S., Nelson, c., C, & & Vivekananthan, P. S. (1968). A multidimensional approach to the structure structure of of personality personality impressions. impressions. Journal Journalof ofPersonality Personalityand and Social Psychology, Social Psychology, the 9, 283-294. 9,

Rosenberg, S., & Sedlak, Sedlak, A. A. (1972). Structural Structural representations representations of of implicit implicit personality personality theory. theory. Rosenberg, In L. Berkowitz Berkowitz (Ed.), (Ed.), Advances in in experimental experimentalsocial socialpsychology psychology(Vol. (Vol.6,6,pp. pp.235235In 297). New New York: York: Academic Academic Press. 297). Rosenthal, R., J. A., DiMatteo, M. M. R., Rogers, P. P. L., & Archer, D. D. (1979). Sensitivity Sensitivity R, Hall, J. R, Rogers, Rosenthal, nonverbal communication. communication.Baltimore: Baltimore: Johns HopkinsUniversity UniversityPress. Press. to nonverbal Johns Hopkins Ross, L. D. The intuitive intuitive psychologist psychologist and and his his shortcomings. shortcomings. In Ross, D. (1977). The In L. Berkowitz Berkowitz (Ed.), (Ed.), Advances experimentalsocial socialpsychology psychology(Vol. (Vol.10, 10,pp. pp.173-220). 173-220).New NewYork: York:Aca AcaAdvances in experimental demic demic Press. Press. Sarbin, T. T. R R. (1954). (1954). Role Role theory. theory. In In G. Lindzey Lindzey (Ed.), (Ed.), Handbook Handbookof ofsocial socialpsychology psychology(Vol. (Vol. Sarbin, 1, pp. pp. 223-258). Cambridge, Cambridge, MA: MA: Addison-Wesley. Addison-Wesley. social context context of of conduct. conduct New New York: York:Praeger. Praeger. Sarbin, T. T. R R. (1982). The social Sarbin, Schank, R. c., C, & & Abelson, Abelson, R R. P. P. (1977). Scripts, Scripts,plans, plans,goals, goals,and andunderstanding. understanding. Hillsdale, Schank, R Hillsdale, NJ: Erlbaum. Erlbaum. NJ: Schneider, D. (1970). Implicit personality theory: A review. Psychological Psychological Bulletin, Bulletin,79,79, 294-309.

Pittenger, J. J. (1977). (1977). Perceiving Perceiving the theface face of of change changein in changing changingfaces: faces:Implications Implications Shaw, R, R., & Pittenger, Shaw, J. Bransford Bransford (Eds.), (Eds.), Perceiving, Perceiving,acting, acting, for aa theory theory of of object object perception. perception. In In R R. Shaw Shaw & J. for knowing: psychology (pp. 103-132). Hillsdale, NJ:NJ: Erlbaum. knowing:Toward Towardananecological ecological psychology (pp. 103-132). Hillsdale, Erlbaum. Showers, C. C. J., J., & & Cantor, Cantor, N. N. (1985). (1985). Social Social cognition: cognition:A A look lookat atmotivated motivatedstrategies. strategies.Annual Annual Showers, Review Review of of Psychology, Psychology,36, 36,275-305. 275-305.

Shweder, R R. A. (1982). (1982).Fact Factand and artifact artifactin intrait traitperception: perception:The Thesystematic systematicdistortion distortionhypothe hypothesis. research (Vol. 11, pp. (Ed.), Progress sis. In B. A. Maher (Ed.), Progressininexperimental experimentalpersonality personality research (Vol. 11, pp. 65-100). New York: Academic Press. Shweder, R R. A., & Bourne, E. J. (1981). Does the concept conceptof ofthe theperson personvary varycross-culturally? cross-culturally? (Cognitive Science Tech. Rep. No. No. 6). 6). Ann AnnArbor Arborand and Chicago: Chicago:University Universityof ofMichigan Michigan and University of Chicago. Shweder, R R. A., & & D'Andrade, R R. G. (1979). Accurate reflection or systematic distortion? A reply to Block, Weiss, and Thome. Journal Psychology, Journalof of Personality Personalityand andSocial Social Psychology,

37, 37, 1075-1084. Simon, H. H. A. A. (1967). (1967). Motivational Motivationaland and emotional emotionalcontrols controlsof ofcognition. cognition.Psychological Psychological Review, Review, 74, 29-39. Siovic, Slovic, P., & & Lichtenstein, S. (1971). Comparison of Bayesian and regression approaches to OrganizationalBehavior Behaviorand andHuman Human the study of information processing in judgment. Organizational Performance, Performance,6,6,649-744. 649-744. Smith, E. E., & «fe Medin, D. L. L. (1981). Categories Categoriesand andconcepts. concepts.Cambridge, Cambridge,MA: MA:Harvard Harvard Smith, University Press. 91,91,392-413. 392-413. Smith, E. E. R R. (1984). (1984). A Amodel modelof ofsocial socialinference inference processes. processes.Psychological PsychologicalReview, Review, SruU, T. K. (1981). Person memory: memory: Some tests of associative storage and retrieval models. Srull, Journal Human Learning andand Memory, 7, 440-463. Journalof ofExperimental ExperimentalPsychology: Psychology: Human Learning Memory, 7, 440-463. Tulving, E. E. (1983). Elementsof ofepisodic episodicmemory. memory.Oxford: Oxford:Oxford OxfordUniversity UniversityPress. Press. Tulving, (1983). Elements

27 CHAPTER '1.7

MENTAL REPRESENTATIONS OF PERSONS

735 735

D. (1974). Judgment under uncertainty: Heuristics and biases. Tversky, A., & Kahneman, D. Science, 185, 185, 1124-1131. 1124-1131. Science, of the implausibility Weiss, D. S., & Mendelsohn, G. A. (1986). An empirical demonstration of of the semantic similarity explanation of how trait ratings are made and what they Personality and andSocial SocialPsychology, Psychology,50, 50,595-60l. 595-601. of Personality mean. Journal of Winograd, T. (1975). (1975). Frame representations and the procedural-declarative controversy. In Representationand andunderstanding: understanding: Studies in cognitive D. Bobrow & A. Collins (Eds.), Representation Studies in cognitive science (pp. (pp. 185-210). 185-210). New New York: York: Academic Academic Press. Press. science Wishner, 1. J. (1960). Reanalysis of "Impressions of personality." Psychological PsychologicalReview, Review,67, 67, 96-112. Cognitiveorganization organizationand andchange: change: information-processing approach. Wyer, R. S. (1974). Cognitive AnAninformation-processing approach. Potomac, MD: MD: Erlbaum. Wyer, R. S. (1989). Social memory and social judgment. In P. R. Solomon, G. R. Goethals, B. R. Stephens (Eds.), Memory: Interdisciplinary Interdisciplinaryapproaches approaches(pp. (pp. C. M. Kelley, & B. (Eds.), Memory: 221-242). New York: Springer-Verlag. cognition,inference, inference,and and attribution, Hillsdale, Wyer, R. S., & Carlston, D. E. (1979). Social cognition, attribution. Hillsdale, NJ: Erlbaum. Wyer, R. S., & Gordon, S. E. (1984). The cognitive representation of social information. In of social cognition cognition (Vol. (Vol.2,2,pp. pp.73-150). 73-150). (Eds.), Handbook of R. S. Wyer & T. K. Srull (Eds.), Hillsdale, NJ: Erlbaum. Wyer, R. S., Jr., Shoben, E. J., Fuhrman, R. W., & Bodenhausen, G. V. (1985). Event Journalof of Personality Personality memory: The temporal organization of social action sequences. Journal Psychology,49, 49,857-877. 857-877. and Social Social Psychology, context. Hillsdale, cognition in its its social social context. Wyer, R. S., & Srull, T. K. (1989). (1989). Memory and cognition NJ: Erlbaum. perception.Pacific PacificGrove, Grove,CA: CA:Brooks/Cole. Brooks/Cole. Zebrowitz, L. A. (1990). Social perception.

This Page Intentionally Left Blank

28 CHAPTER 28

PERSONALITY STRUCTURE THE RETURN RETURN OF OF THE THE BIG BIG FIVE FIVE THE

S. WIGGINS AND PAUL D. D . TRAPNELL TRAPNELL JERRY S. UNIVERSITY OF OF BRITISH BRITISH COLUMBIA COLUMBIA UNIVERSITY

I. INTRODUCTION INTRODUCTION I.

five-factor model of personality The recent resurgence of interest in the venerablefive-factor characteristics appears to reflect a "working "working consensus" consensus"among amongaasubstantial substantialnumber number of investigators on on the the primary importance of the the dimensions of: of: (I) (I) Surgency/ Extraversion, (II) (II) Agreeableness, Agreeableness, (III) (III) Conscientiousness, Conscientiousness, (IV) (IV) Neuroticism, Neuroticism, and and the to Experience/Intellect. These dimensions have recently been the (V) Openness to principal focus of Annual Review Reviewof ofPsychology Psychologychapters chapters(e.g., (e.g.,Digman, Digman,1990; 1990; Wig Wig1991; McCrae, 1992), gins & & Pincus, 1992), special issues of journals (e.g., Costa, 1991; 1993; Wiggins, 1996), and scores of articles edited books (e.g., Costa & & Widiger, 1993; in personality, clinical, and social psychology journals. In the present chapter, we we will focus on earlier writers who have contributed, directly or indirectly, to the five fiveon current factor tradition and on current writers writers who who have have been been associated associated with with distinctive distinctive five-factor model. theoretical perspectives on thefive-factor his historical review of the vicissitudes of personality research methods, In his identified three trajectories of development: continuous, arrested, and Craik (1986) identified the last describing those methods that declined in prominence for aa interrupted, the period of time and reemerged at a later date. If a similar type of historical historical analysis were made of the "Big Five" dimensions of personality, it would reveal several "interruptions" over time. These interruptions reflect, in part, the ambiguities of of publication dates, the relative unavailability of unpublished and technical reports, and model of and other of five-factor model the five-factor in the interest in that interest say that to say fair to is fair it is But it artifacts. But other artifacts. ,_

_

„

PSYCHOLOGY PERSONALITY PSYCHOLOGY OF PERSONALITY ILtNDBOOK OF HANDBOOK

_^_

737 737

PlEss. ACADEMIC PRESS. IY COMIGIIT 0C1 91997 COPYRIGHT 97 B Y ACADEMIC

Au liGHTS OF IEPRODUcnoN IN ANY fORM IESEIVEO. ALL Rionnj OF REPRODUCHON IN ANY FORM RESERVED.

738 738

WIGGINS AND AND TRAPNELL

TABLE II TABLE Selected Citations from from the History History of of the the Five-Factor Five-Factor Model Model of of Personality Personality Selected Decade Decade

Citations Citations

1930s 19405 1940s 1950s 1960s

Odbert (1936) Thurstone (1934); Allport AUport and Odbert Cattell (1943,1945); (1943. 1945); Eysenck Guilford (1948); Fiske (1949) Cattell Eysenck (1947); Guilford Guilford (1959) Cronbach and Meehl (1955); Loevinger (1957); Cattell (1957); Guilford (1963. 1967); Borgatta Tupes and Christal (1961); Tupes and Kaplan (1961); Norman (1963,1967); (1964) Hofstee (1976); Goldberg Adcock (1972); Eysenck (1972); Howarth (1976); Cattell (1973); Hofstee (1977); Tomas (1977); Brokken (1978); Digman (1979) Borkenau (1988); Botwin and Buss (1989); Conley (1985); McCrae and and Costa (1985c); Borkenau Digman and Takemoto-Chock Takemoto-Chock (1981); Goldberg Goldberg (1980,1981,1982); (1980. 1981. 1982); R. Hogan Hogan (1986); John, Goldberg, and Angleitner Goldberg (1989); Trapnell Trapnell and Angleitner (1984); Peabody Peabody and Goldberg Wiggins (1990)

1970s 19805 1980s

personality has waxed waxed and waned over over the past five for aa variety variety of of reasons, reasons, five decades for and that we are currently witnessing a waxing that seems Hkely likely to continue well into the future. A. The 19305 A. The 1930s

In Table I, we have selected a few, from among many, citations that are meant to highlight significant issues that have occured (and reoccured) in the evolution of of the contemporyfive-factor Thefirst citation isisto toaapaper paperby byL.L.L.L.Thurstone Thurstone five-factor model. model. The first citation (1934), to which Goldberg (1993) has recently called attention because of its prepre five-factor model. We Wecite citeThurstone Thurstonehere here science with respect to the contemporaryfive-factor as a seminal source of the multiple-factor multiple-factor methods which have played such an important role in multivariate models of personality traits. Thurstone (1934) also describes a study in which raters were provided with a list of 60 trait trait adjectives adjectives "in **in common use for describing people" and in which each rater was asked to indicate in adjectives that might be used used in in describing describing someone someone he he knew knewwell. well.Thurstone Thurstonefound found five common factors sufficed in accounting for the intercorrelations among that five sufficed adjectives and he reached the optimistic conclusion that "the ''the scientific description of personality might not be so hopelessly complex as it is sometimes thought to parsi be" (p. (p. 14). 14). This may be the earliest example of Big Five enthusiasm for for parsimony. mony. Thurstone realized that his method of factor factor analysis was indeterminant in the sense that it did not identify a unique set of interpretable orthogonal axes.^ axes.1 In part, for this reason, he advocated the identification of clusters of synonyThe issue issue of of factor factor indeterminancy indeterminancy was was subsequently subsequently to to have have an an interesting interesting "interrupted" "interrupted" 1^The development in the history of psychometrics psychometrics (see Steiger, 1979).

28 RETURN RETURNOF OFmE THE BIG FIVE CHAPTER 28 BIG FIvE

739739

of the sphere formed by the five basic factors, as mous adjectives on the surface of traithe put it, "whatever be their nature." Since there are obviously hundreds of trait descriptive terms "in common use," such an enterprise would benefit from a lexicon of trait-descriptive terms that would permit investigators to select and sample terms AUport and Odbert (1936) assembled such a lexicon by on a systematic basis. Allport New International International exhaustively examining the trait-descriptive terms in Webster's New Dictionary (1925) and classifying classifying them under the categories of personal traits, Dictionary temporary states, social evaluations, and metaphorical terms. With reference to observed. Thurstone's (1934) rating study, they observed,

Theoretically it would be possible to apply this ingenious method to a complete list of trait-names, such as that contained in this monograph. One might deter determine the amount of overlap in meaning between all the terms as they are commonly understood and employed. The investigator might then declare that such and such trait-names are roughly synonymous and that only one of them needs to be retained if what is desired is is aavocabulary vocabulary of ofcompletely completely independent independent terms. The trait-names would would be be grouped, grouped,and andonly onlyaasingle singlerepresentative representativewould would be saved for each group. (Allport (AUport&d Odbert, Odbert,1936, 1936,p.p.33). 33).

It was, of course, Cattell (1943) who responded to this suggestion in the forties. Although related, it is important to distinguish between the tradition begun AUport and Odbert. Thurstone's concern was with by Thurstone and that begun by Allport of mind" which, in the development of of multivariate models for capturing "vectors of of personality, would reveal rtveal personality personality structure as manifested in individ individthe realm of ual differences. Historically, this tradition has been characterized by disagreements of technical issues, such as the number of of factors to retain, the on a number of of factor rotation, and the nature of of criteria for establishing appropriate method of differences of opinion are evident in the contrasting the validity of factors. Such differences of Cattell and Eysenck, for example. systems of The tradition begun by Allport AUport and Odbert has been concerned with the development development of of taxonomies taxonomies of of personality personality attributes as reflected, for the most part, taxoin ordinary language (John, Angleitner, & Ostendorf, 1988). In reviewing the taxo of trait-descriptive terms developed developed over the past 50 years, John, Goldberg, nomies of differ in: (a) sampling procedures, and Angleitner (1984) note that such taxonomies differ of reduction, (c) degree of structure, (d) abstractness, and (e) criteria for (b) extent of evaluation. Allport AUport and Odbert, for example, considered all terms in an unabridged Ust, and listed terms alphabetically dictionary, made no attempt to reduce their list, within four broad categories. Possible hierarchies were not considered and no attempt was made to evaluate the final lexicon. Goldberg (1982), in contrast, began Ust from a previous dictionary search, reduced this list considerably, considerably, with a reduced list structured the the taxonomy taxonomy according according to to strict strict rules, rules, generated generated aa large large number number of of structured categories which which mapped mapped onto onto higher-order higher-order dimensions, dimensions, and and evaluated evaluated the the final categories final taxonomy by by several several different different criteria. criteria. taxonomy

740 740

WIGGINS WIGGINS AND AND TRAPNELL TRAPNELL B. B.

The 19408 The 1940s

Taxonomies of trait-descriptive terms may be employed for a variety of purposes by lexicographers, psycholinguists, cognitive psychologists, and other students of of language. They may also serve as a basis for the construction of instruments in the development of multivariate models of personality structure, as has been so ably demonstrated by Cattell. Cattell (1943) considered the AUport-Odbert Allport-Odbert lexicon to be a useful definition ofthe of judgments of of of the "language personality sphere." On the basis of semantic similarity, he reduced the lexicon of approximately 4500 terms to 171 171 synonym groups. Bipolar rating scales were constructed to represent the synonym groups, and from from their intercorrelations in a peer-rating study, 35 clusters were identified identified as the "standard ''standard reduced personality sphere." In subsequent factor factor ana analytic studies, Cattell (1945) identified identified 12 12 primary factors factors underlying the 35 clusters. Project Cattell was a consultant to the Michigan VA Selection Research Project (E. for the selection of graduate (E. L. Kelly & Fiske, 1951), in which procedures for assessment" students in clinical psychology were investigated. The "basic data of assessment" in this project were a set of 22 22 bipolar rating scales selected from from Cattell's 35 standard clusters. The rating scales were used to obtain self-ratings, teammate staff assessment ratings for for all trainees. This design allowed Fiske (1949) ratings, and staff perform the first study of the consistency of primary factor factor structures among to perform from different different sources. Two important findings emerged from from his study: ratings from found (1) "a high degree of consistency exist[ed] between the factorial structures found different rating sources" (p. 344), and (2) (2) the data strongly suggested a 5 (not in different 12) factor factor solution. During this same decade the two major major taxonomic research Guilford and Hans J. Eysenck were beginning to take form. programs of J. P. Guilford Guilford (1948) was instrumental in fostering a more widespread understanding of Guilford of factor analytic approaches to test development and in applying these methods to factor two-factor structural the study of traits. Eysenck's (1947) first description of his two-factor form. model also appeared at this time, although in inchoate form.

C. C. The The 19508 1950s The 1950s was a period of interrupted development in the history of the Big Five model and it was not until the end of that decade that the importance of Fiske's was recognized. Although there was was little research of direct relevance demonstration was to the five-factor model, there were two lines of development in personality assessassess ment that laid the groundwork for a reemergence of the model in later years. The philosophy of science underlying the idea of construct validity (Cronbach & Meehl, 1955; Loevinger, 1957) encouraged a more substantive approach to test construction 1955; than did the widely held empirical perspective associated with the MMPI, and this substantive approach was compatible with dimensional assessment and theories of of personality structure. The fifties was also a period in which factor analytic ap apGuilford (1959) proaches to personality structure flourished. Cattell (1957) and Guilford

CHAPTER 28 28 CHAFfER

RETURNOF OFTIlE THEBIG BIGFIVE FIVE RETuRN

741 741

presented major major reviews of of their their structural research programs and numerous studies presented of the 16 PF and GZTS GZTS appeared. appeared. of

D. The 19608 1960s Ernest Tupes was a member member of of the previously mentioned Michigan V VA Assessment Ernest A Assessment staff, and and in his subsequent subsequent work at Lackland Air Air Force Base he used 20 of of the 22 from the earlier project, in the assessment of of Air Air Force Acad Acadbipolar rating scales, from factor structure of of peer ratings on these 20 variables, emy cadets. In investigating the factor Tupes and Christal (1958) found found a clear and generalizable five-factor five-factor solution which Tupes Surgency/Extraversion, Agreeableness, Conscientiousness, Emo Emothey identified identified as SurgencylExtraversion, tional Stability, and Culture. Moreover, both the individual rating scales and the factor scores were found found to be related to important important criterion vari varifive personality factor Effectiveness Reports (CER), which are ratings of of leadership leadership ables, such as Cadet Effectiveness ability and officer officer potential based on a composite composite of of peer ratings, upperclassmen upperclassmen ability ratings, and ratings by tactical officers. For example, the correlations between factor scores and CERs in three classes of of cadets were as follows: Surgency (.24), factor Agreeableness (.35), Conscientiousness (.60), Emotional Stability Stability (.58), (.58), and and Culture Culture Agreeableness (.35), Conscientiousness Emotional (.53) (Tupes & Kaplan, Kaplan, 1961, 1961, p. 7). (.53) (Tupes p. 7). In what what we would classify classify as the first clear "advocacy" paper, Tupes and investigated the "universal" nature of of the five-factor five-factor solution by Christal (1961) investigated direct comparisons of of factorial factorial results from from eight highly diverse samples of of subjects. direct of these samples were military, miUtary, two were Cattell's university student samples, Four of from the Kelly-Fiske study of graduate students. The trait-rating trait-rating and two were from from Cattell's 35 clusters. The stability variables in all samples had been derived from five-factor solution across diverse samples and conditions was "remarkable": "remarkable": of the five-factor

In many ways it seems remarkable that such stability should be found in an area which to date has granted anything but consistent results. Undoubtedly the consistency has always been there, but it has been hidden by inconsistency of of factorial techniques and philosophies, the lack of replication using identical factor titles. titles. None None of of the the factors factors variables, and disagreement among analysts as to factor identified in this study are new. They They have have been identified many many times times in in previous previous identified analyses, although they have not always been called by the same names. & Christal, Christal,1961, 1961,p.p.12) 12) (Tupes &

No doubt, the impact of of Tupes and Christal's advocacy on the psychometric psychometric community of of the early sixties might have been limited by the fact fact that all of of their form of of technical reports and technical notes issued from from the reports were in the form Laboratory at Lackland AFB and circulated circulated to a distinguished, but small, Personnel Laboratory of civilians. If If it had not been for for a published article by Warren Warren Norman, a group of distinguished civilian whose research research was supported supported by the Personnel Laboratory, Laboratory, distinguished development there might well have been another "interruption" in the historical development of the Big Five. of

742 742

WIGGINS WIGGINS AND AND TRAPNEU TRAPNELL

Norman's (1963) paper is a milestone in the historical development of the five-factor five-factor model because it provided: (1) a clear statement of the rationale and procedures for for developing a well-structured taxonomy of personality attributes; (2) psychometric criticisms of Cattell's earlier factorial work and an advocacy of of analytic orthogonal rotation; (3) an analytic comparison of the generalizability of of the five-factor solution within an appropriate experimental design; and (4) a call self-report measures of the five factors, using peer ratings for the development of self-report as criteria. Although clearly a five-factor enthusiast, Norman shared the skepticism of of earlier investigators regarding the sufficiency sufficiency of these factors and felt felt that "it *'it is time to return to the total pool of trait names in the natural language-there language—there to search for for additional personality indicators not easily subsumed under one or an an582). Acting on his own recommendations, of these five recurrent factors" (p. (p. 582). other of Norman returned Webster's Third ThirdNew NewInternational InternationalDiction Dictionreturned to the newly available Webster's ary (1961) and developed a new master set of 18,125 terms which was rigorously reduced and classified classified (Norman, 1967) to provide the basis for most contemporary taxonomies of personality attributes, including those developed by Goldberg and his collaborators (see John et aI., al., 1988, 1988, pp. 184-189).

E. The 19708 1970s Norman's contributions to the five-factor tradition led some to characterize the model as the "Norman Five" (e.g., Bouchard, Lalonde, & Gagnon, 1988). 1988). From a strictly historical perspective, one might be tempted to characterize the model as the "Cattell Five," but that would be a serious faux pas. For despite its origins in fivethe clusters and rating scales that he developed, Cattell has never accepted the five factor factor alternative to his 20 or more primary factors, nor has he been willing to useful approximations to five of his nine or more second accept the Big Five as useful strata factors (Cattell, 1973). 1973). Although Cattell's reasons for five for disowning the fivefactor factor model may appear appear more methodological than substantive, one should bear content that exists in Cattell's theory in mind the close relation between method and content of personality. More than any other theorist, Cattell relies upon specific methods of factor factor analytic investigation investigation in his quest for for the underlying determinants of of of phenotypic variables (Wiggins, 1984). Cattell's (1973) objections to five-factor solutions, and to other solutions of of underrelatively small dimensionality, are that: (1) correlation matrices have been under factored; (2) factors have been mechanically, usually orthogonally, rotated through clusters of surface variables ("cluster ("cluster chasing"); and (3) the resultant factors are often "pseudo second-order factors," rather than "true" or "grounded" second often stratum, which which isis aa general general psycho psychostratum factors: "We can distinguish between a stratum, factor, and an order, order, logical statement of the breadth and manner of influence of a factor, which is where it appears in a factor-analytic experiment" (pp. 132-133). Cattell's concern is is that that factors factors which which are, are, in in truth, truth, secondary secondary stratum stratum factors factors may may be be mistak mistakconcern level. This This concern concern enly correlated correlated into into aa set set of of primaries primaries and and appear appear at at the the primary primary level. enly

CHAPTER CHAPTER 28 28

FIVE RETURN OFTOE mE BIG FIVE

743 743

reflects his conceptual preference preference for primary factors: "the evidence is that the primaries... scientifically more important and constant and that the secondar secondarprimaries . . . are scientifically ies are more elusive and above all related to the accidental circumstances of the particular group" (p. 107). Cattell's conceptual and methodological preferences are far from univer universally shared, and the seventies was a period in which these preferences preferences were in increasingly challenged by workers from from within and outside of the Cattellian tra tra1972; Eysenck, 1972; 1972; Goldberg, Norman, dition (e.g., Adcock, 1972; Digman, 1972; & & Schwartz, 1972, 1972, 1980; 1980; Howarth, 1976; 1976; Howarth Howarth & & Browne, 1971; 1971; Karson & O'Dell, 1974; Sells, Demaree, & & Will, 1970). 1970). In response to these challenges, Cattell (1973) observed, "There has been a tendency to view these alternative posipossible conclusions as attacks on the establishment, and when that occurs posi tions which are themselves mutually incompatible are mistakenly seen as in alli alliance" (p. 289). Although it is true that these criticisms were too varied in nature to suggest a formal formal "alliance," there does appear to be the common theme that overextraction of factors. During Cattell's earlier work was characterized by an overextraction the next decade, the rather remarkable "return of the the Big Big Five" Five" signaled signaled an an era era the next decade, the rather remarkable "return of in which a new "establishment" threatened to replace the one to which Cattell in which a new "establishment" threatened to replace the one to which Cattell referred. referred.

F. The 1980s From the relatively crude criterion of topical citation counts, it might appear that of, the Big Five model suffered suffered a number of false starts interest in, and advocacy of, over the years and then exploded in the eighties. The development was more gradual, however, and much of it centered around the work and energy of of 1960, ORI has Lewis R. Goldberg at the Oregon Research Institute (ORI). Since 1960, from a few days been a site to which scholars have been invited for periods ranging from to several years. Since his early collaborations with Norman, Goldberg has had an enduring, and appropriately skeptical, interest in providing a firm lexical and psychometric base for the Big Five dimensions, and this interest has been widely communicated to others. In the latter respect, it is of interest to note some of the 30 years. years.^2 In chronological order of visits, they include visitors to ORI over the past 30 Dean Peabody, Jerry Wiggins, Warren Norman, Willem Hofstee, Frank Brokken, Conley, and Peter Peter Arend Tomas, Oliver John, Alois Angleitner, David Buss, James Conley, for citations). Borkenau (see Table I for During the eighties, Goldberg's published work served as an introduction to the Big Five for for a younger younger audience and rekindled the interests of more experi experienced investigators:

After our experience with these scales [Cattell's rating scales] our interest in After model waned, waned, and returned returned only only when when Goldberg Goldberg (1981, (1981, 1982, thefive-factor five-factor model 22 Goldberg, Goldberg, L. L. R., Visits Visitsand andvistors vistorstotoORI. ORI.Memo Memototocolleagues colleagues (March (March 2, 2,1990). 1990).

WIGGINS WIGGINS AND ANDTRAPNELL TRAPNELL

744 744

1983) renewed the search for universal dimensions in trait names, and developed a better instrument. instrument. (McCrae (McCrae & cfe Costa, 1985b, pp., 165) of John Digman Digman in Continuity is also evident in the long-standing interest of the five-factor early studies of five-factor model of personality. Digman's (1963, 1972) early of child personality of factors were sufficient personality ratings suggested suggested that a relatively relatively small number of sufficient for for characterizing that domain, and subsequent reanalyses of of these and adult studies convinced him of the centrality of the Big Five dimensions dimensions (Digman, 1979), which he has continued to advocate for 20 years (e.g., Digman, 1989; 1989; Digman Digman & & Inouye, Inouye, 1986; Digman & Takemoto-Chock, Takemoto-Chock, 1981). In addition to his multivariate demonstra demonstrations of the ubiquitous Digman (1985) has been ubiquitous nature of the five-factor five-factor structure, Digman especially especially concerned concerned with substantive interpretation of of the factors themselves in personality constructs. His previously previously mentioned chapter on the light of earlier personality Annual Review Review of of Psychology Psychology (Digman, (Digman, 1990) is almost personality structure in the Annual reflects nicely the current exclusively concerned with the Big Five literature and reflects Zeitgeist in the field. field. zeitgeist

n. THEORETICAL PERSPECTIVES ON THE THE BIG FIvE FIVE n. contrasting theoretical perspectives, it is often often useful useful to do so in In comparing and contrasting terms of George foci and ranges George Kelly's (1955) distinction distinction between between their differing differing foci of construct or set of of constructs constructs refers of convenience. The focus of convenience of a construct conveniently applied, and typically to the events to which a construct is most conveniently devising a construct. represents the particular events an author had in mind when devising The range of convenience of constructs refers to the extent extent to which constructs prove useful original focus focus of convenience. useful when applied to events outside their original We different theoretical theoretical perspectives on the W e hope to make clear in this section that different Big Five involve somewhat somewhat different different foci and ranges of of convenience. We W e would would also like to endorse endorse Kelly's constructive-alternativistic constructive-alternativistic philosophy philosophy that different different versions of the "truth" may coexist without chaos. In comparing important to do so with traits, it is important comparing theoretical perspectives on traits, reference to an important distinction, first made made by Allport AUport (1937, pp. 299-300), between the two different different usages of that term which have resulted resulted in conceptual ambiguities in the trait literature. Rather than identifying these different different usages identifying these with a particular trait theory, we shall refer to them as "traitt" and "trait2" refer "traiti" "traita" (Wiggins, 1984). The trait1 concept denotes trait attributions, couched in ordinary language, traiti couched which are meant to describe or summarize the pattern of an individual's conduct describe summarize of conduct to date, from a normative perspective and with reference to the likely social outcome perspective reference social outcome of causal or generative generative mechanisms which of that conduct. The trait2 concept denotes causal are meant to explain explain the behavior described in trait1 traiti language (Wiggins, this volume, chap. 4). The importance and clarity clarity of this distinction will will be seen to vary vary among different different theoretical perspectives on the Big Five.

28 CHAPTER 28

RETURNOF OFTIlE THEBIG BIGFIVE FIVE RETURN A. A.

745 745

The Enduring Dispositional View of of the Big Five: Costa and McCrae Somewhat surprisingly, the "breakthrough" that we feel has restored restored confidence confidence in personality psychology was not the proposal of a new conceptual model or models: trait theory. paradigm but a return to one of the oldest models: (Costa & d McCrae, McCrae, 1980, 1980,p.p.6767) (Costa ) Theoretical Orientation Orientation 1. Theoretical

The trait theory of of Costa and McCrae was originally developed developed in the context of of of personality personality and aging. Within that focus of of convenience they they longitudinal studies of felt that a self-report, self-report, multivariate, trait trait model of of personality would would serve to clarify clarify felt of concepts and measures used in the field" field" and integrate the "bewildering variety of (Costa & & McCrae, 1980, p. 68). Their Their subsequent subsequent research research confirmed confirmed this this expectation expectation (Costa impressive degree of of longitudinal stability (McCrae & & Costa, 1990) and revealed an impressive further strengthened strengthened their commitment to the concept of of traits as enduring enduring which further dispositions (Costa & & McCrae, 1980). dispositions other mainstream personality theories, the theoretical perspective of Like other Costa and McCrae places a heavy emphasis upon the trait concept. Within this context, their position on the distinction between traiti traitl and trait traita2 concepts appears to be closer to that of of Allport Allport and, to some extent, Murray than to that of of Cattell (see Wiggins, 1984): "The more of ], the more likely they of a trait people have [trait [trait2], 2 are to show the behavior it disposes them toward, and thus the more frequently behavior frequently we are likely to see it [traitd" [traiti]" (McCrae & Costa, 1990, p. 23). Although Although McCrae McCrae and Costa (1990) explicitly **note "note that our definition definition of of traits says nothing about their origins" (p. 24), their use of of the concept does not always distinguish the two: "The trait names do not refer refer to the underlying underlying physiology, but to the abstract consistencies in the ways people act and experience [traitl] [traiti] and to whatever whatever complex underlying underlying causes they may have [trait2]" [traita]" (p. 25)." For For the most part, however, it is the enduring enduring nature nature of of dispositions to behave and experience in particular particular ways (traitl) (traiti) that makes them so central to, and predictive of, of, the ways in which emerging lives develop. Costa and McCrae rejected rejected the "constructed "constructed consistency" view of of traits (e.g., Mischel, 1968) and adopted adopted a "realist" position (e.g., Loevinger, 1957) that received received strong support from from their demonstration demonstration of of the convergent convergent validity (McCrae, 1982) and longitudinal stability (Costa & McCrae, 1988a) of of both peer and spouse ratings of of self-reporting self-reporting targets. Further, they adopted a generally realist position on the veridicality of of self-report self-report and rejected rejected both "social 1986; McCrae & Costa, 1983) desirability" (Costa & McCrae, 1988b; McCrae, 1986; "impression management" management" (Costa (Costa & McCrae, in press; McCrae McCrae & Costa, and "impression favor of of a self-disclosure self-disclosure 1984) conceptions, on impressive empirical grounds, in favor model (Johnson, 1981). McCrae and and Costa (1984) maintain that a trait theory of of individual differences differences constitutes a legitimate and potentially comprehensive alternative to psychoanalytic, psychoanalytic.

746 746

WIGGINS WIGGINS AND AND TRAPNELL TRAPNELL

behavioristic, and humanistic theories of personality, rather than being an append appendage of or peripheral to those three conceptions of personality (pp. 32-33). Moreover, (pp. 32-33). the demonstrated empirical comprehensiveness of the five-factor model should be viewed as a theoretical advance: The growing consensus on the five-factor five-factor model model has has led led to to great great strides strides in in integrating personality research; we we hope the the NEO-PI can can be equally useful in integrating personality theory. The first first step step in in this this process process isis the the recognition recognition doesreflect reflect aatheory theoryofofpersonality personality. .. .. . . . that the NEO-PI does (Costa in press) (Costa&&McCrae, McCrae, in press) 2. 2. The The NEO Personality Inventory (NEO-PI) (NEO-PI)

The NEO-PI (Costa & McCrae, 1985. 1989, 1992) is an extension extension of of an an earlier three three1985,1989,1992) factor NEO model (neuroticism, extraversion, and openness) that now includes the factor additional Big Big Five dimensions of Agreeableness and Conscientiousness. The earlier structures three-factor model evolved from from a series of studies of the stability of trait structures three-factor across age groups (Costa & McCrae, 1976, 1976,1978,1980). factors 1978, 1980). Three age-invariant factors were first found found in the 16 PF (Cattell, Eber, & Tatsuoka, 1970) and later identified 16 PF (Cattell, Eber, & Tatsuoka, 1970) and later identified in a conjoint conjoint factor factor analysis of the Eysenck Personality Inventory (Eysenck & Eysenck, 1964), the EASI Temperament Temperament Survey (Buss & & Plomin, 1975), and the Eysenck. Experience Inventory (Coan, 1974). In In the theoriginal original NEO NEO Inventory, Inventory, each each ofthe of the three three age-invariant factors contained six facets which were chosen to be representative of age-invariant of previously identified substantive components and which were thought to provide "a useful useful level of generality" generality'* in representing these components (Costa & McCrae, 1986, p. 68). 1986, p. 68). Facet scales of Extraversion and Neuroticism were based, in part, on the structure and content of Buss and Plomin's (1975) EASI and on the results of structural analyses of other inventories (e.g., 16 do 16 PF. PF, GZTS. GZTS, EPI). EPI). The domain of Neuroticism comprised the facets of anxiety, hostility, depression, self comprised selfconsciousness, vulnerability, and impulsiveness. The domain of Extraversion included the facets of gregariousness, attachment, assertiveness, activity, excite excitement-seeking, and positive emotions. Facet scales for Openness were based on three subscales from from Coan's (1974) Experience Inventory-fantasy, Inventory—fantasy, aesthetics, and ideas-which ideas—which were supplemented by three subscales developed by Costa and Mc McCrae-feelings, Crae—feelings, actions, and values. The current five-factor NEO-PI had its origins in a study in which domain scales from of from the three-factor three-factor NEO Inventory were correlated with: (1) a set of adjectival markers of the Big Five developed by Goldberg (1983) and (2) factor factor adjectival from an an extended set of Big Five adjectives which included the scores derived from N, E, and 0 O domain scores converged converged on Goldberg markers. In this study, the N, factor scores, and were found to be corresponding Goldberg scales and adjective factor adjectival measures of Agreeableness and Conscientiousness uncorrelated with the adjectival 1985c). These findings encouraged Costa and McCrae to extend (McCrae & Costa, 1985c). their NEO Inventory by by constructing brief brief questionnaire scales to index the domains of Agreeableness and Conscientiousness. Brief, preliminary preliminary markers markers (McCrae (McCrae & & of Agreeableness and Conscientiousness. Brief,

CHAPTER CHAPTER 28 28

RETURN RETURN OF OF lHE THE BIG BIG FIvE FIVE

747 747

Costa, 1987) were later expanded into a full complement of facet scales for these two domains (Costa, McCrae, & Dye, 1991). 1991). A convenient 60-item 6O-item short form of of the NEO-PI has recently been published under the title "Five-Factor Inventory" (FFI; Costa & McCrae, 1989). These expansions and additions have been incorpo incorpoNEO (Costa & McCrae, 1992). rated in a recent revision of the NED

3. Comprehensiveness Comprehensivenessof ofthe theNEO-PI NEO-PIModel Model 3. Advocates of a "new" multivariate model, even one as long-standing as the fivefive of models in the personal personalfactor model, should not be oblivious of the rich history of for example, the following major major research ity assessment literature. Consider, for traditions: (1) Jung's (1923/1971) theory of psychological types was the first wellwell articulated typology and it has survived to this day in the Myers-Briggs Type McCauUey, 1985); Indicator (MBTI; Myers & McCaulley, 1985); (2) Murray's (1938) taxonomy of of human needs has spawned more assessment instruments than any other perspective, (PRF; Jackson, and it is currently best represented by the Personality Research Form (PRF; 1984); 1984); (3) Eysenck's (1947) long-standing system is among the most prominent PersonaUty on the contemporary scene, and it is operationalized in the Eysenck Personality Questionnaire (EPQ; (EPQ; Eysenck Eysenck & & Eysenck, Eysenck, 1975); 1975); (4) (4) the the empirical empirical tradition tradition in in Questionnaire objective personality personality assessment assessment (Meehl, (Meehl, 1945) 1945) is is still still very very much much alive alive in in the the most most objective widely used used and and studied studied of of all all questionnaires, questionnaires, the the Minnesota Minnesota Multiphasic Multiphasic Personality Personality widely Inventory (MMPI; (MMPI; Hathaway Hathaway & & McKinley, McKinley, 1983); 1983); (5) (5) the the Institute for Personality Personality Inventory Institute for Assessment and and Research Research (IPAR; (IPAR; MacKinnon, MacKinnon, 1948), 1948),' both both historically historically and and cur curAssessment rently, is is aa major major center center for for personality personality assessment assessment research, and the the common common rently, research, and descriptive language language for for assessment assessment which which evolved evolved at at that is represented represented descriptive that institution institution is Block, 1961); 1961); (6) (6) the the interpersonal interpersonal circumplex circumplex tradi tradiby the the California California Q-Set by Q-Set (CQS; (CQS; Block, tion (Freedman, (Freedman, Leary, Leary, Ossorio, Ossorio, & & Coffey, Coffey, 1951) 1951) has has been in recent recent tion been revitalized revitalized in years and and is is implemented implemented by by instruments instruments such such as as the the Interpersonal Interpersonal Adjective Adjective Scales Scales years (IAS; Wiggins, 1995); (7) Holland's highly influential theory of vocational choice (lAS; Wiggins, 1995); (7) Holland's highly influential theory of vocational choice is based based on on the the most most extensively extensively validated validated personality typology and and is is operation operationis personality typology alized by the Self-Directed Search (SDS; Holland, 1985). alized by the Self-Directed Search (SDS; Holland, 1985). Costa and McCrae's claims to comprehensiveness of their five-factor model are largely, and justifiably, based on empirical studies of the relations between the NEO-PI and all of the above instruments, which, as we have noted, are representa representaZondermajor research traditions in personality assessment (Costa, Bush, Zonder tive of the major man, & McCrae, 1986; 1986; Costa & McCrae, 1988a; Costa, McCrae, & Holland, 1984; McCrae & Costa, 1985a, 1989a, 1989b; McCrae, Costa, & Bush, 1986). 1986). Perusal of of meaningful convergences were the just cited references suggests that predicted and meaningful found in all of these studies. This line of research has served to place the Big Five model in general, and the NEO-PI in particular, in the mainstream of both historical conten^porary research in personality assessment. and contemporary It is of course possible that future research will discover other domains of of personality not covered by the scales of the NEO-PI, but in the meantime, no other system has a better claim to comprehensiveness. (Costa (Costa & <&McCrae, McCrae, 1985, 1985,p.p 27 . 27) )

748 748

WIGGINS AND TRAPNELL WIGGINS AND TRAPNELL

B. The The Dyadic Dyadic Interactional Interactional View View of of the the Big Big Five: Five: Wiggins Wiggins B. The interpersonal tradition in personality assessment originated in a different different re research context than did the five-factor model and, until recently, developed quite independently of the Big Five tradition. The history of the interpersonal tradition spans some 40 years (LaForge, 1985) and, like the Big Five model, it has been characterized by both extended "interruptions" over time and a recent widespread of convenience of this resurgence of interest (Wiggins, 1985, 1985, 1996a). The focus of tradition was upon dyadic interactions in psychotherapeutic settings (Leary, 1957) and, as a consequence, it has emphasized the two dimensions of personality that (Surgency/Extravermost clearly implicate interpersonal transactions: dominance (Surgency/Extraver nurturance (Agreeableness). (Agreeableness). The The dyadic-interactional dyadic-interactional perspective perspective on on the the sion) and nurturance five-factor model modelisisguided guided by bythe themetatheoretical metatheoretical concepts concepts of of agency agency and and commu commufive-factor nion (Wiggins, 1991). This perspective assigns a conceptual priority to the first two manifestations of agentic and communal factors of the model and emphasizes the manifestations factors (Wiggins & Trapnell, 1996). concerns within the remaining three factors The interpersonal system of personality diagnosis (Leary, 1957) originated in an attempt to translate certain concepts of Harry Stack Sullivan (1953) into concrete further enriched measurement operations. The conceptual basis of the system was further from the social exchange literature (Carson, 1969; by incorporation of concepts from formulation of complementarity and similarity Foa & Foa, 1974) and by an explicit formulation The conceptual conceptual continuity continuity ofthe of the system system in interpersonal transactions (Kiesler, 1983). The has been maintained, in part, by the remarkable similarities between earlier SuUivanSullivan has ian concepts and recent developments in the field of cognitive psychology (Carson, 1991). The The plausibility of of measurement procedures associated with the interpersonal interpersonal of the fit between the conceptual system has also increased with demonstrations of model and the empirical data (e.g., Wiggins, Phillips, & Trapnell, 1989). The structural model underlying the interpersonal system is is a two-dimensional orthogocircumplex in which variables appear in a circular order around the bipolar, orthogo nal coordinates of dominance/agency/status and nurturance/communion/love. nurturance/communionllove. The circumplex model provides an alternative to the simple-structure model of factor analysis in which all variables are expected to have their principal loadings on one or the other of two orthogonal factors. As a consequence, there is no optimal orientation of the principal axes of a circumplex, because any rotation is as good as any other. The placement of axes through dominance (PA) and nurturance (LM) in Figure 11 is not arbitrary, however, because it is based on a substantial body of of theoretical writings and empirical findings that attest to the fundamental fundamental nature of agency and communion as conceptual coordinates for the understanding and measurement of interpersonal behavior (Wiggins, 1991). Leary (1957) translated key Sullivanian concepts into measurement operations by placing them within a trait! traiti perspective: "The "The basic basicunits unitsof ofpersonality personalitycome comefrom from the protocol language by which the subject's interpersonal behavior is described" (p. 34). The well-established circumplex structure of of interpersonal trait attributions (p. may thus be distinguished from from their possible underlying causes (Wiggins, this volume, chap. 4).

CHAPTER CHAPTER 28 28

749 749

RETURN llfE BIG RETURN OF OFTOE BIG FIVE FIVE AssuredDominant ArrogantCalculating

BC O

PA PA EXTRAV Gregariousp EXTRAV Gregarlous, Extraverted ' ^ Extraverted o NO ^ NO I/ I

I I

Coldhearted

DE DE

Warm- O Agreeable LM LM

o-

.. AGREE AGREE

O FG AloofIntroverted

o

HI HI

JK

UnassumingIngenuous

UnassuredSubmissive

FIGURE 1 FIGURE

Circumplex model of interpersonal behavior.

Although the similarities between the interpersonal circumplex components Surgency/Extraverof dominance and nurturance and the Big Five components of Surgency/Extraver sion and Agreeableness have been recognized for some time (e.g., Goldberg, 1981), it is only recently that the relations between these two traditions have been considconsid in some detail. McCrae and and Costa (1989b) examined the convergences between ered in and self-ratings on the Interpersonal Adjective peer and self-ratings on the NEO-PI and Scales (IAS; (lAS; Wiggins, 1995). Combined factor analysis of the preceding sets of circumplex pattern patternof of lAS IASvariables variablesaround aroundthe the orthogonal orthogonal variables revealed a clear circumplex NEO-PI factors of Extraversion and Agreeableness (McCrae & Costa, 1989b, p. p. 589). As one might expect on substantive grounds, the obtained circumplex pattern of lAS IAS variables was shifted from the standard orientations of dominance (PA) and nurturance (LM) toward the off-diagonal off-diagonal coordinates of extraversion (NO) and ingenuousness (JK). (JK). Another way of looking at this structure is provided in Figure 1, in in which the NEO-PI domain scores of Extraversion and Agreeableness have been projected onto the lAS IAS circumplex. Although the optimal orientation of the two "interpersonal" axes of the five fivefactor model is likely to remain a contentious and nontrivial theoretical issue between interpersonalists and factor analysts, recognition that these two long longstanding traditions are, in many senses, complementary complementarytotoeach eachother otherisiscertain certaintoto 1989b; Trapnell & Wiggins, lead to advances in both traditions (McCrae & & Costa, 1989b; 1990). For example, the interpersonal circumplex model lends itself well to the kinds offine-grained fine-grained distinctions required in thefields fields of psychodiagnosis (Wiggins, 1982) and psychotherapy (Kiesler, 1988). In the former field, Wiggins and Pincus 1988).

WIGGINS AND AND TRAPNELL

750 750

(1989) have demonstrated that the circumplex model captures nicely many of the distinctions made among six of the personality disorders of Axis II of DSM-III (American Psychiatric Association, 1980). 1980). However, these distinctions were fur further clarified five clarified and additional personality disorders illuminated when the full fivefactor factor model of the NEO-PI was employed. In particular, despite the utility of the interpersonal circumplex for psychodiagnostic work, the full range of psychopsycho pathology represented by the personality disorders cannot be understood without taking into account the dimensions of Neuroticism, Conscientiousness, and, to some extent, Openness to Experience of the five-factor model (Wiggins & Pincus, 1989). Because the interpersonal circumplex and the five-factor model are comple complementary to rather than competitive with each other, there is much to be said for an approach to personality assessment that combines the advantages of both models. adjectival measure of To this end, Trapnell and Wiggins (1990) have extended the adjectival of Conscienthe circumplex (IAS) (lAS) to include the additional Big Five dimensions of Conscien instrutiousness, Neuroticism, and Openness to experience. This flexible flexible 124-item instru ment (IAS-B5) has excellent structural properties and promising convergent and discriminant relations with the NEO-PI (Costa & McCrae, 1985) and the Hogan Personality Inventory (R. Hogan, 1986).

C. De The Competency Competency View View of of the the Big Big Five: Five: Hogan Hogan C. Although there are always individual differences, people need attention and approval; status; status; and and predictability. predictability. These These needs needs are are largely largely satisfied satisfied through through approval; social interaction. interaction. The The key key to to attaining attaining status status and and approval approval is is interpersonal interpersonal social competence. As observers, observers, we we think think and and talk talk about about other other people people by by using using trait trait competence. As groups; trait trait words to to evaluate evaluate others others in in terms terms of of their their usefulness usefulness for for our our social social groups; words terms mark mark individual individual differences differences in in social social competence. terms competence. (R>Hogan, Hogan, Carpenter, Briggs, & Hansson, (R. Carpenter, Briggs, & Hansson, 1984, p. 1984, 27) p, 27) 1. Theoretical Orientation

traiti and trait trait22 concepts is central to Hogan's theoretical The distinction between trait} orientation and stems, in part, from from the focus of of convenience of of his theory upon the prediction of effective effective performance performance in work and social settings. Hogan equates different observational perspectives of of actors and observers. this distinction with the different Actors have needs for social approval, status, and predictability which cause them traita2 status: to seek social interactions with others. Such needs are clearly granted trait ". . . these master motives cause and explain social action-why action—why we do it and why it takes its prototypical forms" (R. Hogan, Jones, & Cheek, 1985, 1985, p. 195). Approval, status, and predictability are sought through ritualized interactions with others. The actors's role in this social drama is scripted with reference to the sociability (managing human resources of the group dimensional coordinates of sociability conformity^ (preserving (preserving the the culture culture of of versus managing its technical resources) and conformiry3 ^This Thisdimension dimensionisisnow nowlabeled labeled "prudence," "prudence/*but butthe theearlier earlierlabel labelmight mightmake makemore moresense senseininthe the 3 present context. context.

28 CHAPTER 28

RETURN OF OFmE THEBIG BIGFIVE FIVE RETURN

751 751

the group versus instigating cultural change). These coordinates appear appear to have the status of of the metaconcepts metaconcepts of communion and agency in interpersonal interpersonal theory of communion of his system to the typology (Wiggins, 1991). R. Hogan (1983) notes the similarity of provided by the interpersonal interpersonal circumplex (e.g., gregarious-extraverted gregarious-extraverted type). His for the vocational preference preference typology of of Holland (1985) (e.g., own preference preference is for social type). Specialized Specialized occupational occupational and and social roles provide the vehicles through which for approval, status, and predictability may be pursued in interac interacan actor's needs for tions with others. In the course of of these structured structured interactions, an actor attempts to self-images (trait (trait2) self-presentations. communicate his underlying self-images 2 ) to others through self-presentations. Goffman (1959), Hogan does not view these self-presentations self-presentations as self selfUnlike Goffman efforts to foster foster a favorable favorable impression, dictated by the situational demand conscious efforts of a particular particular role. Rather, he views these presentations as relatively self-conscious self-consciousof free efforts efforts to communicate one's self-view. To emphasize this difference difference in concep concepfree tualization. Hogan (1982, p. 258) adopted adopted the term "self-interpretation" "self-interpretation" tualization, Cheek and Hogan as aa more more accurate accurate representation representation of of their their position. position. Hogan Hogan also also disagrees disagrees with with as Goffman's view view that that "personality" may be be reduced reduced to to self-presentations self-presentations across across Goffman's "personality" may different situations. situations. Instead, Instead, he he postulates postulates aa number number of of stable traita structures structures within within different stable trait2 the individual that are of both biological and social origins (R. Hogan, Jones, & the individual that are of both biological and social origins (R. Hogan, Jones, & Cheek, 1985). Cheek, 1985). Observers evaluate the potential contributions of of an actor to a group by means means of trait attributions, couched in ordinary language, which capture an actor's social of couched capture reputation (trait!). (traiti). In R. Hogan's view, these evaluative trait descriptors evolved reputation of social control necessitated necessitated by the circumstances of of group living. as a means of for the most part, well captured by the Big Five dimensions of of Further, they are, for personality. Thus, for for Hogan, the Big Five are exclusively dimensions employed observers in the evaluation evaluation of of actors' contributions to groups or work organiza organizaby observers tions. This formulation formulation is less radical than it might appear when it is recalled that focus of of convenience of of the original Big Five work was on observer evaluations the focus of fitness fitness for for the occupations of of clinical psychologist psychologist (E. L. Kelly & & Fiske, 1951) of and Air Air Force officer officer (Tupes & Christal, 1958). and competency perspective led him to formulate formulate an alternative six-factor six-factor Hogan's competency interpretation of of the Big Five that is more directly focused focused on qualities related to interpretation successful performance. These factors factors (and their corresponding corresponding Norman Norman numbers) numbers) successful (la). Prudence (III), Likability (II), Ambition (Ib), (lb), Adjustment Adjustment are Sociability (Ia), Intellectance (V). Norman's ExtraversionlSurgency Extraversion/Surgency factor factor was split apart (IV), and Intellectance because "sociability and assertiveness are conceptually distinct, and, in addition, different psychological dynam dynamthey are not very highly correlated. . . . [They] have different associated with status-seeking, whereas sociability is associated ics; assertiveness is associated with affiliativeness affiliativeness and popularity" (R. Hogan, 1983, p. 66). 2. Measurement Measurement Rationale Rationale

Structural model. model In Hogan's view, a typology of of occupational or voca vocaa. Structural from a circumplex model formed formed by the tional specialization can be constructed from

WIGGINS WIGGINS AND AND TRAPNEU TRAPNELL

751. 752

coordinates of sociability and conformity. Further, the remaining four orthogonal factors of of Likability, Ambition, Adjustment, and Intellectance may be used to assess individual differences differences in competence within each of the occupational categories 1983; R. Hogan & Johnson, 1981). identified by the circumplex typology (R. Hogan, 1983; This combined circumplex-Big Five model is structurally, although not substan substantively, similar to that advocated by Trapnell and Wiggins (1990) in the interpretation interpretation of IAS-B5. IAS-B5. However, Hogan does not recommend the use of his own dimensions of sociability and conformity conformity for for defining vocational personality types. Instead, he suggests that sociability and conformity, along with the other four factors of his system, be interpreted interpreted as dimensions of competence within the typology provided by Holland's (1985) hexagonal model (R. Hogan, 1986, 1986, pp. 22-33; see also John Johnson, 1983).

responding. Unlike Costa and McCrae, Hogan adopts a strong self selfb. Item responding. presentational view of item responding: The processes involved in answering inventory items are formally identical to the processes processes underlying underlyingsocial socialinteractions interactions.. . . . .responses responsestotopersonality personalityinvento inventothe ries are are not not self-reports self-reports.. .. .. Rather, Rather, the the responses responses reveal reveal how how the the person person wants wants ries to be be regarded regarded .. .. .. scale scale scores scores derived derived from from an an aggregation aggregation of of item item responses responses to do not not necessarily necessarily reflect reflect underlying underlying traits traits [trait2] [trait2] in in the the respondent. respondent. do Hogan,1986, 1986,p.p.2).2). (K Hogan, (R.

The implications of Hogan's theory of item responding for the construction and of items are less explicit than are those of some other test authors. For selection of example, Jackson's (1986) theory of item responding serves as an explicit rationale for the innovative procedures he developed to minimize social desirability variance for at the item level (Jackson, 1970). 1970). In this context, at least, it is clear that Hogan for minimizing social (Nicholson & Hogan, 1990) would not adopt these procedures for desirability variance in test construction.

3. 3.

The Hogan Personality Inventory (HPI) The

The HPI operationalizes Hogan's socioanalytic trait theory by means of a 310self-report questionnaire. Consistent with that theory's observer-centered, observer-centered, item, self-report competency interpretation of the Big Five dimensions, the HPI targets dimensions ''broad, general importance for for personal and social of reputation thought to possess "broad, effectiveness" (R. Hogan, 1986, 1986, p. 5). There are six HPI global or "primary" scales. scales, Adjustment, and Intellectance, each Ambition, Sociability, Likability, Prudence, Adjustment, factor from from Hogan's six-factor six-factor variant of the Big Five model. corresponding to a factor decomposed into a number of small (3 to In addition, each primary scale may be decomposed clusters" (HICs). Each of the 6 items each) subscales called "homogeneous item clusters" 43 HICs assesses a specific trait judged by Hogan and his research team to be most relevant to an individual's reputation with respect to one of the six HPI trait

CHAPTER CHAPTER 28 28

RETURN RETURN OF OF nIE THE BIG BIG fivE FIVE

753 753

dimensions (e.g., "Math Ability" HIC for Intellectance, "Status-Seeking" HIC for Ambition). Although the HPI purportedly spans the same universe of content as does the NEO-PI (i.e., the Big Big Five), the the HPI HPI apportions apportions this thisuniverse universe into intorather rather different different content domains and facets than does the NEO-PI. For example, whereas the NEO NEOan "openness to PI measures Factor V with six facets found empirically to define define experience" dimension, the HPI measures Factor V with nine HICs judged judged to exemplify exemplify the reputation of "intelligent." These nine HICs define two virtually orthogonal clusters, however, only one of which ("culture") bears a conceptual and empirical relation to NEO-PI Openness; the other cluster ("scholastic ability") Opencorrelates weakly with Conscientiousness and is essentially uncorrelated with Open & Wiggins, 1990). Similarly, Similarly, the the HPJ's HPI's eight eight Prudence Prudence HICs, HICs, selected selected ness (Trapnell & reputation of "prudential" (cautious/responsible), form form two indepen indepento index the reputation dent clusters. Only one of these clusters correlates with NEO-PI Conscientiousness; ("excitementthe other correlates substantially with an NEO-PI Extraversion facet ("excitement seeking") (Trapnell & & Wiggins, 1990). 1990). These and other substantive and structural differences between the HPI and the NEO-PI reflect reflect the divergent trait theoretical differences of positions held by R. Hogan and by Costa and McCrae in their conceptions of the Big Big Five Five dimensions: dimensions: the the NEO-PI NEO-PI reflects reflects aa "traditional" "traditional" factor factor analytic analytic (i.e., (i.e., the Eysenck, Cattell) Cattell) orientation orientation to to the the five factor factor model, model, while while the the HPI HPI reflects reflects an an Eysenck, explicit socioanalytic socioanalytic orientation. orientation. explicit

the HPI HPI 4. Applications Applications o of 4. f the As would be expected from its its focus of convenience, the principal applications of the As HPI have been in the prediction of organizational and occupational performance. performance. perforApplied research with the HPI has typically involved: (1) development of a perfor mance criterion for a general (e.g., salesperson) or specific (e.g., service operations dispatcher) occupational role, (2) correlation of the full set of HPI HICs with that criterion, and (3) selection of a subset of HICs that best predicts the criterion and scoring these items as an "occupational performance scale" for use in personnel selection and other practical assessment applications. 11 studies in which the HPI pre R. Hogan (1986) summarizes findings from 11 predicted such performance ratings, company commendations, performance criteria as supervisor ratings, test),total totaldollar dollarsales salesrevenue, revenue,and andtraining trainingcourse coursecomple complefitnessratings ratings(treadmill (treadmilltest), fitness tions. Findings from from these and other studies (see J. Hogan & & Hogan, 1986) have prompted development of HPI scales for clerical, sales, and managerial potential, employee reliability, stress tolerance, and service orientation (R. Hogan, 1986), 1986), as well as HPI personnel selection scales for occupations such as hospital service worker (Raza, Metz, Dyer, Coan, & & Hogan, 1986), service operations dispatcher (R. Hogan, Jacobson, Hogan, & & Thompson, 1987), line-haul driver (J. Hogan et al., combination driver driver (J. (J. Hogan, Hogan, Hogan, & & Briggs, Briggs, 1984), and and habilitation habilitation aI., 1985), combination therapist (J. (J. Hogan, Hogan, Arneson, Arneson, Hogan, Hogan, & & Jones, 1986). Hogan Hogan has has recently recently incorpoincorpotherapist Jones, 1986).

754 754

WIGGINS WIGGINS AND AND TRAPNELL TRAPNELL

rated Hogan Assessment Systems to further of further develop and market applications of the HPI in government and industry.

D. The Lexical View of the Big Five: Five: Goldberg The most promising of the empirical approaches to systematizing personality differ differences have been based on one critical assumption: Those Thoseindividual individual differences ofof thethe most significance in thein daily of persons encesthat thatare are most significance the transactions daily transactions of with persons with . . . Moreover, this each otherwill will eventually become encoded their language. each other eventually become encoded in theirinlanguage . . . Moreover. this Themore moreimportant important is an corollary: The is an fundamental axiom has a highly significant corollary: transactions, thelanguages more languages individualdifference difference in human individual in human transactions, the more will have awill termhave a term for it. for it. (Goldberg,1981, 1981, 141-142) (Goldberg, pp.pp. 141-142) .

Although there have been "interruptions" in the development of the Big Five model over time, there is one important line of research that has been continuous classification of of for 25 years. In the mid-sixties, Norman developed a hierarchical classification carefully selected trait-descriptive terms which were grouped approximately 1600 carefully into subordinate semantic categories within the broader classes provided by the model.'* Norman's Norman's (1967) (1967) lexicon lexicon of of trait trait descriptors descriptors and and his hispreliminary preliminaryBig Big Big Five mode1.4 Five classification system were the starting points for Goldberg's (1977) subsequent efforts to construct a common lexicon for the major major characteristics of human person personefforts ality.

/. Theoretical Theoretical Orientation Orieniation 1. Goldberg's conceptual approach to the study of individual differences, and the differs considerably from from those of the endur endurmethods whereby he studies them, differs ing dispositional, dyadic interactional, and competency approaches. As a con consequence, his work is not as easily classified with respect to distinctions between trait. and trait traiti traita2 or between range and focus of convenience. As is evident from from the significant or funda fundaquotation above, Goldberg (1981) assumes that the most significant mental dimensions of human interaction may be revealed through the study of *'we should find a uniuni language. The strongest form of this proposition is that "we emergence of the individual differences differences encoded into the set of of emergence of versal order of all the world's languages" (p. 142). 142). Rigorous linguistic studies of the structure function of ordinary trait language usage will identify identify the significant distinc distincand function tions involved in person perception. The generalizability of these distinctions across different language groups will confirm (or disconfirm) their universal signifi signifidifferent cance. Goldberg does not appear to subscribe to any particular theory of traits. The distinction we made earlier in this chapter between the Allportian AUportian personality attributes tradition and the Thurstonian personality structure tradition is useful useful in the present context. Although each tradition may be pursued independently of the ^This This unpublished unpublished research research isis succinctly succinctly summarized sununarized in in John John et et aI. al. (1988. (1988, pp. pp. 184-189). 184-189).

4

CHAPTER CHAPTER 28 28

RETURN BIG FIVE RETURNOF OFTIlE THE BIG FIVE

755755

other, there is a growing consensus that taxonomic research should provide the basis for subsequent test construction procedures. Most current approaches to inventory construction are based on some some sort of special interest, although these of classifications within the universe of content of classifications may be considerably less formal than those discussed by John et al. (1984). Costa and McCrae's NEO-PI is based on a review of of the literature of of earlier scales and inventories, with particular reference to the study of of aging. Wiggins' lAS IAS of trait terms that was a direct was developed from a theoretically based taxonomy of offshoot offshoot of one of Goldberg's earlier taxonomies. Hogan's HPI was based on a competency perspective. Goldberg's review of the earlier Big Five literature from a competency taxonomies are based upon semantic relations uncovered in the natural language, priori distinctions from psychological psychological theory. He hoped that rather than upon a priori such a semantic taxonomy "would serve to decode the Babel that has been created idiosynby researchers using their own labels for constructs, and speaking their own idiosyn cratic tongues" (John et al., aI., 1984, p. 4). To that end, his research efforts to date taxonomic structure have been directed toward "the development of a compelling compelling taxonomic for the the personality-descriptive personality-descriptive terms terms in in the the natural natural languages" (Goldberg, 1982, for languages" (Goldberg, 1982, p. 203). 203). p. focus upon linguistic concepts does not itself Goldberg's focus itself restrict the potential of results from his research program. Indeed, if lexical universals range of application of if lexical fundamental dimensions of are found, and if they are related to fundamental of human interaction, of cross-cultural generalizability. the model would have an extraordinary degree of Similarly, an exclusive focus on ordinary language usage (traiti) (traitl ) is not without of generative mechanisms (traita). implications for the study of (trait2 ). In fact, Goldberg's hnguistic analysis will reveal reveal which phenotypical patterns (1970) main point is that linguistic of explanation and on that point he is in agreement with Allport, AUport, Cattell, are worthy of and others. Taxonomic Research Research 2. Taxonomic In his quest for a compelling compeUing taxonomic structure, Goldberg has strongly emphasized the importance of developing explicit criteria whereby such compellingness may be evaluated objectively. These criteria, in tum, turn, have been the subject of of extensive empirical investigations conducted by Goldberg and members of of his research team. The issues to which these criteria are directed include those pertaining to: (1) grouping, (2) abstractness, (3) structure, (4) generalizability, and (5) comprehen comprehensiveness (John et al, aI., 1984). Within Goldberg's (1980, 1981, 1982) taxonomies, terms are grouped grouped and organized in terms of of their culturally shared meaning, as determined determined by meaning meaningsimilarity ratings of native speakers, dictionary definitions, and co-occurence of of attributions in other and self-ratings. A A distinction is made between between internal struc structure (based on judges' ratings of semantic similarity) and external structure (based of structure on observers' attributions to self and others). In general these two types of converge, although there is a tendency for internal structures to be simpler and more & Goldberg, 1989). Finer distinctions schematic than external structures (Peabody &

756 756

WIGGINS AND TRAPNELL WIGGINS

within semantic subcategories are achieved by Goldberg's extension of Peabody's (1967) model for for separating the evaluative and descriptive components of trait terms. The issue of abstractness abstractness concerns concerns the the breadth breadth or or generality generality of of categories categories employed in a taxonomy and the manner in which differences differences in breadth among trait categories may be represented within a hierarchical structure, and this assump assumption has been closely examined by Goldberg and his co-workers (Goldberg, 1986; Hampson, Goldberg, & John, 1987; 1987; Hampson, John, & Goldberg, 1986). 1986). Their findings suggest suggestthat thatstrict stricthierarchies hierarchiesare areless lessfrequent frequentfor forpersonality personaUtytraits traitsthan thanfor for findings preferred or "basic level" of personality categories of natural objects. Moreover, the preferred differ from from that found found in other domains (John, Hampson, & description appears to differ Goldberg, 1991). structureinvolves involvesthe thechoice choicebetween betweenaadimensional dimensionaland andaacategori categoriThe issue of structure cal approach to trait classification. This is the same issue involved in the choice classification and between dimensional between dimensional and typological person classification preference is clearly for for a and categorical psychiatric classification. Goldberg's preference approach which does justice to the continua implied by trait terms and dimensional approach their modifiers in most languages. A dimensional approach also permits rigorous quantification and the use of multivariate analysis in the construction and evaluation quantification unsurpassed of taxonomies. In the latter respect, Goldberg's empirical analyses are unsurpassed in the field of trait taxonomic research. generalizability of oftaxonomic taxonomic structure structure isisatat the the The issue of cross-language generalizability core of Goldberg's (1981) search for universals in personality lexicons. Although it is much too early to evaluate this aspect of his taxonomies, some very substantial develfirst steps have been taken in this direction. Parallel taxonomies have been devel 1978; Hofstee, 1976; oped in the Dutch language (Brokken, 1978; 1976; Tomas, 1977) and in the German language (Angleitner, Ostendorf, & & John, 1990; 1990; John & Angleitner, 1982). This 1982). This international cooperative cooperative effort effort has has provided provided rich rich data datafor for the the assessment assessment of generalizability: "The use of identical procedures in three district but closely related languages (English, Dutch, German) produces a multilanguage multilanguagemultimethod matrix for for assessing the appropriateness of competing operations" (John (John et aI., al., 1984). The issue of inclusiveness inclusivenesshas hasbeen beencharacterized characterized as as"the "thedegree degreeto towhich whichthe the constructs from from a particular domain of individual differences differences can be represented within the structure postulated for aI., 1984, 1984, pp. 9-10). An for this domain" (John et al., equally important aspect of inclusiveness is the extent to which a truly representative from the natural language was ensured by initial search and culling sample of terms from further on, both of these aspects of inclusiveness procedures. As we will indicate further appear to have been demonstrated for the Big Five model.

3. 3.

Contributions to the Big Five

Interest in the five-factor model derived mainly from the claim that five dimensions for all nontrivial personality might provide an adequate preliminary taxonomy for traits—those whose importance in human interaction has resulted in a descriptive traits-those

CHAPTER 28 28 CHAnER

RETURNOF OFDIE THE BIG FIVE REroRN BIG FIVE

757757

label in the natural language (e.g, dominant), as well as those reflected in the constructs of of personality researchers (e.g., Machiavellianism). The demonstrations by Fiske, Tupes and Christal, Norman, Digman, and others of a recurrent five fivefactor structure among Cattell's 20-35 rating scales were important mainly because Cattell's method of of deriving his rating scales appeared to provide a comprehensive comprehensive sampling of of traits. However, psychometric deficiencies of of the Cattell rating scales themselves (McCrae & Costa, 1985b, p. 164), shortcomings of his taxonomic research procedures (John et al, aI., 1988, pp. 183-184), and the repeated demonstrations of of a five-factor structure on essentially the same set of of variables (Waller & Ben-Porath, 1987) did little to advance the claim of comprehensiveness of of the five-factor model. Goldberg's (1980) factor analysis of of adjective clusters derived from the Nor Norman (1967) trait taxonomy taxonomy provided the first persuasive evidence that five large factors provided a comprehensive lan comprehensive account of trait description in the English language. Goldberg's results, which were remarkably similar to the earlier studies of of the Cattell rating scales, were impressive because: (1) the Norman and Goldberg taxonomies used explicit and reasonably reasonably objective inclusion/exclusion criteria which permitted a more scientifically scientifically defensible defensible claim to comprehensiveness comprehensiveness than did the of variables (75 and 133 Cattell taxonomy; (2) Goldberg factored a much larger set of earUer researchers using the Cattell rating scales; (3) composites scales) than did earlier (small clusters of of synonyms) were factored rather than single variables; (4) semantic Goldberg demon demonconsistency of the clusters was evaluated empirically; and (5) Goldberg strated, in a psychometric tour de force, that the five-factor structure was invariant across varieties of of factor extraction and rotation, number of of factors extracted extracted (6 through 12), method of of variable selection, different different samples, and different different rating (self and peer). In a figure presented in this landmark paper, Goldberg Goldberg also targets (self of personality personality structure attempted a conceptual integration of alternative models of within a "Big Five" framework. framework. Within such a framework, commonalities among Norsemantic (Osgood, Peabody), interpersonal (Leary, Wiggins), lexical (Cattell, Nor man), factor analytic (Guilford), temperament (Eysenck, Buss, & Plomin), and psychodynamic (Block) models of of personality structure were made apparent. Gold Goldpsychodynamic offered the Big Five as a first step in "decoding [the] Babel" of of personality berg offered constructs that had heretofore littered the field. Goldberg's (1980) study strongly advanced the claim to comprehensiveness of of the five-factor five-factor model, at least in regard to trait distinctions encoded in the natural language. The inclusiveness of of a five-dimensional five-dimensional framework framework with respect to the constructs of of personality researchers was later conclusively demonstrated by Costa and McCrae, as we discussed earlier. Goldberg's (1990, 1992) standard markers of (1990,1992) of the Big Five factor structure have facilitated both lexical and assessment research on these dimensions. Most of of Goldberg's research is directed toward providing a firmer linguistic basis for future assessment of the Big Five and other dimensions of personality. Current perspectives on the Big Five range from the "basic research" orientation of Goldberg to the "applied" interests of of Hogan, and they are held by other of investigators whose research interests involve different different mixtures mbctures of these two orien-

758 758

WIGGINS WIGGINS AND AND TRAPNELL TRAPNELL

tations. Goldberg's and, Goldberg's empirical empirical work work has served to t o kkeep e e p the rest of us honest honest and, at the same time, to illustrate that basic and applied research can be complementary same complementary in advancing advancing the field field of personality personality structure.

m. ni. CONCLUSION Although A l t h o u g h characterized by "interruptions," the history history of the Big Five Five dimensions dimensions of of personality personality structure structure suggests suggests a cumulative convergence convergence of thought that consti constitutes the longest, and quite possibly the most important, chapter to date in the history of personality personality structure research. It is clearly not the final final chapter, however, however, history and the most most important important developments developments may may still lie ahead. Chief Chief among among these devel develo p m e n t s may m a y be b e the incorporation incorporation of the structural m o d e l of the Big B i g Five within within opments model variety of theoretical perspectives (Wiggins, 1996). Perhaps Perhaps an increasingly wide variety the ultimate contribution contribution of the Big B i g Five Five model model will be b e the increased increased opportunities opportunities affords for communication communication among among investigators of different different theoretical theoretical persua persuait affords sions in personality, social, and clinical clinical psychology. psychology.

ACKNOWLEDGMENT Preparation of this chapter was was supported by Social Sciences and and Humanities Research Research Preparation Council of Canada Grant 410-87-1322.

REFERENCES M. Dreger (Ed.), Multivariate Multivariate Adcock, C. J. (1972). Multivariate measurement theory. In R. M. personality research personality research (pp. 1-50). Baton Rouge, LA: Claitor's Publishing. Allport, G. W. W. (1937). Personality: Personality: A psychological psychological interpretation. interpretation. New New York: Holt. AUport, Allport, W., & Odbert, H. S. (1936). Trait-names: A psycho-lexical psycho-lexical study. Psychological Psychological AUport, G. W., Monographs, Monographs, 47(1, Whole No. 211). American Psychiatric Association. (1980). Diagnostic statistical manual manual of of mental mental disor disorDiagnostic and statistical ders ders (3rd ed.). Washington, DC: Author. Angieitner, A., Ostendorf, F., & John, O. Ostendorf, P., O. P. (1990). Towards a taxonomy of personality personality Angleitner, A., descriptors in German: A psycho-lexical study. European European Journal Journal of of Personality, Personality, 4, 89-118. Block, J. (1961). The Q-sort Q'Sort method method in personality personality assessment assessment and and psychiatric psychiatric research. research, Spring Springfield, IL: Charles C. C. Thomas. 8-17. Borgatta, E. F. (1964). The structure of of personality characteristics. Behavioral Behavioral Science, Science, 9,9,8-17. Borkenau, P. (1988). The multiple classification of acts and the Big Five factors of personality. Journal of Research Research in Personality, Personality, 22, 337-352. Journal of Bouchard, M., Lalonde, F., & Gagnon, M. (1988). The construct validity of assertion: Contri Contriand Norman's personality factors. Journal Journal of of butions of four assessment procedures and Personality, 56, 763-783. Personality, 56,

CHAPTER CHAPTER 28 28

RETURN OF mE THE BIG FIVE RETURN

759 759

languageof ofpersonality. personality.Meppel, Meppel,Holland: Holland:Krips. Krips. Brokken, F. B. (1978). The language Buss, A. H., & Plomin, R. (1975). A temperament temperament theory theoryof of personality personalitydevelopment. development.New New York: Wiley (Interscience). Carson, R. C. (1969). Interaction Interaction concepts conceptsofofpersonality. personality.Chicago: Chicago:Aldine. Aldine. Carson, R. C. (1991). The social-interactional viewpoint. In M. Hersen, A. E. Kazdin, & A. S. Bellack (Eds.), The clinical clinicalpsychology psychologyhandbook handbook (2nd (2nd ed., ed., pp. pp. 185-199). 185-199).New New York: Pergamon Press. B. (1943). The description of personality. II. II. Basic traits resolved into clusters. Cattell, R. B. Journal of of Abnormal Abnormal and Social Social Psychology, Psychology,38, 38,476-507. 476-507. Psychological (1945). The principal trait clusters for describing personality. Psychological Cattell, R. B. (1945). Bulletin, Bulletin, 42, 42, 129-169. Cattell, R. B., B., Eber, H. W., & Tatsuoka, M. M. (1970). Handbook for for the the Sixteen Sixteen Personality Personality IL: Institute for Personality and Ability Factor (16 PF). Factor Questionnaire Questionnaire (16 PF). Champaign, IL: Testing. B. (1957). Personality Personality and andmotivation motivationstructure structureand andmeasurement. measurement.Yonkers-on Yonkers-onCattell, R. B. Hudson, NY: World Book. Cattell, R. B. (1973). Personality Personalityand andmood moodby byquestionnaire. questionnaire.San SanFrancisco: Francisco:Jossey-Bass. Jossey-Bass. J. M., & & Hogan, R. (1982). Self-concepts, self-presentation, and moral judgments. Cheek, 1. Psychologicalperspectives perspectivesononthe theself self(Vol. (Vol.2,2, In J. Suls & A. G. Greenwald (Eds.), Psychological pp. 249-273). Hillsdale, NJ: Erlbaum. personality.New NewYork: York:Columbia ColumbiaUniversity UniversityPress. Press. Coan, R. W. (1974). The optimal personality. Conley, J. J. J. (1985). Longitudinal stability of personality traits: multitrait-multimethodtraits: A multitrait-multimethod of Personality Personalityand andSocial SocialPsychology, Psychology,49,49,1266-1282. 1266-1282. multioccasion analysis. Journal of Journal Costa, P. T., Jr. (Ed.). (1991). Clinical use of the five-factor model [Special Section]. Journal of Personality PersonalityAssessment, Assessment,57,57,(3). of (3). (1986). Correlations of of Costa, P. T., Jr., Bush, C. M., Zonderman, A. B., & McCrae, R. R. (1986). factor scales with measures of the five-factor model. Journal of of Personality MMPI factor Personality Assessment, 50,640-650. 640-650. Assessment, 50, (1976). Age differences differences in personality structure: A cluster Costa, P. T., Jr., & McCrae, R. R. (1976). of Gerontology, Gerontology,31, 31,564-560. 564-560. analytic approach. Journal of & McCrae, R. R. (1978). Objective personality assessment. In M. Storandt, Costa, P. T., Jr., & clinicalpsychology psychologyof of aging aging(pp. (pp. 119-143). 119-143). I. C. Siegler, & M. F. Elias (Eds.), The clinical New York: Plenum Press. Costa, P. T., Jr., & McCrae, R. R. (1980). (1980). Still stable after after all these years: Personality PersonaUty as a key to some issues in adulthood and old age. In P. B. Baltes & O. G. Brim, Jr. (Eds.), Life span development developmentand and behavior behavior(Vol. (Vol.3,3,pp. pp.65-102). 65-102).New NewYork: York:Academic AcademicPress. Press. Costa, P. T., Jr., & McCrae, R. R. (1985). (1985). The NEO NEO Personality PersonalityInventory Inventorymanual. manual.Odessa, Odessa, FL: FL: Psychological Assessment Resources. Costa, P. T., Jr., & McCrae, R. R. (1986). Major contributions to personality psychology. In S. Modgil & c. 63-72). C. Modgil (Eds.), Hans Eysenck: Eysenck:Consensus Consensusand andcontroversy controversy(pp. (pp. 63-72). Barcombe, Lewes Sussex, England: Falmer. Costa, P. T., Jr., & McCrae, R. R. (1988a). Personality in adulthood: A six-year longitudinal study of self-reports self-reports and spouse ratings on the NEO Personality PersonaUty Inventory. Journal of Personality Personalityand andSocial SocialPsychology, Psychology,54,54, 853-863. of 853-863. Costa, P. T., Jr., & McCrae, R. R. (1988b). From catalogue to classication: Murray's needs and the five-factor model. Journal of 258-265. of Personality Personalityand andSocial SocialPsychology, Psychology,55,55, 258-265. R. R. R. (1989). (1989). The The NEO-PIINEO-FFI NEO-PI/NEO-FFI manual manualsupplement. supplement Odessa, Odessa, Costa, P. T., Jr., & McCrae, R. FL: Psychological Assessment Resources. FL:

760 760

WIGGINS WIGGINS AND AND TRAPNELL TRAPNELL

Jr., & Costa, P. T., Jr., & McCrae, R R. R R. (1992). Revised Revised NED NEO Personality PersonalityInventory Inventory(NED (NEOPI PIOdessa, FL: FL: R) and NEO NED Five-Factor manuaL Five-FactorInventory Inventory(NED-FF/). (NEO-FFI),Professional Professional manual, Odessa, Psychological Assessment Resources. Costa, P. T., Jr., & & McCrae, R R. R R. (in press). The NEO Personality Inventory (NEO-PI). R. Briggs & & J. Cheek (Eds.), Personality Personality measures measures(Vol. (Vol.1).1).Greenwich, Greenwich,CT: CT: In S. R JAI Press. Costa, P. T., Jr., McCrae, R R. R, R., & & Dye, D. A. (1991). Facet scales for Agreeableness and Conscientiousness: A revision of the NEO Personality Inventory. Personality Personality and and Individual Differences, Differences,12,12, 887-898. 887-898. Individual R, & Holland, J. L. (1984). Personality and vocational interests Costa, P. T., Jr., Jr., McCrae, R R. R., in an adult sample. Journal Journalof ofApplied AppliedPsychology, Psychology,69,69,390-400. 390-400. Costa, P. T., Jr., Jr., & & Widiger, T. A. (Eds.). (1993). Personality Personalitydisorders disordersand andthe thefive-factor five-factor model of of personality. personality. Washington, Washington,DC: DC:American AmericanPsychological PsychologicalAssociation. Association. methods: An historical perspective. Journal Journal of of Craik, K. H. (1986). Personality research methods: Personality,54, 54,18-51. 18-51. Personality, Cronbach, L. J., & Meehl, Meehl, P. P. E. E. (1955). (1955).Construct Constructvalidity validityin in psychological psychological tests. tests.Psychological Psychological Bulletin,52, 52,281-302. 281-302. Bulletin, Digman, J. M. (1963). (1963). Principal Principaldimensions dimensionsof ofchild childpersonality personalityas asseen seenin inteachers' teachers*judgments. judgments. Development,34, 34,43-60. 43-60. Child Development, Child Digman, J. M. (1972). The structure of child personality as seen in behavior ratings. In R. M. Dreger (Ed.), Multivariate Multivariatepersonality personalityresearch research 587-611). Baton Rouge, R (pp.(pp. 587-611). Baton Rouge, LA: Claitor's Publishing. Digman, J. M. (1979, November). The The fi five domainsof ofpersonality personalityvariables: variables: Analysis ve major domains Analysis of questionnaire questionnairedata dataininlight lightofofthe thefive fiverobust robust factors emerging from studies of rated of factors emerging from studies of rated characteristics.Paper Paperpresented presentedatatthe theannual annualmeeting meetingofofthe theSociety SocietyofofMultivariate Multivariate characteristics. Experimental Psychology, Los Angeles. Digman, J. M. (1985, November). The The Big Five Five factors factorsof of personality: personality:Some Someefforts effortsafter after meaning. meaning.Paper Paperpresented presentedatatthe theannual annualmeeting meetingofofthe thesociety societyofofMultivariate MultivariateExperi Experimental Psychology, Berkeley, CA. Digman, J. M. (1989). Five robust trait dimensions: dimensions: Development, stability, and utility. Journal Journal of ofPersonality, Personality,57,57,195-214. 195-214. Digman, J. M. (1990). Personality structure: five-factor model. Annual Annual structure: Emergence of thefive-factor Review of Psychology,41, 41,417-440. 417-440. of Psychology, Digman, J. M., M., & Inouye, J.J. (1986). (1986).Further Furtherspecification specificationof ofthe thefive fiverobust robustfactors factorsof ofpersonal personalPsychology, 50, 50, 116-123. ity. Journal Journal of ofPersonality Personalityand andSocial Social Psychology, 116-123. Digman, J. M., & & Takemoto-Chock, Takemoto-Chock, N. N. K. K.(1981). (1981).Factors Factorsin in the thenatural naturallanguage languageof ofpersonal personality: Behav ity: Reanalysis, comparison, comparison, and and interpretation interpretation of of six sbc major majorstudies. studies.Multivariate Multivariate Behavioral Research, Research,16, 16,149-170. 149-170. Kegan Eysenck, H. J. (1947). Dimensions Dimensionsof ofpersonality. personality,London: London:Routledge Routledge&& KeganPaul. Paul. Eysenck, H. J. (1972). Primaries or second order factors: A critical consideration of Cattell's BritishJournal JournalofofSocial Socialand and Clinical Psychology, 265-269. 16 PF battery. British Clinical Psychology, 11, 11, 265-269. Eysenck, H. J., & Inventory. & Eysenck, S. B. G. (1964). The The manual manualof ofthe the Eysenck EysenckPersonality Personality Inventory, London: University of London Press. Eysenck, H. J., & Questionnaire. & Eysenck, S. B. G. (1975). Manual Manualof ofthe theEysenck EysenckPersonality Personality Questionnaire, San Diego, CA: EdITS. Fiske, D. W. (1949). Consistency of the factorial structures of personality ratings from Journalof ofAbnormal Abnormaland andSocial SocialPsychology, Psychology,44,44, 329-344. 329-344. different sources. Journal ofof the Springfield: Charles C Thomas. Foa, U. G., & & Foa, Foa, E. E. B. B. (1974). Societal Societalstructures structures themind. mind, Springfield: Charles C Thomas.

CHAPTER CHAPTER 28 28

RETURNOF OFmE THEBIG BIGFIVE FIVE RETURN

761 761

T., Ossorio, Ossorio, A. G., & & Coffey, Coffey, H. H. S. S. (1951). (1951).The Theinterpersonal interpersonal dimension dimension Freedman, M., Leary, T., 20,20, 143-161. of personality. Journal Journal of ofPersonality, Personalityy 143-161. Goffman, Goffman, E. (1959). The presentation presentationof ofself selfinineveryday everydaylife. life.Garden GardenCity, City,NY: NY:Doubleday Doubleday Anchor Books. analysis ofpersonality scales Goldberg, L. R R. (1970). Why Why measure measurethat thattrait? trait?An Anhistorical historical analysis of personality scales and inventories inventories (ORI (ORI Tech. Tech. Rep. Rep. 10, 10,No. No.3). 3).Eugene: Eugene: Oregon Oregon Research Research Institute. Institute. a taxonomy of trait Goldberg, L. R. (1977, August). Language Languageand andpersonality: personality:Developing Developing a taxonomy of trait descriptive descriptiveterms. terms.Invited Invited address addresstotothe theDivision Division of ofEvaluation Evaluation and andMeasurement Measurement atat the annual meeting of the American Psychological Association, San Francisco. Some ruminations ruminationsabout aboutthe the structure of individual differences: Goldberg, L. R R. (1980, May). May). Some structure of individual differences: Developingaacommon commonlexicon lexicon forthethe major characteristics of human personality. Developing for major characteristics ofhuman personality. PaperPaper presented at the annual meeting of the Western Psychological AS8ocation, Assocation, Hono Honopresented lulu, HI. R. (1981). Language and individual differences: The search for universals in Goldberg, L. R Review of ofpersonality personalityand andsocial socialpsychology psychology personality lexicons. In L. Wheeler (Ed.), Review (Vol. 2, pp. CA: Sage. pp. 141-165). Beverly Hills, CA: R. (1982). From From Ace Ace to to Zombie: Zombie: Some Some explorations explorations in in the the language language of of personal personalGoldberg, L. R personality assessment assessment ity. In C. D. Spielberger & J. N. Butcher (Eds.), Advances in personality (Vol. 1, pp. 203-234). Hillsdale, NJ: Erlbaum. R. (1983, June). The The magic magicnumber numberfifive, plusororminus minustwo: two: Some conjectures ve, plus Some conjectures Goldberg, L. R the dimensionality dimensionalityofof personality descriptors. Paper presented a research seminar. Paper presented at aatresearch seminar, on the personality descriptors. Gerontology Research Center, Baltimore. Gerontology R. (1986). The validity of rating procedures to index the hierarchical level of of Goldberg, L. R Journalof ofMemory Memoryand andLanguage, Language, 323-347. 25,25, 323-347. categories. Journal R. (1990). factor (1990). An alternative "description of personality": The Big-Five factor Goldberg, L. R structure. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 1216-1229. Psychology, 59, 59, 1216-1229. R. (1992). The development of markers for for the Big-Five factor structure. Goldberg, L. R PsychologicalAssessment, Assessment,4, 4, 26-42. Psychological 26-42. Goldberg, L. R R. (1993). (1993).The Thestructure structureof ofphenotypic phenotypicpersonality personalitytraits. traits.American AmericanPsychologist, Psychologist, 48, 26-34. 48, R., Norman, Norman, W. W.T., T.,& &Schwartz, Schwartz,E. E.(1972). (1972).The Thecomparative comparativevalidity validityofofquestion questionGoldberg, L. R, naire data (16 PF scales) and objective test data (O-A Battery) predicting five five peer peerResearch Bulletin, 12, 1-18. rating criteria. Oregon Oregon Research ResearchInstitute Institute Research Bulletin, 12, 1-18. Goldberg, L. R, R., Norman, W. W.T., T.,& &Schwartz, Schwartz,E. E.(1980). (1980).The Thecomparative comparativevalidity validityof ofquestion questionpeernaire data (16 PF scales) and objective test data (O-A Battery) in predicting five five peer 4, 4, 183-194. PsychologicalMeasurement, Measurement, 183-194. rating critera. Applied Psychological Guilford, J. P. (1948). Factor analysis in a test development program. Psychological PsychologicalReview, Review, 55, 79-94. 55, Guilford, J. P. (1959). Personality. Personality.New NewYork: York:McGraw McGrawHill. Hill. Hampson, S. S. E., E., Goldberg, Goldberg, L. L. R, R., & &John, John,O. O.P. P.(1987). (1987).Category Categorybreadth breadthand andsocial socialdesirabil desirabilPersonality,1,1,241-258. 241-258. for 573 personality terms. European Journal of of Personality, ity values for R. (1986). Category breath and hierarchical Hampson, S. E., John, O. P., & Goldberg, L. R structures in personality: Studies of asymmetries in judgments of trait implications. Journal ofPersonality Personalityand andSocial SocialPsychology, Psychology, 37-54. Journal of 51,51, 37-54. & McKinley, McKinley, J.J. C. C. (1983). (1983). The TheMinnesota MinnesotaMultiphasic Multiphasic Personality Inventory R., & Personality Inventory Hathaway, S. R, manual. New York: Psychological Corporation. manual. first stages ofof thethe Groningen taxonomy studystudy of of Hofstee, W. K. D. (1976). (1976). Dutch Dutch traits: traits:The The first stages Groningen taxonomy personality descriptive descriptiveadjectives. adjectives.Unpublished Unpublishedmanuscript, manuscript,University UniversityofofGroningen, Groningen, personality The Netherlands.

762 762

WIGGINS WIGGINS AND ANDTRAPNELL TRAPNELL

Hogan, J., Arneson, S., Hogan, R., & Jones, S. (1986). Development Development and validation validation of of personnel personnel selection selection procedures procedures for for the job job of of habilitation hahilitation therapist. therapist Tulsa, OK: Hogan Assessment Systems. Hogan, J., & Hogan, R. (1986). Hogan Hogan Personnel Personnel Selection Series manual. manual Minneapolis, MN: National Computer Systems. National Development and validation validation of of an employee employee Hogan, J., Hogan, R., & Briggs, S. (1984). Development selection for combination combination drivers. Tulsa, OK: University of Tulsa. selection test for line Hogan, J., Peterson, S., Hogan, R., & Jones, S. (1985). Development Development and validation validation of of a linehaul driver inventory, Tulsa, OK: University of of Tulsa. haul driver selection selection inventory. A socioanalytic socioanalytic theory of personality. In M. M. Page (Ed.), 1982 Nebraska Nebraska Hogan, R. (1983). A Symposium and Research (pp. 55-89). Symposium on Motivation: Motivation: Personality-Current Personality—Current Theory Theory and of Nebraska Press. Lincoln: University of Hogan Personality Personality Inventory Inventory manual. manual Minneapolis, MN: National Com ComHogan, R. (1986). Hogan puter Systems. Hogan, R., Carpenter, B. N., Briggs, S. R., & Hansson, R. O. (1984). Personnel assessment J. J. J. Bernardin Bemardin & & D. A. A. Bownas (Eds.), (Eds.), Personality Personality assessment assessment in personnel selection. In 1. organizations (pp. 21-52). New York: Praeger. in organizations Hogan, R., Jacobson, G., Hogan, J., & Thompson, B. (1987). Development Development and validation validation of a service operations dispatcher dispatcher selection inventory. inventory. Tulsa, OK: Hogan Assessment of service operations Systems. August), The structure structure of of personality. personality. Paper presented Hogan, R., & Johnson, J. A. (1981, August). of the American Psychological Association, Los Angeles. at the annual meeting of (1985). Socioanalytic Socioanalytic theory: theory: An An alternative alternative to to armadillo armadillo Hogan, R., Jones, W., & Cheek, J. M. (1985). psychology. In B. R. Schlenker (Ed.), The self self and and social social life (pp. 175-198). New York: McGraw-Hill. Holland, J. L. (1985). Making Making vocational vocational choices. Englewood Cliffs, NJ: Prentice-Hall. identified in the Howarth, E. (1976). Were Cattell's "personality sphere" factors correctly identified British Journal Journal of of Psychology, Psychology, 67, 213-230. first instance? British Personality: An An Howarth, E., & Browne, J. A. (1971). An item factor analysis of the 16 PF. Personality: International Journal, Journal 2, 117-139. International A sequential system for personality scale development. In C. D. Jackson, D. N. (1970). A topics in clinical and and community community psychology psychology (Vol. 2, pp. Spielberger (Ed.), Current topics 61-96). New York: Academic Press. manual (3rd ed.). Fort Huron, MI: MI: Research Jackson, D. N. (1984). Personality Personality Research Form manual Psychologists Press. Jackson, D. N. (1986). The process of responding in personality assessment. In A. Angle Angleitner & J. S. Wiggins (Eds.), Personality Personality assessment assessment via questionnaires questionnaires (pp. 123-142). New York: Springer-Verlag. John, O. P., & Angleitner, A. (1982). Explorations Explorations into everyday everyday person-descriptions: person-descriptions: Some Some preliminary preliminary data (Research Report of the Trait-Taxonomy Project, No. 1). Department of of Psychology, University of of Bielefeld, Bielefeld, Germany. John, O. P., Angleitner, A., & Ostendorf, F. (1988). The lexical approach to personality: A A historical reveiw of trait taxonomic research. European European Journal Journal of of Personality, Personality, 2, 171-203. Taxonomies John, O. P., Goldberg, L. R., & Angleitner, A. (1984). Better than the alphabet: Taxonomies of personality-descriptive terms in English, Dutch, and German. In H. C. J. Bonarius, of Personality psychology psychology in Europe: Europe: Theoretical G. L. M. Van Heck, & N. G. Smid (Eds.), Personality and empirical empirical developments developments (Vol. 1, pp. pp. 83-100). Berwyn, PA: Swets North America. and

CHAPTER CHAPTER Z8 28

RETURN RETURNOF OFmE THEBIG BIGFIVE FIVE

763 763

John, O. P., Hampson, S. S. E., & Goldberg, L. R R. (1991). The basic level in personality personalitydifferent contexts. Journal of of trait hierarchies: Studies of trait use and accessibility in different 60, 60, 348-361. Personality Personalityand andSocial SocialPsychology, Psychology^ 348-361. Johnson, J. A. (1981). The "self-disclosure" and "self-presentation" "self-presentation" views of item response Psychology, dynamics and personality scale validity. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 40, 40, 761-769. 761-769. Johnson, J. A. (1983). Criminality, creativity, and craziness: Structural similarities in three types of nonconformity. In W. S. S. Laufer Laufer & J. M. Day (Eds.), Personality Personalitytheory, theory,moral moral development, development, and andcriminal criminalbehavior behavior(pp. (pp.81-105). 81-105).Lexington, Lexington,MA: MA:D. D.C.C.Heath. Heath. Psychologicaltypes types(H. (H.G. G. Baynes, Baynes,Trans., Trans.,revised revised by byR R.F.F.C. C.Hull). Hull). Jung, G. C. (1971). Psychological Princeton, NJ: Princeton University Press. (Original work published 1923) (1974). Is the 16 16 PF factorially factorially valid? Journal of of Personality Personality Karson, S., & O'Dell, J. W. (1974). Assessment, Assessment,38, 38,104-114. 104-114. & Fiske, D. W. (1951). The The prediction predictionof ofperformance performanceinin clinical psychology. Kelly, E. L., & clinical psychology. Ann Arbor: University of Michigan Press. psychology of of personal personal constructs constructs(Vol. (Vol.1). 1).New NewYork: York:Norton. Norton. Kelly, G. A. (1955). The psychology for complementarity complementarity in Kiesler, D. J. (1983). The 1982 Interpersonal Circle: A taxonomy for PsychologicalReview, Review,90, 90,195-214. 195-214. human transactions. Psychological D.J. (1988). Therapeutic Therapeuticmetacommunication: metacommunication: Therapist impact disclosure as feedback Therapist impact disclosure asfeedback Kiesler, D. J. (1988). in psychotherapy. psychotherapy. Palo PaloAlto, Alto,CA: CA:Consulting Consulting Psychologist Psychologist Press. Press. LaForge, R R. (1985). The early development of the Freedman-Leary-Coffey Freedman-Leary-Coffey Interpersonal Laforge, Journalof ofPersonality PersonalityAssessment, Assessment, 613-625. System. Journal 49,49, 613-625. Interpersonaldiagnosis diagnosis personality. New York: Ronald Press. ofof personality. New York: Ronald Press. Leary, T. (1957). Interpersonal Psychological Loevinger, J. (1957). Objective tests as instruments of psychological theory. Psychological Reports, 3,3,635-694. 635-694. Reports, Proposalfor for an aninstitute institutefor forpersonality personality assessment research. assessment andand research. Mackinnon, D. W. (1948). Proposal Department of Psychology, Berkeley. Unpublished memo, University of California, Department R. R R. (1982). Consensual validation of of personality traits: Evidence from from self-reports self-reports McCrae, R Journalof ofPersonality Personalityand andSocial Social Psychology, 293-303. Psychology, 43, 43, 293-303. and ratings. ratings. Journal R. R R. (1986). Well-being scales do not measure social desirability. Journal of of McCrae, R Gerontology,41, 41,390-392. 390-392. Gerontology, McCrae, R R. R R. (Ed.). (1992). The five-factor five-factor model: Issues and applications [Special issue]. Journal Journalof ofPersonality, Personality,60(2). 60(2). McCrae, R R. R, R., & Costa, P. T., Jr. (1983). Social desirability scales: More substance than Psychology, 51, 51, 882-888. style. Journal Journalof of Consulting Consultingand andClinical Clinical Psychology, 882-888. dispositions: Personality McCrae, R Emerginglives, lives,enduring enduring dispositions: Personality R. R, R., & & Costa, P. T., Jr. (1984). Emerging in adulthood. adulthood. Boston: Boston:Little, Little,Brown. Brown. McCrae, R R. R, R., & Costa, P. T., Jr. (1985a). Comparison Comparison of of EPI and Psychoticism scales with measures of thefive-factor five-factor model of personality. Personality Personalityand andIndividual IndividualDifferences, Differences, 6, 6, 587-597. McCrae, R R. R, R., & Costa, P. T., Jr. (1985b). Openness to experience. In R R. Hogan & W. H. Jones (Eds.), Perspectives Perspectivesininpersonality personality(Vol. (Vol.1, 1,pp. pp.145-172). 145-172).Greenwich, Greenwich,CT: CT: JAI Press. R. R, R., & Costa, P. T., Jr. (1985c). Updating Norman's "adequate taxonomy": McCrae, R Intelligence and personality dimensions in natural language and in questionnaires. Journalof ofPersonality Personalityand andSocial Social Psychology, 710-721. Journal Psychology, 49,49, 710-721. R. R, R., & Costa, P. T., Jr. (1987). Validation of the five-factor model of personality McCrae, R 52,52, across instruments and observers. Journal Journalof of Personality Personalityand andSocial SocialPsychology, Psychology, 81-90.

764 764

WIGGINS WIGGINSAND ANDTRAPNELL TRAPNELL

McCrae, R R. R, R., & Costa, P. T., Jr. (1989a). Reinterpreting the Myers-Briggs Type Indicator from Personality, from the perspective of the five-factor model of personality. Journal of of Personality, 57, 57, 17-40. McCrae, R R. R, R., & Costa, P. T., Jr. (1989b). The structure of interpersonal traits: Wiggins's circumplex and thefive-factor five-factor model. Journal Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, 56, 56, 586-595. McCrae, R Emerging lives enduring R. R, R., & & Costa, P. T., Jr. (1990). Personality Personalityininadulthood: adulthood: Emerging lives enduring dispositions. dispositions. New New York: York:Guilford Guilford Press. Press. McCrae, R R. R, R., Costa, P. T., Jr., & Busch, C. M. (1986). Evaluating comprehensiveness in personality systems: The California Q-Set and the five-factor model. Journal of of Personality, 4, 430-446. Personality,554, 430-446. Meehl, P. E. (1945). The dynamics of "structured" personality tests. Journal of Clinical of Clinical Psychology, Psychology, 1, 1, 296-303. Personalityand andassessment. assessment.New NewYork: York:Wiley. Wiley. Mischel, W. (1968). Personality Explorations in personality. personality. New NewYork: York:Oxford Oxford University UniversityPress. Press. Murray, H. A. (1938). Explorations Manual:AAguide guidetotothe thedevelopment developmentand anduse use & McCaulley, M. H. (1985). Manual: Myers, I. B., & ofof Myers-Briggs Type TypeIndicator. Indicator.Palo PaloAlto, Alto,CA: CA:Consulting ConsultingPsychologists PsychologistsPress. Press. the Myers-Briggs R. (1990). The The construct construct validity validity of of social social desirability. desirability. American American Nicholson, R. A., & Hogan, R Psychologist,445, 290-292. Psychologist, 5, 290-292. Norman, W. T. (1963). Toward an adequate taxonomy of personality attributes: Replicated factor factor structure in peer nomination personality ratings. Journal Journalof ofAbnormal Abnormal and andSocial Social Psychology, 66, 66,574-583. 574-583. Psychology, Normative operating characteristics Norman, W. 2800personality personalitytrait traitdescriptors: descriptors: Normative operating characteristics W. T. (1967). 2800 for a university university population. population. Ann Ann Arbor: Arbor: University University of of Michigan, Michigan, Department Department of of Psy Psyfor chology, Peabody, D. (1967). Trait Trait inferences: inferences: Evaluation Evaluation and and descriptive descriptive aspects. aspects.Journal Journalof ofPersonal Personal644 ). ity and Social Social Psychology PsychologyMonograph, Monograph,7(4, 7(4,Whole WholeNo. No. 644). Peabody, D., & Goldberg, Goldberg, L. L.R R.(1989). (1989).Some Somedeterminants determinantsof offactor factorstructures structuresfrom frompersonal personality trait-descriptors. Journal 57,57, 552-567. Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, 552-567. Raza, S., Metz, D., Dyer, P., Coan, T., & Hogan, J. (1986). (1986). Development and validation validation of of personnel selectionprocedures proceduresfor forhospital hospitalservice servicepersonnel. personnelTulsa, Tulsa,OK: OK:University University personnel selection of Tulsa. Sells, S. B., Demaree, R R. G., & Will, D. P., Jr. (1970). Dimensions of personality: I. Conjoint Conjoint factor factor structure of Guilford Guilford and Cattell trait markers. Multivariate MultivariateBehavioral BehavioralResearch, Research, 5, 5, 391-422. Steiger, J. H. (1979). Factor indeterminancy 1970's: Some interesting indeterrninancy in the 1930's and 1970's: Psychometrika, 44, 44,157-167. 157-167. parallels. Psychometrika, interpersonaltheory theoryof ofpsychiatry. psychiatry.New NewYork: York:Norton. Norton. Sullivan, H. S. (1953). The interpersonal PsychologicalReview, Review,41, 41,1-32. 1-32. Thurstone, L. L. (1934). The vectors of mind. Psychological lexicographicalstudy studyofofthe thepersonality personalitydomain domain (HeymansBulletins Bulletins Tomas, A. (1977). A lexicographical (Heymans No. HB-77-296-EX). Groningen, The Netherlands: Psychologische Instituten R. U. Groningen. Trapnell, P. D., & Wiggins, J. S. (1990). Extension of the Interpersonal Adjective Scales to of Personality Personality and Social include the Big Five dimensions of personality. Journal of Psychology, Psychology, 59, 59,781-790. 781-790. Tupes, E. C, C., & Christal, R. E. (1958). Stability of personality trait rating factors obtained under diverse conditions. USAF WADC WADC Technical TechnicalNote, Note,No. No.58-61. 58-61.

CHAPTER CHAPTER 28 28 RETURN RETURNOF OFmE THEBIG BIGFIVE FIVE

'

765 765

Tupes, E. c., C, & Christal, R. E. (1961). Recurrent personality factors based on trait ratings. USAF ASD Technical USAFASD TechnicalReport, Report,No. No.61-97. 61-97. C, & Kaplan, M. N. (1961). (1961). Relationships between personality traits, physical Tupes, E. c., AFD proficiency, and cadet effectiveness effectiveness reports of Air Force Academy Cadets. USAF USAFAFD Technical 61-53. TechnicalReport, Report,No. No. 61-53. Waller, N. G., & Ben-Porath, Y. S. (1987). (1987). Is it time for clinical psychology to embrace the 887-889. five-factor five-factor model modelof ofpersonality? personality?American AmericanPsychologist, Psychologist,42,42, 887-889. Wiggins, J. S. (1982). Circumplex models of interpersonal behavior in clinical psychology. In P. C. Kendall & J. N. Butcher (Eds.), Handbook Handbook of of research researchmethods methodsininclinical clinical psychology (pp. 183-221). New York: Wiley. Wiggins, J. J. S. S. (1984). (1984). Cattell's CattelPssystem system from from the theperspective perspective of ofmainstream mainstream personality personality theory. theory. MultivariateBehavioral BehavioralResearch, Research, 176-190. Multivariate 19, 19, 176-190. Wiggins, J. S. (1985). Symposium: Interpersonal circumplex models: 1948-1983 [Commen [CommenJournalof ofPersonality PersonalityAssessment, Assessment, 626-631. tary]. Journal 49, 49, 626-631. J. S. S.(1991). (1991).Agency Agencyand andcommunion communionas asconceptual conceptualcoordinates coordinatesfor forthe theunderstanding understanding Wiggins, J. and measurement of interpersonal behavior. In W. Grove & D. Cicchetti (Eds.), Thinkingclearly clearlyabout aboutpsychology: psychology:Essays Essaysininhonor honorof ofPaul PaulE.E.Meehl Meehl(Vol. (Vol.2,2,pp. pp. Thinking 89-113). Minneapolis: University of Minnesota Press. J.S.S.(1995). (1995).Interpersonal Interpersonal Adjective Scales: Professional Manual. Odessa, FL: PsychoWiggins, J. Adjective Scales: Professional Manual. Odessa, FL: Psycho logical Assessment Resources. Wiggins, J. J. S. S. (1996a). (1996a). An An informal informal history of of the the interpersonal interpersonal circumplex circumplex tradition. tradition. Journal Journal of Personality PersonalityAssessment, Assessment,66,66, 217-233. of 217-233. Thefifive-factor modelofofpersonality: personality:Theoretical Theoretical perspectives. Wiggins, J. S. (Ed.). (1996b). The ve-factor model perspectives. Guilford Press. New York: Guilford Wiggins, J. S., Phillips, N., & Trapnell, P. (1989). Circular reasoning about interpersonal clasbehavior: Evidence concerning some untested assumptions underlying diagnostic clas Journalof ofPersonality Personalityand andSocial Social Psychology, 296-305. sication. Journal Psychology, 56, 56, 296-305. Wiggins, J. S., & Pincus, A. L. (1989). Conceptions of personality disorders and dimensions PsychologicalAssessment: Assessment: Journal of Consulting Clinical PsycholAA Journal of Consulting andand Clinical Psychol of personality. Psychological ogy, 1, 1, 305-316. ogy, Annual Review Wiggins, J. S., & Pincus, A. L. (1992). Personality structure and assessment. Annual of of Psychology, Psychology,43, 43,417-440. 417-440. five Wiggins, J. S., & Trapnell, P. D. (1996). A dyadic-interactional perspective on the fiveve-factor model factor model. In J. S. Wiggins (Ed.). The Thefifive-factor modelof ofpersonality: personality:Theoretical Theoretical perspectives perspectives (pp. (pp.88-162). 88-162).New NewYork: York:Guilford Guilford Press. Press.

This Page Intentionally Left Blank

CHAPTER 29 CHAPTER 29

EXTRAVERSION AND ITS POSITIVE EMOTIONAL CORE DAVID WATSON AND LEE ANNA CLARK THE OF IOWA THE UNIVERSITY UNIVERSITY OF IOWA

I. INTRODUCTION INTRODUCfION The extravert is a very familiar character. To most people, the term "extravert" quickly conjures up an image of one who seeks out and enjoys the companionship who is is poised, confident, and and facile facile in social situations. situations. Other Otherdescrip descripof others—one others-one who tors of the extravert that are perhaps less central, but nevertheless commonly recognized, include bold, assertive, lively, energetic, enthusiastic, and optimistic. opposites—introverts—can be be broadly broadly sketched Conversely, their characterological opposites-introverts-can effective. as more quiet and reserved, more socially aloof, and less interpersonally effective. Introversion-Extraversion (which we will simply call "Extraversion") also is an extremely important concept in trait psychology. For example, an Extraversion factor can be identified in virtually every widely used multidimensional personality PersonaUty Questionnaire (MPQ; Tel(MPQ; Tel inventory, including the Multidimensional Personality (PRF; Jackson, 1984), and the legen, 1982, 1985), the Personality Research Form (PRF; (16PF; Cattell, Eber, & Tatsuoka, 1980). Sixteen Personality Factor Questionnaire (16PF; specifiMoreover, Extraversion factors have been isolated in many instruments not specifi cally designed to assess personality traits, including the California Psychological cally Inventory (CPI; Gough, 1987) and the the Minnesota Minnesota Multiphasic Multiphasic Personality Personality Inventory Inventory Inventory (CPI; Gough, 1987) and (MMPI; Hathaway Hathaway & & McKinley, see Costa, Costa, Busch, & McCrae, McKinley, 1943; 1943; see Busch, Zonderman, Zonderman, & McCrae, (MMPI; 1986). Similarly, aa general general Extraversion Extraversion dimension dimension has has been been identified identified in in an an inven1986). Similarly, inven tory designed designed to to measure measure traits traits and and behaviors behaviors relevant relevant to to personality personality disorder disorder tory (Clark, 1993; Clark, Vorhies, Vorhies, & & McEwen, McEwen, 1994). (Clark, 1993; Clark, 1994). 767 HAIVDBOOK OFOF PERSONALITY PSYCHOLOGY HANDBOOK PERSONALITY PSYCHOLOGY

767

© 1997 BY ACADEMIC PlESS PRESS.. COPYRIGHT C ALL IIGIm Of REPRODUCTION IN ANY fORM RESERVED.

ALL RIGHTS OF REPRODUCHON IN ANY FORM RESERVED.

768 768

WATSON ANDCLARK CLARK WATSON AND

Extraversion has been included as a higher-order factor factor in every every Furthermore, Extraversion major major taxonomic scheme of of personality traits that has been developed during the past 50 years. For For example, extensive interest has recently recently focused focused on a five-factor five-factor model of personality that invariably includes Extraversion of personaUty Extraversion as one of of its constituent constituent dimensions (e.g., Digman, 1990; Goldberg, 1993; 1993; McCrae & Costa, 1985, 1987). Other Other theorists (e.g., Eysenck Eysenck & Eysenck, 1975; Tellegen, 1985) have instead instead pro proposed three-factor three-factor structures, but nevertheless retain retain Extraversion (or a conceptu conceptuconsensual ally comparable dimension) in their models. Thus, although we still lack a consensual taxonomy of of personality personality traits, every every model includes Extraversion Extraversion as one of of its taxonomy constituent factors. factors. In the following following sections we explore the nature, components, and correlates correlates of of Extraversion. Before Before doing so, however, we need to clarify clarify two important important points. concept is often often viewed typologically, with "extraverts" and "introverts" "introverts" First, this concept defining defining discrete and self-limited self-limited categories. However, research research consistently shows Extraversion is a dimension dimension of of individual differences. differences. That is, although extreme extreme that Extraversion found at both ends of of the factor, scores can occur anywhere along a cases can be found individuals obtain intermediate scores on the broad continuum. Moreover, most individuals identifiable as either either introverts or extraverts. trait, and thus are not clearly identifiable Second, as noted earlier, extraversion consistently emerges as a higher-order higher-order disposition in taxonomic schemes of of personality traits. Higher-order Higher-order traits represent represent disposition hierarchy of of dispositions: this is the level at the broadest, most general level in the hierarchy attempt to explain individual individual differences differences with the fewest fewest possi possiwhich personologists attempt broadly applicable, dimensions. Extraversion is, therefore, therefore, a very ble, and most broadly general dimension dimension that is itself itself composed of of more specific, primary traits; its exisexis tence is inferred inferred from from empirically empirically observed covariations among these primary traits. attempt to identify identify the most important important components of of this trait, but our We will attempt focus will ultimately ultimately be on the higher-order higher-order disposition itself. focus n. EARLY CONCEPTIONS OF EXTRAVERSION n of the Trait A. Inconsistent and Contradictory Aspects of Although the various theoretical conceptualizations conceptualizations of of Extraversion all share some Although prominent component component traits (e.g., talkativeness, sociability), a closer inspection of of prominent contradictory features. For exam examthese models reveals some inconsistent and even contradictory description offered offered by Costa and McCrae (1985) closely resembles the ple, the description popular conception of the extravert: "In addition to liking people and preferring preferring popular conception of large groups and gatherings, extraverts are also assertive, active, and talkative; they excitement and stimulation, and tend to be cheerful cheerful in disposition. They are like excitement optimistic*' (p. 10). upbeat, energetic, and optimistic" consider the the portrait offered offered by Eysenck and and Eysenck (1975): In contrast, consider The typical extravert is sociable, likes parties, has many friends, needs to have people to talk to, and does not like reading or studying by himself. He craves

29 EXTRAVERSION EXTRAVERSION CHAPTER 29

769 769

excitement, takes chances, often sticks his neck out, out, acts on the spur of the moment, and is generally an impulsive individual. He is fond of practical jokes, Hkes change; change; he he is is carefree, carefree, easy-going, easy-going, always has a ready answer, and generally likes optimistic, and likes Ukes to 'laugh *laugh and be merry' m e r r y. ' . . .. .. [A]Itogether [A]ltogether his feelings are reUable person. person, (p. (p. 5) not kept under tight control, and he is not always a reliable These two descriptions are entirely consistent in several important important ways. Both Both cheerful, and optimistic, and that that extraverts are sociable, talkative, cheerful, emphasize that they enjoy enjoy change and excitement excitement in their their lives. Beyond Beyond that, however, there there are they some sharp divergences. Most notably, Eysenck Eysenck and Eysenck-but Eysenck—but not Costa and McCrae—argue that extraverts are impulsive, risk-taking, risk-taking, and somewhat unreliable. McCrae-argue Generally Generally speaking, the Eysenckian Eysenckian extravert extravert seems more poorly poorly socialized than individual described described by Costa and McCrae. does the individual An inspection of of other relevant conceptualizations reveals even greater disAn dis Hogan (1983) and Tellegen (1985) have emphasized crepancies. For example, both Hogan that extraverts, in addition addition to being socially facile facile and influential, are ambitious, that achievement-oriented individuals. Thus, despite their shared char charhardworking, and achievement-oriented somewhat removed removed from from the unrelia unreUaacteristics, these individuals ultimately seem somewhat ble, impulsive, and fun-loving person described described by the Eysencks. fun-loving person B, B.

Origin of the Extraversion Concept

reflect how the Extraver ExtraverWhy have these discrepancies emerged? In large part, they reflect sion construct construct has evolved over time, a topic we consider shortly. They also reflect, reflect, however, the multifaceted multifaceted nature of the original theoretical notion itself. The con concept of introversion-extraversion introversion-extraversion was introduced introduced by Jung (1921), although William WiUiam proposed somewhat somewhat similar typological typological schemes James (1907) and others had earlier proposed Guilford & Braly, 1930, Jung's (see J. P. Guilford 1930, for a review of these early models). In Jung's theory, introversion extraversion are not personality traits per se, but instead introversion and extraversion represent represent differing differing attitudes or orientations orientations toward the world. Introverts Introverts are ori oriented toward internal, subjective subjective experience, focusing focusing on their own thoughts, feel feelented ings, and perceptions. Consequently, they tend to be introspective, ruminative, and self-preoccupied, and appear appear aloof, quiet, unsociable, and reserved reserved to others. In self-preoccupied, contrast, extraverts are more more externally and objectively objectively focused; they are more concerned concerned with other other people and the world around around them, and oriented more toward action than thought. They are seen as active, outgoing, and sociable by others. In modem modem dispositional terms this Jungian concept clearly is multidimensional, multidimensional, combining several several essentially essentially unrelated unrelated traits. As expected, extraverts are more more combining described as being more active and impulsive, less dysphoric, sociable, but are also described self-preoccupied than introverts. This multidimension multidimensionand as less introspective and self-preoccupied aUty was quickly recognized recognized by investigators seeking to measure the concept concept empiri empiriality cally (J. P. Guilford Guilford & Braly, 1930). afterward, J. P. Guilford Guilford and Guilford Guilford 1930). Soon afterward, (1934,1936)—using factor analysis-demonstrated analysis—demonstrated that (1934, 1936)-using the then-new then-new technique of factor existing introversion-extraversion introversion-extraversion items could be factored factored into several several distinct traits, traits.

770 770

WATSON WATSON AND ANDCLARK CLARK

including sociability, neuroticism, introspectiveness, impulsivity, and masculinity masculinityfemininity. c. C. Guilford's Model

In succeeding succeeding studies, J. P. Guilford Guilford and his colleagues colleagues concentrated concentrated on devising devising of these specific specific traits; these efforts efforts ultimately ultimately led to creation creation scales to assess each of of the Guilford-Zimmerman Guilford-Zimmerman Temperament Guilford & Zim Zimof Temperament Survey (GZTS; J. P. Guilford Guilford, Zimmerman, Guilford, 1976). The most merman, 1949; see also J. S. Guilford, important finding finding to emerge from from these studies was that the sociability, sociabiUty, negative important emotionality, and introspectiveness/impulsivity introspectiveness/impulsivity items remained remained separate, even at the higher-order higher-order factor factor level. Thus, Guilford Guilford identified identified three higher-order higher-order factors factors of which are especially relevant here. The first, which he calls in these data, two of of three primary traits: Ascendance (dominance versus Social Activity, is composed of SociabiUty (social interest versus aloofness), and General Activity submissiveness), Sociability higher-order dimension clearly can be identified identified (energy versus sluggishness). This higher-order Extraversion construct construct that is the focus of of our discussion. as the Extraversion Guilford applied the term "Introversion-Extraversion" "Introversion-Extraversion" to his Ironically, Guilford higher-order factor factor composed composed of of the Rhathymia Rhathymia (restraint (restraint versus impulsivity) and higher-order Thoughtfulness Thoughtfulness (reflective (reflective versus unreflective) unreflective) scales. Thus, Guilford Guilford viewed viewed the impulsive and unreflective unreflective aspects of of the original Jungian Jungian concept concept as the core of of Extraversion. Ultimately, however, the factor factor names are less important than the finding that the sociability/energy finding sociability/energy and impulsivity/unreflective impulsivity/unreflective aspects of of the Jungian concept concept ultimately define separate factors, even at the broadest broadest level.

D. Eysenck's Model It is in this context that we must view the evolution of of H. J. Eysenck's model of Extraversion. Consistent with Guilford's interpretation of Guilford's interpretation of of the construct's core, Eysenck used the Rhathymia scale as the starting point for Rhathymia for the construction of of his original Extraversion scale, which was contained in the Maudsley Personality Extraversion Inventory (MPI; Eysenck, 1959). His analyses of of 261 Guilford Guilford items ultimately led a scale that was largely composed of him to construct construct composed of questions from from Rhathymia Rhathymia and from the GZTS Sociability scale (see Eysenck & Eysenck, 1969; J. P. Guilford, from Eysenck Guilford, 1975). This item composition composition was essentially retained in a later version of of the scale, which was included in the Eysenck Personality Inventory (EPI; Eysenck & Eysenck, 1968). Extraversion originally contained strong elements of of both both Thus, Eysenck's Extraversion Sociability and Impulsivity. However, Guilford's Guilford's finding finding that that these components are independent of of one another another has since been replicated by several other other largely independent generally acknowledged acknowledged that Impulsivity should be split investigators, and it is now generally off from from Sociability, Energy, and other aspects of of Extraversion Extraversion (e.g., J. P. Guilford, Guilford, off patterns 1975; Zuckerman, Kuhlman, & Camac, 1988). Eysenck observed similar patterns largely removed removed the Impulsivity Impulsivity component component from from his latest in his own data, and so largely

CHAPTER CHAPTER 29 29

EXTRAVERSION EXTRAVERSION

771 771

contained in the Eysenck Personality Personality Questionnaire Questionnaire Extraversion scale, which is contained (EPQ; Eysenck & Eysenck, 1975). Data to be presented subsequently, however, (EPO; Eysenck indicate that the EPO impulsivity-related qualities to EPQ Extraversion Extraversion scale still taps impulsivity-related construct. a greater extent than do other measures of the construct. Unlike most of of the other theoreticians in this area, Eysenck Eysenck always has been primarily interested in the higher-order higher-order factor factor itself, and has never never articulated articulated a systematic structure structure at the primary-trait primary-trait level. More recently, however, he has suggested that Extraversion Extraversion is composed of nine primary primary traits, several of of which suggested clearly overlap with Guilford's Guilford's (Le., (i.e.. Sociable, Dominant, Assertive, Surgent, Active, of which continue to shade toward toward Impulsivity Impulsivity (i.e., SensationLively), but others of (Le., Sensation Carefree) (Eysenck (Eysenck & & Eysenck, 1985). Seeking, Venturesome, Carefree) Catteirs Model E. Cattell's Cattell worked more or less independently independently from from Guilford Guilford and ultiand Eysenck, but ulti identified a very similar higher-order higher-order Extraversion Extraversion dimension. This line of of mately identified AUport and Odbert's (1936) effort effort to compile an exhaustive exhaustive research originated in Allport Ust of of trait-related trait-related terms in the English language. Allport AUport and Odbert Odbert eventually list represented trait dispositions. Cattell settled on a list of 4,504 terms that clearly represented reduced this set to a more manageable manageable pool of (1945, 1946) reduced of 35 clusters through Subsequent factorial factorial studies led Cat Catrational content sortings and cluster analyses. Subsequent construct his set of of 16 primary traits (see Cattell et aaI., l , 1980). tell to construct Guilford, Cattell always was more interested interested in these primary primary traits than Like Guilford, higher-order dimensions. However, However, factor factor analyses analyses of in higher-order of the 16PF consistently confirm that is largely defined confirm the existence of of a higher-order higher-order Extraversion Extraversion factor factor that aI., 1980): A by five primary primary traits (Cattell et al., A (warmhearted, (warmhearted, easygoing easygoing versus reserved, detached), E F (enthusiastic E (dominant, ascendant versus submissive), F (enthusiastic versus taciturn), H (bold, adventurous adventurous versus shy, timid), and O Q22 (socially (socially enmeshed enmeshed versus autonomous). of the Early Models F. Summary and Integration of higher-order dimensions identified identified by Guilford, Guilford, Eysenck, and Cattell are The higher-order strongly strongly related to one another another (e.g., Eysenck Eysenck & Eysenck, 1969) and have several component component traits in common. The most consistently consistently recurring themes are those of of Ascendance and Sociability-in Sociability—in all of these views, extraverts are gregarious, friendly, friendly, dominant, and socially facile. They enjoy enjoy being with other people and are confident confident and comfortable comfortable when interacting with them. Thus, all of of these views strongly emphasize the sociaVinterpersonal social/interpersonal aspects of of the construct. Other Other common features features may also be noted, however. For For example, both Eysenck and Cattell view extraverts as bold and adventurous individuals who seek further involves excitement and stimulation in their lives; this sensation-seeking further Guilford and Eysenck Eysenck both argue risk-taking in Eysenck's model. Furthermore, Guilford that extraverts are active, lively, and full of energy. Finally, both Eysenck and

772 772

WATSON WATSON AND ANDCLAU CLARK

Cattell include a substantial affective affective component component in their their models-they models—they describe extraverts as cheerful, cheerful, optimistic, and enthusiastic. As we will see, these energy and affective affective components have become even more prominent prominent in recent formulations. formulations.

m. in. CONTEMPORARY CONCEPTIONS OF EXTRAVERSION EXTRAVERSION A. Extraversion in Relation to Positive and Negative Affect Affect More recent formulations formulations have retained many of of the traditional primary traits, including a continued recognition of of the social/interpersonal social/interpersonal component. There have also been some significant significant changes, however, the most notable of of these being an increased focus on the positive emotional aspect of the dimension. aspect of understand these changes, we must briefly briefly examine examine the structure of of subjec subjecTo understand self-rated tive emotional experience. Studies have repeatedly demonstrated demonstrated that self-rated mood is characterized characterized by two dominant dimensions that reflect reflect the crucial role of of mood valence in affective affective experience (Mayer & Gaschke, 1988; Watson, 1988b; Watson & 1988; Tellegen, 1985). Specifically, Speciflcally, negatively negatively valenced valenced mood terms strongly co-occur in mood terms co-occur individuals, and and so combine to form form a broad broad factor factor called "Negative Affect" Affect" (NA); similarly, positively valenced valenced mood mood states also tend tend to co-occur, and so jointly compose the higher-order higher-order dimension dimension of of "Positive Affect" Both of of these Affect" (PA). Both dimensions can be assessed either either as a state (i.e., transient transient mood mood fluctuations) fluctuations) or individual differences differences in general affective as a trait (i.e., stable individual affective level). NA represents one's level of subjective distress and NA NA represents of subjective distress and dissatisfaction. dissatisfaction. High NA reflects reflects a wide range of of negative mood states, including fear, fear, anger, sadness, gUilt, guilt, contempt, disgust, and self-dissatisfaction. self-dissatisfaction. In contrast, contrast, PA PA represents a state of of pleasurable arousal, and reflects and effectively reflects feelings feelings of of being actively actively and effectively engaged. High PA is composed composed of of terms reflecting reflecting enthusiasm enthusiasm (e.g., excited, enthusiastic), enthusiastic), joy (happy, ), (happy, delighted delighted),), energy energy (active, energetic), energetic), mental alertness (attentive, interested interested), and confidence confidence (strong, confident). confident). Terms suggesting warm and affiliative affiliative feelings feelings (e.g., friendly, friendly, sociable, warmhearted warmhearted) ) are also strong markers of of the PA PA factor, factor, a point point we consider consider in a later section. It is important important to note, however, that that these affiliative A measures, including all of affiliative terms have been excluded from from most P PA of our own scales. Thus, significant correlations between P A and various measures of social! PA social/ interpersonal interpersonal tendencies (e.g., social activity, Extraversion) cannot be explained explained by overlapping content. content. Measures Measures of of NA NA and PA are largely independent, typically correlating from from -.10 - . 1 0 to - . .25 2 5 with one another (e.g., Watson, 1988b). Moreover, they tend to have distinctly different different correlates (e.g., Clark & Watson, 1988; Watson, 1988a). For our purposes, the most important important findings concern concern their differential differential relations with personality: state P PA scales are significantly correlated with Extraversion, but not A significantly correlated with Neuroticism, whereas state NA scales are substantially substantially correlated correlated with Neuroti Neuroticism, but are generally unrelated to Extraversion Extraversion (Costa & McCrae, 1980, 1984; Emmons & Diener, 1986; Tellegen, 1985; Watson Watson & Clark, Clark, 1984).

CHAPTER 29 29 CHAPTER

EXTRAVERSION

B. B.

773 773

Costa and and McCrae's Model Model

Thesefindings findings have led researchers to explore more fully the relation between P A PA and Extraversion, and consequently to augment the affective affective component of the Extraversion construct. Costa and McCrae (1980, 1984) were the first investigators (1980,1984) to explore these relations systematically, and they subsequently incorporated these findings 1992). As findingsinto intotheir theirmodel modelofofExtraversion Extraversion(Costa (Costa&&McCrae, McCrae,1985, 1985,1992). Asoutlined outlined (1985, 1992), Extraversion is composed of six by Costa and McCrae (1985,1992), six primary traits or facets, several of which are very similar to those of the earlier models (Assertiveness, Excitement-Seeking, and Activity). One modest change is that Sociability is split correlated facets, Gregariousness (desiring the company of others) into two strongly correlated affection toward others). and Warmth (feelings of affection The most significant significant addition, however, is the inclusion of of a facet called "Posi "Posifacet reflects stable individual differences differences in the tendency to tive Emotions." This facet experience positive emotions; not surprisingly, it is significantly correlated with facets state PA scales. It is noteworthy, however, that all of the other Extraversion facets (especially Warmth, Activity, and Assertiveness) are also related to state PA, at & McCrae, 1984). 1984). Moreover, our own data indicate least to some extent (Costa & that Positive Emotions does not necessarily have the strongest correlations with trait PA measures; in fact, we have found that the Activity, Assertiveness, and PA. Positive Emotions facets all correlate similarly with trait P A. C. C. Tellegen's Tellegen's View: View: Positive Positive Emotionality Emotionality Although Costa and McCrae assign a more prominent role to individual differences differences sociability—the preference for, and enjoyment enjoyment of, others' in PA, they still view sociability-the company-as company—as the core of the dimension (McCrae & & Costa, 1987). 1987). Tellegen (1985; Tellegen et al., 1988), however, has proposed a more radical reformulation reformulation of of the construct. In fact, individual differences A play such a prominent role in Telle differences in P PA Tellegen's conceptualization that he has proposed proposed that the higher-order higher-order disposition be renamed "Positive Emotionality." EmotionaUty." In Tellegen's model, model. Positive Emotionality is divided into four four primary facets. One component, Well-Being, Weil-Being, represents the individual differences differences in PA that define the core of the construct. Two other facets facets reflect reflect the interpersonal tendencies that are more traditionally associated with Extraversion: Social Potency is a measure of ascendance or dominance, whereas Social Closeness assesses sociability sociabihty or affilia affiliation. Tellegen's most interesting primary trait is Achievement: high scorers report that they are ambitious, perfectionistic, and willing to work long hours in pursuit of achievement-related achievement-related goals. Thus, Tellegen's model also links Unks this general trait with effectance 1959), feelings feelings of of competence competence and mastery mastery (e.g., (e.g., effectance motivation (White, 1959), Bandura, 1977), and a style of effective effective engagement with one's environment. As noted earlier, eariier, this view of of the construct emphasizes its adaptive, productive, and quaUties much more than the impulsive, reckless, and sensation-seeking socialized qualities individual of the old Eysenckian model.

774 774

WATSON AND AND CLARK CLARK WATSON

D. Hogan's Bipartite Scheme Tellegen et al. (1988) also note that Positive Emotionality splits into two subfactors subfactors call"" Agentic Positive Emotion Emotionin some of their analyses. One subfactor, which they call domiality," is best defined by the Achievement scale, and essentially reflects the domi "Comnance/competence/mastery aspects of the trait. The second subfactor is called "Com nance/competence/mastery defined by Social Closeness; this munal Positive Emotionality," and is best defined represents the warm, affiliative affiliative side of the disposition. Interestingly, this bipartite scheme closely parallels the model proposed by Hogan (1983). Hogan argues that Extraversion ultimately should be divided into *'Surgency" and "Sociability." two distinct and separate dispositions, which he calls "Surgency" of status. Highly surgent individuals are Surgency primarily revolves around issues of ambitious, tenacious, and influential; they are leaders, and set high behavioral for themselves in work and other activities. In contrast, contrast. Sociability is standards for centered around issues of popularity. Sociable individuals are friendly, expressive, and exhibitionistic, and enjoy being around other people. It is noteworthy that these bipartite schemes produce factors that resemble the major major dimensions of interpersonal theory. Wiggins (1979) has demonstrated that two dimensions consistently emerge in studies of interpersonal traits. One defined by assertive assertiveand andself selfdimension can be called Status or Dominance, and is defined confident at one end, versus submissive and self-doubting at the other; this can be identified with Surgency (Hogan) and Agentic Positive Emotionality (Tellegen). identified termed Love or Affiliation. AffiUation. Warm and agreeable The second dimension can be termed agreeable define one pole, with cold and quarrelsome quarrelsome at the other; this is obviously similar SociabiUty (Hogan) and Communal Positive Emotionality (Tellegen). to Sociability attenThese bipartite schemes have much merit and certainly warrant greater atten tion in the future. Ultimately, it may prove proVe advantageous to decompose Extraversion subfactors are into these two components. As we will see, however, although these subfactors (i.e., they contain some unrelated elements), they also in some sense separate (Le., affect) share some common features (most notably, the experience of positive affect) and ultimately recombine at a higher-order level. Thus, we will continue to treat Extraversion as a single, higher-order construct. Extraversion IV. IV.

A N INTEGRATIVE MODEL OF TIlE THE CONSTRUcr CONSTRUCT AN

A. A A Schematic Schematic Model Model A. Taken together, these models cover a very wide substantive range, and it is easy to lose a coherent sense of the underlying construct itself. Therefore, in order to clarify these views, we we present present an an integrative integrative model model of of the the construct construct summarize and clarify in Figure 1. 1. This model is integrative in the sense that it captures the entire range of content that can be subsumed under this higher-order dimension; that is, it of of the major major constituent elements that are included in all of the current contains all of formulations. We also should emphasize, however, that this model represents our

CHAPTER 29 29 CHAPTER

775 77S

EXTRAVERSION EXTRAVERSION

Extraversion! Extraversion/ positive positive emotionality emotionality

Venture some

z

Affiliation

V

Exc Change Seek

zzx

IWarm

Qregar

I

Positive affectivity

Energy

Z 3 ~ ZTx:

Joy

Enthus

Lively

Activ

..

..

...

...

...

...

...

..

..

...

...

..

lAscendancel

zX

Exhib

Domin

..

...

..

...

...

Ambition

^chiev

Endur

FIGURE FIGURE 11 A A schematic schematic model model of of Extraversion showing showing its component traits and subtraits subtraits (Exc Seek, Excitement Excitement-

Seeking; Warm, Warmth; Gregar, Gregariousness; Enthus, Enthusiasm; Lively, Liveliness; Activ, Activity; Exhib,

Exhibitionism; Domin, Dominance; Achiev, Achievement; Endur, Endurance). See text for for more details.

own understanding of the current literature, and does not necessarily reflect a consensual view of the construct. Figure 11 schematically depicts three levels of the construct. At the highest level is the higher-order dimension itself. Here we have proposed two alternative labels. First, for purposes of historical continuity, we have retained the traditional name of of "Extraversion." "Extraversion."Second, Second,we wealso alsohave haveincluded includedTellegen's Tellegen'sproposed proposedalterna alternative, "Positive Emotionality," EmotionaUty," because it emphasizes the positive affective compo component that, in our view, forms the core of the construct. We explore this issue in more detail shortly. At the next level, the higher-order dimension splits into six component traits. facets—Affiliation, Ascendance, Positive Affectivity, and Energy Energy— Four of these facets-Affiliation, are connected to the higher-order dimension by solid lines, indicating that they are viewed as central features of the construct. The remaining two components— components Venturesomeness and Ambition-are con Ambition—are conceived as less central, and so are connected by dotted lines. lines. Consistent with our earlier discussion, Impulsivity is not included as aacomponent componenttrait traitof ofthis thisdimension, dimension,although althoughVenturesomeness Venturesomenesscontains contains some related qualities (Zuckerman et al., aI., 1988). The peripheral assignment of Venturesomeness and Ambition reflects several considerations. First, these components are completely absent from one or more of the major conceptualizations of the construct. Specifically, although Ambition plays an important role in Tellegen's and Hogan's models, it is omitted from all of of the others. Similarly, Venturesomeness is included in the views of Eysenck, Costa and McCrae, and Cattell, but is absent from those of Hogan and Tellegen. Second, these traits are inconsistently related to the other components. For example, VenVen turesomeness and Ambition are essentially unrelated to each other, and neither is strongly or consistently correlated with Affiliation (e.g., Costa & McCrae, 1988). Third, both traits have been shown to load strongly on higher-order dimensions

WATSON WATSON AND AND CLARK CLARK

776 776

other than Extraversion. Zuckerman et al. Excite al. (1988), (1988), for example, found that Excitement-Seeking scales loaded primarily on a higher-order Impulsivity/Sensation Impulsivity/SensationSeeking dimension, although they tended to load secondarily on Extraversion as well. Similarly, Similarly,Costa Costa and and McCrae McCrae (1988, (1988,1992) have argued argued that that Ambition Ambition actually actually 1992) have well. is a component of Conscientiousness, rather than Extraversion. At the lowest level of Figure 1, 1, each of these six components is itself divided subtraits. ItItisis important important to toemphasize emphasizethat that these thesesubtraits subtraitsalmost almostinevitably inevitably into two subtraits. are strongly correlated with one one another. another. For example, example,Warmth Warmthand andGregariousness Gregariousness scales are typically highly correlated (e.g., Costa & McCrae, 1984), 1984), as as are are measures measures of Achievement and Endurance (e.g., Jackson, 1984). 1984). We We have nevertheless listed sub these subtraits separately in order (1) to capture the entire range of content subsumed under each component, component, and and(2) (2)to tosuggest suggestthe thetypes typesofofscales scalesthat thatassess assessthem. them. Moreover, these subtraits sometimes show rather different different associations with the other primary traits. For example, Costa and McCrae For example, Costa and McCrae(1988) (1988)report reportaa.40 .40correla correlation between NEO Gregariousness and PRF Exhibition, suggesting a moderate Affiliation and Ascendance. However, they also report a .10 relation between Affiliation correlation between NEO Warmth and PRF Dominance, suggesting that these components are are essentially essentially unrelated. unrelated. Furthermore, Furthermore, in in Hogan's Hogan's (1983) (1983) bipartite bipartite components model the the Dominance Dominance subtrait subtrait of of Ascendance Ascendance apparently apparently combines combines with with Ambition Ambition model joins with with Affiliation Affiliation to to com comto form form Surgency, Surgency, whereas whereas the the Exhibitionism Exhibitionism subtrait subtrait joins to prise Sociability. Sociability. These Thesefindings indicate the the desirability desirability of of keeping keeping these these subtraits subtraits prise findings indicate separate, at at least least provisionally. provisionally. separate,

B. Traits B. Descriptions Descriptions of of the the Component Component Traits In order to clarify the range of content subsumed under this higher-order higher-order construct, construct, we will briefly describe each of the component traits. 1. Venturesomeness reflects individual differences differences in boldness and adventur adventurousness. High scorers desire change and variety in their lives, lives, and become bored or dissatisfied when it is absent (Change). (Change). They also enjoy exciting activities and seek out intense, stimulating environments (Excitement-Seeking). Affiliation represents the sociability component component that that has has always alwaysbeen beenpromi promi2. Affiliation models. Highly affiliative affiliative persons have warm and friendly friendly nent in Extraversion models. feelings toward others, and place a high value on close interpersonal relationships (Warmth). Moreover,such suchindividuals individualsenjoy enjoythe thecompany companyofofothers, others,and andare arestrongly strongly (Warmth). Moreover, motivated toward frequent frequent social interaction (Gregariousness). 3. Positive Affectivity represents represents stable stableindividual individualdifferences differencesininthe thefrequency frequency frequently feel happy, cheerful, cheerful, and intensity of positive mood states. High scorers frequently and optimistic about their future (Joy). find it easy to become (Joy). Moreover, they find excited and enthusiastic about events or activities in their lives (Enthusiasm). vigor. High High scor scor4. The Energy component assesses individual differences in vigor. ers report having a great deal of energy, and that they move at a quick, rapid pace

29 CHAPTER 29

EXTRAVERSION EXTRAVERSION

777 777

feel that (Liveliness). These individuals also have many interests and hobbies, and feel they they lead full, busy, and interesting lives (Activity). 5. Ascendance reflects reflects differences differences in assertiveness and social visibility. High scorers are dramatic and entertaining, and like to be the center of of attention attention in social situations (Exhibitionism). They are also forceful and decisive; they are good, forceful strong leaders, and enjoy controlling and/or influencing others (Dominance). enjoy influencing reflects differences differences in success-, mastery-, or competence-related competence-related 6. Ambition reflects enjoy meeting challenges and mastering mastering behaviors. Highly ambitious individuals enjoy difficult tasks; they prize excellence and maintain high standards for for their own difficult behavior behavior (Achievement). They are also willing to work work long hours in pursuit pursuit of of persevere even when they are tired tired or frustrated frustrated (Endurance). their goals, and persevere

V. EXPLORING EXPLOMNG THE CORE OF EXTRAVERSION A. The Central Role of Positive Emotional Experience Figure 1 demonstrates that the Extraversion Extraversion construct subsumes a very wide range of psychological characteristics. What common factor factor do these diverse personality personaUty of related at a higher-order higher-order level? Our Our reading reading attributes share that leads them to be related of of the literature suggests a model similar similar to that that proposed proposed by Tellegen Tellegen (1985; al., 1988). That is, we believe that that individual differences differences in positive Tellegen et aI., emotionaUty form form the core of of the dimension, the unifying unifying "glue" that holds these emotionality various aspects together. We should note, however, that in terms of Figure 1, we affective (i.e., (i.e.. Positive Affectivity) Affectivity) and view this core as including both purely affective Energy components. Although Positive Affectivity Affectivity and Energy components can be correlated distinguished conceptually (as in Figure 1) they tend to be strongly correlated for example, example. Table II). Moreover, as noted noted earlier, studies repeat repeatempirically (see, for edly have demonstrated demonstrated that terms reflecting reflecting Joy (e.g., {t,g,, happy, happy, cheerful), cheerful). Enthusiasm Enthusiasm (e.g., excited, interested interested),), and Energy Energy (e.g., energetic, active) active) are all strong markers of of the PA factor. factor. To illustrate the central role of of positive emotionality in the higher-order higher-order construct. Table I presents data collected collected from from a mixed sample of of 234 psychiatric psychiatric construct, correlations among scales from from Tellegen's MPQ that assess patients. Shown are correlations four 1: the abbreviated, ll-item four of of the components shown in Figure 1: 11-item version of of the Well-Being (WB) Scale (see Tellegen, 1982) is a marker Weil-Being marker of of the Positive Affectivity Affectivity component; Social Potency (SP) assesses Ascendance; Social Closeness (SC) mea measures Affiliation; Affiliation; 'and and Achievement Achievement (Ach) taps Ambition. Ambition. Affectivity is significantly significantly correlated correlated with Table I demonstrates that Positive Affectivity of the other other three Extraversion components. Moreover, the Ascendance, each of Affiliation, and Ambition scales all tend to be more highly related to trait P PA Affiliation, A They have a mean correlation of of .32 with Positive than they are to each other: They Affectivity (range = = .22 to .37), but correlate, on average, only .16 among themselves themselves Affectivity = .00 to .31). For example, consistent consistent with previous research research (e.g., Hogan, (range =

AND CLARK WATSON AND

778 778

TABLE TABLE I ABLEII1 Correlations Correlations among among Extraversion Extraversion Component Component Scales Scales in in aa Mixed Mixed Psychiatric Psychiatric Patient Patient Sample Sample

Extraversion component component Extraversion 1. Positive Affectivity 1. 2. Ascendance 3. Ambition 4. 4. Affiliation Affiliation

1

—

.36"'''' .36** .22** .22"'''' .37.... .37**

2

3

— .31** .31 "'''' .17"'''' .17**

.00

—

Note. Note, N = 234. Positive Affectivity = MPQ Well WeilBeing ((abbreviated); Ascendance = MPQ MPQ Social Social abbreviated); Ascendance Potency; Ambition Ambition = MPQ MPQ Achievement; AffiliaPotency; Achievement; Affilia MPQ Social Social Closeness. Closeness. See See text text for for more more de detion = MPQ tion tails. tails. '" p < *p < .05, two-tailed. "'''' **p < .01, p < .01, two-tailed. =

=

=

=

=

1983; Wiggins, 1979), Ascendance and Affiliation are significantly but only weakly .17); both traits, however, are more strongly related related to one another (r = .17); to trait PA (for Ascendance, r = .36; .36; for Affiliation, r = .37). These data are consistent with the position that positive emotional experience forms the core of the higher-order construct. They further imply that removing this common affective element would lower (and perhaps eliminate) the correlations among the other component traits. To test this, we computed partial correlation coefficients, controlling for scores on Positive Affectivity. These results partially supported our view: the partial correlation between Ascendance and Ambition but that that between between Ascendance Ascendance and and Affiliation Affiliation remained significant ((rr = .26, p < .01), 01 ), but dropped to near zero (r = .05, n.s.). Table II presents data collected from 254 Southern Methodist University undergraduates. Again, the MPQ SP, SC, and Ach scales were used to assess Ascendance, Affiliation, and Ambition, respectively. In addition, three measures of Positive Affectivity/Energy Affectivity/Energy were administered. First, subjects completed a trait form of the PA scale from the Positive and Negative Affect Schedule (PANAS; & Tellegen, Tellegen, 1988). 1988).The The PANAS PANAS PA PA scale scale consists consists of of 10 10descriptors descriptors Watson, Clark, & (active, alert, attentive, determined, enthusiastic, excited, inspired, interested, (active, alert, attentive, determined, enthusiastic, excited, inspired, interested, proud, proud, strong) that have been shown to be excellent markers of high PA. In this sample, PA. strong) subjects rated the extent to which they generally experienced each of these affects on aa five-point five-point scale. scale. Watson, Watson, Clark, Clark, and and Tellegen Tellegen (1988) (1988)present present extensive extensive evidence evidence on demonstrating that the PANAS PA scale is a highly reliable and valid measure of demonstrating that the PANAS PA scale is a highly reliable and valid measure of the underlying PA dimension. the underlying PA dimension. We created two other measures from a large pool of true-false items that were used in the development of the General Temperament Survey, an inventory .

.

CHAPTER 29 CuAPTER

779 779

EXTRAVERSION EXTRAVERSION TABLE T A B L EII D

Correlations Correlations among among Extraversion Extraversion Components Components

in an an Undergraduate Undergraduate Sample in Sample Component/scale Component/scale

11

2

3

4

5

Positive Positive Affectivity Affectivity

1. PANAS PANAS PA PA

2. 2. Posaff Posaff

Energy 3. Energy 4. 4. Ascendance Ascendance

Ambition 5. Ambition Affiliation liation 6. Affi

—

* '" .43 .43**

—

64·· ..64** * '" .49 .49**

.44** .44*'" ·· .30 .30** ** .39 .39**

—

** .47 .47** ** .55 .55** *'" .1S .18**

·· .31 .31** ·· .27 .27**

·· .1S .18**

—

.24·· .24** .16·· .16**

—

.01

Note. Note. NN =254. 254.PANAS, PANAS,Positive Positiveand andNegative NegativeAffect Affect Schedule; Schedule;PA, PA,Positive Positive ==

Affect; Posaff, Positive Positive Affectivity Affectivity Scale; Scale; Ascendance Ascendance === MPQ MPQ Social Social Potency; Potency; Affect; Posaff,

Ambition Ambition

See = MPQ MPQ Achievement; Achievement; Affiliation Affiliation = MPQ MPQ Social Social Closeness. Closeness. See

==

==

text for for more more details. details. text *p two-tailed.. • p < .05, two-tailed

**p < .01, .01, two-tailed. two-tailed. •• p <

that will be described subsequently. These scales were specifically designed to Affectivity and and Energy components of Extraversion. The Positive assess the Positive Affectivity Affectivity (Posaff) scale consists of 21 items directly relevant to the experience of PA; sample items include "I get excited when I think about the future," and "I do." The Energy scale is composed of of 17 items assessing enjoy nearly everything I do." perceived energy and vigor; sample items are "People would describe me as a pretty energetic person," person," "I live a very full life," and "Other people sometimes pre�ty have trouble keeping up with the pace I set." Thus, the students were assessed on five of the six components shown in Figure 1. 1. Table II presents correlations among these scales, and it can be seen that again. Ascendance, these results largely replicated those shown in Table I. Once again, were weakly weakly related related to one another, with a mean intercor intercorAffiliation, and Ambition were = .01 .01 to .24). All three components were, however, more relation of only .14 (range = consistently related to the trait P PA markers: Ascendance, Ambition, A and Energy markers: .42, .42, and .21 .21,, respectively, with the and Affiliation had mean correlations of .42, three P AlEnergy scales, which whichwere werethemselves themselvesstrongly stronglyinterrelated interrelated(mean (meanrr == .50). .50). PA/Energy In In Table II, II, we again see evidence that Ascendance, Affiliation, and Ambition PA/Energy AlEnergy than they are to each other. As before, are more highly related to trait P PA we tested the centrality of P A by computing partial correlations, this time using Posaff scale as the PA marker. When Posaff's Posaff s influence was statistically elimithe Posaff elimi nated, the correlations among the other components all became low and nonsignifi nonsignificant, ranging from .11 to -.08. .08. Similar results were obtained when the other two PA/Energy scales were were partialled partialled out. P AlEnergy scales out. -

780 780

CLARK WATSON AND CUU B. B.

Relations Relations between between Measures Measures of of Positive Positive Emotionality Emotionality and and Extraversion Extraversion

As was discussed earlier, studies have consistently shown that Extraversion is sigsig nificantly related to state PA scales. These correlations are only low to moderate in magnitude, however, typically ranging from .20 to .50 (Costa & McCrae, 1980, 1 980, 1984; Emmons & Diener, 1986; 1984; 1986; Tellegen, 1985; 1985; Watson & Clark, 1984). 1984). These data confirm that Extraversion Extraversion and state P PA confirm A are related, but, by themselves, do not demonstrate that PA is a central feature of the construct. To examine this issue, we must instead consider Extraversion in relation to PA individual differences trait P A measures. If, as we are arguing, stable individual differences in positive PA emotionality compose the core of Extraversion, then trait P A measures should be related to more traditional measures of the higher-order higher-order disposition. Meyer strongly related of 231 and Shack (1989) report results that support this contention. In a sample of correlated .66 with an eight-item trait P PA A students, the EPQ Extraversion scale correlated predicted from from previous research research (Tellegen, 1985; scale. Moreover, as would be predicted Neu.63 with EPQ Neu Watson & Clark, 1984) a parallel measure of trait NA correlated .63 roticism. To explore these relations further, of trait PA, trait further, we administered markers of KA were NA, Extraversion, and Neuroticism to 528 undergraduates. Trait PA and NA assessed using the general form of the PANAS scales. The PANAS PA scale has already been described. The PANAS NA scale was similarly derived, and also factor markers (afraid, {afraid^ashamed, ashamed,distressed, distressed, guilty, hostile, irritable, guilty, hostile, irritable, consists of 10 factor jittery, nervous, nervous, scared, scared,upset). upset). jittery, Extraversion and Neuroticism were measured using scales developed developed by Gold Goldberg (1983). On the basis of extensive analyses of English-language trait names 1981,1982), adjectival pairs to assess each of (Goldberg, 1981, 1982), Goldberg selected eight adjectival of the "Big Five" personality traits composing the prominent five-factor model of of personality (see Digman, 1990; 1987). In 1990; Goldberg, 1993; 1993; McCrae & Costa, 1985, 1985,1987). this sample, we used Goldberg's Surgency scale to measure Extraversion, and his Emotional Stability versus Neuroticism scale to assess Neuroticism. Subjects rated themselves on each of these trait pairs using a five-point scale, where 11 = = very = very much like much like trait A, 3 = = about average on this dimension, and 5 = trait B (for (for more information information regarding these scales, see McCrae & Costa, 1985). Correlations among these measures are presented in Table III, and it can be seen that the expected pattern emerged: The PANAS PA scale correlated .61 .61 with Goldberg's Surgency and only .21 with Neuroticism, whereas the PANAS NA -.21 scale correlated .53 .53 and -.26 with Neuroticism and Surgency, respectively. Table III also presents correlations between the individual Goldberg trait pairs and the PANAS scales. For our purposes, the most noteworthy finding is that significantly correlated with the PANAS PA scale. all of the Surgency items are significantly demonstrate that the association between trait PA and Extraversion Extraversion Thus, these data demonstrate is both strong and very broad based. Mathews (1989) has examined these relations more thoroughly. In her study, 231 higher-order dimensions 231 students completed a battery of tests related to the higher-order

CHAPTER CHAPTER 29 29

781 781

EXTRAVERSION EXTRAVERSION TABLE III

Correlations PA and and NA NA Scales Scales and and Goldberg's Measures of Correlations between between Trait Trait PA Goldberg's Measures of

Surgency Surgency and and Neuroticism Neuroticism

Correlations with with Correlations Goldberg scale/item scale/item Goldberg Surgency Surgency Unenergetic versus energetic Passive Passive versus versus active active

Timid Timid versus versus bold bold

Quiet Quiet versus versus talkative talkative

Retiring Retiring versus versus sociable sociable

Submissive Submissive versus versus dominant dominant

Conforming Conforming versus versus independent independent Humble versus proud

Neuroticism Neuroticism

At At ease ease versus versus nervous nervous Secure Secure versus versus insecure

Emotionally Emotionally stable stable versus versus unstable unstable Not envious versus envious, jealous

Even-tempered Even-tempered versus versus temperamental temperamental Relaxed versus high-strung

Unemotional Unemotional versus versus emotional emotional Objective Objective versus versus subjective subjective

PANAS PANAS PA PA

** .61 .61** ** .5 8 .58** ** .48 .48** 43** ..43** .37** 37** .36** 36**

.3 1 ** 31** 26** ..26**

.25** .25**

.26 ** -.26** .25 ** -.25**

** -.38 -.38** .27 ** -.27** ·* -.16 -.16**

PANAS PANAS NA NA

-.21** .21 ** .22 ** -.22** -.11 -.11**

.25 ** -.25** .08 -.08 ** -.14 -.14** .09* -.09* .01 -.01 ** -.15 -.15**

.53** .53** 46** ..46** .40* .40**·

.06 -.06

.38 •• 38** .35·* 35** ·· .27 .27**

-.01 .09 .09** .09· -.09* -

.08

.22** .22** 7· · .1 .17**

Note. Note. NN - 528. 528.The Thestronger strongercorrelation correlationfor foreach eachitem itemisisininboldface. boldface.PANAS, PANAS, =

Positive and Negative Negative Affect Affect Schedule; Schedule; PA, PA, Positive Positive Affect; Affect; NA, NA,Negative NegativeAffect. Affect. Positive and

< .05, .05, two-tailed. •*p p < two-tailed. * .p < .01, two-tailed. **p < .01, two-tailed.

of ExtraversionIP Extraversion/PA, Neuroticism/NA, and Disinhibition versus Constraint. Subjects A, NeuroticismlNA, completed trait versions of the two PANAS scales, as well as Goldberg's (1983) measures of Surgency, Neuroticism, and Conscientiousness (e.g., reliable versus addiundependable, organized versus disorganized, and careful versus careless). careless). In addi tion, they were assessed on the EPQ Extraversion, Neuroticism, and Psychoticism scales. It should be noted that, despite its name, EPQ Psychoticism is essentially a measure of Disinhibition, tapping individual differences in sensation-seeking, sensation-seeking, impulsivity, nonconformity, socialization, and behavioral restraint (see Tellegen, 1985; Zuckerman 1985; Zuckerman et aI., al., 1988). Finally, subjects completed a preliminary version of the General General Temperament Temperament Survey (GTS; Clark & Watson, 1990). 1990). The The GTS GTS consists consists of of three three factor-analytically factor-analytically derived scales (answered in a true-false format) that assess higher-order dimensions similar to those identified by Eysenck and Eysenck (1975) and Tellegen (1985). Specifically, the Positive Temperament scale broadly assesses individual differences differences in positive emotionality; in terms of Figure 1, it taps content relevant to both the 1,

AND CLARK WATSON AND

782 782

TABLE IV IV Varimax-Rotated Factor Loadings of the Personality Affect Scales and Trait Affect Loading on Scale

Factor 1

GTS Negative Temperament EPQ Neuroticism

.93

Goldberg Neuroticism PANAS NA GTS Positive Temperament Goldberg Surgency EPQ Extraversion

PANAS PA GTS Disinhibition EPQ Psychoticism Goldberg Conscientiousness

Factor 2

Factor 3

.90 .84

.83 .83

.87 .86

.83 .83 .79

.87 .74 .83 -.83

Note, N N = 231. 231. Loadings Loadingsbelow below1.301 |.30| are areomitted. omitted.GTS, GTS, General GeneralTemperament Temperament Note. =

Survey; Survey; EPQ, EPQ,Eysenck EysenckPersonality PersonalityQuestionnaire; Questionnaire;PANAS, PANAS,Positive Positiveand and Neg Negative Affect Schedule; NA, Negative Affect; PA, Positive Affect. Adapted Affect Schedule; from Mathews (1989).

Positive Affectivity Affectivity and the Energy components. Sample items include "I get pretty excited when I'm starting a new project," and "I have more energy than most of the people I know." In contrast, the Negative Temperament and Disinhibition scales are general measures of trait NA and behavioral disinhibition versus concon trol, respectively. (For details regarding the development, reliability, and validity of the GTS, see Watson & 1993; Watson, Clark, McIntyre, & Clark, 1993; Mclntyre, & & Hamaker, 1992).^ 1992).1 Mathews (1989) subjected these personality personaUty and affect measures to a principal components analysis. As expected, three strong factors emerged with eigenvalues 74.5% of the total variance. greater than 1.00; 1.00; together, these factors accounted for 74.5% Three factors therefore were orthogonally rotated using varimax. The resulting IV. The first and third factors clearly represent loadings are presented in Table IV. NA/Neuroticism and Disinhibition versus Constraint, respectively. The second trait NAlNeuroticism factor is, however, of most interest for our present discussion: The trait PA and Thedata datareported reportedininthis thischapter chapterand andininMathews Mathews(1989) (1989)are arebased basedon onpreliminary preliminaryversions versionsof of 1^The 19 and 30 items, respectively. The the GTS Positive Temperament and Disinhibition scales, containing 19

finalversion versionofofPositive PositiveTemperament Temperamentconsists consistsofof27 27items, items,which whichincludes includestwo two12-item 12-itemsubscales subscalesthat that final Thefinal version of of Disinhibi DisinhibiAffectivity and Energy components shown in Figure 1. The assess the Positive Affectivity final version tion includes includes 35 items. These preliminary preliminary versions versions are are very very highly highly correlated correlated with with the thefinalized scales tion 3S items. finalized scales scale—which contains 28 items-was items—was and yield very similar results. The final final Negative Temperament scale-which

used in all analyses.

CHAFfER CHAPTER 29 29

EXTRAVERSION EXTRAVERSION

783 783

Extraversion scales were all highly related to one another, and so jointly compose furthermore, that both types of scales load a common dimension. It is notable, furthermore, factor. equally strongly on this factor. Thus, we again see evidence that measures of of Extraversion Extraversion and positive emo emocorrelated with one another. Taken tionality are highly correlated Taken together, the data we have affirm our contention that positive emotionality reviewed strongly affirm emotionality is a central core component of of Extraversion Extraversion (see also Watson & Clark, 1992; Watson Watson et al., component aI., 1992). Affective and Interpersonal Correlates of of the Disposition C. Affective If positive emotionality truly forms forms the core of of the higher-order If higher-order disposition, then of more tradi traditrait PA measures should have external correlates similar to those of Extraversion scales. That is, the two types of of measures should be comparably comparably tional Extraversion personality variables. As a comprehensive related to non nonpersonality comprehensive discussion discussion of of this issue is of our review, we will focus focus instead on two types of of variables beyond the scope of social/interpersonal that are especially relevant to our argument: (1) indices of of sociaVinterpersonal behavior (which are emphasized emphasized in more traditional traditional conceptualizations conceptualizations of of the con conbehavior of positive emotional experience (which (which are highlighted struct) and (2) measures of in our model). Watson (1988a) examined these relations in a sample of of 71 students, each each of of Watson mood and activities questionnaire questionnaire once once per day over over a 7whom completed a daily mood per day week period (M = = 44.4 44.4 observations per subject). Mood Mood was assessed using 24 week items that yielded PA and NA factor factor scores. To measure social activity, the subjects subjects indicated the number of of hours (to the nearest half-hour) half-hour) that that they had spent with friends that day. Finally, they also completed completed personality personality measures measures both before before and friends after the daily rating period. after PA significantly related Consistent with our model, various trait P A markers were significantly of state PA, but also to social activity. For For example, the not only to average levels of abbreviated A, and .34 with abbreviated WB scale from from the MPQ correlated correlated .44 with mean mean P PA, mean social activity, over the rating period. It is also noteworthy noteworthy that socially based traits showed the same pattern; for MPQ SP scale (which, as noted for example, the MPQ earlier, related to both both mean PA (r = earUer, is a measure of of Ascendance) Ascendance) was equally related = .31) and mean social activity (r = = .29). Given these results, it is not surprising that mean P A and social activity were significantly PA significantly related related to each other other (r = = .28). Watson, Clark, and McIntyre Mclntyre (1990) conducted conducted a more comprehensive comprehensive analysis of weekly mood and social of these relations. In this study, 79 subjects completed completed a weekly = 12.3 assessments per subject). activities questionnaire over a 13-week period (M (Af = Mood Mood was assessed using the PANAS PA and NA NA scales; respondents rated rated their feelings from from the previous week. In addition, they indicated indicated how frequently frequently they of 15 categories of of social activity activity during the past week, had been involved in each of four-point scale (0 = = not at all, 11 = = once, 2 = = twice, 3 = = three or more using a four-point of social activity and tapped times). The items sampled broad classes of tapped a wide range of interpersonal interpersonal behavior (e.g., "romantic activity or or dating," "going to/giving a of

784 784

AND CLARK WATSON AND

party," "having a serious discussion"). Responses to these items were then summed party," to produce an overall index of social behavior. Prior to the weekly rating period, subjects were assessed on various Extraver Extraversion/trait PA markers, including EPQ Extraversion, Goldberg Surgency, GTS PosiPosi tive Temperament, and a General form of the PANAS PA scale. Table V presents correlations between these scales and mean PA, NA, and social activity scores computed over the 13-week rating period. Correlations among the mean scores are NA/ also shown in the table. Finally, Table V includes single markers of trait NAI Neuroticism (the General form of the PANAS NA scale) and Disinhibition versus Constraint (GTS Disinhibition) for comparison purposes. These results closely replicate those of Watson (1988a). As before, mean social activity was significantly related to mean PA (r = .35). A more striking finding,however, however,was wasthat thatthe theExtraversion/trait Extraversion/traitPPA scaleswere weresimilarly similarlyrelated related finding, A scales to mean levels of both positive emotions and interpersonal behavior. It is especially Extraversion was no more more highly highlycorrelated correlatedwith withsocial socialactivity activity noteworthy that EPQ Extraversion (r = .31) than with PA (r = ..30). 30) . Similarly, the GTS Positive Temperament scale had virtually identical correlations with PA and social activity (rs = .39 and .38, respectively). Finally, note that the trait PA measures (Positive Temperament and PANAS PA) were more highly related to social activity (mean r = = .34) than were the traditional Extraversion scales (EPQ Extraversion and Goldberg Surgency; = .24). .24). (These data are examined in a slightly revised form in Watson et mean rr = al., 1992). TABLE TABLEV V Afifect and and Social Activity Scores and and Selected Markers of Extraversion, Correlations between Mean Affect and Disinhibition Disinhibition Neuroticism, and Correlations with Scale

M Social Activity M Negative Affect Affect

ExtraversionIP A markers Extraversion/PA

Temperament GTS Positive Temperament

PANAS PA Extraversion EPQ Extraversion Goldberg Surgency Other personality scales PANAS NA

GTS Disinhibition Disinhibition

Positive Affect Affect M Positive .35** .35**

02 ..02

M Social Social Activity

Affect M Negative Affect

—

— —

-.18 -.18

.39** .39*·

.38·· .38**

-.15

.67·· .67** .30** .30·· .35·· .35**

.30·· .30** .31** .31·· .17

.02 -.02

-.11

-.09

-.07 -.07

.03 -.03

-.14 .14 09 -.09

-

-.

..47** 47··

.05

- 79. 79. GTS, General General Temperament Survey; PANAS, Positive and Negative Affect Affect Schedule; Note. N = Affect, EPQ, Eysenck Personality Personality QuestioMaire; Questionnaire; NA, Negative Negative Affect. Affect. See Watson, Clark, PA, Positive Affect, McIntyre, and Hamaker (1992) for more details. Mclntyre, and

< .05, two-tailed two-tailed.. *pp < •

< .01, .01, two-tailed. **p •• p <

CHAPTER 29 29 CHAPTER

78S 785

EXTRAVERSION VERSION EXTRA

In addition to the personality and affect affect measures described previously previously (see Table IV), subjects in the Mathews (1989) study also answered a series of Table of questions regarding various aspects of of their current lifestyle, emphasizing in particular their regarding interpersonal relationships. Correlations between between these items social behavior and interpersonal Extraversion/trait PA scales (GTS Positive Temperament, PANAS PA, and several Extraversion/trait EPQ Extraversion, Goldberg Surgency) are shown in Table VI. As before, single markers of of trait NA/Neuroticism NAlNeuroticism (GTS Negative Negative Temperament) and Disinhibition (GTS Disinhibition) also are included for comparison purposes. findings, the trait P A and Extraversion scales generally PA Consistent with earlierfindings, have similar lifestyle and behavioral behavioral correlates. For example, EPQ Extraversion and GTS Positive Temperament both are positively correlated with the number of of leadership roles assumed, frequency numfrequency of partying, number of of close friends, and num ber of dating partners, and are (negatively) related to the percent of weekend time spent alone and the severity of of speech anxiety. One interesting difference, however, is that the EPQ Extraversion scale-unlike scale—unUke the other markers of of the higher-order construct—has significant correlations with alcohol, drug, and cigarette construct-has low but significant use. Furthermore, Table VI indicates that these variables are more strongly, broadly, and appropriately related to the Disinhibition dimension. It therefore appears that the impulsive/disinhibited component of the old Eysenckian model has not been entirely eliminated from the EPQ Extraversion scale. TABLE TABLE VI VI

Correlations between between Selected Selected Extraversion, Neuroticism, and Disinhibition Disinhibition Markers and Various Measures of of Current Current Lifestyle Lifestyle and Interpersonal Behavior Correlations Correlations with EPQ

Goldberg

GTS

GTS

PosTemp

PANAS PA

Extra

Surgency

NegTemp

Disinhibition

.26** .19** .19** .19 ** -.19**

.32** .32** .17* .17* -.17 -.17**

.25** .33** .17* -.17*

GTS Item No. of of leadership roles Frequency of of partying

% % of of weekend weekend nights nights alone alone

No. of of close friends friends

No. of of dating partners Degree of of speech anxiety Frequency of of

Alcohol consumption

Drunkenness

Drug use

Cigarette smoking

.20** .20** .19** .19** .14'" -.14* .07 .07 -.01

.07 .07

.02

.1S** .18** .15'" .15* -.11

.09

.03 .03 .04 .04 -.01 -.01

.30** .30** .19**" .19'

-.21'" -.21** .16* .16*

·· .21 .21** ·· .1S .18** · .14 .14*

..34** 34** .24** .09 -.09 .14 .14 .13 .13 .2S"'''' -.28** .12

.13 .13 · .16 .16* .04

-.07 JJ7 .05 .05 .09

-.07 -.18** -.1S"'''' .24"'* .24** -.01

.03 .03

-.08 .OS

.06 .06

.05 .05 .17* .17* .12 -.12 .OS .08 .01 .01 .11 -.11 ·· .26 .26** ·· .3S .38** ·· .36** .36

.32** .32"

Note. Note. N = 217. 217. OTS, GTS, General Temperament Survey; Survey; PosTemp, Positive Positive Temperament; NegTemp, =

PANAS, Posi PosiNegative Temperament; EPQ, Eysenck Personality Questionnaire; Extra, Extraversion; PANAS,

Affect Schedule; P PA, See text for more details. tive and Negative Affect A, Positive Affect. See •*p p < .05, .05, two-tailed two-tailed.. .01, two-tailed. two-tailed. **p •• p < .01,

786 786

WATSON WATSON AND AND CURK CLARK

VI. CONSTRUCT VI. ORIGINS OF mE THE HIGHER.ORDER HIGHER-ORDER CONSTRUCT

A. Positive Affect as an Outcome Variable A. We have demonstrated that positive emotional experience plays a central role in the Extraversion construct. This raises the further further issue of why Extraversion has positive emotionality as its core. That is, why do these diverse primary traits all PA? PA correlate with P A? Perhaps the most intuitively obvious explanation is that P A may represent a consistent outcome of Extraversion and its component primary extraverts may behave behave in inways waysthat thatare areconducive conduciveto toexperienc experienctraits. In other words, extraverts ing higher levels of P PA. A. Compared to introverts, extraverts spend more time with (Affiliation), and because of their greater persuasiveness and social other people (Affiliation), from their social facility (Ascendance), are likely to derive greater reinforcements from interactions. Furthermore, extraverts are more active (Energy) and report being willing to work long hours in pursuit of their goals (Ambition); because of this, they may obtain more career- and competence-related rewards in their lives. Finally, extraverts actively seek out exciting and pleasurable experiences (Venturesome (Venturesomeness), which should further further increase their general PA level. This conceptual model guided most of the early research linking Extraversion & McCrae, 1980, 1980,1984), 1984), and it is surely true to a considerable and PA (e.g., Costa & extent. For example, our own research strongly suggests that certain types of social activity (especially physically active, informal, and epicurean activities) lead to & Watson, 1988; 1988; Watson, 1988a). transient increases in state PA (Clark &

B. B. Positive Positive Affect Affect as as aa Motivating Motivating Force Force However, PA cannot be viewed simply as a behavioral outcome. Considerable research also indicates that it is a powerful powerful energizing force that motivates a broad range of activity. For instance, several studies have shown that high levels of state P A are associated with enhanced affiliative PA affiliative feelings and an increased preference preference for for social, prosocial, and physically strenuous activities (Cunningham, 1988b; Rossi & & Rossi, 1977). Moreover, increased PA levels are related to actual increases in social and prosocia! prosocial behavior (Cunningham, 1988a; Cunningham, Steinberg, & & Grev, 1980; 1985). Thus, it is not simply the case that social 1980; Shaffer Shaffer & & Smith, 1985). xperiences are pleasurable, and hence lead to higher P interaction and mastery eexperiences A PA levels; to some extent, PA also motivates one to pursue these activities. In other words, PA is both a cause and a result of behavior, especially social behavior.

€• Delineating Delineating aa Common Common Biobehavioral Biobeliavioral System System C. Thus far, we have discussed PA and the other component traits of Extraversion as if they were distinct but related constructs that mutually influence one another. is that these components are all reflections of a single integrated Another possibility is differences in system. Tellegen (1985) has argued that this trait reflects individual differences

CHAYrER CHAPTER 19 29

EXTRAVERSION EXTRAVERSION

787 787

the behavioral activation system (BAS; Fowles, 1980; 1980; Gray, 1970), which is thought to control active approach and avoidance behaviors in response to signals of of reward. According to Tellegen, Extraversion/trait A levels may therefore represent individ Extraversion/trait P PA individual differences differences in sensitivity to pleasurable stimuli. Furthermore, it is possible that the other components of Extraversion (e.g., gregariousness, dominance) also are components under the control of the BAS. If so, then the affective and behavioral components of Extraversion both may be manifestations of common, underlying individual differences in active, pleasure-seeking behavior. In support of this idea, Depue, differences Extraversion/Positive Luciana, Arbisi, Collins, and Leon (1994) have shown that Extraversion/Positive Emotionality is related to individual differences in the reactivity of the mesolimbic dopaminergic system. Along these same lines, Depue and his colleagues (Depue & Iacono, lacono, 1989; & Spoont, 1987) have proposed a compelling biobehavioral view Depue, Krauss, & of the construct. Depue et ai. al. (1987) note that almost all of the component features of of Extraversion systematically systematically covary in bipolar disorder. That is, manic states are of associated with heightened energy and hyperactivity; elated and euphoric mood; increased social interest, and more generally, elevations in interest level and hedonic capacity; enhanced excitement-seeking; and heightened feelings of confidence and depressive episodes are characterized by sad and depressed depressed optimism. Conversely, depressive of energy; social withdrawal and anhedonia; avoidance of of strong mood; low levels of stimulation; poor concentration; and feelings of of pessimism and ineffectuality. ineffectuality. In different levels of Extraversion. other words, mania and depression reflect strikingly different Depue et ai. al. (1987) further demonstrate demonstrate that these various traits covary within normal individuals as well. For example, subjects who are currently reporting high levels of perceived energy will also simultaneously report feeling alert, socially interested, confident, high in positive affect, and so on. This temporal covariation strongly suggests that these component traits all reflect the operation of a single, integrated biobehavioral system, which Depue and Iacono lacono (1989) call the "behav "behavioral facilitation system" (BFS). In their view, the BFS promotes enjoyable and efficacious efficacious interactions with the environment. It appears to do so in several related ways. First, the P A component provides a general motivation or incentive to engage PA in goal-directed behavior. Second, increased levels of energy and alertness/concen alertness/concentration provide the physical and mental resources necessary for competent perfor performance. Finally, heightened feelings feeUngs of self-confidence self-confidence and optimism lead to an increased expectancy of of successful outcomes. Furthermore, variability is seen as an inherent feature of the BFS. Thus, individuals can be viewed in terms of of both their mean BFS level and their characteris characteristic degree of of fluctuation around this central tendency. Most individuals exhibit fairly modest variations in their BFS level. Extraverts are those who generally show high BFS activity, whereas introverts typically display lower BFS levels. Individuals suffering suffering from bipolar disorder, however, show much more extreme fluctuations on the BFS dimension (Depue et al, aI., 1987). The models proposed by Tellegen and Depue go beyond the existing data and are clearly somewhat somewhat speculative. Nevertheless, they have considerable value

788

WATSON WATSON AND AND CLARK CLARK

in terms of integrating the current findings and in directing future future investigation in this area. area. More Moregenerally, generally,models modelsproposing proposingintegrative integrativestructures structuresand andsystems systemsoffer offer a more sophisticated-and sophisticated—and ultimately more more satisfying-view satisfying—view of the the construct.

VB. Vn. IMPLICATIONS AND CONCLUSIONS

A. Implications Implications of of the the Changing Changing View View of of Extraversion A. Extraversion Theoretical conceptualizations of Extraversion have gradually but systematically evolved over the past 75 of 75 years. years. This evolution necessarily involves elements of both continuity and change. The continuity is largely reflected in the interpersonal components of the trait: Whatever else has been included in their models, virtually all of the major theorists in the the area have haveviewed-and viewed—and continue continueto to view-extraverts view—extraverts as gregarious and socially ascendant individuals. At the same time, the construct has changed markedly over the years. years. Most notably, many of the older conceptualizations emphasized that extraverts were unreflective, reckless, impulsive, and unreliable. Generally speaking, these views who were were poorly poorly somewhat unsocialized unsocialized individuals individuals who tended to portray extraverts as somewhat society. In In contrast, contrast, recent recent models models have have adapted to the restrictions of contemporary society. Thus, tended to stress the more adaptive and productive aspects of the disposition. Thus, extraverts, in addition to their social facility and gregariousness, are increasingly being ambitious, ambitious, hardworking, hardworking, and and dominant. dominant. Recent Recent conceptualizations conceptualizations viewed as being own) have also stressed the positive affective affective component of the the trait. (including our own) These models emphasize that extraverts are happy, enthusiastic, confident, active, and energetic. More Morefundamentally, fundamentally, ititnow now appears appearsthat thatExtraversion Extraversionessentially essentiallytaps taps and differences in affectively affectively rewarding performance: compared to introverts, individual differences effectively and pleasurably engaged in various extraverts view themselves as more effectively aspects of their lives. reformulated view viewstrongly stronglysuggests suggeststhat thatExtraversion Extraversionshould shouldplay playaamore more This reformulated diagnosis. Meehl (1975) was one of the prominent role in clinical assessment and diagnosis. first first theorists theoriststotorecognize recognizethe thepotential potentialclinical clinicalvalue valueofofthis thistrait. trait.Meehl Meehlargued argued "hedonic capacity" -that that significant, heritable individual differences differences exist in *'hedonic capacity"—that is, in one's basic capacity for experiencing pleasure. According to Meehl, many patients who who seek help will do do so so not not because because of of excessive excessive levels levels of of anxiety anxiety or or some some other negative affect, affect, but because of a general absence of joy and pleasure in their lives. Interestingly, Meehl suggested that Cattell's Surgency scale (which is an lives. excellent marker of his higher-order Extraversion factor) was a likely candidate differences in hedonic capacity. for assessing individual differences an Subsequent research has supported MeehPs Meehl's argument that Extraversion is an important dimension in clinical phenomena. For example, as noted earlier, Depue et al. (1987) have shown that this construct plays a central role in bipolar disorder. Similarly, several studies have shown that trait PA has important implications for & Watson, 1991; understanding anxious and depressive phenomena (e.g., Clark & 1991;

CHAPTER 29

EXTRAVERSION

789 789

Tellegen, 1985; 1985; Watson, Clark, & Carey, 1988). 1988). Specifically, measures of trait NA of both anxiety and depression, are strongly correlated with symptoms and diagnoses of indicating that NA is an important general correlate of psychiatric disorder (see 1984). In contrast, trait PA is clearly and consistently related also Watson & Clark, 1984). only to symptoms and diagnoses of depression, suggesting that low P A may be a PA critical factor factor in distinguishing depressive from from anxious states. Other results corroborate Meehl's (1975) argument that low trait PA levels frequently motivate individuals to seek professional professional help. Clark et al. (1994) exam examfrequently differences among three groups of of subjects: (1) unselected college ined personality differences university counseling center, undergraduates, (2) students seeking counseling at a university disorder units at a state and (3) inpatients on the substance abuse and personality disorder significantly higher on a trait NA factor hospital. The inpatients scored significantly factor than did demonstrating that NA is a general concomitant of the other two groups, again demonstrating of counseHng psychiatric disorder. Interestingly, however, both the inpatient and the counseling center groups groups scored scored lower lower than than the the normals normals on on aa trait trait PA PA factor. factor. center of the ways in which Extra ExtraThis brief review is only meant to suggest some of version/Positive Emotionality is related to important important clinical phenomena. The critiversionlPositive criti cal point is that the clinical implications of this construct should be explored much fully and systematically in the future. Furthermore, the available evidence more fully Extraversion/PA strongly indicates that ExtraversionlP A measures should be routinely included in clinical assessment.

B. B.

Directions for for Future Future Research Research Directions

In this chapter we have attempted to summarize and integrate the major major theoretical Extraversion. Clearly, Clearly, however, however, many many important important questions questions currently currently remain remain views of Extraversion. unanswered. We will conclude our discussion by noting three areas that warrant construct special consideration. First, future research needs to explore the core of the construct in greater detail. We have argued that individual differences differences in positive emotionality constitute the central core of the disposition, but more research is is obviously required on this issue. We believe that naturalistic studies examining how these components covary in the daily lives of individuals will prove particularly useful useful in this regard aI., 1987). (see Depue et al., Second, in constructing the schematic model shown in Figure 1, we were struck by the relative absence of research specifically specifically investigating relations among the basic components of the trait. Needed are more systematic, data-based analyses of the interrelations among the primary facets of Extraversion. Further research along these lines may well indicate that some of the current components should be combined or dropped, and perhaps may also suggest that currently unrecognized facets be added. We emphasize, however, that such research necessarily must emem ploy precise, valid, and specific measures of these facets. Many of of the older facet scales (for (for example, those in the CPI and GZTS) were constructed without regard for discriminant validity, and as a consequence are simply too broad and heteroge heterogefor useful in component-level research. neous to be useful

790 790

WATSON WATSON AND AND CLAB CLARK

future research research should should explore the contention contention that Extraversion Extraversion be be Finally, future divided into two distinct distinct subfactors, one that involves successful successful adaptation adaptation through through divided satisfying interpersonal and another another that that entails adaptation adaptation through through satisfying interpersonal relationships, and d o m i n a n c e , mastery, and and achievement achievement (Hogan, 1983; 1983; Tellegen Tellegen et aaI., l , 1988). These These dominance, two modes m o d e s of of adaptation are conceptually conceptually quite distinct, and and this bipartite scheme scheme two adaptation are therefore has considerable considerable appeal. Nevertheless, as noted noted earlier, both of of these therefore p r o p o s e d subfactors subfactors contain contain a common common positive affective affective element; consequently, consequently, proposed they are a r e ultimately ultimately (though not strongly) related to one another higher-order (though not another at a higher-order B e c a u s e of of this, we continue to believe that Extraversion Extraversion remains viable as level. Because higher-order construct. construct. a single higher-order

REFERENCES Allport, AUport, O. G. W., & Odbert, Odbert, H. S. (1936). Trait-names: A A psycholexical study. Psychological Psychological Monographs, (1 , Whole No. 211). Monographs, 47 47{h Bandura, A A. (1977). Self-efficacy: Self-efficacy: Toward a unifying unifying theory of of behavioral change. Psychologi Psychologi-

cal Review, Review, 84, 191-215.

Cattell, R. B. (1945). The principal trait clusters for for describing personality. Psychological Psychological

Bulletin, Bulletin, 42, 129-161.

description and and measurement measurement of of personality. personality. Yonkers-on-Hudson, Yonkers-on-Hudson, Cattell, R. B. (1946). The description NY: World World Book. NY: Cattell, R. B., Eber, H. W., & Tatsuoka, M. M. (1980). Handbook Handbook for for the Sixteen Sixteen Personality Personality Questionnaire (16PF), for Personality and Ability Testing. Questionnaire (16PF). Champaign, IL: Institute for Clark, L. A A. (1993). Manual Manual for for the Schedule Schedule for for Nonadaptive Nonadaptive and and Adaptive Adaptive Personality Personality (SNAP). Minneapolis: University of of Minnesota Press. (SNAP). Clark, L. A., Vorhies, L., & McEwen, J. L. (1994). Personality disorder symptomatology symptomatology from from the five-factor five-factor model perspective. In P. T. Costa, Jr. & T. A A. Widiger (Eds.), Personality model of ofpersonality personality (pp. 95-116). Washington, Personality disorders disorders and and thefive-factor the five-factor model DC: American Psychological Psychological Association. Clark, L. A., A, & Watson, D. (1988). Mood and the mundane: Relations between daily life events and self-reported self-reported mood. Journal Journal of of Personality Personality and and Social Social Psychology, Psychology, 54,

296-308.

Clark, L. A., & Watson, D. (1990). The The General General Temperament Temperament Survey. Survey. Unpublished Unpublished manu manuscript, script. Southern Southern Methodist Methodist University, Dallas, TX. TX. Clark, L. A., & Watson, D. (1991). Theoretical and empirical issues in differentiating differentiating depres depresClark, from anxiety. In J. Becker Becker & A. Kleinman (Eds.), Advances Advances in mood mood disorders: disorders: sion from Psychological aspects (pp. 221-245). Hillsdale, NJ: Erlbaum. Vol. 1. Psychological B., & McCrae, R. R. (1986). Correlations of Costa, P. T., T , Jr., Busch, C. M., Zonderman, A A. B., of factor scales with measures of the five-factor five-factor model of of personality. Journal Journal of of MMPI factor

Personality Assessment, Assessment, 50, 640-650. Personality Costa, P. T of extraversion and neuroticism on T.,, Jr., & McCrae, R. R. (1980). Influence of Journal of of Personality Personality and and Social Social subjective well-being: Happy and unhappy people. Journal Psychology, 38, 668-678. Psychology, Costa, P. T of wellbeing. T.,, Jr., & McCrae, R. R. (1984). Personality as a life-long determinant of Emotion in adult adult development development (pp. 141-157). In C. Z. Malatesta & C. E. Izard (Eds.), Emotion Beverly Hills, CA: Sage.

29 CHAPTER 29

791 791

EXTRAVERSION EXTRAVERSION

Costa, P. T., Jr., & McCrae, R R. R R. (1985). (1985). The NEO NEO Personality Personality Inventory Inventorymanual. manual,Odessa, Odessa, FL: FL: Psychological Assessment Resources. R. R R. (1988). From catalog to classification: Murray's needs and Costa, P. T., Jr., & McCrae, R Journalof ofPersonality Personalityand andSocial Social Psychology, 258-265. the five-factor five-factor model. Journal Psychology, 55,55, 258-265. Personality Inventory Inventory(NEO-Pl (NEO-PIJr., & & McCrae, R R. R R. (1992). Revised NEO Personality Costa, P. T., Jr., Five-FactorInventory Inventory(NEO-FFI) (NEO-FFI)professional professional manualOdessa, Odessa,FL: FL: R) and NEO Five-Factor manual. Psychological Assessment Resources. Cunningham, M. R R. (1988a). Does happiness mean friendliness?: Induced mood and hetero heteroPersonalityand andSocial SocialPsychology PsychologyBulletin, Bulletin, 283-297. sexual self-disclosure. Personality 14 14, 283-297. expectanCunningham, M. R R. (1988b). What do you do when you're happy or blue?: Mood, expectan Motivation and Emotion, 12, 12, 309-331. cies, and behavioral interest. Motivation R., Steinberg, Steinberg, J., J.,& &Grev, Grev,R R.(1980). (1980).Wanting Wantingto toand andhaving havingtotohelp: help:Separate Separate Cunningham, M. R, of Personality Personalityand and motivations for positive mood and guilt-induced helping. Journal of Social Psychology, Psychology,38, 38,181-192. 181-192. Social R. A., & Iacono, lacono, W. G. (1989). Neurobehavioral aspects of affective affective disorders. Depue, R Annual Review of of Psychology, Psychology,40, 40,457-492. 457-492. Annual Review Depue, R R. A., Krauss, S. P., & Spoont, M. R R. (1987). A two-dimensional threshold model affective disorder. In D. Magnusson & A. A. O Ohman (Eds.), Psychopa Psychopaof seasonal bipolar affective hman (Eds.), thology: An An interactional interactionalperspective perspective(pp. (pp. 95-123). Orlando, FL: Academic Press. thology: 95-123). Orlando, FL: Academic Press. A." Luciana, M., M., Arbisi, P., Collins, P., & Leon, A. (1994). Dopamine and the Depue, R R. A.„ structure of personality: Relation of agonist-induced dopamine activity to positive Psychology, 67, 67, 485-498. emotionality. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 485-498. Digman, J. M. (1990). Personality structure: Emergence of the five-factor model. Annual 1, 417-440. Review Review of of Psychology, Psychology, 441, 417-440. R. A., & Diener, E. (1986). Influence of impulsivity and sociability on subjective Emmons, R Psychology, 50 50, 1211-1215. well-being. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 1211-1215. Manualof ofthe theMaudsley MaudsleyPersonality Personality Inventory. San Diego, CA: EducaEysenck, H. J. (1959). Manual Inventory. San Diego, CA: Educa tional and Industrial Testing Service. & Eysenck, S. B. G. (1968). Manual Manual of ofthe theEysenck EysenckPersonality Personality Inventory, Inventory. Eysenck, H. J., & San Diego, CA: Educational and Industrial Testing Service. Eysenck, H. J., & Eysenck, S. B. G. (1969). Personality Personality structure structureand and measurement. measurement.San San Diego, CA: Knapp. Personality Questionnaire. Eysenck, H. J., & & Eysenck, Eysenck, S. S.B. B.G. G. (1975). (1975).Manual Manualof ofthe theEysenck Eysenck Personality Questionnaire. San Diego, CA: Educational and Industrial Testing Service. Eysenck, H. J., & Eysenck, M. W. (1985). York: (1985). Personality Personalityand andindividual individualdifferences. differences.New New York: Plenum Press. Fowles, D. C. (1980). The three arousal model: Implications of Gray's two-factor two-factor learning Psychophysiology, theory for heart rate, electrodermal activity, and psychopathy. Psychophysiology, 17, 17, 87-104. Goldberg, L. R R. (1981). (1981). Language and individual differences: The search for for universals in personality lexicons. In L. Wheeler (Ed.), Review Review of ofpersonality personalityand andsocial socialpsychology psychology (Vol. 2, 2, pp. 141-165). Beverly Hills, CA: Sage. R. (1982). (1982).From Fromace aceto tozombie: zombie:Some Someexplorations explorationsininthe thelanguage languageofofpersonality. personality. Goldberg, L. R In C. D. Spielberger & J. N. Butcher (Eds.), Advances in personality personalityassessment assessment(Vol. (Vol. 1, 1, pp. 203-234). Hillsdale, NJ: Erlbaum. five, plus oror minus two: Some conjectures Themagical magicalnumber number five, plus minus two: Some conjectures R. (1983, June). June). The Goldberg, L. R thedimensionality dimensionality personality descriptions. Paper presented a research seminar. Paper presented at a at research seminar, on the ofof personality descriptions. Gerontology Research Center, Baltimore. ,

,

792 792

WATSON CLARK WATSON AND AND CLARK

Goldberg, L. R R. (1993). The structure of phenotypic personality personality traits. American American Psychologist, Psychologist, 48, 26-34. Administrator's Guide, Guide. Palo Alto, Gough, H. G. (1987). California Psychological Psychological Inventory Inventory Administrator*s CA: Consulting Psychologists Press. Gray, J. A. (1970). The psychophysiological basis of introversion-extraversion. Behaviour of introversion-extraversion. Behaviour Research Research and and Therapy, 8, 249-266. Guilford, J. P. (1975). Factors and factors of personality. Psychological Psychological Bulletin, Bulletin, 82, 802-814. Guilford, Guilford, Guilford, J. P., & Braly, K. W. (1930). Extroversion Extroversion and introversion. Psychological Psychological Bulletin, Bulletin, 27, 96-107. R. B. (1934). An analysis of of the factors factors in a typical test of of Guilford, J. P., & Guilford, R Psychology, 28, 377-399. introversion-extroversion. Journal Journal of of Abnormal Abnormal and and Social Social Psychology, Guilford, J. P., & Guilford, R R. B. (1936). Personality factors factors S, E, and M and their measure measureGuilford, Journal of of Psychology, Psychology, 2, 109-127. ment. Journal Guilford-Zimmerman Temperament Temperament Survey: Survey: Guilford, J. P., & Zimmerman, W. S. (1949). The Guilford-Zimmerman Manual Beverly Hills, CA: Sheridan Supply. Manual. Zinmierman, W. S., & Guilford, J. P. (1976). The The Guilf Guilford-Zimmerman ord-Zimmerman Guilford, J. S., Zimmerman, Temperament Survey Survey Handbook. Handbook, San Diego, CA: Educational Educational and Industrial Test TestTemperament ing Service. &. McKi!1ley, McKinley,J.J.C. C.(1943). (1943). The TheMinnesota Minnesota Multiphasic Multiphasic Personality Personality Inventory. Inventory, Hathaway, S. R., R, & Minneapolis: University of Minnesota Press. Hogan, R R. T. (1983). A socioanalytic theory of of personality. In M. Page (Ed.), 1982 Nebraska Nebraska Symposium on on Motivation Motivation (pp. 55-89). Lincoln: University University of Nebraska Press. of Nebraska Symposium Personality Research Research Form Form Manual Manual (3rd ed.). Port Huron, MI: Research Jackson, D. N. (1984). Personality Psychologists Press. Pragmatism, New York: Henry Holt. Holt. James, W. (1907). Pragmatism. Psychological types. New York: Harcourt, Brace. Jung, C. G. (1921). Psychological Validating new trait scales in the context context of of three- and and five-f five-factor Mathews, K. K. (1989). Validating actor models of of personality. personality. Unpublished Unpublished master's thesis, thesis. Southern Southern Methodist Methodist University, models Dallas, TX. TX. Mayer, J. D., & Gaschke, Y. N. (1988). The experience experience and meta-experience meta-experience of of mood. Journal Journal of of Personality Personality and Social Social Psychology, Psychology, 55, 102-111. R. R, R., & Costa, P. T., Jr. (1985). Updating Norman's "adequate "adequate taxonomy": Intelli IntelliMcCrae, R gence and personality personality dimensions in natural language and in questionnaires. Journal Journal of of Personality Personality and Social Social Psychology, Psychology, 49, 710-721. McCrae, R R. R, R., & Costa, P. T., Jr. (1987). Validation Validation of of a five-factor five-factor model of of personality personality Journal of of Personality Personality and and Social Social Psychology, Psychology, 52, across instruments and observers. Journal 81-90. Bulletin of of the the Menninger Menninger Clinic, Clinic, Meehl, P. E. (1975). Hedonic capacity: Some conjectures. Bulletin 39, 295-307. R. (1989). of mood and personality: Meyer, G. J., & Shack, J. R (1989). The structural convergence of Journal of of Personality Personality and Social Psychology, Psychology, Evidence for old and new "directions." Journal and Social 57, 691-706. social time: Mood Mood patterns patterns by menstrual Rossi, A. S., & Rossi, P. E. (1977). Body time and social of the week. Social Social Science Science Research, Research, 6, 273-308. cycle phase and day of Shaffer, Shaffer, D. R, R., & Smith, J. (1985). Effects Effects of preexisting moods on observers' reactions to helpful helpful and nonhelpful nonhelpful models. Motivation Motivation and and Emotion, Emotion, 9, 101-122. Brief manual for for the Differential Differential Personality Personality Questionnaire. Questionnaire, Unpublished Tellegen, A. (1982). Briefmanual manuscript. University of Minnesota, Minneapolis. manuscript,

CHAPTER CHAPTER 29 29

EXTRAVERSION EXTRAVERSION

793 793

Tellegen, A. (1985). (1985). Structures of mood and personality and their relevance to assessing anxiety, with an emphasis on self-report. In A. H. Tuma & J. D. Maser (Eds.), Anxiety and the anxiety anxiety disorders disorders(pp. (pp.681-706). 681-706).Hillsdale, Hillsdale,NJ: NJ:Erlbaum. Erlbaum. Tellegen, A., Lykken, D. T., Bouchard, T. J., Jr., Wilcox, K. J., Segal, N. L., & Rich, S. Personality similarity similarity in in twins twins reared reared apart apart and together. Journal Journalof of Personality Personality (1988). Personality Social Psychology, Psychology,54,54,1031-1039. 1031-1039. and Social Watson, D. (1988a). Intraindividual and interindividual analyses of positive and negative affect: Their relation to health complaints, perceived stress, and daily activities. Journal Journal 54,54, 1020-1030. of Personality Personalityand andSocial SocialPsychology, Psychology, 1020-1030. of Watson, D. (1988b). The vicissitudes of mood measurement: Effects Effects of varying descriptors, time frames, and response formats on measures of positive and negative affect. Journal of Personality 55,55, 128-141. Personalityand andSocial SocialPsychology, Psychology, 128-141. Watson, D., & & Clark, Clark,L. L.A. A.(1984). (1984).Negative Negativeaffectivity: affectivity:The Thedisposition dispositiontotoexperience experienceaversive aversive emotional states. states. Psychological PsychologicalBulletin, Bulletin,96,96, 465-490. 465-490. Watson, D., & Clark, L. A. (1992). On traits and temperament: General and specific factors Journalof ofPersonal Personalof emotional experience and their relation to thefive-factor five-factor model. model. Journal ity, 60, 60, 441-476. ity, Behavioral disinhibition disinhibition versus versus constraint: constraint: A A dispositional dispositional Watson, D., & Clark, L. A. (1993). Behavioral Wegner & &J.J.W. W.Pennebaker Pennebaker (Eds.), (Eds.),Handbook Handbook of ofmental mentalcontrol control perspective. In D. M. Wegner (pp. 506-527). New York: Prentice-Hall. affectivity and their Watson, D., Clark, L. A., & Carey, G. (1988). (1988). Positive and negative affectivity of Abnormal Abnormal Psychology, Psychology, 97, 97, relation to anxiety and depressive disorders. Journal of 346-353. Watson, D., Clark, L. A., & McIntyre, Mclntyre, C. W. (1990). [General versus specific relations activity]. Unpublished raw data. between temperament and social activity]. Mclntyre, C. W., & Hamaker, S. (1992). Affect, Affect, personality, and Watson, D., Clark, L. A., McIntyre, Journalof ofPersonality Personalityand andSocial Social Psychology, 1011-1025. Psychology, 63, 63, 1011-1025. social activity. Journal brief Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative affect: The PAN PANAS Personality AS Scales. Journal of of Personality Social Psychology, Psychology,54, 54,1063-1070. 1063-1070. and Social Psychological Watson, D., & Tellegen, A. (1985). (1985). Toward a consensual structure of mood. Psychological Bulletin,98, 98,219-235. 219-235. Bulletin, Psychological White, R. W. (1959). Motivation reconsidered: The concept of competence. Psychological Review, 66, 66,297-333. 297-333. Review, Wiggins, J. S. (1979). A psychological taxonomy of trait-descriptive terms: The interpersonal interpersonal Psychology, 37,37, 395-412. domain. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 395-412. Zuckerman, M., Kuhlman, D. M., & Camac, C. (1988). What lies beyond beyond E and N? Factor of analyses of scales believed to measure basic dimensions of personality. Journal of 54,54, 96-107. Personality Personalityand andSocial SocialPsychology, Psychology, 96-107.

This Page Intentionally Left Blank

CHAPTER 30 CHAPTER 30

AGREEABLENESS: A A DIMENSION OF PERSONALITY WILLIAM WILLIAM G. G . GRAZIANO GRAZIANO TEXAS TEXAS A&M A & M UNIVERSm UNIVERSITY

NANCY EISENBERG EISENBERG ARIZONA ARIZONA STATE STATE UNIVERSm UNIVERSITY

This chapter is devoted to agreeableness as a dimension of personality. This review is composed of three parts. First, we we will briefly review conceptualizations and definitions of the dimension, and summarize the history of research on the dimen dimension. Second, we we will consider theoretical perspectives on agreeableness. Finally, we will focus on a special case of agreeableness, the prosocial personality.

I. CONCEPTUALIZATIONS OF AGREEABLENESS I.

A. Historical Historical Review Review of of Labels Labels for for Agreeableness Agreeableness A. What is agreeableness? In the past, a basic dimension has has been recognized, but it has received different may be disagreement disagreement on the different labels from theorists. There may origins and labels, but descriptions of the basic dimension for the phenomena of of agreeableness show remarkable communalities. For example, Adler (1938/1964) Gemeinschafts suggested that successful successful resolution of all three problems requires Gemeinschaftsgefuhl, or "social interest," manifested manifested in such traits as as cooperation and empathy, gefUhl, identification with others. In keeping with the psychoanalytic apselflessness, and identification IlANDBOOK OFOF PERSONALITY PSYCHOLOGY HANDBOOK PERSONAIITY PSYCHOLOGY79S

795

CoPYII GHT C COPYRIGHT 0 1997 1997 BY BY ACADEMIC ACADEMIC Puss PRESS.. ALL liGHTS OF lEPlODUCIlON IN ANY FOlII WEiVED.

ALL ucins OF REPRooucnoN m ANY FORM RESERVED.

796 796

GRAZIANO GRAZIANO AND AND EISENBERG EISENBERG

proach to attachment, Horney Horney (1945) linked the positive approach to others as part part of of dependency dependency in response to feelings feelings of of inadequacy. inadequacy. Within the psychometric tradition, Fiske (1949) labeled the dimension "con "conformity/' In their their reanalysis reanalysis of of six major major studies, Digman and Takemota-Chock Takemota-Chock formity." "friendly compliance compliance vs. hostile noncompliance." Hogan Hogan (1981) suggested the label "friendly offered the label "likability." Digman and Inouye (1986) later suggested that (1983) offered friendly compliance is similar, if not identical, to the "love-hate" "love-hate" their dimension of of friendly dimension in circumplex models of of personality (e.g., Leary, 1957). More recently, dimension Ostendorf (1993) suggested suggested that the meaning of of agreeableness agreeableness is Johnson and Ostendorf probably determined determined by how a researcher chooses to rotate factor factor axes relative to probably other personality personality dimensions. Depending on the rotation, agreeableness may be other meaningfully seen as either either "possessing a pleasant disposition" or "conforming "conforming to meaningfully others' wishes." In the Wiggins circumplex model of of personality personality (e.g., Wiggins, 1980; Wiggins & & personality dimension of of "warm-agreeable" is explicitly Broughton, 1985), the personality of altrusim, but the dimension is seen as not linked to the motivational orientation of other motivational orientations (e.g., cooperation, martyrdom) and orthogonal to other other dispositions, such as being gregarious and unassuming. Given the scope of of this particular dimension dimension and the sweep of of conceptualiza conceptualizaGiven tions, the potentially relevant research research literature is large. We were forced forced to establish estabhsh for inclusion in this review. The primary focus focus will be on more restrictive criteria for empirically based work on "natural language" individual differences differences in recent, empirically adults (Goldberg, 1981). Special attention attention will be given to those differences differences associ associdimension of of agreeable-disagreeable, as formu formuated with the "Big Five" bipolar dimension lated by Norman (1963) and Tupes and Cristal (1961). B. B,

Agreeableness in in the Natural Language

A major source of A major of information information about personality comes from from language. Cattell Cattell observed that over the centuries, the "pressure "pressure of of urgent necessity" has (1957) observed induced people to generate verbal symbols for for every important aspect of of an individ individbehavior that is likely to affect affect interpersonal interpersonal interaction. If If agreeable agreeableual person's behavior important dimension dimension in the natural language of of personality, then it should ness is an important appear with some frequency frequency in vocabulary vocabulary and language use. appear AUport and Odbert Odbert published their monumental monumental psycholexical psycholexical analysis In 1936 Allport of trait names. These authors examined 17,953 terms in the 1925 edition of of Webster's Webster's of New International International Dictionary Dictionary that were descriptive of of personality or personal behav behavNew major purpose purpose of divided into four four separate lists, or columns. The major ior. The terms were divided from censorial, evaluative evaluative the columns was to separate "neutral" words (Column 1) from Odbert believed words (Column 3). Column 1 words symbolized what Allport and Odbert & Odbert) traits of of personality. personality. were "real" "real" (quotes by Allport & 2,3, regarded as less value neutral By contrast, words in Columns 2, 3, and 4 were regarded purposes. Column 3 is especially relevant. relevant. and less control to personality. For present purposes, It was the longest of of the four four lists and contained character character evaluations such as

CHAPTER CHAPTER 30 30

AGREEABLENESS AGREEABLENESS

797 797

"agitating," "amiable," "agreeable," and "appealing." Allport AUport and Odbert were uncertain about the status of these words as trait names; they appeared to be value estimates. The authors even went so far far as to suggest that words in Column 3 "should be avoided by psychologists unless they are prepared to deal with the subject of social judgment judgment.. . .."" (p. vii). It is important to recognize one important point. The Allport-Odbert Allport-Odbert list is the major major source of terms for linguistic analysis of personality terms. In their attempt to reduce the large number of Allport-Odbert Allport-Odbert trait words into a more manageable set, many subsequent researchers have largely limited Umited themselves to the words in Column 1. 1. This occurred despite the fact that the reliability of the original classifi classification was not very high by modern modem standards (see Allport & Odbert, 1936, pp. pp. 34-36). Consider, for for example, Cattell's (1957) approach. Having limited himself to Column 1, 1, Cattell proceeded to reduce the list to 171 terms by having subjects rate the words on meaning. Cluster analyses further further shortened the list to a "standard reduced personality sphere" of 42-46 clusters. This reduced sphere has been used repeatedly (e.g., Bond & & Forgas, 1984, 1984, Digman & & Takemoto-Chock, 1981; 1981; Norman, 1963; Tupes & Cristal, 1961) and forms the foundation 1963; foundation for for much work on what is now known as the Big Five dimensions of personality. (Of (Of course, not all studies rely on Cattell's reduced sphere; see John [1990] and John, Angleitner, & Ostedorf Ostedorf [1988] for for reviews and thoughtful thoughtful critiques; see also Peabody & Goldberg [1989] for for a discussion of problems in Cattell's item selection.) Given this state of factor-analytic studies of affairs, one might not have expected factor-analytic of personality based on the standard reduced personality sphere to uncover a major dimension devoted primarily to "social evaluation." Nonetheless, when Digman Takemoto-Chock (1981) reanalyzed data from from six and Takemoto-Chock six major, large-scale studies, the vs. hostile hostilenon-compliance." non-compliance." factorto toemerge emergewas waslabeled labeled"friendly "friendlycompliance compUancevs. first factor

n. THEORETICAL PERSPECTIVES ON AGREEABLENESS D.

A. Natural Natural Language Language Perspectives Perspectives on on Agreeableness Agreeableness A. 1. Evolutionary Analyses 1. In his classic paper, Goldberg (1981) extended rigorously the analysis of personality language outlined earlier by Cattell. The heart of cap of Goldberg's analysis can be captured in a single sentence: "Those individual differences differences that are of the most significance significance in the daily transactions of persons with each other will eventually become encoded in our language" (pp. 141-142). A corollary is that the more important an individual difference difference is in human transactions, the more languages will have a term for it. Goldberg suggests that each of the Big Five dimensions concerns answers to five universal questions individuals ask about strangers they X agreeable agreeable(warm (warmand andpleasant) pleasant) are about to meet. One of thefive five questions is, "Is X or disagreeable (cold and distant)?" (p. 161).

798 798

GRAZIANO GRAZIANO AND AND EISENBERG EISENBERG

Also building on the Big Five, but employing a theoretical tack different different from from that that used by Goldberg, Hogan (1983) takes a bioevolutionary bioevolutionary approach. Hogan Hogan argues that personality personality languages has its origins in group processes. He argues that certain individual differences differences (e.g., cooperativeness) were more important in promoting promoting the the survival of of the the group than were others (e.g., aesthetic sensitivity) period of Survival-related individual differences differences during the long period of human human evolution. Survival-related should be especially conspicuous in the language of of personality description. Hogan explicitly argues that the evolutionary acquisition of of personality language terms is mediated through social consensus. It is reputational consensus in dispositional attribution that is important; that is, people who know or watch an individual will attribution of come to an agreement agreement on the person's tendencies. In this analysis, the language of individual differences differences evolved as a vehicle for for assessing social consensus about an approach to social per perindividual's value to a group (cf. Baron's, 1988, ecological approach ception). offered by Goldberg and Hogan Hogan are speculative; furthermore, furthermore, The pictures offered Evolutionary analyses can become complicated complicated they are painted with a broad broad brush. Evolutionary by many many tangential tangential issues (e.g., frequency-dependent frequency-dependent selection); here we consider only the most direct issue, the evolution of of sensitivity to a general dimension. For of heritability of individual differences differences in agreeableness agreeablenessa more detailed analyses of related behaviors, see Graziano (1994) and Loehlin (1992, pp. 56-64). related If we recognize that 99% of of human evolution evolution occurred when humans lived If in hunting/gathering hunting/gathering bands consisting of of approximately approximately 30 individuals, and if we recognize recognize that cooperation cooperation is seen as an essential attribute attribute in such groups groups (e.g., Konner, 1975), then then it is plausible that that an individual's agreeableness might be a dimension attention (Cosmides dimension receiving special attention (Cosmides & Tooby, 1992; Graziano, Jensen-Campbell, Todd, Todd, & Finch, in press). It is not not implausible that that dispositioned disagreeableness and selfishness selfishness could disagreeableness and could lead to exclusion from from social groups groups ((Jensen-Campbell, Jensen-Campbell, Graziano, & West, 1995). Additional analyses of of the dynamics of group inclusion inclusion and and exclusion exclusion are offered of offered by social psychologists. Social Psychological Analyses 2. Social Psychological Analyses What might be the functional functional value of of individual agreeableness to the group? One What explanation explanation is related to the need need for for group action. Groups Groups ordinarily ordinarily have tasks to accomplish, and and the accomplishment accomplishment is easier when group members hold a for accomplishing it. In the language language of of the consensus on the task and a means for early group-dynamics researchers, group group locomotion is more easily easily achieved the early of the group (e.g., Festinger, 1950). more cohesive the members of If an individual holds a nonnormative nonnormative viewpoint viewpoint about group action, social If brought to bear pressures will be brought bear on the deviant to induce group consensus, and thus cohesion. Pressure will appear appear first first in communication communication patterns. Group Group members members direct many many conmiunications of gaining will direct communications toward the deviant, with the intent of conformity (e.g., Dabbs & Ruback, 1987; Schachter, 1951). If If this kind of of pressure conformity stubbornly in the nonnormative nonnormative fails to gain compliance, and the individual persists stubbornly

CHAPTER 30

AGREEABLENESS

799 799

for both the individual and opinion, then there would be important consequences for the group. If a consensus forms that the deviant will not change toward the normative position, then communications to the deviant will drop off off drastically. The deviant will no longer be viewed as a member of the group (e.g., (e.g.. Wilder, 1986). 1986).The The ultimate ultimate consequence is that the individual will be effectively effectively eliminated from the group. Such an outcome is is clearly costly for both the group and the individual. From the group's perspective, the personal resources the individual might have provided to confiscated and compliance might be the group are now lost. Property might be confiscated forced, but these actions are costly compared with obtaining a "willing conversion." From the individual's perspective, rejection has occurred. If an individual holds a deviant viewpoint and is cut off off from the group, he or she may no longer be allowed to draw on group resources, and the individual's very survival may be questionable. As regards the group, it will now be smaller, and possibly poorer in resources, but the group as a whole will be more cohesive and will be better able to accomplish goals consensually defined as important. Thus group goal attainment, group cohesion, and individual influenceability influenceability are interconnected. One of the best documented conclusions in the groups' literature is that in high cohesive groups, individuals tend both to exert influence on other group members and to be susceptible to other individual's influence attempts (Col(Col lins & Raven, 1969). If this line of theoretical speculation is valid, then it is possible to see how agreeableness might have evolved as an important dimension in determining an for evaluating evaluating group members could individual's value to a group. That is, norms for functional be expected to include dimensions about agreeableness. If the norms were functional and had some adaptive significance, then they could be perpetuated through social transmission over very long periods (e.g., Campbell, 1988; 1988; Jacobs & Campbell, 1961). Individuals would be evaluated through the norms, and noncompliance with 1961). the norm could lead to exclusion. If there were a consensus that an individual was chronically unwilling to go along with others, was uncooperative, and gave few individual resources to the group, then that individual's potential contribution to the group would be small. small. Such an individual would be a force against cohesion and group locomotion. Outcomes of individual group members will not be maximized, at least in task-based groups, when cohesion is low (e.g., Bjerstedt, 1961; 1961; Schachter, 1951; cf. McGrath, 1964). 1951; 3. Convergence ConvergenceininSocial SocialAssessment AssessmentofofAgreeableness Agreeableness 3. For attributions of group members along a dimension of agreeableness to serve a useful function, members of the group must reach a consensus on such attributions. To assess convergence in attributions, one must measure concordance between two or more raters who view common targets. Norman and Goldberg (1966) provided clear evidence that peers do converge in their judgments. Costa, McCrae, and their colleagues have used older, nonstudent adults participating in the Baltimore Longitudinal Study to probe this hypothesis. In a paper focusing focusing specifically specifically on

800

GRAZIANO AND EISENBERG EISENBERG

the agreeableness dimension, Costa, McCrae, and Dembroski (1989) found a .30 intraclass correlation among 884 pairs of 344 peers raters for 101 101 targets on the agreeableness dimension. The corresponding intraclass correlation for extraversion was .53. None of the divergent correlations exceeded .15 in absolute value. McCrae and Costa (1987) reported peer intraclass correlations on agreeableness ranging from .28 to .38 (NEO scale), and from .36 to .43 (adjective factor scores). The corresponding intraclass correlations for extraversion were .38 to .52 (NEO scale) and .37 to .59 (adjective factor scores). A small number of studies, however, report fewer encouraging results for and Malloy (1988) used the Social Relations Model agreeableness. Albright, Kenny, and approach to partition variance attributable to raters and targets. Albright et al. a1. found that perceiver variance emerged on judgments of all traits, but the highest perceiver variance appeared on traits about agreeableness. Of the five dimensions assessed, the lowest proportion of target variance (4%) appeared on judgments of agreeableness. The highest proportion of target variance appeared on traits of extraversion (approximately 27%). 27%). Albright et a1. al. concluded that the construct of of agreeableness showed little or no interrater agreement. Watson (1989) computed peer rater convergence as a function of number of peer raters. The Spearman Brown reliability of peer rating of agreeableness ranged from a low of .21 .21 for one rater to a high of .57 for five raters. The corresponding values for extraversion ranged from .36 (one rater) to .73 (five raters). One interpretation of these studies is that agreeableness is less easily seen than extraversion, particularly in minimally acquainted peers. Watson's data suggest, however, that as the number of raters used to form consensus increases, significant convergence can be achieved. We will return to the issue of salience and level of convergence in peer evaluation when we consider the Wiggins circumplex interpretation of agreeableness. Smith and and Kihlstrom Kihlstrom (1987) (1987)conducted conductedfive fiverigorous rigorouslaboratory laboratorysocial socialcognition cognition studies to probe the hypothesis that the Big Five dimensions function as schemas. More specifically, each of the trait dimensions of the Big Five might be understood as a cognitive structure that organizes the lexicon of relevant but lower order traits. As a clear cognitive structure, agreeableness seemed to fare worse than the remaining dimensions of the Big Five. Peabody and Goldberg (1989) reported research that is potentially relevant to the interpretation of the Smith and and Kihlstrom Kihlstrom program program of research. research. These These authors authors note that potential bias that can can occur when samples are restricted to homogeneous targets, such as a rater and his or her friends. Peabody and Goldberg found that such restriction reduces the sizes of all factors, but especially that of agreeableness. From the perspective of Peabody and Goldberg, the tightly controlled program of of laboratory studies conducted by Smith and Kihlstrom involved internal judgments; agree there was relatively little room for variations in in stimulus materials involving agreeableness. Taken together, the bulk of the evidence provides qualified support for the natural language approach articulated by Goldberg (1981) and expanded by Hogan (1983). The strongest support support appears appearsin in more morenaturalistic naturalisticstudies studieswith withlarge largesamples samples

CHAPTER CHAPTER 30 30

AGREEABLENESS AGREEABLENESS

801 801

1988; Graziano, provided with an adequate range of trait words (e.g., Costa et al., ai., 1988; & Hair, 1996; & Hair, 1996; 1996; Graziano, Jensen-Campbell, Steele, & Jensen-Campbell, & Norman & & Goldberg, 1966). 1966). There is convergence across raters in the assessment self-ratings significantly with self-ratings of agreeableness, and the peer consensus correlates significantly of agreeableness. Weak or disconfirrning disconfirming evidence appears in less naturalistic labora laboratory studies (e.g., (e.g.. Smith & & Kihlstrom, 1987). In both kinds of studies, agreeableness is less easy for for naive perceivers to assess, particularly when the assessment is based ai., 1988; 1988; Watson, on a small number of observations of strangers (e.g., Albright et al., for judgments 1989). The precise mechanism of social cognition that is responsible for probably does not operate as an of agreeableness remains unclear. Agreeableness probably omnibus cognitive schema in perceivers through which the relevant lower order & Kihlstrom, 1987). traits in others are organized (Smith & It is interesting that agreeableness can be seen at all, given the obstacles it difficult faces. Intuitively, one might think that agreeableness would be especially difficult to assess in strangers. The situational press surrounding interaction with strangers, or at least the kinds of interactions assessed in most psychological research, is one that requires mildly positive behavior. There is a very serious restriction in range on the relevant agreeableness cues. Other dimensions might be easier to assess. assessed through the tidiness of a stranger's stranger's For example, conscientiousness might be assessed clothing or hair; extraversion might be assessed through the stranger's desire to talk over and beyond the demands of polite interaction. Furthermore, agreeableness affective evaluation, and may be a broader, more diffuse diffuse may be predominantly an affective reaction. As such, we might expect intuitively that it would be harder to identify identify reliably, especially in situations involving limited interaction. On the whole, these intuitions are off off the mark: Agreeableness can be seen, even in strangers. 4. Circumplex f Agreeableness CircumplexAnalysis Analysis oof Agreeableness

major Wiggins' (1991) approach to natural language and personality considers two major themes that appear to underlie underUe many descriptions of of interpersonal behavior. The first theme or dimension of interpersonal behavior is "agency," the condition of of being a differential differential individual, and its manifestation manifestation in striving for for mastery and power. The second theme is "communion," *'communion," the condition of of being part of a larger spiritual or social community, and its manifestation in striving for manifestation for intimacy, union, and solidarity with that larger entity. For present purposes, we will restrict ourselves to the part of Wiggins' conceptualization that is most relevant to the dimension of of agreeableness and its links to natural language and group processes. Wiggins (1991) demonstrates that agency and communion are two distinct orthogonal dimensions, not bipolar ends of a single dimension. Within a given group, seeing all possible combinations of agency and communion should be possible, with as "blends." To illustrate, a sociable, sociable, exhibitionist exhibitionist interpersonal behaviors appearing as person represents a blend of high agentic and high communal orientation, while a communal and low orideferential, trusting person represents aa blend of high communal low agentic ori entation.

802 802

GRAZIANO GRAZIANO AND AND EISENBERG EISENBERG

in which natural language forms of address reflect the Wiggins discusses ways it� operation of of agency and communion in different different languages. One illustration involves a "nonreciprocal power semantic" semantic" (Brown & & Gilman, Oilman, 1960). 1960). Specifically, Specifically, the singu singular form of of address in many languages requires a socially important choice. If the "tu" in person being addressed is a subordinate, the appropriate choice is T (e.g., "tu" *'du" in German). Oerman). If the Latin, Italian, French, and Spanish; "tea" in Russian; and "du" **vos" in Latin, person is a superior, however, the appropriate choice is V (e.g., "vos" acknowledg"vous" in French, and "vea" in Russian). Choices of this sort are public acknowledg ments of of power and dominance, and within the framework of Wiggins' analysis of agency. reflect the interpersonal behaviors of Earlier forms of of language apparently had no clear rules for differential differential use of of T and V among equal status peers. Gradually, Oradually, persons of equal status came to address each other with T as an expression of the "solidarity semantic." That is, language use evolved to allow expression of feelings of solidarity, intimacy, and similarity. In Wiggins' terms, language use evolved to reflect not only the dynamics of power and dominance ("agency"), but also the feelings of intimacy and group of cohesiveness ("communion"). Apparently, this pattern is not restricted to Indo IndoEuropean languages (e.g.. (e.g., White, 1980). The approach taken by Wiggins may suggest some refinements refinements to the natural language approach taken by Goldberg Ooldberg (1981) and Hogan (1983). Whether the differences appear in a single individual or are separated separated agentic and communal differences (e.g.. Bales, 1958), both characteristics characteristics into a task leader and a socioemotional leader (e.g., seem important to effective effective group functioning (Raven & & Rubin, 1983, p. 501). Precisely where does agreeableness fit on the agency-communion circumplex? McCrae and Costa (1989b) probed this question by jointly factoring the Wig Wiggins revised Interpersonal Adjective Scales (IAS-R) (lAS-R) with peer and spouse ratings on NEO-PI for a sample of of 315 adults. Their analyses suggest that the Big Five dimensions of of Extraversion and Agreeableness represent rotations of the agency and communion dimensions (or vice versa; factor analysis does not recognize the primacy of of any particular rotational position of any axis.) If one visualizes agency as an axis running from north (high agency) to south (low agency) and communion from east (high communion) to west (low communion), agreeableness runs from southeast (agreeable) to northwest (disagreeable). In other words, agreeableness represents a low-agentic-high-communal low-agentic-high-communal orientation. McCrae and Costa (1989b) claim certain advantages for the five-factor model essenover the circumplex. First, Wiggins' circumplex model and its dimensions are essen tially interpersonal and describe the relations between two people. According According to McCrae and Costa, concepts like love and warmth and communion may be adequate understandfor research on social psychology, but may not be the best concepts for understand specifically, a dimension like agree agreeing enduring dispositions in individuals. More specifically, ableness may include more than interpersonal elements (e.g., styles of cognition and affect). The The relative relative diversity diversity of of interpersonal behaviors might might be be parsimoniously parsimoniously affect). interpersonal behaviors of interrelated interrelated dimensions dimensions but but by by aa small small set set of of orthogonal orthogonal explained not not by by "blends" explained "blends" of underlying causes. causes. underlying

30 CHAPTER 30

AGREEABLENESS AGREEABLENESS

803 803

There is an intuitive appeal to this reductionistic reductionistic argument. Before Before it is ac accepted however, some problems should be noted. First, it is simply incorrect to assume that "the belief belief that most variables should load on a single factor factor follows follows from from the parsimony principle that traits are more likely to have one cause than two" (McCrae & Costa, 1989b, p. 592). In fact, the principle of of parsimony is conditional: Given an array of of accounts, all of which are comparably comparably adequate to explain a phenomenon, phenomenon, we should should accept the simplest. At this juncture, it is prema premature ture to assume that that the five-factor five-factor solution provides a simpler account of of communal communal behavior behavior or its cause than does the circumplex. Nor is it clear why "traits are more more likely to have one cause than two," or for for that matter 20 (cf. Ahadi & Diener, 1989). Causal identification identification is a complex business, dependent dependent on such diverse variables as level of of observation, mode of of analysis, and substantive question being asked (Cook & Campbell, 1979, pp. 10-36; Houts, Cook, & Shadish, 1986). (Cook 1986). The principle of parsimony is irrelevant to the choice between the two approaches here. of parsimony irrelevant attempt to partition partition a dimension like agreeable agreeableSecond, it may not be wise to attempt interpersonal and non-interpersonal non-interpersonal components. It is true that the agree agreeness into interpersonal allegedly non-interpersonal non-interpersonal variables variables ableness dimension has correlations with such allegedly of feeling (e.g., McCrae & & Costa, 1989a). Still, if if as the Myers-Briggs dimension of interpersonal aspects of of this dimension, then to what coherent set we remove the interpersonal of processes or behaviors might this dimension predict? Furthermore, removing removing of interpersonal elements severs. severs one promising link Unk in the nomological network network the interpersonal of agreeableness agreeableness involving the evolution of of natural language and group behavior of (e.g., Cattell, 1957; & Ostendorf, 1993). The The 1957; Goldberg, 1981; Hogan, 1983; Johnson & for the natural language approach focuses on the interpersonal interpersonal basic rationale for of personality; presumably, differences differences that are of the most significance significance in aspects of of persons with each other become become encoded in language the daily transactions of (Goldberg, 1981, 1981, p. 142). (Goldberg, p. 142). of lab studies that Third, the circumplex approach might help explain outcomes of raise troubling troubUng questions about about the precise mechanisms underlying the operation of of the Big Five. For example, the Smith and Kihlstrom (1987) studies suggested that agreeableness did not operate operate as an organizing cognitive schema for for relevant lower order traits, and that the Big Five dimensions in general lacked "semantic coher coherence." If dimensions are conceptualized conceptualized as a circumplex (Le., (i.e., not all dimensions dimensions are orthogonal), however, results like those reported reported by Smith and Kihlstrom be become more explicable. Emotional-Motivational Perspectives B. Biologically Based, Emotional-Motivational on Agreeableness Agreeableness / . Biological-Aff Biological-Affectect Perspective Perspective as a Complement Complement to to aa 1. Natural pective Natural Language-Cognition Language-Cognition Pers Perspective considered agreeableness largely from from a cognitive perspective. That That So far, we have considered focused on agreeableness in terms of of people's knowledge of of agreeableness, is, we focused

804 804

GRAZIANO AND AND EISENBERG EISENBERG

in both themselves and others. The natural language approach is itself itself cognitive in orientation in that it emphasizes the interdependence among language, social perception, and personality. There are, however, three potential limitations to this perspective. First, the cognitive approach is primarily descriptive, not explanatory. That is, it describes how persons might organize impressions of of agreeableness, but it is less clear on the mechanisms that induce some persons to be more chronically agreeable than others. Second, even within the framework framework of of a natural language approach, the word "agreeable" connotes an affective element in the evaluation of motivational dispositions in the target of of evaluation. Affective Affective and the perception of of agreeableness are probably not described (much less and motivational elements of explained) adequately by models that focus on cognition, language, and rational evaluations (e.g., Adelmann & Zajonc, 1989; Zojonc & Markus, 1984). Third, the cognitive approach has not established clear connections to the focuses literature linking positive emotions to prosocial behavior. That literature focuses on relatively short-term effects and situationally induced emotions, but it is possible that chronic emotional states have a similar relation to positive social behavior preclude (e.g., Cunningham, 1985,1986; 1985, 1986; Eisenberg et al, ai., 1989). Space limitations preclude review of the literature on emotions and prosocial behavior here. Instead, we we a review of possible links between chronic emotions and individual will focus on explanations of differences in agreeableness. See Watson and Clark (this volume, volume. Chap. 29) for a differences of positive emotional processes from a personality personality perspective. detailed analysis of of course, no reason to assume that cognitive and affective/motivational affective/motivational There is, of of agreeableness-related agreeableness-related behaviors are mutually exclusive alternatives, or models of al, 1989; Tellegen, 1985). even incompatible (e.g., Eisenberg, 1986; Eisenberg et aI., 2. Temperament Temperament and and Agreeableness Agreeableness

Individual differences differences in agreeableness in adults may have a temperamental basis. At the very least, there is evidence for the long term stability of disagreeable disagreeable behavior across the life span. Ill-tempered boys become men who are described as undercontrolled, undercontroUed, irritable, and moody (Caspi, Bem, & Elder, 1989). In discussing their temperamental dimension of of emotionality, emotionality. Buss and Plomin (1984) suggest that in comparison with unemotional people, emotional people become distressed when confronted with emotional stimuli, and they react with higher levels of emo emotional arousal. As a consequence, they may be harder to soothe when stressed. Rowe and Plomin (1977) found a correlation of of - . .42 4 2 between ratings of children's emotionality and those of of their soothability. In one especially unpleasant form of disagreeableness, childhood aggression, chronic differences differences in emotional responding may play a key role. Perry and Perry (1974) found that chronically aggressive children react more aggressively when provoked and "require" more suffering suffering from their victims before ending an attack than do nonaggressive children. Apparently, chronically aggressive children are not easy to placate or to soothe. Buss and Plomin (1984) suggest that the underlying arousal in emotionality is due to an overactive of the autonomic nervous system. sympathetic division of

CHAPTER 30 30 CHAPTER

AGREEABLENESS

805 805

As a result of certain biological predispositions, individuals may be likely to trajectory leading to more or less agreeableness follow a particular developmental trajectory (for (for an excellent general review of the biological bases of temperament, see Roth Roth1989). At this juncture, it is not clear how such mechanisms might operate, bart, 1989). differences in agreeableness may emerge as part of the ontogeny of of but individual differences associsystems of excitation and inhibition. If we can assume that agreeableness is associ affect, then models Unking linking brain lateralization ated with the inhibition of negative affect, and inhibition are relevant to this discussion. 3. 3. Neurology and Agreeableness

Kinsbourne and Bemporad (1984) suggest a multiaxial, ontogenic model to explain the development of self-regulatory self-regulatory processes. In this model, the left left frontotemporal frontotemporal cortex controls action over external change, including the planning and sequencing of acts (the "go" system). The right frontotemporal frontotemporal cortex controls internal emo emotional arousal (the "no go" system). The two systems operate synergistically, using information provided by posterior centers. Damage to the right orbital frontal frontal information cortex is associated with emotional disinhibition. Rothbart (1989) notes a parallel modulation of social behavior in accor accorline of thinking in Luria's work in that the modulation frontal lobes. Damage to the dance with the context is associated with the right frontal frontal area is associated with inaction and apathy. The normative left dorsolateral frontal left. evidence suggests that the right hemisphere develops sooner than the left. Fox and Davidson (1984) suggest that there are differences differences in hemispheric specialization for for affect. The left left hemisphere is associated with positive affect affect and specialization approach, while the right hemisphere is associated with negative affect affect and avoid avoidance. Toward the end of the first year of life, development of commissural transfer transfer permits left hemisphere inhibition of right hemisphere function. Two consequences affect and the possibility of behavioral alternations are the inhibition of negative affect between approach and avoidance. The two models outlined here are normative-developmental in focus focus and describe a supposedly universal pattern of neurological ontogeny. However, if if there were individual differences differences in the timing or completeness of any of these lateralization processes (as there surely must be), then there would be implications for for origins of agreeableness. That is, during ontogeny individuals may differ differ in the strength or timing of their left left hemispheric connections, or in commissural transfer. transfer. With these differences, differences, there would be corresponding differences differences in emotional expression and in the inhibition of negative affect. Differences Differences in expression in turn would lead to different From a developmental developmental perspective, perspective, different socialization socialization experiences. From long even if the delayed ontogeny of of inhibition were temporary, there could be longterm consequences. In one example, mothers report decreases in feelings of of attach attachment for for their 3-month-old 3-month-old infants if crying and other forms of negative affect affect do not decrease (Le., (i.e., come under inhibitory control), as they do in most infants infants (Rob (Robson & Moss, 1970). 1970). Longitudinal research on disruptions in mother-infant mother-infant attach attachment has shown patterns of persistent disagreeableness such as aggression and

806

GRAZIANO GRAZIANO AND AND EISENBERG EISENBERG

noncompliance, particularly in boys (e.g., Renken, Egeland, Marvinney, Mangls Manglsdorf, & & Sroufe, 1989). A second speculative candidate involves neuroregulatory amines (Panskepp, 1986). shifts 1986). These neurochemical systems apparently operate globally to influence shifts in vigilance and tendencies to act. For the present purposes, work by Cloninger is funcmost relevant. Specifically, Cloninger (1987) speculates that norepinephrine func tioning is related to reward dependence, which includes such behavior as being emotionally dependent dependent (versus coolly detached), warmly sympathetic (versus tough toughminded), sentimental, and sensitive to social cues. The bulk of the research on focused on psychiatric disorders, with relatively little neuroregulatory amines has focused (for a more detailed treatment, see work on normal adult personality processes (for Rothbart, 1989). 4. voidance and Agreeahleness Agreeableness 4. Approach-A Approach-Avoidance

Another explanation for agreeableness involves conflict between approach and avoidance motives (cf. Graziano, Jensen-Campbell, & & Hair, 1996; 1996; Jensen-Campbell, & Hair, 1996). Infante Infante and and Rancer Rancer (1982) (1982) offer offer aa model model of of argumentative argumentativeGraziano, & ness with two independent motives of approach and avoidance of arguments. That differ in feelings of excitement in advocating positions on controversial is, people differ differ in their issues, and in attacking others' positions (approach). People also differ motivation to avoid arguments. The motive to avoid arguments is seen as a debilitat debilitatfactor, weakening the tendency to approach arguments by the anxiety associated ing factor, with arguing. The motives to approach and to avoid arguments are independent, so that it is possible to be high on both motives, low on both, or high on one but low on on the the other. other. low favorable In this approach, the chronically argumentative person experiences favorable excitement and has a strong tendency to approach arguments, while feeling no inhibition nor tendency to avoid arguments. The chronically nonargumentative person shows the opposite pattern of approach and avoidance. Infante and Rancer hypothesize that the expression of dispositional argumentativeness is moderated by the perceived probability of success in a particular argument and the importance (incentive value) of success in winning that argument. In this system, persons with similar levels of the two motives (e.g., high approach and high avoidance) should different levels of of be more susceptible to situational influences than persons with different the two motives. That is, persons high in approach but low in avoidance will be likely to argue across a range of incentive and probability of success conditions; persons high in both motives will be more responsive to variations in the incentives different situational contexts (see Perry, Williard, & & and probabilities of success in different [1990] for for aa similar similaranalysis analysis of of incentive incentive and and expectation expectation effects effects in in the theselection selection Perry [1990) of victims victims in in ' children's aggression). of Infante and Rancer developed a 20-item scale and reported alpha reliabilities for the approach and avoidance components and resultant difference difference score ranging for from .86 to .91. .91. The correlation between self-ratings and friends' evaluations of from of

CHAPTER CHAPTER 30 30

AGREEABLENESS AGREEABLENESS

807 807

argumentativeness (approach) was .54. .54. The approach score component correlated -.39 .30 with a choice to participate in a debate and .39 to avoid a debate. From the present perspective, it is possible that the individual differences differences identified identiJSed by Infant Infant and his colleagues are part of the larger construct of agreeableness. 5. 5.

Hostility Hostility and Agreeableness

hostilYet another explanation of agreeableness involves emotions associated with hostil ity. In their efforts efforts to identify identify the "active ingredient" in the Type A link to coronary al. (1988) consider the variable Potential of Hostility, as heart disease, Costa et a1. Structured Interview (SI). This study study also provided provided scored from Rosenman's (1978) Structured a wealth of information about correlates of agreeableness, as measured by the NEO-PI. Costa et a1. al. note that much of the confusion confusion surrounding the role of of anger and hostility in cardiovascular disease is probably due to the fact that there are hostihty. The experience of hostility is not both neurotic and antagonistic forms of hostility. the same as the expression of hostility. Across two replicating samples of college {N = = 208), self-reports of the experience of anger were correlated over students (N -I-.60 with NEO-Neuroticism, but approximately -.33 +.60 - .33 with NEO-Agreeableness. correlated .00 with Neuroticism, and Self-reports of anger expression, however, correlated approximately -.40 with Agreeableness. Peer ratings of the focal subjects showed essentially the same pattern, but correlations involving peer ratings of subjects' experience of anger were somewhat lower. In general, both the self-reported experi experience and the expression of hostility hostihty are negatively related to Agreeableness; this evaluation is corroborated in peer reports. More interesting, perhaps, are the correlations of various forms of hostility from the SI with personality. Across both samples, all four four forms of hostility taken from (Hostile Content, Hostile Intensity, Hostile Style, and Potential for for Hostility) had correrelatively small but significant negative correlations with Agreeableness. The corre lations involving peer ratings, however, were generally nonsignificant. Costa et al. a1. note that the college samples showed considerable restriction in range on the hostility variables; variables; with a fuller fuller range, larger correlations might have been seen (see Matthews, 1988). Costa et a1. 1988, there were no conclusive data 1988). al. note that in 1988, linking agreeableness to coronary heart disease. At the very least, data from from their study raise some intriguing questions. Costa et a1. suggest that in their efforts al. efforts to identify the "toxic component of Type A," researchers should not restrict them identify themselves to a narrow view of antagonistic behaviors, and should supplement their use of the standard SI with measures of agreeableness. The literature we have outlined in this section suggests that there are probably important links Unks between emotionaVmotivational emotional/motivational processes and individual differ differences in agreeableness. It also suggests that these emotional processes may have a biological base, and are further modified by by life life experiences as persons persons move move through through the life span. Differences in agreeableness are probably related to important socio Differences socioemotional and health outcomes. Clearly, the interconnection among these variables is worthy of future research.

808 808

GRAZIANO GRAZIANO AND AND EISENBERG EISENBERG

Ill. in. THE PROSOCIAL PERSONALITY A. Definition of Prosociality and Scope of Analyses In this section, we examine prosocial tendencies as one form of agreeableness. Prosocial behavior typically is defined defined as voluntary behavior intended to benefit benefit another (regardless of whether the behavior is motivated by altruism or baser forms of motivation, such as rewards and social approval). Thus, it overlaps considerably "sympawith natural language trait words associated with agreeableness such as "sympa thetic," "generous," "kind," "helpful," and "considerate" (see Goldberg, 1992). helpful ways for a variety of Because people may act in kind, considerate, and helpful of often impossible to assess individuals' motives for for their prosocial reasons and it is often actions, we focus on the broader category of prosocial behavior rather than solely on altruistic behavior (which is a type of prosocial behavior). However, it is possible that altruism and constructs such as sympathy, other-oriented moral reasoning, and perspective taking are more closely related to agreeableness than is prosocial behavior (Penner & Fritzsche, 1993). In this section, we briefly summarize the literature concerning the existence differences in prosocial proclivities. The evidence for for the role of stable individual differences of situational variables in prosocial behavior is persuasive (see Dovidio, 1984; Krebs & Miller, 1985) but is less central to this volume than are data concerning therefore, we we do do not not review review the enormous literature Uterature the role of personality variables; therefore, function of a variety of situational demonstrating that prosocial behavior varies as a function factors. Rather, we focus primarily on the literature concerning stable individual differences in prosocial responding and the possible bases of those dispositions differences (much of which is developmental in focus). To examine the role of dispositional factors in prosocial prosocial behavior, we review research concerning several issues: issues: (1) the role of biology in prosocial tendencies; (2) the relation of prosocial tendencies to social learning variables believed to induce individual differences differences in prosocial responding; (3) consistency in prosocial responding across time; (4) consistency in prosocial responding across situations; and (5) the relation of prosocial behavior to various person/personality person/personality variables. Next we breifly sample studies stemming from an interactional breifly interactional perspective. Given our space constraints, our reviews are illustrative rather than detailed; nonetheless, we try to present an overview of the various types of data that are relevant to determining the role of personality in altruism.

B. gical Bases B. The The Biolo Biological Bases of of Prosocial Prosocial Behavior Behavior 1. L Sociobiology

Recent interest in sociobiology has stimulated much discussion concerning the for the evolution evolutionary bases of prosocial behavior. A variety of mechanisms for (Wynneof altruism in humans have been proposed, including group selection (Wynne Edwards, 1962), kin selection (Hamilton, 1964), reciprocal altruism (Trivers, 1971),

CHAPTER 30 30 CHAPTER

809

AGREEABLENESS AGREEABLENESS

& Wells, 1984), all of which posit reasons and genetic similarity (Rushton, Russell, & why people who assist others would be more likely than less prosocial persons to ensure the survival of their genes in the gene pool (see Boorman & & Leavitt, 1980; & Miller, 1990). In recent variations Cunningham, 1985-1986; Eisenberg, Fabes, & on this theme, some psychologists and biologists have suggested that it is the interaction of cultural and biological factors that result in prosocial behavior being 1983; Hill, 1984; 1984; MacDonald, adaptive (in the reproductive sense; e.g., Batson, 1983; 1984). 2. Heritability Heritability 2.

Evidence of innate, inherited differences in individuals' prosocial tendencies would provide strong support for the assertion that there is indeed a prosocial prosocial personality. personality. However, most of the research and theorizing on the role of genetics in prosocial behavior concerns the existence of a genetic basis of prosocial behavior in the human species, not the existence of biologically based mechanisms that might be prosocial tendencies. tendencies. Thus, Thus,the thesociobiologists sociobiologists the source of individual differences in prosocial and psychologists interested in the genetic basis of altruism have done little work bearing directly on the issue of personality differences in prosocial tendencies. Most of the limited work on inherited differences in prosocial tendencies concerns the construct of empathy. Stimulated by sociobiological ideas, Batson (1983) and M. Hoffman (1981) have proposed that the capacity for empathy is the biological substrate upon which human altruism is built. Empathy (and sympathy) & Miller, in press) as well as conceptually has been empirically (Eisenberg Fabes, & 1980; Feshbach, 1978; 1978; M. L. Hoffman, 1984; 1984; Staub, 1978) (Batson, 1987; 1987; Blum, 1980; with prosocial prosocial behavior; behavior;therefore, therefore,ififHoffman Hoffmanand andBatson Batsonare arecorrect, correct,geneti genetilinked with vicarious emotional emotional responsivity to others others could could cally based individual differences in vicarious account for individual differences in prosocial behavior. Consistent with the perspective that dispositional differences in both empathy and prosocial behavior have a biological basis, several groups of investigators have 44 to .72) in studies of twins' self obtained high estimates of heritability (from .44 self1976; Matthews, reported empathy and prosocial behavior (Loehlin & & Nichols, 1976; Batson, Hom, & Rosenman, Rosenman, 1981; 1981;Rushton, Rushton,Fulker, Fulker,Neale, Neale,Nias, Nias,&&Eysenck, Eysenck,1986). 1986). Horn, & These data must be interpreted with caution, however, because investigators fre frequentlyfind find higher relations between scores of identical twins than between scores to fraternal twins when whenself-report self-reportindices indicesare areused usedinstead insteadofofother othertypes typesofofmeasures measures to assess aspects of personality (Plomin, 1986). 1986). However, recent work with very young children suggests that empathy-related responding is indeed partially genetigeneti cally based, based, particularly particularly the the emotional emotional components components of of empathy empathy (Zahn-Waxler, (Zahn-Waxier, Rob Robinson, & & Emde, 1992). In summary, the few existing studies on the heritability of empathy and proso prosocial tendencies provide evidence consistent with the view that there are stable responding. However, However, additional additional research involv involvindividual differences in prosocial responding. ing behavioral indices of prosocial tendencies is needed. .

GRAZIANO GRAZIANO AND AND EISENBERG EISENBERG

810 810

C. C. The Social Learning Basis of Prosocial Prosociai Behavior The literature concerning the influence of cultural and specific child-rearing tech techniques is too voluminous to review in this chapter and has been reviewed extensively elsewhere (see Eisenberg & Mussen, 1989; 1989; Moore & Eisenberg, 1984; 1984; Radke RadkeYarrow, Zahn-Waxler, & Chapman, 1983). 1983). In general, however, researchers have found found that prosocial prosociai responding is systematically related to both living in certain types of cultures and being exposed to specific types of socialization techniques. Specifically, children who are routinely expected to assist in caring for others and in tasks important important to the existence of the family are more prosocial prosociai than children from from cultures in which such expectations are weak or absent (e.g., (e.g.. Whit Whiting & Whiting, 1975). 1975). Moreover, people are relatively likely to engage in prosocial prosociai actions when they have been exposed to altruistic models, other-oriented other-oriented preachings, preachings, provided with high and inductive (reasoning) modes of discipline; if they have been provided prosociai activities; and if they have moral standards and opportunities to engage in prosocial influences in a warm, supportive context. Findings supporting been exposed to such influences these conclusions have been obtained in laboratory studies and in correlational found in studies of real-life real-life altruists (e.g., rescuers in Nazi studies, and have been found freedom riders in the southern part of the Germany [Oliner & Oliner, 1988] and freedom United States [Rosenhan, 1970]). Although not all researchers have obtained this pattern offindings, findings, the overall pattern is consistent consistent enough to conclude that variations in the learning context are associated with relatively enduring individual differences differences in prosocial prosociai responding. Variations in learning experiences generally are believed to engender individual differences in values, motives, sociocognitive capacities, knowledge about helping, differences affective responses-differences responses—differences associated with variations in self-perceptions, and affective prosociai tendencies. prosocial D. D. Consistency Consistency of of Prosocial Prosociai Responding Responding across across Time Time and Situations Situations and One of the most obvious ways to study the stability of individual differences differences in prosocial prosociai tendencies is to examine relative consistency in those tendencies over time and situations. If there are stable differences differences in individuals' characteristic levels prosociai behavior over time and situations, it is likely that this stability is due of prosocial functioning (see in part to aspects of the individual's personality or sociocognitive functioning West & Graziano, 1989). Those investigators who have obtained longitudinal data have frequently frequently prosociai tenden tendenfound evidence of modest stability in individuals' relative levels of prosocial cies. The evidence of stability is perhaps weakest in studies of young children. For low, positive but nonsignificant nonsignificant correlations found low, example, Dunn and Munn (1986) found 24 months of age age for for a composite of observed sharinglhelpinglcomforting sharing/helping/comforting from 18 to 24 for giving appropriately. Similarly, Eisenberg, Wolchik, Goldberg, behaviors and for in 11- to 2-yearEngel, and Pasternack (1992) examined consistency over 6 months in

CHAPTER CHAPTER 30 30 AGREEABLENESS AGREEABLENESS

811 811

olds' spontaneous and requested prosocial behaviors with mothers and fathers and obtained very modest evidence of consistency. The only correlations that were significant were for boys' (but not girls') requested prosocial behaviors with fathers fathers and for boys' spontaneous behaviors with mothers. The relatively sparse evidence of consistency in prosocial responding in the early years is not surprising given the major changes in sociocognitive capabilities (e.g., role taking, moral reasoning) and other skills, as well as socializers' behaviors, during the first years of of life. Stability in prosocial tendencies likely increases with age in childhood. Dunn and Munn (1986), in an observational study of siblings sibUngs in their homes, found that older siblings (approximately 3 to 6 years of age) were more consistent over a 6month period in their helping, sharing, and comforting behavior than were their 1- to 2-year-old siblings. Moreover, other researchers have obtained evidence of moderate stability in prosocial responding in the preschool and school years. Block found that preschoolers who had been described by their nursery and Block (1973) found deschool teachers as generous, helpful and cooperative, empathic, considerate, de pendable, and responsible at age 4 were more likely at age 5 than their peers to share a prize they had earned with another child who did not have time to earn a prize. Similarly, Baumrind (reported in Mussen & Eisenberg-Berg, 1977) found observthat social responsible, prosocial behavior in the preschool years (as rated by observ significantly correlated with similar behavior when in elementary school ers) was significantly 5 to 6 years later. In addition, Eisenberg et al. (1987) found that donating to charity was consis consis7-8 years to age 9-10 years, and from 9-10 to 11-12. Helping (e.g., tent from age 7-8 helping pick up paper clips or spilled papers), which was assessed at ages 9 to 10 self-reported and 11 11 to 12, also was relatively stable over this 2-year period. Further, self-reported prosocial behavior and mothers' reports of children's prosocial behavior were both consistent over 4 years, as was helping by doing extra tasks for the experimenter (Eisenberg, Carlo, Murphy, & Van Court, 1993; Eisenberg, Miller, Shell, McNal McNalley, & Shea, 1991). Finally, Bar-Tal and Raviv (1979) found that peers' and teachers' sociometric ratings of sixth graders' altruism were relatively stable over a 2-year period; however, sociometric ratings in sixth grade were not significantly significantly related to self-reported willingness to help 2 years later. An alternative index of stability in prosocial responding is the quality of the individual's reaction to a needy or distressed person (rather than simple quantity of a given behavior). In this regard, regard, Radke-Yarrow and Zahn-Waxler Zahn-Waxler (1984) obtained obtained mothers' detailed reports of their 1- to 2-year-olds' naturally occurring reactions when they observed others in distress or need. They found that two-thirds of the children were stable over a 5-year period in mode of of reaction. For example, if children responded emotionally, with avoidance, or with a cognitive, nonemotional nonemotional response at age 2, they were likely to do so at age 7. There are very few data concerning the stability of of prosocial behavior in adolescents and adults. Oliner and Oliner (1988), in a retrospective study of Europe Europeans who had previously rescued Jews from the Nazis in World War II, found that

812 812

GRAZIANO AND AND EISENBERG EISENBERG GRAZIANO

rescuers were more likely than peers who did not engage in rescuing activities to interreport involvement in several prosocial activities during the year before their inter view (e.g., feeding the sick or aged or visiting the ill; making telephone calls for a group or cause; or helping raise money for a group or cause). Thus, using a sample of re of verified altruists, the Oliners obtained evidence of consistency in prosocial responding over 3 to 4 decades. On a less grand scale, Small, Zeldin, and Savin-Williams (1983) found that prosocial behavior, as well as observed prosocial behav behavpeers' ratings of adolscents' prosocial iors, were quite stable over a 3- to 4-week 4-week period (rs ranged from .48 to .99). In another study involving adolescents, Davis and Franzoi (1991) obtained fairly high school students' self-reports of sympathetic concern over correlations between high school 3-year period (rs for for the total total sample ranged from .48 to .64 or .64 to .81 a 2- or 3-year when corrected for measurement error); similar findings have been obtained by (1991,1993). Eisenberg and her colleagues (1991, 1993). Given the fact that sympathetic concern 1987; Davis, is positively related to behaviors that appear to be altruistic (Batson, 1987; (cf. 1983; see Eisenberg et al, aI., in press) and is an index of other-oriented concern (cf. al., 1996), the Davis and Franzoi data also can be viewed as Jensen-Campbell et aI., evidence of stability in altruistic responding. evidence Some of of the strongest evidence of consistency consistency in prosocial responding across settings comes from studies conducted by Savin-Williams and his colleagues. They of adolescents at summer camping outings over periods of of observed four groups of weeks (at wilderness travel programs and travel camp programs; see Savin-Williams, 1981; Small et al., Small, and Zeldin, 1981; aI., 1983; 1983; Zeldin, Small & Savin-Williams, 1982; see Savin-Williams, 1987). Individual differences differences in prosocial behavior were clearly recognized by peers after after only 4 days, and these perceptions remained stable over weeks (Zeldin et aI., al, 1982). Moreover, observed levels of of prosocial behavior were highly consistent across situations (Small et aI., al., 1983). Obtained correlations reported in studies of of cross-situational or cross-time consistency are likely minimal estimates of of the true correlations. As was evidenced in Rushton's (1980) reanalysis of of the Hartshorne and May data, the use of of aggregated indices would doubtless increase the size of of the intercorrelations intercorrelations in relevant research (see Epstein, 1979). Moreover, the intercorrelations would be expected to be higher if all the indices of prosocial behavior were actually indices of altruism (and, conse consequently, were due to higher level motivations such as sympathy and internalized internalized values). In summary, the evidence suggests that there is moderate stability in individu individuals' prosocial responding after the preschool years. However, the correlations over time vary considerably in strength. It is impressive, nonetheless, that researchers often often have obtained any evidence of consistency consistency considering that children's interpre interpretations 'and and understandings of the nature of kindness change with age (see Eisenberg, 1986), as do their competencies with regard to helping (e.g., Peterson, 1983). More Moreof these over, the evidence for consistency is strengthened by the fact that some of of prosocial behavior rather than merely self selfstudies involved observational indices of affected by memory distortions and self-presentational self-presentational report indices (which may be affected

CHAPTER CHAPTER 30 30

AGREEABLENESS AGREEABLENESS

813

concerns). In addition, it is likely that estimates of of consistency in most studies were underestimated because few researchers have corrected stability coefficients coefficients for measurement error.

£. The The Relation Relation of of Prosocial Prosocial Responding Responding to to Person Person Variables Variables E. If there are stable individual differences prasociality, one would expect prosocial differences in prosociality, different personal characteristics than less prasocial prosocial persons to exhibit somewhat different persons. In particular, prosocial (and particularly altruistic) persons would be exex pected to exhibit high levels of those characteristics that have been conceptually linked to other-oriented, moral responding-for responding—for example, role rale taking, sympathy and empathy, high-level moral reasoning, valuing of others, feelings of responsibility toward others, others, and and the the tendency tendency to to ascribe ascribe responsibility responsibility for for others to the the self self (see (see toward others to Eisenberg et al., in press; Schwartz & Howard, 1984; Staub, 1974, 1978, 1986; Eisenberg et aI., in press; Schwartz Howard, 1984; Staub, 1974, 1978, 1986; Underwood & Moore, Moore, 1982). 1982). Moreover, Moreover, because because prosocial prosocial behavior often involves involves Underwood behavior often not only an other-orientation but also the ability to enact helping actions, it is not only an other-orientation but also the ability to enact helping actions, it is reasonable to expect correlations between indices of prosocial behavior and individreasonable to expect correlations between indices of prasocial behavior and individ ual competence competence and and control, control, including including social social competence competence and ual and internal internal locus locus of of control. control. There do seem to be some associations between prosocial prasocial behavior and those personality characteristics conceptually linked to altruism, although these relations often not strong or very consistent. For example, prosocial responding has are often & been positively associated with a communal orientation (Clark, Ouellett, Powell, & nurturance (e.g., (e.g., Romer, Romer, Gruder, Cruder, & & Lizzardo, Lizzardo, 1986; 1986; Rushton, Rushton, Chris ChrisMilberg, 1987), nurturance & Fekken, 1981; 1981; Rushton, Littlefield, & & Lumsden, 1986), social interest (Cran (Cranjohn, & al, 1981), social social extensivity extensivity (Oliner & Oliner, Oliner, 1988), 1976; Rushton et aI., dall & Harris, 1976; al., 1989; 1989; Oliner & Oliner, 1988), and ascription social responsibility (Eisenberg et aI., 1989; Staub, 1974; of responsibility for others to the self (Eisenberg et al., aI., 1989; 1974; see review 1981,1984). 1984). Personal norms regarding helping, presumably in Schwartz & Howard, 1981, constructed from from internalized moral values, also have been linked with prosocial behavior in several studies (see Eisenberg, 1986; Pomazal & Jaccard, 1976; Schwartz & altruis & Howard, 1984; 1984; Zuckerman & & Reis, 1978), as has endorsement of altruistic values (Larrieu & Mussen, 1986; aI., 1981). Moreover, in a review 1986; Rushton et al., of the research on the personality correlates of community mental health volunteers, found that the following person Allen and Rushton (cited by Krebs & Miller, 1985) found personality traits were characteristic of the volunteers: internal locus of contral, social of control, responsibility, inner directedness, achievement via independence, self-control, flex flexfor intimacy, and nurturance. ibility, superego strength, self-acceptance, capacity for The relations of prosocial behavior to sociocognitive capabilities and affective affective for the traditionally aspects of responding are perhaps more consistent than relations for studied personality characteristics. For example, individual differences in sympathyl sympathy/ empathy (Eisenberg et aI., al, in press) perspective-taking ability (Underwood & UnderMoore, 1982), 1982), and level of moral reasoning (Blasi, 1980; Eisenberg, 1986; Under wood & Moore, 1982) have all been positively linked with prosocial responding in

814 814

GRAZIANO AND GRAZIANO AND EISENBERG EISENBERG

of the literature. Moreover, this linkage might be stronger when one consid considreviews of ers multiple mUltiple sociocognitive and affective affective skills. For example, Knight, Johnson, Carlo, and Eisenberg (1994) found found that children who were high in sympathy, perspective perspective taking, and the understanding of of money were highest in donating money to hospital hospitalized children. Sociability and assertiveness are personality characteristics that are more con consistently related to prosocial behavior than are most traits. Specifically, they are performance of of prosocial acts positively associated with children's and adolescents' performance that involve social initiative and direct interaction interaction with others (e.g., Eisenberg, Cameron, Tryon, & Dodez, 1981; Eisenberg, Pasternack, Cameron, & Tryon, 1984; Murphy* 1937). A A certain certain level of assertiveness assertiveness may Hampson, 1984; Larrieu, 1984; Murphy; necessary for for people to spontaneously spontaneously approach approach others who need assistance be necessary (Midlarsky & Hannah, 1985). Given that high levels of of spontaneous prosocial prosocial behavior have been associated with high sociability, moral reasoning, and other of competent competent social behavior, children who tend to perform perform helping and measures of spontaneously may fall into the group of of agreeable people with sharing behaviors spontaneously extraversion or emotional stability (as identified identified by Johnson & Ostendorf, Ostendorf, high extraversion 1993). of assertiveness and dominance have been been associated In contrast, low levels of compliant prosocial behavior (e.g., assisting in response to a request; request; with children's compliant 1981,1984; Among Eisenberg et al., aI., 1981, 1984; Eisenberg & Giallanza, 1984; Larrieu, 1984). Among preschoolers, non nonassertive children are viewed as easy targets by their peers, and assertive are asked to share or assist more often often than are their more assertive peers (Eisenberg l , 1981; 1981; Eisenberg, McCreath, & Ahn, 1988). Thus, the types of of prosocial et aaI., performed by persons varying in assertiveness and sociability probably differ differ acts performed somewhat somewhat in terms of of their social significance and motivational bases, and are differentially performed differentially performed depending on the social context. We now turn to this issue of of the interaction interaction between personal characteristics and environment. environment.

F. The Interaction of of Person and Environmental Variables Situational variables appear appear to interact with person characteristics in a variety of of ways (see Snyder & Ickes, 1985). Because this chapter's intent is to focus on of the myriad agreeableness as a person variable, we do not attempt to describe all of possible person-situation person-situation interactions in prosociality. Rather, we acknowledge these interactions with illustrative examples. researchers have found found that specific types of of helping contexts engender Some researchers in people with particular particular dispositional characteristics. For example, prosocial action in Meter (1972) asked students to volunteer volunteer for for a variety of of Gergen, Gergen, and Meter prosocial activities ranging from from counseling needy students to participating in an experiment involving unusual states of of consciousness. They found found that students experiment different dispositional characteristics and motives choose different different helping with different seemed to be a match between between students' dispositions dispositions and activities, and that there seemed

CHAPTER CHAPTER30 30

AGREEABLENESS AGREEABLENESS

815 81S

the ways in which they chose to help. It is likely different personality Ukely that somewhat different traits are associated with planned versus informal helping. In another particularly illustrative study, Romer et al. a1. (1986) found found that re redispositional nurturance and need for succor succorceptive givers, that is, people high in dispositional ance (the tendency to seek aid and support from others, especially when in need), were predisposed to help when compensation was expected. In contrast, adults who were high in nurturance and were more inner sustaining (independent) were more compensation was not not expected (and likely than receptive givers to help when compensation compensation was expected). People low in helped less than receptive givers when compensation nurturance and high in succorance helped least, regardless of of compensation. The results of of studies such as these are consistent with other data suggesting that disposi dispositional characteristics affect affect individuals' estimates of the costs and benefits of prosoof benefits of proso cial action in helping contexts (e.g., Penner, Michael & Brookmire, 1979). The results of these and other studies suggest that the association between underestimated frequently frequently by dispositional factors and prosocial behavior may be underestimated of this problem researchers who do not attend to moderating variables. Awareness of of prosocial behavior (e.g., Eisen Eisenis evident in theoretical work on heuristic models of of most berg, 1986, Figure 1). It is important to recognize, however, that the focus of empirical work in the area has been on prosocial behavior, not solely altruistic affect nonaltruistic probehavior. Many dispositional and situational variables that affect pro influence altruistic responding. The social behaviors would not be expected to influence person-centered variables (e.g., moral association between prosocial behavior and person-centered reasoning, sympathetic tendencies) would increase if if altruistic behavior only were considered (e.g., Rholes & Bailey, 1983).

CONCLUSIONS IV. SUMMARY AND AND CONCLUSIONS IV. of personality. Agreeableness This chapter examined agreeableness as a dimension of is probably best conceptualized as a general latent variable that summarizes more specific tendencies and behaviors (e.g., being kind, considerate, likable, cooperative, helpful). We used the natural language approach developed by Goldberg (1981) to organize the diverse findings in the literature. The central proposition of of the natural language approach is that individual differences differences that are of of the most signifi significance in the daily transactions of persons with each other will eventually become encoded in our language. Agreeableness should certainly qualify qualify as an individual difference having significance for people's daily transactions. We then discussed ways in which theoretical work by Hogan (1983) and Wiggins (1991) might be linked to the natural language approach. In particular, we noted how personality language may have its origins in the need for human groups to take concerted efficient when individual members are action (Hogan, 1983). Group action is most efficient willing to conform to group norms and to suspend their own individual concerns for the good of the group.

GRAZIANO GRAZIANO AND AND EISENBERG EISENBERG

816 816

With a few notable exceptions, the bulk of the literature is consistent with these theoretical ideas. Across a range of studies, agreeableness emerges in the natural language descriptions of of the self self and peers. Furthermore, there is evidence that self-rating self-rating and peer evaluations converge in assessing agreeableness. Such results are remarkable if we assume that the original research materials developed by Allport and Odbert and refined by Cattell probably biased outcomes against uncovering an agreeableness dimension. The precise mechanism linking assessments of agreeableness to specific behaviors remains unclear. Careful Careful laboratory work suggests that agreeableness probably does not operate as an omnibus cognitive schema organizing lower level traits (Smith & Kihlstrom, 1987). saUent as an individual Other work suggests that agreeableness may be less salient difference, particularly among minimally acquainted people, than dimensions such Watson, 1989). 1989). Assessments Assessments as extraversion or dominance (e.g., Albright et al., aI., 1988; 1988; Watson, of of agreeableness may be more global, more affect-laden, or more diffuse than other of assessments. kinds of Promising lines of of research have emerged linking agreeableness to motiva motivadifferences in tional processes and affect. Wiggins (1991) suggests that individual differences agreeableness might be part of of a motivational system in which people strive for intimacy, union, and solidarity with the groups to which they belong (or seek differences in to belong). Basic biobehavioral research suggests that individual differences self-regulatory processes agreeableness in adults may have their origins in affective self-regulatory in childhood. In particular, individual differences differences in the pattern of inhibition of negative affect affect may be related to the development of agreeableness (e.g., Fox & Davidson, 1984), and these may be related to health, especially especially cardiovascular disease (Costa et al., 1988). aI., Prosocial behavior can be conceptualized as a form of agreeableness. Recent research suggests that there may be important dispositional components to prosocial behavior, and these may be seen even in young children. Precise identification identification of these dispositions has been inhibited by problems of differentiating differentiating among social motives, and by weak measures of of altruism as an outcome and as a disposition. A A moderfurther problem is that researchers have focused on main effects and not on moder ated relations. Despite conceptual conceptual complexities and despite efforts to suppress its appearance, a construct approximated with the label agreeableness continually imporreappears in personality research. Its pervasiveness is best explained by its impor tance for understanding personality and interpersonal behavior.

ACKNOWLEDGMENT Part of this work was supported by NSF Grant BNS 8705780 to William Graziano, and by NSF Grants BNS 8807784 and DBS-9208375, National Institute of Child Health & & Develop Development Career Development Award K04 HD 00717, and National Institutes of Health Grant K02 MHOO903 MH00903 to Nancy Eisenberg. The authors express their appreciation to Steve Briggs,

30 CHAPTER 30

AGREEABLENESS AGREEABLENESS

817 817

Charles Halverson, Lauri Jensen Campbell, Delroy Paulhus, Louis Penner, Niels Waller, for comments on an earlier version of this chapter. David Watson, and Steve West for

REFERENCES Adelmann, P. K., & Zajonc, R R. B. B. (1989). (1989). Facial efference efference and the experience of emotion. Review of of Psychology, Psychology,40, 40,249-280. 249-280. Annual Review Annual Adler, A. (1964). Social interest: interest:AAchallenge challengetotomankind. mankind.New NewYork: York:Capricorn. Capricorn.(Original (Original work published 1938) Ahadi, S., & Diener, E. (1989). Multiple determinants and effect Personality effect size. Journal of Personality and Social Social Psychology, Psychology,56, 56,398-406. 398-406. Albright, L., Kenny, D. A., & Malloy, T. E. (1988). (1988). Consensus in personality judgments at Journalof ofPersonality Personalityand andSocial Social Psychology, 387-395. zero acquaintance. Journal Psychology, 55,55, 387-395. (1983). Personality characteristics of community mental health Allen, N. J., & Rushton, J. P. (1983). Journalof ofVoluntary VoluntaryAction ActionResearch, Research, 36-49. volunteers: A review. Journal 12,12, 36-49. Psychological AUport, G. W., & Odbert, H. S. (1936). Allport, (1936). Trait names: A psycho-lexical study. Psychological Monographs,47(1, 47(1,Whole WholeNo. No.211). 211). Monographs, Bales, R R. F. (1958). (1958). Task roles and social roles in problem solving groups. In E. E. Maccoby, psychology (3rd (3rd ed., ed., pp. pp. Readings in social psychology T. M. Newcomb, & E. L. Hartley (Eds.), Readings 437-447). New York: Holt, Rinehart. R. M. (1988). An ecological framework framework for for establishing a dual mode theory of social Baron, R knowing. In D. Bar-Tal & A. A. W. W.Kruglanski Kruglanski(Eds.), (Eds.),The Thesocial socialpsychology psychologyof ofknowledge knowledge (pp. 48-82). Cambridge, England: Cambridge University Press. Bar-Tal, D., D., & Raviv, A. (1979). Consistency in helping-behavior measures. Child Develop Development, 50, 50, 1235-1238. ment, Batson, C. D. (1983). Sociobiology and the role of religion in promoting prosocial behavior: Journalof ofPersonality Personalityand andSocial Social Psychology, 1380-1385. An alternative view. Journal Psychology, 45, 45, 1380-1385. Batson, C. D. (1987). Prosocial motivation: Is it ever truly altruistic? In L. Berkowitz (Ed.), Advancesin inexperimental experimental social psychology (pp. 65-122). Orlando, Academic Press. Advances social psychology (pp. 65-122). Orlando, FL:FL: Academic Press. Bjerstedt, A. (1961). Preparation, process, and product in small group interaction. Human Relations,114, 183-189. Relations, 4, 183-189. Blasi, A. A. (1980). (1980). Bridging Bridgingmoral moralcognition cognition and and moral moralaction: action:A A critical critical review reviewof of the theliterature. literature. Psychological 1-45. PsychologicalBulletin, Bulletin,88,88, 1-45. developmentand andthe theprovenance provenanceofofthought: thought: & Block, J. H. (1973, January). Ego development Block, J., & studyof ofego egoand andcognitive cognitivedevelopment development young children (Grant A longitudinal longitudinal study in in young children (Grant No.No. MH16080). Progress report for the National Institute of Mental Health. MHI6080). Friendship, altruism altruismand andmorality. morality.London: London:Routledge Routledge& & Kegan Paul. Blum, L. A. (1980). Friendship, Kegan Paul. & Forgas, J. P. (1984). Linking person person perception to behavior behavior intentions Bond, M. H., & across cultures: The The role role of of cultural cultural collectivism. collectivism. Journal Journalof of Cross-Cultural Cross-CulturalPsychology, Psychology, 15, 15, 337-352. geneticsof ofaltruism. altruism.New NewYork: York:Academic AcademicPress. Press. R. (1980). The genetics Boorman, S. A., & Levitt, P. R Oilman, A. (1960). (1960). The pronouns of power and solidarity. In T. A. Sebreok Brown, R., R, & Gilman, (Ed), Style in language language (pp. (pp.253-276). 253-276).Cambridge: Cambridge:Technology Technology Press. Press. (Ed.), developing personality traits. Hillsdale, Buss, A., & & Plomin, R R. (1984). Temperament: Temperament:Early Early developing personality traits. Hillsdale, NJ: Erlbaum.

818 818

GRAZIANO AND AND EISENBERG GRAZIANO EISENBERG

Campbell, D. T. (1988). Descriptive epistemology: Psychological, sociological, sociological, and and evolution evolutionary. In E. S. Overman (Ed.), Methodology Methodology and and epistemology epistemologyfor forthe thesocial socialsciences: sciences: Selected Selectedpapers papers(pp. (pp.435-486). 435-486). Chicago: Chicago:University University of of Chicago ChicagoPress. Press. Bem, D. J., & Caspi, A., Bern, & Elder, Elder, G. G. H., H., Jr. Jr. (1989). (1989). Continuities Continuities consequences consequences of of interactional interactional styles across the life course [Special issue]. Journal Journal of ofPersonality, Personality,57,57,375-406. 375-406. andand measurement. Yonkers-on Cattell, R. B. (1957). Personality Personalityand andmotivation: motivation:Structure Structure measurement. Yonkers-onHudson, NY: World Book. mood, relationship Clark, M. M. S., S.,Ouellette, Ouellette, R., R.,Powell, Powell,M. M.C, C,&&Milberg, Milberg,S.S.(1987). (1987).Recipient's Recipient's mood, relationship Psychology, 53, 53, 94-103. type, and helping. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 94-103. Cloninger, C. R. (1987). Neurogetic adaptive mechanisms in alcoholism. Science, Science, 236, 236, 410-416. B. E., E., & &Raven, Raven,B. B.H. H.(1969). (1969).Group Groupstructure: structure:Attraction, Attraction,coalitions, coalitions,communication, communication, Collins, B. & E. Aronson (Eds.), Handbook Handbookof ofsocial socialpsychology psychology(2nd (2nd and power. In G. Lindzey & MA: Addison-Wesley. ed., Vol. 4, pp. 102-204). Reading, MA: Quasi-experimentation: Design analysis issues Cook, T. D., & & Campbell, D. T. (1979). Quasi-experimentation: Design andand analysis issues settings,Chicago: Chicago:Rand RandMcNally. McNally. for field settings. Cosmides, L., L., & & Tooby, Tooby,J.J.(1992). (1992).Cognitive Cognitive adaptations adaptationsfor forsocial socialexchange. exchange.In InJ.J.H. H.Barkow, Barkow, & J.J. Tooby Tooby (Eds.), (Eds.), The Theadapted adaptedmind: mind: Evolutionary psychology Evolutionary psychology andand the the L. Cosmides, & generationof ofculture culture(pp. (pp.163-228). 163-228).New NewYork: York:Oxford OxfordUniversity UniversityPress. Press. generation Antago Costa, P. T., Jr., Jr., McCrae, R. R., & & Dembroski, T. M. (1989). Agreeableness versus Antagonism; Explication of a Potential Risk Factor for CHD. In A. Siegman & & T. M. Dem Demnism; Insearch searchofofcoronary-prone behavior (pp. 41-63). Hillsdale, Erlbaum. coronary-prone behavior (pp. 41-63). Hillsdale, NJ:NJ: Erlbaum. broski (Eds.), In & Harris, M. M. D. (1976). Social interest cooperation and altruism. Journal Journalof of Crandall, J. E., E., & Individual IndividualPsychology, Psychology,32, 32,50-54. 50-54. Cunningham, M. M.R. R.(1985-1986). (1985-1986).Levites Levitesand andbrother's brother'skeepers: keepers:AAsociobiological sociobiologicalperspective perspective Humboldt Journal Journalof ofSocial SocialRelations, Relations, 35-67. 13,13, 35-67. on prosocial behavior. Humboldt & Ruback, R. B. (1987). (1987). Dimensions of group process: Amount and Dabbs, J. M., Jr., & structure of vocal interaction. In L. Advances in inexperimental experimentalsocial social L. Berkowitz (Ed.), Advances psychology (Vol. 20, pp. 123-137). San Diego, CA: Academic Press. Davis, M. H. (1983). Measuring individual differences multi differences in empathy: Evidence for a multiPsychology, 44, 44, 113-126. dimensional approach. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 113-126. (1991). Stability and change in adolescent self-consciousness Davis, M. H., & & Franzoi, S. (1991). self-consciousness 25,25, 70-87. and empathy. Journal Journalof ofResearch ResearchininPersonality, Personality, 70-87. specification five robust factors ofof personal Digman, J. J. M., M.,&&Inouye, Inouye,J.J.(1986). (1986).Further Further specificationofofthe the five robust factors personal50,50, 116-123. ity. Journal Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, 116-123. Digman, J. J. M., M.,& &Takemoto-Chock, Takemoto-Chock,N. N.K. K.(1981). (1981).Factors Factorsininthe thenatural naturallanguage languageofofpersonal personality: ity: Re-analysis, comparison, and interpretation of six major studies. Multivariate MultivariateBe Behavioral havioral Research, Research,1 616, 149-170. , 149-170. altruism: An An empirical empirical and and conceptual conceptual overview. overview. Dovidio, J. J. (1984). Helping behavior and altruism: in experimental experimentalsocial socialpsychology psychology(Vol. (Vol.17,17,pp. pp.361361In L. Berkowitz (Ed.), Advances in 427). New York: Academic Press. Dunn, J., & & Munn, Munn, P. (1986). Siblings and the development of prosocial behaviors. Interna InternationalJournal Journalof ofBehavioral BehavioralDevelopment, Development, 9, 265-284. tional 9, 265-284. Eisenberg, N. (1986). Altruistic and behavior. Hillsdale, NJ:NJ: Erlbaum. Altruisticemotion, emotion,cognition cognition and behavior, Hillsdale, Erlbaum. & Dodez, R. (1981). Socialization of prosocial Eisenberg, N., Cameron, E., E., Tryon, K., & DevelopmentalPsychology, Psychology,17, 17,113-182, 773-782. behavior in the preschool classroom. Developmental & Van Court, P. (1996). Prosocial Prosocialdevelopment developmentininlate late Eisenberg, N., Carlo, G., Murphy, B., & adolescenceand andearly earlyadulthood: adulthood: longitudinal study. Paper submitted editorial Paper submitted for for editorial adolescence AA longitudinal study. review.

CHAPTER CHAPTER 30 30

AGREEABLENESS AGREEABLENESS

819 819

Eisenberg, N., Fabes, R R. A., & Miller, P. A. (1990). The evolutionary and neurological roots Hoffman (Eds.), Crime in biological, biological,social, social, of prosocial behavior. In L. Ellis & H. Hoffman and moral contexts contexts (pp. (pp. 247-260). 247-260). New New York: York: Praeger. Praeger. Eisenberg, N., Fabes, R R. A., & Miller, P. A. (in press). The evolutionary and neurological roots of prosocial behavior. In L. Ellis & H. Hoffman Evolution, the brain, Hoffman (Eds.), Evolution, and in in biosocial criminology. and criminal criminal behavior: behavior:Explorations Explorations biosocial criminology, Eisenberg, N., & Giallanza, S. (1984). The relation of mode of prosocial behavior and other proprietary behaviors to toy dominance. Child Study Journal, Journal, 14, 14,115-121. 115-121. Eisenberg, N., McCreath, H., & Ahn, R R. (1988). Vicarious emotional responsiveness and prosocial behavior: Their interrelations in young children. Personality Personalityand andSocial SocialPsy Psychology Bulletin, Bulletin, 14, 14,298-311. 298-311. Eisenberg, N., Miller, P. A., Schaller, M., Fabes, R R. A., Fultz, J., Shell, R, R., & Shea, C. (1989). The role of sympathy and altruistic personality traits in helping: A re-examination. Journal 41-68. Journalof ofPersonality, Personality,57,57, 41-68. R., McNalley, S., S., & & Shea, Shea,C. C.(1991). (1991).Prosocial Prosocialdevelopment development Eisenberg, N., Miller, P. A., Shell, R, Developmental Psychology, Psychology,27, 27,849-857. 849-857. in adolescence: A longitudinal study. Developmental & Mussen, P. (1989). The roots of ofprosocial prosocialbehavior behaviorininchildren. children,Cambridge, Cambridge, Eisenberg, N., N., & England: Cambridge University Press. Eisenberg, N., Pasternack, J. F., F., Cameron, E., & Tryon, K. (1984). The relation of quantity Development, 55, 55, and mode of prosocial to moral cognitions and social style. Child Development, 1479-1485. R., Pasternack, J., Lennon, R, R., Beller, R, R., & Mathy, R R. M. (1987). Eisenberg, N., Shell, R, DevelopmentalPsy PsyProsocial development in middle childhood: A longitudinal study. Developmental chology, 23, 23,712-718. 712-718. chology, Parental Eisenberg, N., Wolchik, S. A., Goldberg, L., Engel, I., & Pasternack, J. F. (1996). Parental values, reinforcement, reinforcement, and young children's prosocial behavior: A longitudinal values, and young children's prosocial behavior: A longitudinal study.study. Manuscript submitted for editorial review. 1. On predicting most of the people most of Epstein, S. (1979). The stability of behavior: 1. of the time. time. Journal Psychology, 37, 37, 1097-1126. Journalof ofPersonality Personalityand andSocial Social Psychology, 1097-1126. Feshbach, N. D. (1978). Studies of empathic behavior in children. In B. A. Maher (Ed.), Progress in experimental experimentalpersonality personalityresearch research(Vol. (Vol.8, 8,pp.pp.1-47). 1-47).New NewYork: York:Aca AcaProgress demic Press. PsychologicalReview, Review,57, 57,271-292. 271-292. Festinger, L. (1950). Informal social communication. Psychological Fiske, D. W. W.(1949). (1949).Consistency Consistencyof ofthe thefactorial factorialstructure structureof ofpersonality personalityratings ratingsfrom from different different Journalof ofAbnormal Abnormaland andSocial SocialPsychology, Psychology,44,44, 329-344. sources. Journal 329-344. R. J. (1984). Hemispheric substrates of affect: affect: A developmental Fox, N. A., & Davidson, R &R R. J. Davison (Eds.), The Thepsychology psychologyof ofaffective affectivedevelopment development model. In N. A. Fox & (pp. 353-382). Hillsdale, NJ: Erlbaum. (1972). Individual orientations to prosocial Gergen, K. J., Gergen, M. M., & Meter, K. (1972). Journalof ofSocial SocialIssues, Issues,28,28, 105-130. behavior. Journal 105-130. R. (1981). Language and individual differences: The search for universals in Goldberg, L. R Review of ofpersonality personalityand andsocial socialpsychology psychology personality lexicons. In L. Wheeler (Ed.), Review HUls, CA: Sage. (Vol. 2. pp. 141-165). Beverly Hills, R. (1992). The The development development of of markers markers for for the the big-five big-five factor factor structure. structure.Journal Journal Goldberg, L. R of 4, 4, 26-42. of Personality Personalityand andSocial SocialPsychology, Psychology, 26-42. (1994). The development development of agreeableness as a dimension of personality. Graziano, W. G. (1994). R. P. Martin (Eds.), The developing developingstructure structure In C. F. Halverson, G. A. Kohnstamm, & R

820 820

GRAZIANO GRAZIANO AND AND EISENBERG EISENBERG

of to to adulthood (pp. 339-354). Hillsdale, of temperament temperamentand andpersonality personalityfrom frominfancy infancy adulthood (pp. 339-354). Hillsdale, NJ: Erlbaum. Graziano, W. G., Jensen-Campbell, L. A., & Hair, E. C. (1996). (1996). Perceiving interpersonal conflict and reacting to it: The The case case for for agreeableness. agreeableness. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, Psychology^70(4). 70(4). Graziano, W. G., Jensen-Campbell, L. A., Steele, R. G., & Hair, E. C. (1996, February andand provincial in in Texas. 7). Unknown Unknown words in self-reported self-reportedpersonality: personality:Lethargic Lethargic provincial Texas. Unpublished manuscript, Texas A&M University, College Station. Graziano, W. G., Jensen-Campbell, L. A., Todd, M., & Finch, J. F. (in press). Interpersonal Interpersonal attraction from from an evolutionary psychology perspective. In J. A. Simpson & D. T. Kenrick (Eds.), Evolutionary social psychology. psychology. Hillsdale, Hillsdale, NJ: NJ: Erlbaum. Erlbaum. of Theoretical Hamilton, W. D. (1964). The genetical evolution of social behavior. Journal of Theoretical Biology, 7, 7, 1-52. Biology, factors Hampson, R. B. (1984). Adolescent prosocial behavior: Peer group and situational factors associated with helping. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 153-162. Psychology, 46, 46, 153-162. of Social and Biological Hill, J. (1984). Human altruism and sociocultural fitness. fitness. Journal of Biological Structures,7,7,17-35. 17-35. Structures, Hoffman, M. L. (1981). (1981). Is altruism part of human nature? Journal of of Personality Personalityand andSocial Social Hoffman, Psychology, 40, 40,121-137. 121-137. Psychology, Hoffman, M. L. (1984). Interaction of affect affect and cognition in empathy. In C. E. Izard, J. Hoffman, Zajonc (Eds.), Emotions, Emotions, cognition, cognition, and behavior behavior (pp. 103-131). Kagan, & R. B. Zajonc Cambridge, MA: Cambridge University Press. Hogan, R. T. (1983). A socioanalytic theory of personality. In M. Page (Ed.), Nebraska Motivation:Personality-Current Personality-Current theory research 58-89). Symposium on onMotivation: Symposium theory andand research (pp. (pp. 58-89). LinLincoln: University of Nebraska Press. Horney, conflicts. New NewYork: York:Norton. Norton. Homey, K. (1945). Our inner conflicts. (1986). The person-situation debate: A Houts, A. C, c., Cook, T. D., & Shadish, W. R., Jr. (1986). Personality,54, 54,52-105. 52-105. issue]. Journal of of Personality, critical multiplist perspective [Special issue]. Infante, D. A., & Rancer, A. S. (1982). A conceptualization and measure of argumentative argumentativeness. Journal 46,46, 72-80. Journalof ofPersonality PersonalityAssessment, Assessment, 72-80. c., & Campbell, D. T. (1961). The perpetuation of an arbitrary tradition through Jacobs, R. C, Abnormal and Social several generations of a laboratory micro-culture. Journal of of Abnormal Psychology, Psychology,62, 62,649-658. 649-658. Jensen-Campbell, L. A., Graziano, W. G., & Hair, E. C. (1996). Personality Personality and and relationships relationships as moderators of interpersonal conflict in adolescence. Merrill-Palmer Merrill-PalmerQuarterly, Quarterly,42,42, 148-163. Jensen-Campbell, L. A., Graziano, W. G., & West, S. G. (1995). Dominance, prosocial orientation, and female preferences: Do nice guys finish of Personality finish last? Journal of Personality Social Psychology, and Social Psychology,68, 68,427-440. 427-440. John, O. P. (1990). The big five-factor taxonomy: Dimensions of personality in the natural language and in questionnaires. In L. A. Pervin (Ed.), Handbook of of personality personality theory theory and research research(pp. (pp.261-277). 261-277).New NewYork YorkGuilford Guilford Press. Press. John, O. P., Angleitner, A., & Ostedorf, F. (1988). (1988). The lexical approach to personality: European Journal of of Personality, A historical review of trait taxonomic research. European Personality, 2, 171-203. Johnson, J. A., & Ostendorf, F. (1993). (1993). Clarification of the five factor factor model with the circnmpltx.Journal JournalofofPersonality Personality and Social Psychology, abridged big five-dimensional five-dimensional circumplex. and Social Psychology, 55, 55, 824-835.

CHAPTER 30 30 CHAma

821 821

AGREEABLENESS AGREEABLENESS

Kinsbourne, M., & Bemporad, B. B. (1984). Lateralization of emotion: A model and the evievi dence. In N. A. Fox & dence. &R R. J. Davidson (Eds.), The Thepsychology psychologyof of affective affectivedevelopment development (pp. 259-292). Hillsdale, NJ: Erlbaum. Knight, G. P., Johnson, L. G., Carlo, G., & Eisenberg, N. (1994). A multiplicative model of of the dispositional antecedents of a prosocial behavior: Predicting more of the people 66(1), 178-183. more of the time. Journal Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, 66(1), 178-183. Konner, M. (1975). Relations among infants and juveniles in comparative perspective. In M. Lewis & L. A 100-130). A. Rosenblum (Eds.), Friendship Friendship and peer relations relations (pp. 100-130), New York: Wiley. Krebs, D. L., & Miller, D. T. (1985). Altruism and aggression. In G. Lindzey & E. Aronson (Eds.), Handbook Handbook of of social social psychology psychology (3rd (3rd ed., ed.. Vol. Vol. 2, 2, pp. pp. 1-71). 1-71). New New York: York: Ran Random House. Larrieu, J. A. (1984, March). Prosocial andand sex:sex: Predictors of children's Prosocialvalues, values,assertiveness, assertiveness, Predictors of children's naturalistic naturalistic helping. helping.Paper Paper presented presented atat the the biennial biennial meeting meeting ofof the the Southwestern Southwestern Society for the Research in Human Development, Denver, CO. Larrieu, J., & Mussen, P. (1986). Some personality and motivational correlates of children's prosocial behavior. Journal 529-542. Journalof of Genetic GeneticPsychology, Psychology,147, 147, 529-542. Leary, T. (1957). Interpersonal personality. New York: Ronald Press. Interpersonaldiagnosis diagnosisofof personality. New York: Ronald Press. Loehlin, J. C. (1992). Genes Newbury Park, Genes and environment environmentininpersonality personalitydevelopment. development. Newbury Park, CA: Sage. R. C. (1976). Heredity, Heredity, environment, environment, and personality. personality. Austin: C., & Nichols, R Loehlin, J. C, University of Texas Press. of the development development of altruism. MacDonald, K. (1984). An ethological-social learning theory of Sociobiology, 5,5, 97-109. 97-109. for human sociobiology. Ethology and Sociobiology, Implications for A. (1988). Coronary heart disease and Type A behaviors: Update on an Matthews, K. A alternative to the Booth-Kewley and Friedman (1987) quantitative review. Psychologi Psychological Bulletin, Bulletin,104, 104,373-380. 373-380. A, Batson, C. D., Horn, J., & Rosenman, R Matthews, K. A., R H. (1981). "Principles in his ": The heritability of empathic nature which interest him in the fortune of others . . ..": Journalof ofPersonality, Personality,49,49, concern for others. Journal 237ITI-lAl. 247 . McCrae, R. R., & Costa, P. T., Jr. (1987). (1987). Validation of the five-factor model of personality across instruments and observers. Journal Journalof of Personality Personalityand andSocial SocialPsychology, Psychology,53,53, 81-90. R. R, R., & Costa, P. T., Jr. (1989a). Reinterpreting the Myers-Briggs Type Indicator McCrae, R from the perspective of the five-factor model of personality. Journal of of Personality, from Personality, 57, 17-40. 57, R., & Costa, P. T., Jr. (1989b). The structure of interpersonal interpersonal traits: Wiggins' McCrae, R. R, Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, circumplex and thefive-factor five-factor model. Journal 56, 586-595. 56, psychology: A A brief introduction. introduction. New York: Holt, Rine RineMcGarth, J. E. (1964). Social psychology: hart & Winston. Midlarsky, E., & Hannah, M. E. (1985). Competence, reticence, and helping by children and adolescents. Developmental DevelopmentalPsychology, Psychology,21,21,534-541. 534-541. The development development of of altruism. altruism. In In G. G. Whitehurst Whitehurst (Ed.), (Ed.), Moore, B. S., & Eisenberg, N. (1984). The of child development (pp. 107-174). Greenwich, CT: CT: JAI Press. Annals of behavior and and child childpersonality. personality.New NewYork: York:Columbia ColumbiaUniver UniverMurphy, L. B. B. (1937). Social behavior sity Press. -

822

GRAZIANO AND GRAZIANO AND EISENBERG EISENBERG

P., & helping: TheThe development of of & Eisenberg-Berg, N. (1977). Caring, Caring,sharing, sharing,and and helping: development Mussen, P., prosocial behavior behavior in in children. children,San SanFrancisco: Francisco:Freeman. Freeman. Norman, W. T. (1963). Toward an adequate taxonomy of personality attributes: Replicated factor structure in peer peer nomination nomination personality personality ratings. ratings.Journal Journalof ofAbnormal Abnormaland andSocial Social Psychology, Psychology, 66, 66,574-583. 574-583. & Goldberg, L. R. (1966). Raters, ratees, and randomness in personality Norman, W. T., & 4, 681-691. structure. Journal Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, 4, 681-691. of of Jews in in Nazi Oliner, S. P., & & Oliner, P. M. (1988). The The altruistic altruisticpersonality: personality:Rescuers Rescuers Jews Nazi Europe, New York: Free Press. Europe. Panskepp, J. (1986). The neurochemistry of behavior. Annual Review of of Psychology, Psychology, 37, 37, Annual Review 77-107. Peabody, D., & & Goldberg, Goldberg,L. L.R. R. (1989). (1989).Some Somedeterminants determinantsof offactor factorstructure structurefrom frompersonal personalJournalof ofPersonality Personalityand andSocial Social Psychology, 532-567. ity-trait descriptors. Journal Psychology, 57, 57, 532-567. & Fritzsche, B. A. (1993, August). Measuring Measuringthe theprosocial prosocialpersonality: personality: Four Penner, L. A., & Four construct validity validitystudies. studies.Paper Paper presented presented atat the the annual annual meeting meeting of of the the American American construct Psychological Association, Toronto, Canada. & Brookmire, D. A. (1979). Pre- and anti-social behavior as Penner, L. A., Michael, D. E., & and personality and situational variables. variables. Multivariate Multivariate a function of cost estimates and Research, 4, 111-124. ExperimentalClinical ClinicalResearch, Experimental 4, 111-124. & Perry, Perry, L. L. C. C.(1974). (1974). Denial Denial of ofsuffering suffering in in the the victim victim as as aastimulus stimulusto toviolence violence Perry, D. G., & Development,45, 45,55-62. 55-62. in aggressive boys. Child Development, & Perry, L. C. (1990). Peers perceptions of the consequences that Perry, D. G., Williard, J., & Development,61, 61,1310-1325. 1310-1325. Child Development, victimized children provide aggressors. Child focus on on children's children's Peterson, L. (1983). Influences of age, task competence, and responsibility focus altruism. Developmental DevelopmentalPsychology, Psychology,19,19,141-148. 141-148. psychology. Hillsdale, NJ:NJ: Erlbaum. Plomin, R. (1986). Development, Development,genetics, genetics,and and psychology, Hillsdale, Erlbaum. & Jaccard, J. J. (1976). An information approach to altruistic behavior. Pomazal, R. J., & Journal of ofPersonality Personalityand andSocial SocialPsychology, Psychology, 317-326. Journal 33,33, 317-326. Radke-Yarrow, M., & & Zahn-Waxler, Zahn-Waxier, C. C. (1984). Roots, motives, and patterns in children's prosocial behavior. In E. Staub, D. Bar-Tal, J. Karylowski, & J. Reykowski (Eds.), Development perspectives (pp.(pp. Development and maintenance maintenanceofofprosocial prosocialbehavior: behavior:International International perspectives 81-99). New York: Plenum Press. Radke-Yarrow, M., Zahn-Waxler, C, C., & & Chapman, M. (1983). Children's prosocial disposi disposipsychology:Vol. Vol, 4. 4, of child psychology: tions and behaviors. behaviors. In P. H. Mussen (Ed.), Handbook of (pp.(pp. 469-545). New York: Wiley. Socialization, social development Socialization,personality, personality,and and social development 469-545). New York: Wiley. Raven, B. H., & & Rubin, J. Z. (1983). Social Socialpsychology psychology(2nd (2nd ed.). ed.). New New York: York: Wiley. Wiley. Renken, B., Egeland, B., B., Marvinney, D., D., Manglsdorf, S., & Sroufe, L. A. (1989). Early Early childhoodantecedents antecedentsofofaggression aggression passive-withdrawal in early elementary childhood andand passive-withdrawal in early elementary schoolschool 257-282. [Special ofof Personality, 57, 57, [Special issue]. issue].Journal Journal Personality, 257-282. W., & & Bailey, S. (1983). The effect effect of level of moral reasoning in consistency Rholes, S. W., Social Cognition, Cognition,2,2,32-48. 32-48. between moral attitudes and related behaviors. Social & Moss, H. A. (1970). Patterns and determinants of maternal attachment. Robson, R. K., & Journal 976-985. Journalof ofPediatrics, Pediatrics,77,77, 976-985. & Lizzardo, T. (1986). A person-situation approach to altruistic Romer, D., Gruder, C. L., & Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, 1001-1012. 51,51, 1001-1012. behavior. Journal Rosenhan, D. L. (1970). The natural socialization of altruistic autonomy. In J. Macaulay & Altruism and helping helpingbehavior behavior(pp. (pp.251-268). 251-268). New New York: York:Aca AcaL. Berkowitz (Eds.), Altruism demic Press.

CHAPTER 30 50 CHAPTER

AGREEABLENESS

823 823

Rosenman, R. H. (1978). The The interview interview method method of of assessment assessment of the coronary-prone coronary-prone behavior behavior pattern. In T. M. Dembroski, S. M. Weiss, J. L. Shields, S. G. Haynes, & M. Feinleib (Eds.), Coronary-prone Coronary-prone behavior behavior(pp. (pp. 55-69). 55-69). New New York: York: Springer-Verlag. Springer-Verlag. Rothbart, M. K. (1989). Biological processes in temperament. In G. A. Kornstramm, J. Bates, & M. K. Rothbart (Eds.), Handbook Handbook of of temperament temperamentin inchildhood childhood(pp. (pp. 77-110). 77-110). Sussex, England: Wiley. Rowe, D., & Plomin, R. (1977). Temperament in early childhood. Journal of Personality of Personality Assessment, Assessment, 41, 41, 150-156. Rushton, J. P. (1980). Altruism, Englewood Cliffs, NJ: Altruism, socialization, socialization,and andsociety, society, Englewood Cliffs, NJ:Prentice-Hall. Prentice-Hall. Rushton, J. P., Chrisjohn, R. D., & Fekken, G. C. (1981). The altruistic personality and the self-report ferences, 2,2,1-11. self-report altruism scale. Personality Personalityand andIndividual IndividualDif Differences, 1-11. Rushton, J. P., Fulker, D. W., Neal, M. c., C, Nias, D. K. B., & Eysenck, H. J. (1986). Altruism and aggression: The heritability of individual differences. Journal of of Personality Personalityand and Psychology, 50, 50,1192-1198. 1192-1198. Social Psychology, Rushton, J. P., Littlefield, C. H., & Lumsden, C. J. (1986). Gene-culture coevaluation of of Proceedingsof ofthe theNational National complex social behavior: Human altruism and mate choice. Proceedings Academy of of Sciences Sciencesof of the the U.S.A., U.S.A.,83, 83,7340-7343. 7340-7343. Rushton, J. P., Russell, R. J. H., & Wells, P. A. (1984). (1984). Genetic similarity theory: Beyond kin selection. Behavior Genetics, Genetics,14, 14,179-193. 179-193. Adolescence: An An ethological ethological perspective. Savin-Williams, R. C. (1987). Adolescence: perspective. New York: Springer-Verlag. Savin-Williams, R. C, C., Small, S. A., & Zeldin, R. S. (1981). Ethology and Sociobiology, Sociobiology, 2, 167-176. conmiunication. Journal of of Abnormal Abnormal and Schachter, S. (1951). Deviation, rejection, and communication. Psychology, 46, 46,190-207. 190-207. Social Psychology, S. H., & Howard, J. A. (1981). (1981). In J. P. Rushton & R. M. Sorrentino (Eds.), Schwartz, S. helping behavior behavior (pp. 189-211). Hillsdale, NJ: Erlbaum. Altruism and helping Schwartz, S. H., & Howard, J. A. (1984). Internalized values as motivators of altruism. In perspectives E. Staub, D. Bar-Tal, J. Karylowski, & J. Reykowski (Eds.), International lnternational perspectives (pp. 229-255). New York: Plenum Press. on positive development (pp. Small, S. A., Zeldin, R. S., & Savin-Williams, R. C. (1983). In search of personality traits: A multimethod analysis of naturally occurring prosocial and dominance behavior. Journal of of Personality, Personality,51, 51,1-16. 1-16. Smith, S. S., & Kihlstrom, J. F. (1987). When is a schema not a schema? The Big Five traits Cognition, 5,5,26-57. 26-57. as cognitive structures. Social Cognition, Personality and and social socialbehavior. behavior. In In G. G. Lindzey Lindzey & & E. E. Aronson Aronson Snyder, M., & Ickes, W. (1985). Personality Handbook of of social social psychology psychology (3rd (3rd ed. ed. Vol. Vol. 2, 2, pp. pp. 1-71). 1-71). New New York: York: Ran Ran(Eds.), Handbook dom House. Helping aa distressed distressed person: person: Social, Social,personality, personality, and and stimulus stimulus determinants. determinants. Staub, E. (1974). Helping experimentalsocial socialpsychology psychology(Vol. (Vol.7,7,pp. pp.293293In L. Berkowitz (Ed.), Advances in experimental 341). New York: Academic Press. 341). Social and andpersonal personalinfluences influences(Vol. (Vol. Staub, E. (1978). Positive social behavior and morality: Social 1). New York: Academic Press. development of altruism and aggresStaub, E. (1986). A conception of the determinants and development aggres sion: Motives, Zahn-Waxier, E. M. Cummings, & sion: Motives, the self, and the environment. In C. Zahn-Waxler, lannotti (Eds.), Altruism and and aggression: aggression:Biological Biological and social origins 135-164). R. Iannotti and social origins (pp.(pp. 135-164). Cambridge, England: Cambridge University Press.

824 824

GRAZIANO GRAZIANO AND AND EISENBERG EISENBERG

Tellegen, A. (1985). Structures of mood and personality and their relevance to assessing anxiety, with an emphasis on self-report. In A. H. Tuma & J. D. Maser (Eds.), Anxiety Anxiety and anxiety disorders disorders (pp. (pp.681-706). 681-706). Hillsdale, Hillsdale, NJ: NJ: Erlbaum. Erlbaum. Trivers, R. L. (1971). The evolution of reciprocal altruism. Quarterly Quarterly Review Review of of Biology, Biology, 46, 46, 35-57. (1961). Recurrent actors based Tupes, E. c., C, & Cristal, R. E. (1961). Recurrent personality personalityffactors based on on trait trait ratings ratings (USAF ASD Tech. Rep. No. No. 61-97). Lackland Air Force Base, TX: U.S. Air Force. Underwood, B., & Moore, Moore, B. B. (1982). (1982). Perspective-taking Perspective-taking and and altruism. altruism.Psychological PsychologicalBulletin, Bulletin, 91, 143-173. 91, Watson, D. (1989). Strangers' Strangers* ratings of the five robust personality factors: Evidence of a surprising convergence with self-report. Journal of of Personality Personalityand andSocial SocialPsychology, Psychology, 57, 120-128. 57, West, S. G., & Graziano, W. G. (1989). Long-term stability of personality: An introduction. of Personality, Personality,57, 57,175-193. 175-193. Journal of White, G. M. (1980). Conceptual Conceptual universals universals in in interpersonal interpersonal language. language.American AmericanAnthropolo Anthropologist, 82, 82, 759-781. gist, Whiting, J. W. W. M. M. (1975). (1975). Children Childrenof ofsix sixcultures. cultures,Cambridge, Cambridge, MA: Harvard MA: Harvard Whiting, B. B., & Whiting, University Press. Wiggins, J. S. (1980). Circumplex models of interpersonal behavior. In L. Wheeler (Ed.), personality and social psychology (Vol. 2, pp. 265-294). Beverly Hills, of personality Review of CA: Sage. J. S. S. (1991). (1991).Agency Agencyand andcommunion communionasasconceptual conceptualcoordinates coordinatesfor forthe theunderstanding understanding Wiggins, J. and measurement of interpersonal behavior. In D. Cicchetti & 8L W. W . Grove (Eds.), criticallyininpsychology: psychology:Essays Essaysin in honor Paul Meehl (pp. 89-113). New Thinking critically honor ofof Paul E.E. Meehl (pp. 89-113). New York: Cambridge University Press. Wiggins, J. S., Broughton, R. (1985). (1985). The Interpersonal Circle: A structural model for the integration of personality research. In R. Hogan & W. H. Jones (Eds.), Perspectives Perspectives in personality personality (Vol. (Vol. 1,1,pp. pp. 1-47). 1-47). Greenwich, Greenwich, CT: CT:JAI JAI Press. Press. Wilder, D. A. (1986). Social categorization: Implications for the creation and reduction of of intergroup bias. In L. Berkowitz (Ed.), Advances in experimental experimentalsocial socialpsychology psychology (Vol. 19, pp. 293-356). Orlando, FL: 19, pp. FL: Academic Press. Wynne-Edwards, V. C. (1962). Animal Animal dispersion dispersion in relation relation to social behavior. behavior. New York: Harper. c., Robinson, J. L., & Emde, R. N. (1992). The development of empathy in Zahn-Waxler, C, twins. Developmental DevelopmentalPsychology, Psychology,28, 28,1038-1047. 1038-1047. Zajonc, R. B., & & Markus, Markus, H. H. (1984). (1984). Affect Affect and andcognition: cognition:The Thehard hardinterface. interface. In In C. C.E. E.Izard, Izard, J. Kagan, & R. B. Zajonc (Eds.), Emotions, cognition, cognition, and behavior behavior (pp. 73-102). Cambridge, England: Cambridge University Press. Zeldin, R. S., Small, S. A., & Savin-Williams, R. C. (1982). Prosocial interactions in two mixed-sex adolescent groups. Child Development, Development,53, 53,1492-1498. 1492-1498. for predicting altruistic Zuckerman, M., & Reis, H. T. (1978). Comparison of three models for behavior. Journal of of Personality Personalityand andSocial SocialPsychology, Psychology,36, 36,498-510. 498-510.

CHAPTER CHAPTER 31 31

CONCEPTIONS AND CORRELATES OF OPENNESS TO EXPERIENCE JR, ROBERT R. MCCRAE AND PAUL T. COSTA, JR. GERONTOLOGY RESEARCH RESEARCH CENTER, CENTER, NATIONAL NATIONAL INSTITUTE INSTITUTE ON ON AGING, AGING, GERONTOLOGY NATIONAL INSTITUTES INSTITUTES OF OF HEALTH, HEALTH, BALTIMORE, BALTIMORE, MARYLAND MARYLAND NATIONAL

Personality traits are normally defined as dimensions of individual difference, and are oftenfirst by noting groups of individuals who who are conspicuously conspicuously first recognized by they are an attempt to under underdifferent. Much of personality psychology has been devoted to an stand psychopathology, because phobics, hypochondriacs, and suicides are so disdis tressingly different from the rest of us. Artists and poets form another group long held to be different, if not deviant. They are are remarkable for their specific artistic talents, talents, but they they are are also also characterized characterized by a set of mental, emotional, and attitudinal characteristics that set them apart (MacKinnon, 1962). Think of Leonardo da Vinci, of Beethoven, or of Whitman: They are all dreamers with keen imaginations, seeing possibilities that others miss. They are sensitive and passionate, with a wide and subtle range of emotional reactions. They are adventurous, bored by familiar famiUar sights, and stifled by routine. They have an insatiable curiosity, as if they retained into adulthood the child's wonder at the world. world. And And they theyare are unorthodox, unorthodox,free-thinking, free-thinking,and and prone proneto toflout floutcon convention. As neurotics can be used as exemplars of high scorers on the dimension of Neuroticism, so artists can be considered prime examples of individuals high in Openness to Experience. Few people have the gifts needed to be a creative artist, but many people have the dispositions. Indeed, recent research suggests that OpenOpen ness to Experience is one of the fundamental dimensions of personality (McCrae, HANDBOOK HANDBOOK OF OF PERSONALITY PERSONAUTY PSYCHOWGY PSYCHOLOGY

825 825

826 826

McCRAE MCCRAE AND AND COSTA

1993-1994,1994), to an an understanding of the artistic temperament, 1993-1994, 1994), relevant not only to but also to such diverse issues as social attitudes, hypnotizability, career changes, and moral reasoning. And although it is convenient to use the artist as an exemplar of Openness and and to refer to "open" "open" and and "closed" "closed"individuals, individuals,ititmust mustbe beremembered remembered that Openness refers to a continuum of individual differences in processing experi experience, and that the majority of people are intermediate in Openness. In this chapter we review the empirical literature fundamenHterature on Openness as a fundamen focuss is on the conceptualization of tal dimension of personality, but our major focu Openness. We will argue that Openness cannot be understood as the culture that is acquired through education or good breeding, nor as intellect or any other cognitive ability. Instead, we will suggest that Openness must be viewed in both structural and motivational terms. Openness is seen in the breadth, depth, and permeability of of consciousness, consciousness, and and in in the the recurrent recurrent need need to to enlarge enlarge and and examine examine ex expermeability perience. perience. A caution to the reader is in order: order: The concept of Openness appears to be unusually difficult to grasp. Among personality psychologists, it is the most & John, 1992), and controversial of the five basic factors of personality (McCrae & lay raters appear to have preconceptions about Openness that are inconsistent with psychological definitions (Funder & Sneed, 1993). But data provide abundant support for the construct, and an increasing number of psychologists have adopted support it (e.g., John, 1990). 1990). Here is a construct that must be approached with an open mind and and aa willingness willingness to to learn learn new new ways ways of of thinking thinking about about people. mind I. OPENNESS AS A BASIC DIMENSION OF PERSONALITY Isolated parts of the broad domain of Openness have long been recognized in psychology. Authoritarianism (Adorno, Frenkel-Brunswik, Levinson, & & Sanford, 1950/1969) was once a major research topic for personality psychologists; explor exploratory behavior has fixture of animal research since has been a fixture since the the 1950s 1950s (Bedyne, (Berlyne, 1955); 1955); Rogers' (1961) theory of psychotherapy was based on generating conditions to enhance openness to feelings; private self-consciousness (Fenigstein, Scheier, & Buss, 1975) has recently attracted much attention. However, these traits have rarely been seen as aspects of a broader and more basic dimension of Openness. A few researchers have pointed to such a dimension, although with somewhat different different labels. There appear to have been four relatively independent discoveries of the dimension we call Openness. 1. Working from Cattell's (1946) distillation distillation of the personality sphere sphere as as repre represented in natural language traits and in psychological tests, Fiske (1949) and later Tupes and Christal (1961/1992) and Norman (1963) reported five robust factors hn, Angleitner, & which have come to be called the Big Five (Jo (John, & Ostendorf, 1988) five-factor model model of personality (McCrae & & Costa, Costa, 1987). and form the basis of thefive-factor Cultureby byboth bothTupes Tupesand andChristal Christaland andNorman; Norman;Fiske Fiske The fifth factor was called Culture

CHAPTER CHAPTER 31 31

OPENNESS OPENNESS TO TO EXPERIENCE EXPERIENCE

827 827

had also considered the label Cultured Culturedbefore before adopting adopting the thephrase phrase Inquiring InquiringIntellect. Intellect factor is focused focused on intelli intelliAs typically construed within the Big Five tradition, this factor gence or intellectual activity, but includes cultural sophistication and imagination. his analyses of English language adjectives adjectives Goldberg (1981) found found a similar factor in his Intellect; Hogan (1986), strongly influenced influenced by Big Five research, and called it Intellect; Intellectanceininhis hispersonality personality inventory. inventory. included a measure of Intellectance 2. Tellegen and Atkinson (1974) began with an analysis of measures which had been empirically linked to hypnotic susceptibility. In a joint analysis with measures of ego resiliency and control, they found three replicable factors which they interpreted as Stability (the opposite pole of Neuroticism), Introversion (versus Extraversion), and "openness to absorbing and self-altering self-altering experiences," or AbAb sorption; only Absorption was related to hypnotic susceptibility. The Absorption factor was was defined defined by by scales scales measuring measuring reality reality absorption, absorption, fantasy fantasy absorption, absorption, dissofactor disso ciation, devotion-trust, autonomy, and openness to experience. Tellegen and Atkin Atkinciation, devotion-trust, autonomy, and openness to experience. Tellegen and son explicitly expUcitly noted noted the the breadth breadth of of their their Absorption Absorption factor: factor: "it "it exemplifies exemplifies the the son substantive divergence and structural convergence that is suggestive combination of combination of substantive divergence and structural convergence that is suggestive of aa major major dimension" dimension" (p. 273), and and they they described described cognitive cognitive and and motivational motivationalof (p. 273), affective components. They concluded that Absorption was best interpreted as affective components. They concluded that Absorption was best interpreted as aa capacity capacity for for absorbed absorbed and and self-altering attention, found in peak and mystical self-altering attention, found in peak and mystical experiences, hypnosis, hypnosis, and and artistic artistic creativity. creativity. experiences, 3. Coan (1974) was concerned with what he called the optimal personality, and he examined characteristics identified identified in a wide range of personality theories. He drew upon the work of Fitzgerald (1966) to measure the scope of awareness. Fitzgerald had been concerned with questionnaire assessment of the psychoanalytic concept of regression in service of the ego (Kris, 1952), and many of his items concerned regressive behavior and experience. Coan added other questions with a concerned less pathological cast and found found a general factor of openness in an analysis of the items. Coan reported that his Experience Experience Inventory Inventory items items were were correlated correlated with with items. Coan reported that his "measures that suggest emotional sensitivity, aesthetic interests, liberalism, and independence" and "a certain intellectual and emotional flexibility" (pp. 80-81). Because both Fitzgerald (1966) and Tellegen and Atkinson (1974) had drawn on of A As, the earlier work of s, O'Hara, and Munger (1962), these research lines are not strictly independent. It is noteworthy, however, that Tellegen and Atkinson focused on the depth and intensity of attention, whereas Coan was impressed by the scope of awareness in Open individuals. for age differences 4. Looking for differences in personality structure, Costa and McCrae (1976) clustered the scales of the Sixteen Personality Factor Questionaire (16PF; Neuroticism and Extraversion Cattell, Eber, & Tatsuoka, 1970). In addition to Neuroticism found that scales B B (intelligence), I (tender-mindedness), M (imagina (imaginaclusters, they found tion), and Ql formed a loose cluster in some age groups. They Q1 (liberal thinking) formed interpreted this this cluster cluster as as Openness Openness to to Experience, Experience, and and continued continued research research on on the the interpreted dimension using a modification of Coan's scales. An Experience Inventory (EI; dimension using a modification of Coan's scales. An Experience Inventory (EI; Costa & McCrae, McCrae, 1978) was created created to to measure measure Openness Openness in in the the areas areas of of fantasy, fantasy, Costa 1978) was aesthetics, feelings, actions, ideas, and values. When jointly factored with 16PF aesthetics, feelings, actions, ideas, and values. When jointly factored with 16PF

828

MCCRAE AND COSTA McCRAE

scales, a reasonably clear Openness factor was found, defined by 16PF B, M, and 01 scales and Ql and Experience Inventory Fantasy, Aesthetics, Actions, Ideas, and Values scales. The EI scales were revised and ultimately incorporated in the Revised NEO Personality Inventory (NEO-PI-R; Costa & & McCrae, 1992a), a questionnaire & measure of the five-factor model. model. Factor Factoranalytic analyticstudies studies of ofthe the NEO-PI-R NEO-PI-R (Costa (Costa& thefive-factor McCrae, 1992b; 1992b; Costa, McCrae, & & Dye, 1991) show that the six Openness scales consistently define a separate factor in men and women, in young adults and old, and in self-reports and observer ratings. Because the total NEO-PI Openness score is significantly and substantially related to Tellegen and Atkinson's Absorption, Goldberg's Intellect, and Norman's Culture (McCrae & Costa, 1985a), it appears that the lexical tradition and the ego regression traditions have converged in the identification of a broad and basic dimension of personality. five-factor model have also supported supported personaUty. Other studies of thefive-factor this conclusion. For example, when the 100 items of Block's (1961) California Q QSet (CQS) were factored, one of of the five factors contrasted "Values intellectual matters," "Rebellious, non-conforming," and "Unusual thought processes" with matters," values," "Judges "Judgesin in conventional conventionalterms," terms,"and and"Uncomfortable "Uncomfortable "Favors conservative values," with complexities"; this factor correlated .62 with NEO-PI Openness scores (McCrae, Costa, & & Busch, 1986). Similarly, Amelang and Borkenau (1982) found Unabhdngigkeitder derMeinungsbildung Meinungsbildung (Independence Judga factor they called Unabhitngigkeit (Independence of of Judg ment) in analyses of questionnaires and adjectives in a German sample. A wealth of more specific correlates of Openness have been identified; some of these are summarized in Table I (see also McCrae, 1993-1994).

n. n. TRAOmONAL TRADITIONAL CONCEmONS CONCEPTIONS OF OPENNESS Convergence on an empirical l�vel level has not been matched by convergence on a conceptual level. Indeed, there is not even widespread agreement on the label to use for this dimension (Saucier, 1992). Goldberg (1981) and Digman and Inouye (1986) preferred the and thethe correspond the term Intellect; Intellect;Norman Norman(1963) (1963)used usedCulture,' Culture; and correspondCosta, 1989b) is identified as Sensation versus ing factor (McCrae & Sensation versusIntuition Intuitionin in & & McCaulley, McCauUey, 1985). It is under underthe Myers-Briggs Type Indicator (MBTI; Myers & to sum sum up up one of the the broadest constructs in personality psychology difficult to standably difficult in a single word, but the choice of labels is important. As Digman (1987) pointed out, researchers like Guilford, Eysenck, and Cattell assumed that intellectual interinter ests were a reflection of intelligence and could best be measured by intelligence tests. Measures of the disposition of Openness were thus relatively neglected by these influential factorists. OpennesstotoExperience Experiencehas hasitsitsdisadvantages, disadvantages,too. too.Especially Especiallywhen when The term Openness abbreviated as Openness, Openness,ititmay maysuggest suggestthe therather ratherdifferent differenttrait traitofofinterpersonal interpersonal self-disclosure (Jourard, 1964). Openness may also suggest a passive openness or self-disclosure or uncritical receptivity, which is clearly inappropriate. Open people actively seek

CHAPTER CHAPTER 31 31

TO EXPERIENCE OPENNESS TO

829 829

TABLE II TABLE Selected Correlates Correlates of of Self-reports Self-reports on on the the NEO NEO Personality Personality Inventory Inventory (NEO-PI) (NEO-PI) Openness Openness Selected Scale in in Baltimore Baltimore Longitudinal Longitudinal Study Study of of Aging Aging Samples Samples Scale

Criterion Criterion Observer ratings Observer

Spouse-rated NEO-PI NEO-PI Openness Openness

N

rr

144 144

.60

Mean Mean peer-rated peer-rated NEO-PI NEO-PI Openness Openness

213

.60

Aesthetically reactive

254

.40 .33 -.41 -.41 40 -.40

CQS CQS items

Skilled in in play play and and humor humor Judges Judges in in conventional conventional terms terms Favors conservative conservative values values Favors Bipolar Bipolar Adjective Adjective Scales Scales Uncurious-Curious Uncurious-Curious Uncreative-Creative Uncreative-Creative Imperceptive-Perceptive Imperceptive-Perceptive Simple-Complex Simple-Complex SDS SDS Occupations Occupations Anthropologist Author Journalist Sculptor/sculptress Revised CPI CPI ScalesG Scales'* Social Presence Empathy Achievement via Independence Flexibility PRF Needs Change Sentience Understanding ACL Creative Personality Scale GZTS Thoughtfulness MBTI Sensation-Intuition Sensation-Intuition Haan Coping Scales Intellectuality Logical Analysis Regression in Service of of the Ego Suppression Sensation Seeking Scales V Thrill and Adventure Seeking Experience Seeking Disinhibition of Impulses Boredom Susceptibility

254 254

254

-

.

375

.30 .30

375 375

.29 .29

375 375

.34

375

.35

275

275

.36 .43 .38 .31

348 348

.42 .42

275

275

348

348 348

.43 .41 .42

296

.40 .40

296

296

310 275

468 348

348

348 348

348

312

312

312 312

.55 .54 .46 .35 .73

.45 .48 .34 .12 -.12 .34 .55 .28 .20 .20

Note. .05. CQS, CQS, California Q-Set; Q-Set; CPI, CPI, California California Psychological Inventory; Inventory; Note, All ps < .05.

PRF, Personality Personality Research Research Form; Form; GZTS, GZTS, Guilford-Zimmerman Guilford-Zinmierman Temperament Temperament Survey; PRF,

MBTI, MBTI, Myers-Briggs Myers-Briggs Type Type Indicator; Indicator;SDS, SDS, Self-Directed Self-DirectedSearch. Search.(Data (Dataadapted adaptedfrom fromCosta Costa&&

& Holland, HoUand, 1984; 1984; McCrae, McCrae, 1987, 1987, 1990; 1990; McCrae McCrae & & & & Piedmont, Piedmont, 1993). & G**Correlations Correlationsare arewith withOpenness Opennessfactor factorscores. scores. McCrae, 1988a, 1988a, 1988b; 1988b; Costa, Costa, McCrae, McCrae, McCrae,

Costa, 1985b, 1985b, 1987; 1987; McCrae, McCrae, Costa, Costa, & & Busch, Busch, 1986; 1986; McCrae, McCrae, Costa, Costa, Costa,

830 830

MCCRAE AND ANDCOSTA COSTA McCRAE

experience and are apt to be particularly reflective reflective and thoughtful thoughtful about about the out experience ideas they encounter. encounter. Whatever label we select will be insufficient insufficient to communicate the construct construct For that we must rely on exemplars, like artists and poets; on an analysis of of fully. For cf. Briggs, 1989); the elements or facets facets that combine to form form the broad domain ((cf. and on a network network of of empirical correlates and outcomes associated with high or low useful to have a standing on the dimension. In addition, however, it would be useful conceptual definition, definition, a theory of of openness, that can help explain explain why people people conceptual differ in Openness. Let us turn now to a consideration of of some possible ways of of differ construing Openness. construing

A. Openness as Culture The term Culture Culture was selected by Tupes and Christal (196111992) (1961/1992) because it seemed to sum up the elements "intellectual, cultured," "esthetically "esthetically fastidious," "imagina "imaginative," and "polished," in contrast contrast to "boorish," "practical, logical," and "clumsy, awkward." Given these definers, the label seems apt; they have a peculiarly high highreflect differences differences in social class brow cast that suggests that the dimension may reflect and breeding. If If this characterization had been confirmed confirmed by subsequent subsequent research, the topic of of Openness Openness would perhaps have belonged belonged in a handbook handbook of of sociology, not personality psychology. However, Tupes and Christal reported reported that this was of the five factors factors in their analyses, and subsequent studies have the least clear of suggested that the elements of of polish and sophistication are far far less central to the dimension than intellectual intellectual and aesthetic interests and imagination imagination ((McCrae, McCrae, 1990; McCrae & Costa, 1987; 1987; Peabody Peabody & Goldberg, 1989). 1989). of personality is the result result of of The label Culture suggests that this aspect of education-particularly education—particularly the liberal education that has long been the central ideal of of Western Western universities. Among Among the distinctive elements of of this approach approach are expo exposure to a broad broad range of of ideas, cultivation of both arts and sciences, and encourage encouragement of of a critical attitude with regard to accepted values and assumptions. Clearly, a liberal education will be most congenial to those who are by disposition open open to experience. education is indeed broadening is an empirical The extent to which a liberal education question; there is some evidence in support of of this premise ((e.g., Freede.g., Webster, Freed education itself itself probably plays only a minor role in the man, & Heist, 1962). But education development of of Openness. In a national survey of of nearly 10,000 men and women development of 35, a short scale measuring measuring Openness showed showed only a modest correla correlaover the age of of education, rr = = .28, p p < .001 ((Costa al., 1986). Education was tion with years of Costa et aI., neither necessary nor sufficient sufficient for for Openness. About About one-third of of respondents with education or less scored above the median on Openness; conversely, conversely, an eighth-grade education about one-third of of respondents respondents with some college education scored scored below the about open without having had the benefits benefits of of formal formal educa educamedian. Individuals who are open tion may may be be culturally culturally unsophisticated-they unsophisticated—they may may be be deeply deeply moved moved by by the the verses verses tion on greeting greeting cards-but cards—but from from aa psychological psychological perspective, perspective, they they are are open open nonetheless. nonetheless. on

CHAPTER 31 31 CHAPTER

OPENNESS TO EXPERIENCE

831 831

(attendConversely, highly educated individuals may show the trappings of culture (attend ing museums, lectures, and concerts) without a deep appreciation of the experiences these events provide. family influences influences Formal education, of course, is not the only source of culture; family might also be important in modeling and encouraging breadth of interest and tolerant attitudes. As Rogerians might expect, loving and lenient parents tend to have children who are higher in Openness, but the association is very modest in magnitude (McCrae & Costa, 1988). 1988). By contrast, there was evidence of a strong heritable component of Openness in a study of adult Swedish twins (Bergeman et al, 1993), and Tellegen and his colleagues (1988) reported similar findings for aI., findings for Absorption in an American sample. These studies suggest that it may be wise to reverse the causal interpretation of the association of education with Openness: The intellectual interests of open men and women may lead them to seek higher levels of education. B. B.

Openness as Cognitive Ability

Perhaps the most popular alternative label for the dimension we have called Open OpenIntellect, which whichisisdefined defined asas"the *'the ness to Experience is some form of the word Intellect, reason.. . . [[and the] capacity for for knowledge and understanding" ability to learn and reason and the] consider(Morris, 1976, 1976, p. 682). 682). This definition suggests that the trait domain under consider abiUties. Studies of trait adjectives show ation is best viewed as a set of cognitive abilities. intelligent,perceptive, perceptive,knowledgeable, knowledgeable, analytical among that such terms as intelligent, andand analytical are are among the definers definers of of the the factor-indeed, factor—indeed, they they are are the the chief chief definers definers in in some some studies studies the (Angleitner & & Ostendorf, Ostendorf, 1989; 1989; Goldberg, Goldberg, 1989). 1989). The The interest interest that that open open individuals individuals (Angleitner have in in aa wide wide range range of of experiences experiences might might be be understood understood as as the the result result of of their their have facility in in handling information; certainly certainly intellectual intellectual interests interests tend tend to to follow follow abilifacility handling information; abili ties. Further, Further, studies studies have have shown shown that that Openness, Openness, alone alone of of the thefive factors, is is positively positively ties. five factors, related to psychometric measures of intelligence and other cognitive abilities related to psychometric measures of intelligence and other cognitive abilities (e.g., (e.g., McCrae, 1987). 1987). The The heritability heritability of of Openness Openness might might be be explained explained by by the the heritability heritability McCrae, of intelligence. intelligence. Psychologists Psychologists have have spent spent more more time time and and effort effort studying studying intelligence intelligence of Olild than any other trait; by adopting the term Intellect, personality psychologists than any other trait; by adopting the term Intellect, personality psychologists ccould claim this vast literature as their own. Openness could be construed as intelligence claim this vast literature as their own. Openness could be construed as intelligence itself, or, or, as as Cattell Cattell suggested, as the the reflection reflection of of intelligence intelligence in in the the personality personality itself, suggested, as sphere. sphere. Intellect Despite these temptations, there are five reasons to reject the label Intellect and the interpretation interpretation it suggests:

1. Factor analytic studies of natural language adjectives are inconclusive and factor suggest that rated intelligence may mark not one but two factors. As in all factor analyses, the nature of the factors depends chiefly on the variables included. Re Refactor searchers like Borgatta (1964), convinced a priori of the interpretation of the factor (intelligent,rational rational as intelligence, included markers reflecting this interpretation (intelligent, logical,clear clearminded, minded,alert, alert, mature). Researchers with a broader conception and logical, mature). Researchers with a broader conception of of

832 832

MCCRAE AND AND COSTA

the factor included variables such as imaginative, andand artistic, imaginative,prefer prefervariety, variety,original, original, artistic, and found a correspondingly different factor (e.g., McCrae & Costa, 1985b). One & way to avoid the possible biases of variable selection is to attempt to obtain a representative sample of adjectives. Goldberg's (1989) work in this regard has led him to a factor he characterizes as Intellect; Peabody's (1987) analyses of semantic similarity judgments recovered a factor he identified as Openness to Experience. Both these efforts were based on the assumption that the English language adeade quately represents all important personality traits in single adjectives, but that assumption has been questioned (McCrae, 1990). For example, the phrase "prefers variety," which corresponds to the widely researched trait of novelty-seeking or need for variety (Maddi & & Berne, 1964), apparently has no counterpart in natural language adjectives. Further, studies that include ability terms like intelligent intelligent typically typicallyfind findthat that these items have substantial secondary loadings on the Conscientiousness factor. Table II confirms this by showing loadings for variables related to rated intelligence on both Openness and Conscientiousness factors. Rated intelligence appears to be related to both factors in peer ratings, self-reports, and semantic similarity judgjudg ments. Conceptually, this is perfectly reasonable. Individuals may be considered reasons: they may be intellectually curious, intelligent for either (or both) of two reasons: efficient, well-organized, competent, imaginative, and inventive, or they may be efficient, too broad, broad, and careful in their work. From this perspective, the label Intellect Intellect isis too because it encompasses and confounds aspects of two basically independent dodo mains. mains. 2. In another respect, respect. Intellect Intellect is too narrow a label. Even if we include intellectual interests along with intellectual intellectual abilities, abilities, the range range of of phenomena known known empirically to correlate with Openness would hardly be suggested by the term. Who would guess that individuals high in a factor labeled Intellect would be more easily hypnotized (Tellegen & Atkinson, 1974), more variable variable in in mood mood (Wessman (Wessman & & Ricks, 1966), 1966), or more "skilled in play and humor" (cf. Table I)? Open people are not only able to grasp new ideas, they enjoy doing so. The merely intelligent tend to have highly developed interests in specialized fields fields in which they excel; open people have a wide and ever-increasing range of interests. Further, these interests extend beyond intellectual pursuits. Open people want to taste different food, to see new new sights, sights, to to reconsider reconsider their theirvalues, values,to to develop developelaborate elaborate fantasies. Cognitive abilities may in some degree facilitate this exploration of the world, but they are neither necessary nor sufficient for it. Need for variety, tolerance of ambiguity, and preference for complexity all represent motivational aspects of Openness. In In addition, open people can can be be charac characterized by their nontraditional attitudes, their rich and complex emotional lives, and their behavioralfiexibility. flexibility. Like the other four basic dimensions of personality, Openness is a broad constellation of traits with cognitive, affective, and behavioral manifestations. It cannot be reduced to a single underlying ability. 3. The empirical association of Openness with psychometric measures measures of of intel intelligence is too weak to imply equivalence between the two constructs (McCrae,

CHAPTER 31

833 833

OPENNESS OPENNESS TO TOEXPERIENCE EXPERIENCE TABLE TABLE II II

Loadings of Rated Intelligence Variables on Openness Openness and Conscientiousness Conscientiousness

Factors in Selected Studies

Factor Study

N

Openness

Conscientiousness

Norman (1963) 74 74

Sample Sample C ratings

215

Sample D ratings

241 241

84

315 315

25 25

63 63

25 25

Female self-reports self-reports

189 189

Male self-reports self-reports

189

— —

25 25

Self-reports Self-reports

498

<40

41 41

44

Internal Internal analyses

—

66

54

Self-reports (Table 6)

192

51 51

19

Self-reports Self-reports (Table 8)

95

36

35

Self-reports (Table 9)

157 215 175 175

42

33 33

Borgatta (1964) Female ratings Male ratings

144

Conley (1985)

McCrae McCrae

& Costa (1985b, (1985b, 1987)

Peer ratings Peabody (1987) Goldberg Goldberg (1989)

Self-reports (Table 7)

Self-reports, Study 6 Self-reports, Study 7 Median

738

95 95

66

47 10 10 60

63 63

44

14 14

59

18 18

45

21 21

51 51

33 33

Note. Note. Decimal points are omitted. Conley (1985) did not report an Openness

Peabody*s (1987) data are based on similarity similarity judgments made by four four factor. Peabody's raters.

1993-1994). In a sample of men from the BLSA, correlations of .22 and .20 were found between NEO-PI Openness and WAIS Vocabulary and Total Army Alpha scores, respectively; further, when measures of personality and cognitive ability are factored jointly, six factors, not five, are recovered, with measured intelligence forming a distinct factor (McCrae, 1994; 1994; McCrae & & Costa, 1985a, 1985b). The IQ, and average participant in the BLSA receives high scores on measures of IQ, somewhat larger correlations would probably be seen in unselected samples. But given the reliability of psychometric measures, even correlations of .30 or .40 would would mean that most of the valid variance in intelligence is not related to Openness to Experience. The one form of cognitive ability that does show somewhat stronger correlacorrela 1987). Correlations around .40 tions with Openness is divergent thinking (McCrae, 1987). were consistently seen between a total divergent thinking score and a variety of

834 834

MCCRAE MCCRAE AND ANDCOSTA COSTA

self-report self-report and rating measures measures of Openness. We should not, not, however, discount discount the possibility may reflect reflect motivational features features of possibility that high scores scores on such tests may people rather rather than ability: Curious and and imaginative imaginative people people may maybecome become more more open people flexible and and fluent fluent thought. In any case, if an ability involved in tasks that require flexible Creativity would make a better interpretation of Openness were to be advanced, Creativity label than Intellect. The identification identification of Openness with Intellect Intellect may may also be misleading misleading with assessment, because because it suggests that Openness might be measured by regard to assessment, and validities of cognitive tests psychometric tests. Given the relative reliabilities and self-report questionnaires or ratings, this is a tempting alternative, and histori historiversus self-report cally it has been extremely influential. Cattell, for example, included a measure of Eysenck (1991) intelligence in his 16PF instead of asking questions about intellect. Eysenck also considered that this domain was adequately covered by cognitive measures. Of course, if Openness were equivalent equivalent to intelligence, this would be an appropriate Of T o the extent that Openness is something else, this approach ensures an decision. To incomplete assessment of personality. Figure 11 summarizes the relations between Openness, Intellect, Intelligence, Conscientiousness described in the preceeding sections. Both the breadth and and Conscientiousness

CONSCIENTIOUSNESS

F1G1JD FIGURE11 Schematic Schematicrepresentation representation of ofrelations relations among amongOpenness Opennessto toExperience, Experience,Measured MeasuredIntelligence, Intelligence,

Conscientiousness, and Intellect. Both Openness and Intellect are modestly related to Measured Intelli Intelligence. Some aspects of Intellect (perceptive, fore (perceptive, curious) curious)are areshared sharedwith withOpenness; Openness;others others(logica� {logical^ foresighted) are adventurous, empathic) are shared sharedwith withConscientiousness. Conscientiousness.Many Manyelements elementsininOpenness Openness(libera� (liberal, adventurous, empathic) are not not included included in in Intellect. Intellect. are

CHAPTER CHAPTER 31 31

OPENNESS OPENNESS TO TO EXPERIENCE EXPERIENCE

835

its superiority the relatively greater independence of the Openness construct suggest its to Intellect as a basic dimension of personality. 4. But there are also other, pragmatic reasons for preferring the term Openness. Openness, Intellect and Openness have very different different evaluative connotations. Individuals who are closed to experience would probably accept this designation; they are content to be down-to-earth and may be proud of their traditionalism. By contrast, no one wants to be called stupid. stupid. The highly evaluative term Intellect presents difficulties assess difficulties when feedback feedback on personality is provided, as in client-centered cUent-centered assessment (Costa & McCrae, 1989; 1989; McReynolds, 1985). It may also suggest to psycholo psychologists that Openness is superior to Closedness. In fact, there are many advantages advantages— both to the individual and to society-to Both innovation innovation society—to being closed to experience. Both and conservation conservation are necessary processes in any culture and any individual life. 5. Finally, the identification identification of Openness with Intellect effectively effectively short cir cirpersonaUty and intelligence. intelUgence. No one is likely to do research on cuits research on personality development of intelligence, the question of whether Intellect contributes to the development former but if we distinguish Openness from intelligence, we can ask whether the former affects or is affected by the latter—a question that might have profound conseaffects affected latter-a profound conse quences for for developmental and educational psychology. The heuristic value of of distinguishing such concepts is seen in the work of Welsh (1975) on the related constructs of origence and intellectence. On empirical, conceptual, and heuristic grounds, then, it seems that Openness for this factor than Intellect. This phrasing also spares us some is a better label for empirical embarrassments. For example, open individuals frequently entertain ideas & Meier, that we would not readily associate with intelligence. Epstein (Epstein & 1989) has developed a scale measuring "beliefs in esoteric and dubious phenomena, such as astrology and the existence of ghosts" (p. 51). 51). In a college sample, this Esoteric Thinking scale was substantially correlated with NEO-PI Openness, rr = = .47, N iV = = 59, 59, p < < .001. far easier to see these beliefs beUefs as an outcome of open open.47, .001. It is far reflection of intelligence. intelUgence. mindedness than as a reflection

m. HI. ALTERNATIVE CONCEmONS: CONCEPTIONS: STRUCTURAL AND MOTIVATIONAL A. A. Openness Openness as as Psychic Psychic Structure Structure As the consensual validation vaUdation of Openness ratings across observers demonstrates, is rather easily inferred inferred from observable speech and behavior (McCrae & & Openness is fundamentally Openness is a matter of inner experience, a Costa, 1989a). But fundamentally mental phenomenon phenomenon related to the scope of awareness or the depth and intensity of consciousness. It is therefore therefore not surprising that much of the conceptual basis from the work of dynamic psychologists, such as Adorno of this dimension comes from conet al. (1950/1969), Kris (1952), (1952), and Rogers (1961), for whom the concept of con sciousness was central.

836 836

MCCRAE MCCRAEAND ANDCOSTA COSTA

As Coan (1974) noted, psychoanalysts have typically focused on psychopathol psychopathology and on the restrictions in awareness brought about by the defensive processes of of repression. A A particular causal sequence is suggested by this model: intrapsychic conflkts conflicts lead to repression, which leads to limitations in the scope of of awareness. Removing Removing the conflicts should thus increase openness. This view is close to the one held by Rogers (1961), who viewed Openness as an outcome and reflection of mental health. Appealing as this formulation may be to both psychoanalysts and humanistic psychologists, it has two significant significant problems. First, Openness is unre unrelated to Neuroticism Neuroticism and most measures of mental health, meaning that poorly adjusted individuals are as likely to be open as are well-adjusted individuals (McCrae & Costa, 1985a). Second, it is difficult difficult to explain the generality and pervasiveness of of Openness from this perspective. In most psychoanalytic thought, defense mechanisms mechanisms operate on specific specific conflicts or anxieties. It is understandable that an early trauma might leave a particular blind spot, but why should it also lead to conservative political views and indifference indifference to art and beauty? A A more plausible dynamic model can be inferred from the writings of of Frenkel FrenkelBrunswik in The Authoritarian Authoritarian Personality Personality (Adorno et aI., al, 1950/1969). This book was ambitious—perhaps too ambitious-attempt ambitious—attempt to integrate political philosophy, an ambitious-perhaps personality theory. Certainly no one today social attitude research, and dynamic personality of the F Scale as a measure of of Openness to Experience, would advocate the use of but there can be little doubt that the high scorers on the Ethnocentrism and F scales studied by Adorno et al. were closed to experience (although they were also probably highly antagonistic). The CQS definers of low Openness (McCrae et aI., al., favors 1986) in particular show an uncanny resemblance to authoritarian features: favors conservative judges in conventional conventional terms; uncomfortable uncomfortable with complexities,' complexities; conservative values; judges moralistic; sex-role sex-role stereotyped stereotyped behavior,' behavior; and even productive productive (one of the few desir desirmoralistic,' characteristics attributed to authoritarians). Conversely, Frenkel-Brunswick Frenkel-Brunswick able characteristics noted that "there seems to be a general tendency on the part of of low scores [non [nonexperience—emotional, cognitive, authoritarians] to expose themselves to broad experience-emotional, of having to modify modify [their] preconceived notions and perceptual—even perceptual-even at the risk of of having to sustain conflicts" (p. 464). The hypothesis that Openness is inversely inversely and of related to authoritarianism is supported by evidence that Altemeyer's (1981) Right RightWing Authoritarianism scale is related to total NEO-PI-R Openness, r = .57, = -.57, and all its facets, rs = = -.29 - . 2 9 to -.63, -.63, N N = = 722, P p < < .001 (Trapnell, 1994). Frenkel-Brunswik Frenkel-Brunswik interpreted her findings in terms of lower defensiveness defensiveness on the part of of open people, an ability to allow into consciousness unacceptable or undesirable undesirable impulses. It is the denial of these impulses and affects in authoritarians of prejudice, through the operation of of projection that leads to the development of and externalization. "The resultant break between the conscious and unconscious of the high scorers, as compared with the greater fluidity layers in the personality of of transition and of intercommunication between the different different personality strata in of the low scorers, appears to have the greatest implications for their personality patterns" (p. 474).

CHAPTER31 31 OPENNESS OPENNESSTO TOEXPERIENCE EXPERIENCE CHAmR

837 837

However, Frenkel-Brunswik seems to have made a subtle but important disdis authoritarians tinction in the causal sequence envisioned. She did not suggest that authoritarians conflicts in particular ways because of have more conflicts, but that they deal with conflicts the structure of their consciousness. We might hypothesize that open men and women would intellectualize their conflicts, whereas closed men and women would repress or deny them. Projection may be used by closed individuals because they simpUstically in terms of good and bad, and it is easier to assume see the world simplistically self is good. Under this interpretation, authoritarianism that others are bad and the self closedness) of defense, but one of of the determinants of of the form (or closed ness) is not the result of of defense used. Repression and projection would then be seen not as causes of political conservativism and aesthetic insensitivity, but as correlates that share the same underlying cause: closedness to experience. There is some empirical evidence for this position. Haan (1965) devised de defense mechanism mechanism scales by empirically contrasting MMPI item responses of of individu individucUnically judged high or low in the use of various defenses; she created parallel als clinically coping mechanism scales from CPI items. Among the defense scales, scales. Openness to positively related to Intellectualizing IntellectuaUzing and negatively related to Experience was positively Denial (Costa, Zonderman, & McCrae, 1991). As Table I shows, Openness is also positively related to coping scales measuring Intellectuality and Logical Analysis, and negatively related to Suppression. Table I shows that Openness is also related to Regression in Service of the Ego, a concept advanced by Kris (1952) to explain artistic creativity. Although defense mechanism, it is clear that Regression in Service of of the Ego described as a defense is defined in structural rather than defensive terms. Kris suggested that some individuals mdividuals can loosen the boundaries that separate mature, reality-oriented second secondhinking seen in dreams ary process thinking from the prelogical, primary process tthinking and psychotic delusions. This form of of regression is adaptive, because primary process thinking is the source of of creativity: the conventional associations between ideas and images are temporarily abandoned, leaving the mind free free to try new associations. The artist then returns to secondary process thinking to select the useful useful products of this freer freer association and adapt them to the requirements of of reality. The consequences of of permeable cognitive structures are not always adaptive. Hartmann, Russ, Oldfield, Siven, and Cooper (1987) studied chronic nightmare sufferers. They reported that their subjects were likely to be artists or students, and were described as being open, vulnerable, and defenseless "on the thin-boundary or permeable-boundary of the continuum permeable-boundary end of continuum in all senses in which that term is used" (p. 56). (Hartmann, 1991, has gone on to develop his ideas about boundaries in the mind that provide a modern modem psychodynamic perspective on Openness.) Similarly, recent studies have shown links between Openness Openness and certain forms of of cognitive aberration. West, Widiger, and Costa (1993) found that, among college students, NEO-PI-R NEO-PI-R Openness was associated with Perceptual Aberration and especially Magical Ideation scales (Chapman, Chapman, & Raulin, 1978; Eck Eckblad & Chapman, 1983). Table III shows correlations of of NEO-PI-R NEO-PI-R scales with measures of dissociation (Bernstein & Putnam's, 1986, DES; Riley's, 1988, QED) QED)

838 838

McCRAE MCCRAE AND AND COSTA COSTA TABLE TABLE m m Correlations between NEO-PI-R Openness Scales and Measures of of Dissociation and Eccentric Perceptions NEO-PI-R Scale

DES

QED QED

EP EP

01: Fantasy Ol:

.37··· .37***

.58··· .58***

.24·· .24**

"· 32*** .32

.42··· .42***

02: 0 2 : Aesthetics 03: 0 3 : Feelings

.22· .22*

.33··· .33***

.30··· .30***

.15

.30··· .30***

.32··· .32***

04: 0 4 : Actions

05: 0 5 : Ideas

06: 0 6 : Values

Total Openness

.13 .13

-.04 -.04 .30··· .30***

.14 .14

.10

" .49· .49***

47*** .47··· .12 .12 .03 -.03 .47"'·· .47***

Note. N N = 127. DES, Dissociative Experiences Experiences Scale; Scale; QED, Questionnaire of of Experiences of of Dissociation; EP, Eccentric Perceptions. Data cited by permission from D. Watson, J. Harrison, and A. K. Slack, 1993, of dissociation dissociation and their relation to general [Measures of traits of of personality). personality]. Unpublished raw raw data, data. Univer Univerof Iowa. sity of Iowa. *p < .05 .05.. •• **p < .01. ***p < < .001. • p < p < .01. ""'p .001. =

and eccentric perceptions as measured by the Schedule for Non-Adaptive and Adaptive Personality (Clark, 1993). Very open people appear to have some of the characteristics of schizo typal thinking; whether these are adaptive or maladaptive schizotypal will probably depend on other aspects of personality and on the individual's soso cial environment. Perhaps the the most mosthighly highlydeveloped developedversion versionof ofthe thestructural structuralmodel modelofofOpenness Openness is found in the work work of of Rokeach Rokeach (1960), (1960), who whoargued arguedthat thatideas, ideas,beliefs, beliefs, and andattitudes attitudes were structured differently differently in inopen openand andclosed closedindividuals. individuals.Highly Highlydogmatic dogmaticindividu individuals were thought to have compartmentalized thinking in which inconsistent beliefs were isolated and discrepant information was summarily rejected. Individuals low in dogmatism were able to tolerate ambiguity and could gradually shift attitudes as the the weight weight of ofevidence evidence accumulated. accumulated. Using Using Coan's Coan'sOpenness Openness scale, scale,Wyrick Wyrick(1969) (1969) showed that open individuals in in fact acknowledged more frequent revisions in attitudes than did closed individuals. It appears that one useful and important way to characterize Openness is in terms of the structure of consciousness. Open individuals have access to more thoughts, feelings, and impulses in awareness, and can maintain many of these simultaneously. Tolerance of ambiguity, emotional ambivalence, and perceptual synasthesia are all hallmarks of the open person. The capacity for absorption, for of this this structure. structure. For Forthe the closed closed individual, individual, deeply focused attention, may be a result of ideas, feelings, feelings, and and perceptions perceptions are are relatively relatively isolated isolated and and must must compete compete for for full full ideas,

31 CHAPTER 31

TO EXPERIENCE OPENNESS TO

839 839

attention. For the open individual, all these elements may be simultaneously in awareness, providing a deeper and more intense experience.

B. B.

Openness as as Need Need for for Experience Experience Openness

A structural account of Openness may be necessary, but it does not seem to be sufficient. Open people are not the passive recipients of a barrage of experiences they are unable to screen out; they actively seek out new and varied experiences. Openness involves motivation, needs for variety (Maddi & Berne, 1964), 1964), cognition (Osberg, 1987), sentience, and understanding (Jackson, 1984). 1984). This active pursuit of experience can be seen in all the facets of Openness. Closed individuals may have daydreams, but they are likely to be conventional and repetitive and serve functions of escape from from stress (McCrae, 1982) or mere wish fulfillment. fulfillment. The the functions characterized by novelty and elaboration and daydreams of open individuals are characterized for actions: Any Any reasonable reasonable are motivated by their intrinsic interest. The same is true for adult (including some who are merely high in Agreeableness) would be willing to taste a new dish; the truly open go in quest of varied cuisines. Fiske (1949) highlighted the active curiosity of open individuals by naming corresponding factor factor Inquiring Intellect. Philosophical arguments are boring to his corresponding closed individuals because they have no practical value; they are interesting to open people because they are intellectually challenging and because they may lead to of new and surprising conclusions: Both the process of exploring and the novelty of discovery appeal to open people. Open individuals tend to endorse liberal political and social values (McCrae, in press) because questioning authority is a natural extension of their curiosity. The same willingness to pursue questions of value leads to higher moral development (Lonky, Kaus, & Roodin, 1984) and to the artist's rejection of convention. bohemian rejection Perhaps the clearest evidence of open individuals' need for experience per se is found found in their appreciation of the arts. At least since Kant's Critique Critiqueof ofJudgment, Judgment, it has been generally recognized that the aesthetic experience is disinterested: There is no practical reason to be concerned with the fate of tragic heroes-no heroes—no tangible benefit benefit from listening to symphonies. The only function function art serves is "to clarify, clarify, intensify, or otherwise enlarge our experience" (Canaday, 1980, p. 5), and this is the quintessential aim of open men and women. Zuckerman's (1979, 1984) extensive research and theorizing on Sensation Seeking is surely relevant to an understanding of the motivational aspects of Open Openness. As Table I shows, all the Sensation Seeking scales are significantly significantly related to Openness, particularly Experience Seeking. Zuckerman's (1984) suggestion that "novelty, in the absence of threat, may be rewarding through the activation of of noradrenergic neurons" (p. 413), points to a possible neurochemical basis for Openness. We have argued that open people are characterized both by a particular permeable structure of consciousness and by an active motivation to seek out the unfamiliar. It seems probable that the structure is the result of the motivation,

840 840

MCCRAE MCCRAE AND ANDCOSTA COSTA

rather rather than vice versa. In the absence of of a need for for new experience, an open structure would would not provide any clear advantage; indeed, it would expose the individual to distracting thoughts, troubling impulses, and cognitive inconsistencies inconsistencies for experience provides an incentive to tolerate ambigu ambigu(cf. Maddi, 1968). The need for exploratory drive may overcome its need ity and dissonance, just as an animal's exploratory for security. In both cases, the evolutionary evolutionary function function seems clear: Greater Greater experience experience for ultimately provides a basis for for better better adaptation. adaptation.

IV. ApPLICATIONS rv. FURTHER RESEARCH AND AND APPLICATIONS researched and least understood understood It may seem odd to assert that Openness is the least researched of of the five fundamental fundamental dimensions of of personality. After After all, there have been decades of research research on psychological defenses, defenses, authoritarianism, authoritarianism, hypnosis, creativity, and of the need for for variety. However, However, these diverse lines of of research research have not been inte integrated by the conception conception of of Openness as a fundamental fundamental domain domain of of personality personality which is reflected reflected in each. As a result, there has been little cross-fertilization cross-fertilization of of ideas literature has been fragmented. fragmented. The power of of the construct of of Openness in and the literature interpreting these areas can be seen in the explanations it provides for for unanticipated unanticipated interpreting results. Without Without it, how would we explain the curious finding finding that private self selfconsciousness is related to belief belief in paranormal paranormal phenomena (Davies, 1985)? Or Or the fact that the MBTI Sensation scale is negatively negatively related to Zuckerman's Sensation Sensation fact 43, N N = = 170, P p < < .001, in BLSA BLSA participants? participants? Once Once we we understand understand Seeking, r = - . .43, .001, in of these variables variables reflects reflects an aspect aspect of of Openness, the associations are clear. that each of Personality for round of of research research on these topics, psychology is poised for a new round Personality concept of of Openness. New instruments instruments (e.g., Costa & McCrae, 1992a; guided by the concept Trapnell & Wiggins, 1990) provide validated validated measures of of Openness and some of of Trapnell facets in both self-report self-report and observer rating formats. Scales measuring related its facets absorption and private self-consciousness self-consciousness can be used more intelli intellitraits such as absorption understood as aspects of a broader and more fundamental fundamental con congently if they are understood struct. -

Questions A. Some Research Questions important and pervasive a dimension of of individual difference difference that Openness is so important figure routinely in the research of of personality psychologists. Researchers Researchers it should figure benefit by including measures of of Openness in their studies in other fields can also benefit research on (McCrae, in press). Social psychologists should assess openness in research formation and change. Educational psychologists should consider consider Openness Openness attitude formation moderator variable in assessing assessing the value of of different different teaching methods. Indus Indusas a moderator of Openness in their trial and organizational psychologists should include measures of personnel selection batteries. Behavioral geneticists should should study its heritability heritability personnel (Bergeman et al., mvestigate its role in health (Bergeman aI., 1993), and health psychologists should investigate information seeking seeking and behavior behavior change. Cognitive psychologists should should examine information

CHAPTER 31 31 CHAPTER

OPENNESS OPENNESS TO TO EXPERIENCE EXPERIENCE

841 841

the relation of Openness to field independence, cognitive complexity, and other cognitive styles (Tetlock, Peterson, & Berry, 1993). of future of In an earlier article (McCrae & Costa, 1985a), we argued that the future Openness lay in the investigation of its effects effects across the life span in such areas as vocational career career and family life. We know that open individuals have Artistic and Investigative interests (Costa, McCrae, & Holland, 1984) and that they make more midcareer shifts (McCrae & Costa, 1985a). There There are are characterizations characterizations of of rigid rigid and and flexible flexible managers (Kahn, Wolfe, Quinn, Snoek, & Rosenthal, 1964) that suggest how variations in Openness might be seen in organizational settings. But much satisfaction, remains to be learned about how Openness influences career paths, job satisfaction, effects of Openness Openness and retirement planning. Similarly, we know little about the effects on the course of intimate and personal relationships. We know that Openness is adulthood (Costa & McCrae, 1988a), so we should be able to approach stable in adulthood longitudimany of these questions through retrospective studies; every prospective longitudi nal study should certainly include baseline measures of Openness.

B. B.

Applications in in Psychiatry Psychiatry and and Clinical Clinical Psychology Psychology Applications

psychiPersonality traits have always been considered important for the diagnosis of psychi atric disorders, and extreme and maladaptive variants of some aspects of personality classified as personality disorders in DSM-IV (American Psychiatric Associa Associaare classified tion, 1994). Wiggins and Pincus (1989) have shown that scales measuring these understood in terms of the five-factor model. Histrionics, for disorders can be understood example, are extremely high in Extraversion; Avoidants are high in Neuroticism and low in Extraversion. OpenNone of the DSM-IV personality disorders is uniquely associated with Open showsthat that aspects aspects of of Openness Openness ness, but a careful consideration of diagnostic criteria shows affect of 1994). The restricted affect of are relevant to several disorders (Costa & Widiger, 1994). self-aggrandizing fantasy fantasy of Narcissists, and the behavioral rigidity Schizoids, the self-aggrandizing of Compulsives are all significant significant clinical features that may well be related to Openness. In his reconceptualization of personality disorders, Millon (1986) has suggested that each disorder is characterized by a particular intrapsychic structure, and as we have seen, seen. Openness is a major major determinant of psychic structure. However, Openness has has a history of being overlooked, and it is is also worthwhile to consider that there may be personality disorders not identified identified in DSM-IV which represent represent pathological forms of Openness. Some individuals are so rigid in their adherence to tradition and so unwilling to accept change that they are unable to adapt to inevitable social changes. When combined with very low levels of of is also possible Agreeableness, this closedness may take on an antisocial character. It is that excessively high levels of Openness (particularly in the absence of comparably high levels of intelligence and Conscientiousness) may constitute a personality disorder. Such individuals may be so easily drawn to each new idea or belief belief that they are unable to form a coherent and integrated life structure.

MCCRAE AND ANDCOSTA COSTA McCRAE

842 842

Even where Openness is not relevant to the diagnosis of of a psychiatric disorder, disorder, it may have important important implications for for psychotherapy psychotherapy (Miller, 1991). Open individu individuals may be more receptive receptive to the idea of of therapy itself itself and more tolerant tolerant of of imagina imaginative forms forms of of therapy, such as Gestalt or hypnotherapy. Closed Closed individuals are are more likely to prefer prefer concrete and practical practical suggestions. Biofeedback Biofeedback and directive therapies may prove more successful successful with them. We began by suggesting that artists can be seen as exemplars of Openness, just as neurotics neurotics are exemplars exemplars of of Neuroticism. People, however, are not one onedimensional exemplars; they are individuals individuals who vary on at least five dimensions dimensions of of personality. People who consult psychologists and psychiatrists about their problems dispositions bring more than these problems to the therapy; they also bring other dispositions that shape their lives and condition their responses to therapy. Individuals who are respond quite imaginative, sensitive, empathic, flexible, inquisitive, and tolerant will respond differently from from those who are practical, down-to-earth, rigid, and dogmatic. Clini Clinidifferently account in designing the appropriate appropriate treatment treatment cians need to take Openness into account for each client. for

REFERENCES Adorno, T. W., W., Frenkel-Brunswik, E., E., Levison, Levison, D. D. J., J., & & Sanford, Sanford, R R. N. N. (1969). (1969). The Theauthoritar authoritarpersonality. New NewYork: York: Norton. Norton.(Original (Originalwork workpublished published 1950) 1950) ian personality. Altemeyer, B. (1981). Right-wing Right-wingauthoritarianism. authoritarianism.Winnipeg: Winnipeg:University UniversityofofManitoba ManitobaPress. Press. ber die faktorielle Struktur und externe Amelang, M., & Borkenau, P. (1982). O tJber exteme Validitlit Validitat einiger Fragebogen-Skalen Fragebogen-Skaien zur Erfassung Erfassung von Dimensionen der Extraversion und emotionalen LabilitiU Labilit^t [On the factor structure and external validity of some question questionnaire scales measuring dimensions of extraversion and neuroticism]. ZeitschriJt Zeitschrift fUr filr Differentielle 3, 3, 119-146. Differentielle und und Diagnostische DiagnostischePsychologie, Psychologies 119-146. American Psychiatric Association. (1994). Diagnostic Diagnosticand andstatistical statisticalmanual manualofofmental mentaldisor disorders (4th ed.). Washington, DC: DC: Author. Angleitner, A., & Ostendorf, F. (1989, July). July). Personality Personalityfactors factorsvia viaselfself-and andpeer-ratings peer-ratings based on a representative Paper presented at at descriptive terms. based representativesample sampleof ofGerman Germantrait trait descriptive terms. Paper presented the First European Congress of Psychology, Amsterdam. A s, A., 0*Hara, O'Hara, J. W., & Munger, M. P. (1962). The measurement of subjective experiences As, presumably related to hypnotic susceptibility. Scandinavian Scandinavian Journal Journal of of Psychology, Psychology, 3, 5, 47-64. Bergeman, C. S., Chipuer, H. M., Plomin, R, R., Pedersen, N. L., McClearn, McCleam, G. E., Nesselroade, J. R., R, Costa, P. T., Jr., & McCrae, R R. R R. (1993). (1993). Genetic and environmental effects on openness to experience, agreeableness, and conscientiousness: An adoption/twin study. of Personality, Personality,61, 61,159-179. 159-179. Journal of Berlyne, D. E. (1955). The arousal and satiation of perceptual curiosity in the rat. Journal of Comparative Comparativeand andPhysiological PhysiologicalPsychology, Psychology, 238-246. of 48,48, 238-246. & Putnam, Putnam,F. F.W. W.(1986). (1986).Development, Development,reliability, reliability,and andvalidity validityof ofaadissocia dissociaBernstein, E. M., & of Nervous and Mental Disease, Disease, 1174, 727-735. 74, 727-735. tion scale. Journal of (1961). The TheQ-sort Q-sortmethod methodininpersonality personalityassessment assessment and psychiatric research. SpringBlock, J. (1961). and psychiatric research. Spring field, IL: Charles C. Thomas.

31 CHAPTER 31

OPENNESSTO TOEXPERIENCE EXPERIENCE OPENNESS

843 843

The structure structure of of peronality peronality characteristics. characteristics. Behavioral BehavioralScience, Science,9,9,8-17. Borgatta, E. F. (1964). The 8-17. Briggs, S. R. (1989). The optimal level of measurement for personality constructs. In D. M. Buss & N. Cantor (Eds.), Personality 1990's (pp.(pp. 246-260). New York: Personalityresearch researchfor forthe the 1990*s 246-260). New York: Springer-Verlag. Canaday, J. (1980). What is art? artl New York: Knopf. ofof personality. Yonkers-on-Hudson, Cattell, R. B. (1946). The The description descriptionand andmeasurement measurement personality. Yonkers-on-Hudson, NY: World Worid Book. Cattell, R R. B., Eber, H. W., & Tatsuoka, M. M. (1970). The handbook for the Sixteen Personality Personality Factor FactorQuestionnaire. Questionnaire.Champaign, Champaign,IL:IL:Institute InstituteforforPersonality Personalityand andAbil Ability Testing. J., Chapman, Chapman, J. J. P., P.,& &Raulin, Raulin,M. M.L.L.(1978). (1978).Body Bodyimage imageaberration aberrationininschizophre schizophreChapman, L. J., Journalof ofAbnormal AbnormalPsychology, Psychology,87, 87,399-407. 399-407. nia. Journal The Schedule Schedulefor forAdaptive Adaptiveand andNon-Adaptive Non-Adaptive Personality. Minneapolis: Clark, L. A. (1993). The Personality. Minneapolis: University of Minnesota Press. personality.New NewYork: York:Columbia ColumbiaUniversity UniversityPress. Press. Coan, R. W. (1974). The optimal personality. multitrait-multimethodConley, J. J. (1985). Longitudinal stability of personality traits: A multitrait-multimethod Journalof ofPersonality Personalityand andSocial Social Psychology, 1266-1282. multioccasion analysis. Journal Psychology, 49, 49, 1266-1282. Costa, P. T., Jr., & McCrae, R. R (1976). Age differences R. (1976). differences in personality structure: A cluster of Gerontology, Gerontology,31, 31,564-570. 564-570. analytic approach. Journal of Costa, P. T., Jr., & McCrae, R. R. (1978). Objective personality assessment. In M. Storandt, clinical psychology psychology of of aging aging(pp. (pp. 1119-143). 19-143). I. C. Siegler, & M. F. Elias (Eds.), The clinical New York: Plenum Press. R. R. (1988a). Personality in adulthood: A six-year longitudinal Costa, P. T., Jr., & McCrae, R study of self-reports self-reports and spouse ratings on the NEO Personality Inventory. Journal of Personality Personalityand andSocial SocialPsychology, Psychology, 853-863. 853-863. of 54,54, R. R R. (1988b). From catalog to classification: Murray's needs Costa, P. T., Jr., & McCrae, R Journalof ofPersonality Personalityand and Social Psychology, 258-265. five-factor model. Journal Social Psychology, 55, 55, 258-265. and thefive-factor Jr.,& &McCrae, McCrae,R R.R. R.(1989). (1989).The TheNEO-PIINEO-FFImanual NEO-Pl/NEO-FFI manualsupplement. supplement.Odessa, Odessa, Costa, P. T., Jr., Assessment Resources. FL: Psychological Assessment R. (1992a). Revised Revised NEO NEO Personality PersonalityInventory Inventory(NEO-PJ (NEOPICosta, P. T., Jr., & McCrae, R. R Five-FactorInventory Inventory(NEO-FFl) (NEO-FFI)professional professional manual. Odessa, Odessa, FL:FL: R) and NEO Five-Factor manual. Psychological Assessment Resources. T., Jr., Jr.,& &McCrae, McCrae,R R.RR.(1992b). (1992b).Trait Traitpsychology psychologycomes comesofofage. age.InIn Sonderegger Costa, P. T., T.T. B.B. Sonderegger NebraskaSymposium SymposiumononMotivation: Motivation: Psychology aging 169-204). (Ed.), Nebraska Psychology andand aging (pp.(pp. 169-204). LinLincoln: University of Nebraska Press. Costa, P. T., Jr., McCrae, R R. R, R., & Dye, D. A. (1991). Facet scales for agreeableness and conscientiousness: A revision of the NEO Personality Inventory. Personality Personalityand andIndi Individual Differences, Differences,12, 12,887-898. 887-898. Costa, P. T., Jr., McCrae, R R, & Holland, J. L. (1984). R. R., (1984). Personality and vocational interests Journalof ofApplied AppliedPsychology, Psychology,69,69,390-400. 390-400. in an adult sample. Journal B., & Larson, R. R., Zonderman, A. B., Barbano, H. E., Lebowitz, B., Costa, P. T., Jr., McCrae, R D. M. (1986). Cross-sectional studies of personality in a national sample: 2. Stability Psychology and and Aging, Aging, 1,1, 144-149. 144-149. in neuroticism, extraversion, and openness. Psychology Costa, P. T., Jr., & Widiger, T. A. (Eds.). (1994). (1994). Personality Personality disorders disordersand andthe thefive-factor five-factor of personality. personality. Washington, DC: model of DC: American Psychological Association. Costa, P. T., Jr., Zonderman, A. B., B., & McCrae, R. R R. (1991). (1991). Personality, defense, coping, Cunmiings, A. L. Greene, & K. H. and adaptation in older adulthood. In E. M. Cummings,

844 844

MCCRAE AND COSTA

Karraker (Eds.), Life-span Perspectives on stress andand coping Life-spandevelopmental developmentalpsychology: psychology: Perspectives on stress coping (pp. 277-293). Hillsdale, NJ: NJ: Erlbaum. (pp. M. F. (1985). Self-consciousness and paranormal belief. Perceptual Davies, M. Perceptualand and Motor MotorSkills, Skills, 60, 60, 484-486. Digman, J. M. (1987, August). Eysenck, Eysenck, Guilford, Guilford, and andCattell Cattelland andthe theBig BigFive Five factors factors ofof personality. personality. Paper Paperpresented presentedatatthe theAmerican AmericanPsychological PsychologicalAssociation AssociationConvention, Convention, New York. M., & & Inouye, Inouye, J.J. (1986). (1986).Further Furtherspecification specification of ofthe the five fiverobust robustfactors factorsof ofpersonal personalDigman, J. M., Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, 116-123. 50,50, 116-123. ity. Journal & Chapman, J. L. (1983). Magical ideation as a measure of schizotypy. Journal Journal Eckblad, M., & of 51,51, 215-225. of ConSUlting Consultingand andClinical ClinicalPsychology, Psychology, 215-225. & Meier, P. (1989). Constructive thinking: thinking: A broad coping variable with specific Epstein, S., & Psychology, 57, 57, 332-350. components. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 332-350. 16, 5, or 3?-Criteria 3?—Criteria for a taxonomic Eysenck, H. J. (1991). Dimensions of personality: 16, paradigm. Personality Personalityand andIndividual IndividualDifferences, Differences, 773-790. 12,12, 773-790. Fenigstein, A., Scheier, M. F., & & Buss, A. H. (1975). Public and private self-consciousness: self-consciousness: Journalof of ConSUlting Consultingand andClinical Clinical Psychology, 522-528. Psychology, 43, 43, 522-528. Assessment and theory. Journal Fiske, D. W. (1949). Consistency of the factorial structures of personality ratings from Journalof ofAbnormal Abnormaland andSocial SocialPsychology, Psychology,44,44, 329-344. different sources. Journal 329-344. T. (1966). Measurement of openness to experience: A study of regression in Fitzgerald, E. T. Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, 4, 655-663. 4, 655-663. the service of the ego. Journal C, & & Sneed, C. C. D. D. (1993). (1993).Behavioral Behavioralmanifestations manifestationsof ofpersonality: personality:An Anecological ecological Funder, D. c., Journal of of Personality Personalityand andSocial Social Psychology, Psychology, 64,64, approach to judgmental accuracy. Journal 479-490. Goldberg, L. R. (1981). Language and individual differences: The search for universals in Review of ofpersonality personalityand andsocial socialpsychology psychology personality lexicons. In L. Wheeler (Ed.), (Ed.), Review (Vol. 2, pp. 141-165). Beverly Hills, CA: Sage. Goldberg, L. R. (1989, June). June). Standard Standard markers markers of of the Big Five Five factor structure. structure. Paper Paper presented at an Invited Workshop on Personality Language, Groningen, The Nether Netherlands. Haan, N. (1965). Coping and defense mechanisms related to personality personahty inventories. Journal Journal 373-378. of of ConSUlting ConsultingPsychology, Psychology,29,29, 373-378. ofof personality differences. Hartmann, E. (1991). Boundaries Boundariesininthe themind: mind:AAnew newpsychology psychology personality differences. New York: York: Basic Books. Hartmann, E., Russ, D., Oldfield, M., Sivan, I., & & Cooper, S. (1987). Who has nightmares? The personality of the lifelong nightmare sufferer. Archives Archives of General General Psychiatry, Psychiatry, 44, 44, 49-56. Hogan, R. (1986). Hogan MN: National PersonalityInventory Inventorymanual. manualMinneapolis, Minneapolis, MN: NationalCom ComHogan Personality puter Systems. Jackson, D. N. Form manual (3rd(3rd ed.).ed.). PortPort Huron, MI:MI: Research N. (1984). (1984). Personality PersonalityResearch Research Form manual Huron, Research Psychologists Press. John, O. P. (1990). The The "Big "Big Five" Five"factor factortaxonomy: taxonomy:Dimensions Dimensionsofofpersonality personalityininthe thenatural natural language and in questionnaires. In L. A. Pervin (Ed.), Handbook Handbookof ofpersonality personalitytheory theory and research research(pp. (pp.66-100). 66-100).New NewYork: York:Guilford GuilfordPress. Press. John, O. P., Angleitner, A., & & Ostendorf, F. (1988). The lexical approach to personality: European Journal Journalof of Personality, Personality, A historical review of trait taxonomic research. European 2, 171-203. Jourard, S. M. (1964). The transparent transparentself. selfPrinceton, Princeton,NJ: NJ:Van VanNostrand. Nostrand.

CHAPTER 31 31 OPENNESS OPENNESSTO TOEXPERIENCE EXPERIENCE

84S 845

R. L., Wolfe, D. M., Quinn, R R. P., Snoek, J. D., & Rosenthal, R R. A. (1964). Organiza OrganizaKahn, R tional ininrole conflict andand ambiguity. NewNew York: Wiley. tional stress: stress:Studies Studies role conflict ambiguity. York: Wiley. Kris, E. (1952). inin art.art. New York: International Universities Press. (1952). Psychoanalytic Psychoanalyticexplorations explorations New York: International Universities Press. Lonky, E., Kaus, C. C. R, R., & &Roodin, Roodin,P. P.A. A. (1984). (1984).Life Lifeexperience experienceand andmode modeof ofcoping: coping:Relation Relation to moral judgment in adulthood. Developmental Developmental Psychology, Psychology,20, 20,1159-1167. 1159-1167. MacKinnon, D. W. (1962). The The nature nature and and nurture nurture of of creative creative talent. talent. American American Psychologist, Psychologist, 17, 17^484-495. 484-495. Maddi, S. R R. (1968). The pursuit of consistency and variety. In R R. P. Abelson, E. Aronson, W. 1. J. McGuire, T. M. Newcomb, M. J. Rosenberg, & P. H. Tannenbaum (Eds.), Theories A sourcebook (pp.(pp. 267-274). Chicago: Rand McNally. Theoriesof ofcognitive cognitiveconsistency: consistency: A sourcebook 267-274). Chicago: Rand McNally. Maddi, S. R, R., & Berne, N. (1964). Novelty of productions and desire for novelty as active and passive forms of the need for Personality,22, 22,270-277. ll^^-lll. for variety. Journal of of Personality, McCrae, R R. R R. (1982). Age differences differences in the use of coping mechanisms. Journal JournalofGerontol of Gerontology, 37, 37, 454-460. ogy, R. R R. (1987). Creativity, divergent thinking, and openness to experience. Journal McCrae, R of Personality Personalityand andSocial SocialPsychology, Psychology, 1258-1265. of 52,52, 1258-1265. R. R R. (1990). Traits and trait names: How well is openness represented in natural McCrae, R EuropeanJournal JournalofofPersonality, Personality, 4, 119-129. languages? European 4, 119-129. McCrae, R R. R R. (1993-1994). Openness to experience as a basic dimension of personality. Imagination,Cognition, Cognition,and and Personality, 39-55. Imagination, Personality, 13, 13, 39-55. R. R R. (1994). Openness to experience: Expanding the boundaries of Factor V. McCrae, R EuropeanJournal JournalofofPersonality, Personality, 8, 251-272. European 8, 251-272. R. R R. (in press). Social consequences of experiential openness. Psychological PsychologicalBul BulMcCrae, R letin. letin. R. R, R., & Costa, P. T., Jr. (1985a). Openness to experience. In R R. Hogan & McCrae, R W. H. Jones (Eds.), Perspectives Perspectivesininpersonality personality(Vol. (Vol.1, 1,pp. pp.145-172). 145-172).Greenwich, Greenwich,Cf: CT: JAI Press. R, & Costa, P. T., Jr. (1985b). Updating Norman's "adequate taxonomy": McCrae, R R. R., Intelligence and personality dimensions in natural language and in questionnaires. Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, 710-721. Journal 49, 49, 710-721. McCrae, R R. R, R., & Costa, P. T., Jr. (1987). Validation of of the five-factor model of personality across instruments and observers. Journal 52,52, Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, 81-90. R. R, R.,& &Costa, Costa,P.P.T., T.,Jr. Jr.(1988). (1988).Recalled Recalledparent-child parent-childrelations relationsand andadult adultpersonality. personality. McCrae, R Journalof ofPersonality, Personality,56,56, 417-434. Journal 417-434. R. R, R., & Costa, P. T., Jr. (1989a). Different Different points of view: Self-reports and ratings McCrae, R advances in the assessment of personality. In J. P. Forgas & M. J. Innes (Eds.), Recent advances in social socialpsychology: psychology:An Aninternational international perspective 429-439). Amsterdam: Elsevier. (pp.(pp. 429-439). Amsterdam: Elsevier. in perspective R. R., McCrae, R R, & Costa, P. T., Jr. (1989b). Reinterpreting the Myers-Briggs Type Indicator from the perspective of the five-factor model of personality. Journal of of Personality, from Personality, 57, 17-40. 57, McCrae, R R. R, R., Costa, P. T., Jr., & Busch, C. M. (1986). (1986). Evaluating comprehensiveness comprehensiveness California Q-Set and the five-factor model. Journal of in personality systems: The California of Personality,54, 54,430-446. 430-446. Personality, McCrae, R R. R, R., Costa, P. T., Jr., & Piedmont, R R. L. (1993). (1993). Folk concepts, natural language, and psychological constructs: The California Psychological Inventory and the fivefive Journalof ofPersonality, Personality,61,61,1-26. 1-26. factor model. Journal

McCRAE MCCRAE AND AND COSTA COSTA

846 846

McCrae, R. R., & John, O. P. (1992). An introduction to the five-factor model and its applications. Journal Journalof of Personality, Personality,60,60,175-215. 175-215. McReynolds, P. (1985). (1985). Psychological assessment and clinical practice: Problems and pros prospects. In J. N. Butcher & c. G. D. Spielberger (Eds.), Advances in personality personalityassessment assessment (Vol. 4, pp. 1-30). Hillsdale, NJ: Erlbaum. psychotherapeutic utility of the five-factor model of personality: A Miller, T. (1991). The psychotherapeutic clinician's experience. Journal 57,57, 415-433. Journalof ofPersonality PersonalityAssessment, Assessment, 415-433. (1986). Personality prototypes and their diagnostic criteria. In T. Millon & & G. L. Millon, T. (1986). Towards the the DSM-IV Klerman (Eds.), Contemporary Contemporarydirections directionsininpsychopathology: psychopathology: Towards DSM-IV (pp. 671-712). New York: Guilford Guilford Press. W. (Ed.). (1976). The TheAmerican Americanheritage heritagedictionary dictionary of the English language, Boston: of the English language. Boston: Morris, W. Houghton-Mifflin. Houghton-Mifflin. 8L McCaulley, McCauUey, M. H. (1985). Manual: Manual:AAguide guidetotothe thedevelopment developmentand anduse useofof Myers, Myers, I. B., & the Myers-Briggs MyerS'Briggs Type TypeIndicator. Indicator.Palo PaloAlto, Alto,CA: CA:Consulting ConsultingPsychologists PsychologistsPress. Press. Norman, W. T. (1963). (1963). Toward an adequate taxonomy of personality attributes: Replicated Norman. factor structure in peer nomination personality ratings. Journal Journalof ofAbnormal and Social Social factor Psychology, 66, 66,574-583. 574-583. Psychology, Osberg, T. M. (1987). (1987). The convergent and discriminant validity of the Need for for Cognition Osberg. Scale. Journal Journalof ofPersonality PersonalityAssessment, Assessment, 441-450. 51,51, 441-450. Peabody, D. (1987). (1987). Selecting representative trait adjectives. Journal of of Personality Personality and and Peabody. Psychology,52, 52,59-71. 59-71. Social Psychology, Peabody, D D., & Goldberg. Goldberg, L. R. (1989). Some determinants of factor factor structures from from Peabody. Journalof ofPersonality Personalityand and Social Psychology, S5^l-S(fl. Social Psychology, 57, 57, 552-567. personality-trait descriptors. Journal Riley, K. C. (1988). (1988). Measurement Measurement of dissociation. Journal of of Nervous and Mental Disease, Disease, Riley. 176, 176, 449-450. 449-450. Rogers. becoming aaperson: person:AAtherapist's therapistsview viewof ofpsychotherapy. psychotherapy,Boston: Boston: Rogers, C. R. (1961). On becoming Houghton-Mifflin. Houghton-Mifflin. Books. . Rokeach, M. (1960). The open and closed mind. mind. New York: Basic Books. Saucier. Saucier, G. (1992). (1992). Openness versus intellect: Much ado about nothing? European Journal of Personality,6,6,381-386. 3S1-3S6, of Personality, Tellegen. Tellegen, A A., & Atkinson. Atkinson, G. (1974). Openness to absorbing and self-altering self-ahering experiences ("absorption"). ("absorption'*), a trait related to hypnotic susceptibility. Journal of of Abnormal Psychol Psychology, ogy, 83, 83, 268-277. L.. & Rich, Rich. S. Tellegen. Tellegen, A A., Lykken. Lykken, D. T T., Bouchard. Bouchard, T. JJ., Jr Jr., Wilcox, K. JJ., Segal. Segal, N. L., of Personality Personality (1988). Personality similarity in twins reared apart and together. Journal of and Social Social Psychology, Psychology,54, 54,1031-1039. 1031-1039. Tetlock, P. E E., Peterson. Peterson, R. S., & Berry. Berry, J. M. (1993). Flattering and unflattering unflattering personality Tetlock. of Personality Personality and and portraits of integratively simple and complex managers. Journal of Social Social Psychology, Psychology,64, 64,500-511. 500-511. of Trapnell, P. D. (1994). (1994). Openness versus Intellect: A lexical left tum. turn. European Journal of Trapnell. Personality, Personality,8,8,273-290. 273-290. Trapnell. Trapnell, P. D D., & Wiggins. Wiggins, J. S. (1990). Extension of the Interpersonal Adjective Scales to of personality. Journal of of Personality Personality and Social include the Big Five dimensions of Psychology, Psychology,59, 59,781-790. 781-790. Tupes. Tupes, E. c.. C, & Christal, R. E. (1992). Recurrent personality factors based on trait ratings. of Personality, Personality,60, 60,225-251. 225-251.(Original (Originalwork workpUblished published1961) 1961) Journal of Webster, H H., Freedman. Freedman, M M., & Heist. Heist, P. (1962). Personality changes in college students. In Webster. and social interpretation of of (Ed.), The The American Americancollege: college:A Apsychological psychological and social interpretation N. Sanford (Ed.). the higher learning learning(pp. (pp.811-846). 811-846).New NewYork: York:Wiley. Wiley. .•

.•

.•

.•

.•

.•

.•

.•

.•

.•

.•

CHAPTER 31 CHAPTER

OPENNESS OPENNESS TO TO EXPERIENCE EXPERIENCE

847 847

G. S. (1975). Creativity Creativityand andintelligence: intelligence: personality approach. Chapel UniverWelsh, O. AA personality approach. Chapel Hill:Hill: Univer sity of North Carolina, Institute for Research in Social Science. Wessman, A. E., & Ricks, D. F. (1966). Mood and personality. personality. New York: Holt, Rine Rinehart & Winston. West, K. Y., Widiger, T. A., & Costa, P. T., Jr. (1993). The placement of of cognitive cognitive and and perceptual ve-factor model Unpublished manu perceptual abberations abberationswithin withinthe thefifive-factor modelofofpersonality. personality. Unpublished manuscript, script. University of Kentucky, Lexington. Wiggins, J. S., & Pincus, A. L. (1989). Conceptions of personality disorders and dimensions and Psychol of personality. Psychological JournalofofConsulting Consulting andClinical Clinical PsycholPsychologicalAssessment: Assessment:AAJournal ogy, ogy, 1, i, 305-316. Wyrick, L. C. (1969). Correlates Correlatesof of openness opennesstotoexperience. experience.Unpublished Unpublishedmaster's master'sthesis, thesis. University of Arizona, Tucson. seeking: Beyond Beyond the the �ptimal optimallevel levelof of arousal. arousal,Hillsdale, Hillsdale, Sensation seeking: Zuckerman, M. (1979). Sensation NJ: Erlbaum. Behav Zuckerman, M. (1984). Sensation seeking: A comparative approach to a human trait. Behavioral and and Brain Brain Sciences, Sciences^7, 7,413-471. 413-471. ioral

This Page Intentionally Left Blank

32 CHAPTER 32

CONSCIENTIOUSNESS AND INTEGRITY AT WORK JOYCE HOGAN UNIVERSITY OF OF TULSA TULSA UNIVERSITY

DENIZ SONES DENIZ S. ONES UNIVERSITY OF OF MINNESOTA MINNESOTA UNIVERSITY

I. I. OVERVIEW OVERVIEW

Conscientiousness refers to conformity and socially prescribed impulse control. All comprehensive studies of natural-language personality descriptions-beginning descriptions—beginning with Allport AUport and Odbert (1936)-identify (1936)—identify a Conscientiousness dimension. dimension. Norman's Norman's (1963, 1967) peer rating studies of the structure of trait terms provide a taxonomic (1963,1967) foundation for organizing contemporary inquiries about personality structure; Con Conanaly scientiousness is one of five components of the taxonomy. Goldberg's (1990) analyses of Norman's trait lists repeatedly confirmed thefive-factor five-factor structure; he coined the expression "Big Five" to describe this structure (see Goldberg, 1993, for a concise summary). Goldberg's research is persuasive; because we now know the concise structure of of the trait lexicon, we can celebrate a major contribution to social scisci ence research. Ironically, while personality psychologists were busy exploring the structure of the trait lexicon and identifying identifying stable individual differences differences in interpersonal of behavior, the use of of personality assessment decUned declined among applied psychologists. Skepticism regarding the usefulness of personality measurement reached a peak during the 1960s. Two critiques were particularly influential. influential. Thefirst first was Mischel's lLtNDBOOK OF OF PERSONALITY PERSONALITY PSYCHOLOGY PSYCHOLOGY HANDBOOK

849 849

CoPYRIGHT © C 1997 ACADEMIC PRESS COPYRIGHT 1997 BY BY ACADEMIC PRESS..

ALL ,1G1fIS OF IEPIODUcnON IN ANY FORM RESERVED. AU RIGHTS OF REPRODUCHON IN ANY FORM RESERVED.

850 850

HOGAN AND HoGAN AND ONfS ONES

(1968) book, which which claimed claimed that that (1) (1) there there isisno noevidence evidencethat thatpersonality personalityisisconsistent consistent across situations, and (2) personality measures explain only a trivial amount of of variance in social behavior. The second, a review by Guion and Gottier (1965), concluded that there was no evidence for the validity of personality personaUty instruments. These claims spawned considerable research which ultimately resulted in a reversal critics' conclusions. Nevertheless, the shadow of skepticism stills exists and of the critics' some applied psychologists continue to endorse these critiques of personality asas sessment. The recent literature on the Big Five personality factors provides compelling evidence for its structural robustness; it is the basis for the resurgent interest in personality assessment. The evidence indicates that personality structure is consisconsis 1981; Johnson Johnson & &Ostendorf, Ostendorf, tent across different theoretical frameworks (Goldberg, 1981; 1993), using different assessments assessments (e.g., (e.g., Conley, Conley, 1985; 1985;Costa Costa&&McCrae, McCrae,1992; 1992;Lorr Lorr&& (e.g.. Bond, Nakazato, & Shiraishi, 1975; Bor1975; Bor Youniss, 1973), in different cultures (e.g., 1989; Digman & & Takemoto-Chock, 1981), and using ratings & Ostendorf, 1989; kenau & & Inouye, 1986; McCrae & Costa, obtained from different sources (e.g., Digman & Norman, 1963; 1963;Norman Norman&&Goldberg, Goldberg,1966; 1966;Watson, Watson,1989). 1989).Substantial Substantialempiri empiri1987; Norman, cal evidence exists for the five-factor five-factor structure of peer descriptors (Cattell, 1943, 1946,1947; 1949; Norman, 1963; 1963; Tupes Tupes & & Christal, Christal, 1961). 1961). Borgatta's Borgatta's (1964) (1964) 1946, 1947; Fiske, 1949; methods of of data data accumula accumulafinding acrossfive five methods research extends the robustness of thefinding tion. We view the Big Five as a useful nosology, not as a theory nor an explanation. Nevertheless, there is some disagreement about the nature and meaning of the constructs, the scope of the taxonomy, and the degree to which the dimensions are fundamental and incisive (cf. Block, 1993; 1993; Hough, Eaton, Dunnette, Kamp, & 1993; Waller & Ben-Porath, 1987). Among 1990; Johnson & & Ostendorf, 1993; McCloy, 1990; inpersonality personaUtyassessment assessment isisbased basedon onqualita qualitarenewed interest interest in applied psychologists, renewed tive (Goldberg, 1992; 1992; R. R. Hogan, Hogan, 1991; 1991;Schmidt, Schmidt,Ones, Ones,& &Hunter, Hunter,1992) 1992)and andquantita quantitative (Barrick & & Mount, 1991; 1991; Hough et aI., al., 1990; 1990; Ones, Viswesvaran, & & Schmidt, 1993; Tett, Tett,Jackson, Jackson,& &Rothstein, Rothstein,1991) 1991)reviews reviewswhich whichconclude concludethat thatwhen whenpersonal personality research is organized in terms of the Big Five factors, personality is consistently related to job performance criteria. The Conscientiousness dimension led the personality assessment revival in applied psychology. There are at least two reasons for this. First, lack of conscienconscien tiousness is a major problem in the workplace. Conscientious employees are good organizational citizens; delinquent employees, in contrast, are nonproductive and erode the economic health of an organization. Employers beg the psychological community for effective assessments of "honesty" "honesty" and -which are and "integrity" "integrity"—which are their their The demand demand persists persists and and there there isis no no sign sign that that itit will will abate. abate. Second, Second, empirical empirical words. The findingssupport supportthe thevalidity validityofofConscientiousness Conscientiousnessmeasures measuresfor forpredicting predictingcounter counterfindings productive behavior and job performance. Some personality measures that were developed to predict organizational delinquency criteria are widely used (e.g., 1972; J. Hogan & & Hogan, 1989; 1989; Paajanen, 1985). Meta-analyses including Gough, 1972; significant relations general measures of Conscientiousness show consistent and significant with all job performance criteria in the occupations studied (Barrick & Mount,

CuAPl'ER 32 CHAPTERS!

CONSCIENTIOUSNESS CONSOENTIOUSNESS

851 851

1991; 1991; Tett et aI., al, 1991). These validity validity generalization generalization studies studies indicate indicate that the the Big Big Five Five dimensions of personality, Conscientiousness is the only dimension of personality to show consistent validities across organizations, jobs, and situations. In this chapter, we review three theoretical explanations of the Conscientious Conscientiousness construct, including psychoanalytic theory, role-taking and folk concepts, and socioanalytic theory. We then describe the manner in which Conscientiousness has been assessed. We next review the empirical findings that support the validity of Conscientiousness measures. Finally, we offer offer some advice for for thinking about Conscientiousness—in terms of measurement and interpretation. Conscientiousness-in

n. n. CONCEPTUALIZING CONSCIENTIOUSNESS

A. Psychoanalytic Psychoanalytic Theory Theory A. In personality psychology, Freud provided an important early discussion of the Conscientiousness construct. He explained Conscientiousness in terms of the super superego—conscience—the first structure to develop in personality. The superego deterego-conscience-the deter mines one's attitude toward authority and, according to Freud, the superego is crucial because all of development concerns coming to terms with authority. Freud not only recognized the importance of the conscience, but he also anticipated a modem dilemma: that too much Conscientiousness can be as undesirable as too modern little. Although today's applied appUed psychologists seem to be concerned only with too Uttle, Freud was also concerned with the problems of too much-i.e., much—i.e., the problems little, caused by a rigid, omnipotent, and punitive superego that define the extreme high end of Conscientiousness. relaUnlike the other personality structures, the superego depends on social rela authorfundamental determinant determinant of a person's disposition toward author tionships, and the fundamental ity is the resolution of the Oedipus complex. The Oedipus complex, a universal constellation of unconscious wishes and fantasies, involves interaction, conflict, and negotiation with one's parents. Whether its resolution is positive or negative depends depends on the relationship between the child and his or her parents. Further, Freud argued that the view of childhood as a period of innocence free free of sexual corruption was mythical. Rather, children are primitive, undisciplined, and born of bom in a state of "polymorphous perversity." Many of the child's pleasure-producing activities arouse parental disapproval and only a small number of behaviors are socially approved (Fancher, 1990, 1990, p. 374). 374). Over time, the family or caretakers channel these primitive tendencies into acceptable expressions and civilized behavior. The nature of this long process of social development where the parents attempt to curb unruly childhood sexuality is fundamental fundamental to all subsequent relations with parents and authority. The insights of Freud suggest that Conscientiousness is a product of the superego that develops from resolving conflict between childhood sexuality and parentally guided forces of socialization. One of Freud's more accurate theoretical

HOGAN HoGAN AND AND ONES ONES

852 852

insights is that Conscientiousness begins in the process of resolving conflicts with authority. Relations with the parents determine relations with other authority figures figures in life-military life—military superiors, employers, mentors, and experts. B. B.

Role-Taking Role-Taking and and Folk Folk Concepts Concepts

The historic chasm between personality theory and personality measurement bebe comes particularly apparent in the 25 years between Freud's last writings and of the validity vaHdity of personality measures. Before Before 1950, the modern demonstrations of conventional wisdom of criminology (Sutherland, 1951) was that delinquents and nondelinquents could not be differentiated differentiated on the basis of personality. According nondelinquents to Gough and Peterson (1952), the most notable review of the time (Schuessler & Cressey, 1950) concluded, after after examining 113 studies, that personality measures could not distinguish criminals from noncriminals. From a role-taking perspective and based based on on "intuitive "intuitive grounds," Gough and and Peterson Peterson developed developed aa pool pool of of 64 and grounds," Gough 64 items that strongly differentiated delinquent from nondelinquent males, females, items that strongly differentiated delinquent from nondelinquent males, females, and army army personnel, personnel, calling calling into into question question social social science science research research findings from the the and findings from previous 25 years. previous 25 years. four themes characterized Most interesting, however, was Gough's insight that four items—role-taking deficiencies, resentment, alienation, and re rethe discriminating items-role-taking belliousness. These themes were retained in the items included in the Socialization California Psychological Inventory (Gough, 1957, 1987), perhaps the scale of the California most well-validated broad bandwidth measure of Conscientiousness available. From a role-taking theory of psychopathy and folk concepts, concepts, Gough (1960) proposed that people are normally distributed along a continuum of socialization so that some are unusually scrupulous and conscientious, most are normally rule-compliant, and some are hostile to society's rules and conventions. On empirical grounds, excessive hostility is associated with criminal and delinquent behavior. However, Gough's for criminal and delinquent behavior is that the "psychopath" fails to explanation for deficient role-taking capacity anticipate social expectations and therefore exhibits a deficient during social interaction. Insensitivity to expectations and rules seems to result from an egotistical inability to understand the effects effects of one's behavior on others. from However, this argument does not explain why delinquents or psychopaths are insensitive to the rules in the first place.

C. Socioanalytic Socioanalytic Theory Theory C. for the empirical relations Socioanalytic theory (R. Hogan, 1983) also can account for while explaining the importance of and individual differences differences in the Conscientious Conscientiousness dimension. Hogan's theory contends that (1) people evolved and still live in groups; (2) every group is characterized by a status hierarchy and those with status will make the rules for for people living in the group; (3) people are consciously or unconsciously motivated by status and social acceptance, which are prerequisite for reproductive success in the group; (4) social interaction is the process by which

CHAPTER32 32 CONSCIEN11 CoNsaEmiousNESS CHAPTER 0USNESS

853 853

all human needs and goals, including reproductive success, are met; and (5) the fundamental because the needs for process of social interaction and social life is fundamental status—getting ahead-and ahead—and social acceptance-getting acceptance—getting along—can status-getting along-can cross purposes. People interact in terms of their identities, which are constructed unconsciously from attempts to achieve status and social acceptance in the peer group. Individual differences arise, in part, because some people are more skilled in their social performance and more attentive to processes that will support that identity than others. Following socioanalytic theory, theory. Conscientiousness Conscientiousness is part of of an identity choice-an choice—an interpersonal strategy for dealing with the members of of one's group. In childhood, one might receive attention and approval for for being tidy, compliant, and dependable; a child is likely Ukely to repeat activities that bring such approval from authority. However, by adulthood, the processes by which one supports a Conscien Conscientious identity are unconscious. It is easy to understand how Conscientiousness promotes survival in the group, and survival in today's organization. People engage consistent with their identity; those who want others to see in activities that are consistent them as Conscientious will show up for work on time, complete assignments accu accurately, mow their lawns, and keep a balanced checkbook. People who earn the reputation of being Conscientious do not make waves, do not challenge authority, like rules, and avoid arguments, ambiguities, and altercations. differences in Conscien ConscienBut how does socioanalytic theory explain individual differences tiousness, particularly deviancy? These may be only one strategy for supporting a Conscientious identity, and that is by complying with the rules, customs, norms, and expectations of the group; through such behavior, one is regarded by peers of behavioral and co-workers as "conscientious." However, there are any number of predispositions that lead to a delinquent reputation. For example, among these are taking on a "tough guy" identity that facilitates status in a deviant group such as a gang. Ot�ers Others might be acts acts reflecting reflecting alienation, disaffection, dissolution, hostility hostility toward authority, impulsiveness, and vengefulness. Still another source of individual variation is self-deception self-deception about the congruence between one's actions and group norms. Consider the employee who routinely lectures co-workers on business ethics, but continually steals time and resources from the company company without any cognizance of of the contradiction. As Freud suggested, some of of us are unaware of the meanings of our actions, and in this case self-deception sets the stage for self-defeat. self-defeat. How can we reconcile the socioanalytic theory of Conscientiousness with the Big Five model? Socioanalytic theory insists that two definitions of of personality must be considered-personality considered—personality from the view of the actor and personality from the view of the observer. Hogan suggests that personality from the view of of the actor is a personal, intrapsychic evaluation of what a person is like "way down deep." of goals, intentions, fears, motives, and beliefs; much of of this It probably consists of scientific content is not observable and therefore will not be easily amenable to scientific study. However, personality from the view of the observer is based on an actor's behavior and coded in terms of trait words which describe that person's reputation. Reputations are reasonably reliable across observers and time. Observers describe

8S4 854

HoGAN AND AND ONES ONES HOGAN

actors' behavior using trait terms-responsible, con terms—responsible, dependable, and careful, or, conversely, irresponsible, chaotic, and careless. Reputation is encoded in trait words, trait words have a well-defined mathematical structure, and these trait words are the substance of the Big Five model. How do we get to Conscientiousness? Because people evolved in groups, there were pressures to get along as well as to get ahead. Trait words are the descriptive categories observers use to evaluate others during inevitable social interactions. As evaluative categories, these words reflect the amount of status and acceptance observers are willing to grant an actor; these trait descriptors become one's public reputation and have consequences for group success. Trait words can be organized in terms of the Big Five personality factors and these reflect the qualities quaUties and contributions which that person can be expected to bring to the group. group. The Big Five Conscientiousness dimension is concerned with a person seeming responsible and trustworthy, characteristics that are fundamental for maintaining a group.

m. in. ASSESSING CONSCmNTIOUSNESS CONSCIENTIOUSNESS In the 50 years of personality research that began in the 1930s, all major inventories have contained some scale level assessment of the Conscientiousness construct, broadly defined. Although test authors had different different agendas when constructing their instruments, it is noteworthy that, regardless of purpose, they included an assessment of Conscientiousness. In thinking about these scales, Cronbach's (1960) application of Shannon and Weaver's (1949) distinction between bandwidth and fidelity with thethe Socialization scale ofof thethe CPI which Gough appropriate.We Webegan began with Socialization scale CPI which Gough fidelityisisappropriate. and Peterson (1952) developed to distinguish delinquents and nondelinquents. The broadbandwidth bandwidth and andan anarray arrayof ofexternal external complexity of the scale's content gives it aa broad correlates, from voting behavior to incarceration. Investigations of the internal structure of the CPI Socialization Socialization scale scale indicate that it isis composed composedof offour fourhierarchi hierarchically ordered subfactors. These are hostility toward rules and authority, thrillthrill seeking impulsiveness, social insensitivity, and alienation (e.g., Rosen, 1977). The broad bandwidth of the Socialization scale necessarily reduces itsfidelity fidelity in predict predictcom ing any single relevant behavior. However, we contend that the richness and complexity of the Socialization scale is also appropriate for measurement of the broad Conscientiousness construct. Most criteria that applied psychologists aim to predict are complex in nature, with many factors interacting to cause the behavior of of interest. A good example is the criterion of job performance in industrial/organiza industrial/organizational psychology. To predict and explain such complex criteria, complex and rich predictors work best (Ones, Mount, Barrick, & & Hunter, 1994). The Big Five heuristic provides a systematic systematic way way to toidentify identify Conscientiousness Conscientiousness measures included in omnibus personality inventories. Based on the work of Costa, McCrae, and their colleagues, a number of published studies report correlations between the NEO-PI (Costa & & McCrae, 1992) and other well-known personality

CHAPTER 32 32 CHAmR

8SS 855

CONSQENTIOUSNESS CONSCIENTIOUSNESS

inventories published between 1930 and 1970. Table I presents selected correlational results for the NEO Conscientiousness scale and scales from other inventories. From this, it is apparent that the Conscientiousness construct is complex; there are at least three themes underlying the Table I correlations. Although these analyses Conscientiousness scale as the factor marker, other researchers also use the NEO Conscientiousness have discovered the complex essence of this dimension (Barrick & Mount, 1991). scale. Interpersonal Style Inventory Impulse First, from the CPI Self-Control scale, Judging/Perceiving Type, a compo compoControl scale, and Myers-Briggs Type Indicator JudginglPerceiving of control emerges. These correlates suggest that Conscientiousness, in part, nent of cautiousconcerns a lack of impulsiveness and spontaneity, and a disposition toward cautious Preferness and criticality. Second, from the Order scales of the Edwards Personal Prefer component ence Schedule (EPPS) and the Personality Research Form (PRF), the component of orderliness, tidiness, and compUlsiveness compulsiveness emerges. These relations suggest that of Conscientiousness is also associated with being organized, neat, and methodical. Third, from the CPI Achievement via Conformity scale, the EPPS Endurance scale, and the PRF Orientation toward Work versus Play dimension (Skinner, Jackson, & component of of hard work and perseverance perseverance emerges. These Rampton, 1976), 1976), a component TABLE I

Correlations between the NEO Conscientiousness Scale and Selected Personality Scales Scale

Measure

California Psychological Psychological Inventory Inventory (McCrae, (McCrae, Costa, Costa, & California & Piedmont, 1993)

Self-Control

Good Impression

Achievement via Conformance Conformance

& & Costa, 1989)

Interpersonal Style Style Inventory Inventory (McCrae (McCrae Interpersonal

Edwards Personal Personal Preference Preference Schedule Schedule (Piedmont, (Piedmont, Edwards McCrae, McCrae,

& Costa, Costa, 1992) &

Personality Research Research Form Form (Costa (Costa Personality

& McCrae, McCrae, 1988) &

JudginglPerceiving Judging/Perceiving (male)

.49 -.49

Impulse Control Control Impulse Stability Stability Order Order

& Rampton, Rampton, 1976) &

.46 -.46

.71 .71

.43 .43

.68 .68

.63 .63

Achivement Achivement

.46 .46

Endurance Endurance

.42 .42

Cognitive Structure Structure Cognitive Order Order

Jackson, Jackson,

.39 .39

Endurance Endurance

Harm Avoidance Avoidance Harm

Personality Research Research Form Form (Skinner (Skinner scales: scales: Skinner, Personality

37 .37

.

.40 -.40

JudginglPerceiving Judging/Perceiving (female)

& Costa, Costa, 1991) &

.31 .31

.32

Flexibility

Norm Favoring (vector 2) Myers-Briggs (McCrae

NEO NEOrr

.33 .33

.39 -.39

.60

Desirability Desirability

.46 .46

Orientation Orientation toward toward Work

.60

versus versus Play

8S6 856

HoGAN AND AND ONES ONES HOGAN

correlates suggest that Conscientiousness concerns responsible work orientation, where a person works hard because it is the right thing to do-as do—as opposed to a person who is ambitious but not necessarily conscientious. In a joint joint factor factor analysis of the the Comrey Comrey Personality Personality Scales Scales (CPS), (CPS), the the Eysenck Eysenck Personality Inventory, and the Sixteen Personality Factors (16PF), Noller, NoUer, Law, and Comrey (1987) interpreted the first first factor as Conscientiousness. Although it contained components similar to control and orderliness identified above, above, the the analy analysis also revealed the theme of conformity. The 16PF-G conformity scale defined the factor, with the CPS-C conformity scale loading .65 and the Eysenck lie scale .39. These results suggest that measures of the Conscientiousness construct loading .39. also can reflect tendencies toward rule compliance, obedience, and conventional integrity. These same themes appear when interpreting scores on the CPI Socializa Socialization scale. Since 1980, we we have havewitnessed witnessedthe thedevelopment development of ofaanew newgeneration generationof ofmultidi multidipersonaUty inventories designed to assess some or all the Big Five factors mensional personality as an explicit measurement goal. These inventories include the NEO-PI (Costa & & Mount, 1993), McCrae, 1992), the Personal Characteristics Inventory (Barrick & Goldberg's adjective markers (Goldberg, 1992), the Hogan Personality Inventory &, Hogan, Hogan, 1992), 1992),Lorr Lorrand andYouniss's Youniss's (1973) (1973) Interpersonal Interpersonal Style Style (HPI; (HPI; R. Hogan & Personahty Questionnaire (Tellegen, 1982), and Inventory, the Multidimensional Personality (lAS-R; Wiggins, 1991). All of these the Interpersonal Adjective Scales-Revised (IAS-R; scale of of assess assess "conscientiousness." "conscientiousness." Given Given the the different different orien orieninventories contain a scale is not notsurprising surprisingthat thatthe theinterpretation interpretationof of"consci "conscitations of the inventory authors, it is & Ostendorf, 1993). For entiousness" is inconsistent across instruments (Johnson & example, the NEO NEC Conscientiousness scale reflects orderliness and persistence (McCrae & 1992), whereas the Goldberg adjectives concern dependability, & Costa, 1992), carefulness (Goldberg, 1992), 1992), and andthe theHPI HPIPrudence Prudencescale scalereflects reflects responsibility, and carefulness impulse control, professed probity, preference for predictability, and virtuousness (R. Hogan & & Hogan, 1992). A substantial body of evidence shows that many omnibus measures of personperson ality contain a dimension of Conscientiousness. These measures have varying dede grees of Conscientiousness saturation. Beyond a core interpretation that these measures concern conformity and dependability, there are nuances within each measure that can be interpreted only through their nomological network with other measures. Test-test correlates are useful, but they provide only a limited view of a construct and they are aa necessary but insufficient condition for establishing validity. Analyses such as those reported in Table I need to be expanded to include whichpermit permitbroader broaderunderstanding understandingand andinter interof test-nontest relations, which evaluations oftest-nontest & Nicholson, Nicholson, 1988; 1988; pretation of the Conscientiousness construct validity (R. Hogan & Landy, 1986). In a meta-analysis, examining the construct validity of the Big Five dimensions of personality, personality. Ones, Schmidt, and Viswesvaran (1994) found that Conscientiousness-related scales from mainstream personality inventories correlate {N = = 288,512; 288,512; K = = 226). However, when the correlations .47 among themselves (N between Conscientiousness scales from personality inventories explicitly based on

CHAPTER32 32 CONSClEN11 CoNsaENnousNEss CHAPTER 0USNESS

857 857

the Big Five (HPI Prudence scale, Goldberg's Conscientiousness adjective checklist, and Personal Characteristics Inventory's Conscientiousness Conscientiousness scale) were examined, the average disattenuated correlation was .71. There are seemingly narrow bandwidth measures of of Conscientiousness that serve a specific specific purpose in industrial psychology. These are measures of of "integrity," where the assessment concerns honesty-dishonesty honesty-dishonesty (Murphy, 1993, p. 115). Tradi Traditionally, a distinction is made between tests that inquire directly about honest behavior and attitudes (e.g., "I stole more than $5,000 from from my last employer") and tests that use questions mapping onto the integrity construct. Questions on these measures are similar and, in some cases, identical to inventory items on Conscientiousness scales. Tests that inquire directly about honesty are labeled "overt" (Sackett, Burris, & & Callahan, 1989) or "clear purpose" (Murphy, 1993, p. 117), while tests where items are used to make inferences about the candidate's honesty are labeled "personality-based" (Sackett et al., aI., 1989) or "veiled purpose" (Murphy, 1993, p. 117) integrity mtegrity tests. We will focus on the latter type. Murphy (1993, p. 127) points out that it is difficult difficult to distinguish veiled purpose integrity tests from personality inventories. He also contends that what these tests measure is not well established. We are less skeptical than Murphy because we believe that the meaning of these measures comes from the pattern of their external correlates. Examples of personality-based integrity measures used in workplace testing and in personnel research include the Personal Outlook Inventory (Science Research Associates, 1983), Personnel Reaction Blank (Gough, 1972), Personnel Decisions, Inc., Employment Inventory (Paajanen, 1985), and HPI Reliability ReUability scale (J. Hogan & & Hogan, 1989). Evidence for interpreting the validity of of these integrity measures comes from external or nontest sources such as supervisor's reports of of job behavior, records of employee behavior, self-reports of work incidents and biographical experiences, and peer or co-worker evaluations. In addition, the meaning of of these measures can be inferred from the pattern of their relations with other well-validated instruments. Organizational users are interested in what tests mean in terms of predicting counter counterproductive job performance-that performance—that is, identifying persons whose scores suggest that they might behave in a dishonest or irresponsible way. For example, in a number of concurrent validation studies, scores on the HPI ReUability Reliability scale were associated with concentrations of blood alcohol levels of persons arrested for drunken driving ((rr = .62; Y. Nolan, Johnson, & = -.62; & Pincus, 1994), excessive absences from work (r = -.49; R. Hogan, Jacobson, Hogan, & & Thompson, 1987), work discharges (r = .28; J. Hogan, Hogan, & -.28; & Briggs, 1984), counseling counseling for aberrant behavior ((rr = .18; Raza, Metz, Dyer, Coan, & -.18; & Hogan, 1986), and, conversely, commendations ((rr = .51; .51; J. Hogan et aI., al, 1984). Ones et a1. al. (1993) conducted a comprehensive meta-analysis of of integrity tests and found that the criterion-related validity for predicting supervisory ratings of job performance was .41 (N (N = 7550; K K = 23). Ones et a1. al. (1993) also found that personality-based personality-based integrity tests predict externally measured counterproductive behaviors with an operational validity of of .29 (N {N = K = 62). In terms of test-test relations, the HPI Reliability Reliability scale correlated 93,092; K =

=

=

=

=

=

HOGAN HoGAN AND AND ONES ONES

858 858

Con with CPI Self-control, Good Impression, Socialization, and Achievement via Conformity scales with rs of .70, .49, .49, .46, and .42, respectivley (J. Hogan & & Hogan, 1989). Note the similarity in this pattern of relations to that found in correlations between the CPI scales and the NEO Conscientiousness scale. From peer ratings 128) who completed the HPI ReliabiUty Reliability scale, the Adjective of persons (N (A^ = = 128) Checklist (Gough &
rv. EMPIRICAL EVIDENCE FOR mE THE VALIDITY IV. OF CONSCIENTIOUSNESS

Meta-analyses evaluating the construct validity of Conscientiousness measures are beginning to be published. This is an important advance over earlier quantitative reviews of personality and job performance where measures of various constructs & Barthol, Barthol, 1953; 1953;Schmitt, Schmitt,Gooding, Gooding,Noe, Noe,&&Kirsch, Kirsch,1984). 1984). were aggregated (Ghiselli & unpub For example, Barrick and Mount (1991) reviewed the published and unpubof lished literature from 1952 to 1988 to identify criterion-related validity studies of studies that that met met their inclusion inclusion criteria; criteria; these these personality measures. They found 144 studies contained 162 samples and a total of 23,994 cases. They classified the studies by occupation and criterion type. The occupational groups consisted of professionals, police, managers, sales, and skilled/semiskilled workers, and these accounted for 24% of the samples, respectively. The criterion types consisted of 5,13,41,17, 13, 41, 17, and 24% 5, job proficiency, training proficiency, and personnel data, and these accounted for 68,12, and 33% 33% of of the the samples, samples, respectively. respectively. There There was some overlap overlap in in criterion criterion 12, and 68, type available for the the samples. type available for samples. was no noempirical empirical means means to toclassify classify the the various various personality personality scales scales Because there was into Big Five dimensions, Barrick and Mount asked six subject matter experts to usedin inthe the144 144studies. studies.Scales Scaleswere wereplaced placedinto intoone oneofofsixsixcategories categories classify the scales used labeled Neuroticism, Extraversion, Openness to Experience, Agreeableness, labeled Neuroticism, Extraversion, Openness to Experience, Agreeableness, ConCon and Miscellaneous. Miscellaneous. Raters Raters used usedinformation informationprovided providedin intest testmanuals manuals scientiousness, scientiousness, and or in in research research studies studies to to make make classification classification judgments. judgments. Agreement Agreement between between four four or of six raters was used for final classification decisions, and as an example of six raters was used for final classification decisions, and as an example the the

859

CHAPTER CHAPTER32 32 CONSCIENTIOUSNESS CONSQENTIOUSNESS

Conscientiousness classifications for 18 inventories and personality scales appear in Table IL II. Barrick and Mount applied the meta-analytic procedures specified by Hunter and Schmidt (1990a) to examine the validity of the five personality personaUty dimensions for (1) each occupational group, (2) the three criterion types, and (3) objective versus subjective criteria. Focusing only on Conscientiousness, the results indicated that Conscientiousness scales were valid predictors for all occupational groups evaluated. The estimated true score correlations for professionals, police, managers, sales, and skilled/semiskilled occupations were .20, .22, .22, .23, and .21, .21, respectively. Also, Conscientiousness was consistently valid across all criterion types with estimated n TABLE II Conscientiousness Scale Classifications from the Barrick Barrick & & Mount (1991) Meta-analysis Inventory Inventory

Subscale Responsibility Responsibility

Gordon Personal Personal Profile Profile Gordon

Vigor Vigor

Gordon Personal Personal Inventory Inventory Gordon

California Psychological Psychological Inventory Inventory California

Achievement via via Independence Independence Conformance Conformance Achievement Status Status

Responsibility Responsibility

Edwards Personal Preference Preference Schedule

Endurance Endurance Order Order

Achievement Achievement

Adjective Checklist Adjective Guilford-Zimmerman Temperament Temperament Survey Guilford-Zimmerman Self-Descriptive Inventory Inventory

Thematic Apperception Apperception Test Test Thematic

Achievement Achievement

Order

Restraint Restraint

Decisiveness Decisiveness

Achievement Motivation

& Psychologist Psychologist Ratings Ratings &

Jackson Personality Personality Inventory Inventory Jackson Personality Research Research Form Personality

Initiative Initiative

Achievement Achievement Vigor Order

Organization

Responsibility

Achievement

Endurance Order

Impulsivity

16 Personality Personality Factor Factor Questionnaire Questionnaire 16

Conscientious

Multidimensional Personality Personality Questionnaire Questionnaire

Hard Work

Manifest Manifest Needs Needs Questionnaire Questionnaire

Achievement

Omnibus Omnibus Personality Personality Inventory Inventory

Hogan Personality Personality Inventory Inventory Hogan Comfrey Comfrey Personality Personality Scales

Differential Differential Personality Personality Questionnaire Questionnaire

Impulse Expression

Impulsiveness

Prudence

Ambition

Orderliness

Control Control

Achievement

860 860

HOGAN HoGAN AND AND ONES ONES

true correlations of .23, .23, .23, .23, and .20 for job proficiency, training proficiency, and Finally, in inevaluating evaluatingthe thevalidity validityof ofpersonality personalitymeasures measures personal data, respectively. Finally, for predicting objective versus subjective criteria, Barrick and Mount pointed out that subjective criteria are used about twice as frequently as objective criteria and, generally, true score correlations are higher for subjective ratings. Specifically, for Conscientiousness the estimated true correlation for subjective ratings was .26, whereas the correlation for various types of objective data was .14. Barrick and Mount concluded that the most significantfindings findings of of their metameta analysis concerned the Conscientiousness dimension. It was consistently valid for the five occupational groups and the three criterion types evaluated. They interpret work, people who who possess possess persistence persistence and and responsibil responsibiltheirfindings findings to mean that at work, bet ity and who "exhibit a strong sense of purpose" will perform or be evaluated better than those who do not (p. 18). Barrick and Mount generalize to the larger world of work, stating that "it is difficult to conceive of a job in which the traits world associated with with the the Conscientiousness Conscientiousness dimension dimension would would not contribute to to job associated not contribute job sucsuc cess" (pp. 21-22). Their advice to practitioners is that when the goal is to predict job performance performance based based on on personality personality assessment, assessment. Conscientiousness Conscientiousness measures measures are job are all jobs. jobs. More More recently, recently, Barrick Barrick the ones ones most most likely likely to to yield yield valid valid predictions predictions across across all the and Mount (1993) reported reported that that as as job job autonomy autonomy increases, increases, the the criterion-related criterion-related and Mount (1993) validity of of Conscientiousness Conscientiousness measures measures also also increases. increases. That That is, is. Conscientiousness Conscientiousness validity becomes more more important important for for predicting predicting job job performance performance as as autonomy autonomy becomes becomes becomes in jobs. jobs. On On aa related related note, note, Barrick, Barrick, Mount, Mount, and and Strauss Strauss (1993) (1993) used used more prevalent more prevalent in structural equations equations modeling modeling to to investigate investigate the the joint joint impact impact of of goal goal setting setting and structural and personality on on job job performance. performance. They They found found that that individuals individuals high high on on Conscientious Conscientiouspersonality ness set set goals goals and and persist persist in attaining them, them, and, and, consequently, consequently, perform perform well well on on the in attaining the ness job. So one one reason reason why why Conscientiousness Conscientiousness predicts predicts job performance is is because because job. So job performance Conscientious individuals plan to organize their work, spend more time on their job tasks, and persist at performance, all of which result in more job knowledge & Hunter, 1992). and superior supervisory ratings of job performance (Schmidt & Mount, Personnel PersonnelPsychology Psychologypublished publisheda asecond secondmeta metaFollowing Barrick and Mount, analysis of personality measures as predictors of job performance. This evaluation, conducted by Tett et al. (1991), concerned the same basic research questions raised by Barrick Barrick and and Mount Mountbut butincluded includedan aninvestigation investigationof ofrelevant relevantmoderator moderatorvariables. variables. Tett et al. reviewed approximately 500 research abstracts published since 1968 concerning personality assessment and job job performance. performance. Using Using explicit criteria for inclusion, they identified 86 studies with 97 97 independent samples and 13,521 valid cases. Studies were coded by two trained raters according to 12 key objective characteristics (e.g., exploratory versus confirmatory research, applicant versus inin 94% agreement. agreement. Personality Personality mea meacumbent subjects), and their coding resulted in a 94% classified using eight categories-Big categories—Big sures used as predictors in these studies were classified miscellaneous traits. traits. Classifications Classifications dimensions. Locus of Control, Type A, and miscellaneous Five dimensions, from personality measures evalu evaluof Big Five dimensions used factor analysis results from ated by by Costa Costa and and McCrae (1988). (1988). Meta-analytic Meta-analytic procedures procedures specified specified by by Hunter Hunter ated and Schmidt Schmidt (1990a, (1990a, 1990b) 1990b) were were used used to to estimate estimate true true correlations. correlations. and

CHAPTER 32 32 CONSCIENTIOUSNESS CoNsaEisnousNESS CHAPTER

861 861

The results indicated a mean correlation, corrected for both predictor and criterion unreliability, of .24 between job performance and all personality. That means correlation increased to .29 when the analysis included only confirmatory confirmatory research strategies. In addition, they found mean validities of .38, .30, .30, and .27 between personality measures and job performance studies that included job analyses, applicant subjects, military subjects, and published data, respectively. The corrected mean correlation between Conscientiousness measures and job perfor performance was .18, with the 95% 95% confidence interval ranging from -.11 to .35. Tett et a1. al. reemphasized the validity of personality measures for predicting job performance, particularly where the research strategy is confirmatory and where measures chosen are based on job analysis results. Nevertheless, ongoing controversy exists about methodological and statistical variations introduced to their meta-analysis that difficult (cf. Ones, Mount make the precise estimates of criterion-related validities difficult et al., aI., 1994). ConscientiousOnes's (1993) comprehensive analysis of personality measures, Conscientious ness measures, and integrity tests represents the most extensive research to date on the construct validity of measures of Conscientiousness. Ones's focus is on test-test understanding measures of integrity, which she does through analyses of test-test as well as test-job performance measures. As suggested by Landy (1986) and R. interHogan and Nicholson (1988), these comparisons allow us to understand and inter pret the Conscientiousness construct. Ones's primary research question concerned meawhere the personality trait "integrity" falls under the Big Five factors. Most mea sures that assess integrity are preemployment integrity or honesty tests, which responsibiUty, long-term job test publishers claim evaluate such characteristics as responsibility, commitment, consistency, proneness to violence, moral reasoning, hostility, work & ethics, dependability, depression, and energy level (cf. O'Bannon, Goldinger, & Appleby, 1989). These descriptions suggest that Conscientiousness is the general identify construct underlying integrity tests and integrity tests are largely designed to identify construct—irresponthe characteristics associated with the negative pole of the construct-irrespon sibility, rule violation, and hostility. Ones (1993) identified more than 100 studies reporting correlations between integrity tests and temperament measures. These studies suggest that integrity measures tend to correlate with each other and with personality-based measures of Conscientiousness. However, when other personality scales are included in the analyses, a pattern of relations with the Big Five Agreeableness and Emotional Stability StabiUty factors also emerges (Collins & Schmidt, 1993; Nolan, 1991). So, integrity tests, evaluated in terms of of the Big Five model, are primarily related to Conscien Conscientiousness and secondarily to Agreeableness. Also, integrity tests have substantial correlations with Emotional Stability. To understand what is measured by integrity tests. tests, Ones (1993) asked six specific research questions: (1) Are overt integrity tests correlated with each other? personality-based integrity tests correlated with each other? (3) Do both (2) Are personality-based personality-based integrity tests measure the same underlying construct? overt and personality-based (4) Do integrity tests correlate with Big Five measures? (5) Do integrity tests derive

862 862

HOGAN HoGAN AND AND ONES ONES

their criterion-related validity from the Conscientiousness dimension? (6) What is the estimated predictive validity of integrity tests? To answer these questions, Ones used two modes of data collection. The primary mode included test scores and demographic information from a student sample and job applicant sample (n {n = = 1,365) on the London House Personnel Selection Inventory, Stanton Survey, and Reid Report (all overt integrity tests according to Ones, 1993); 1993); the thePDI PDIInc. Inc.Employ Employscale). Per Perment Inventory, Hogan Personality Inventory (including the Reliability scale), sonnel Reaction Blank, and Inwald Personality Inventory (all personality-based andthe thePersonal PersonalCharacteristics CharacteristicsInventory Inventoryand andGoldberg's Goldberg'sAdjective Adjective integrity tests); and Checklist (both personality inventories). Between 300 and 500 students completed each inventory over 30 sessions of data collection. The secondary mode of data surveyof ofpublished publishedand andunpublished unpubUshedreports reportsof ofcorrelations correlationsbetween between collection was aasurvey overt integrity tests, personality-based integrity tests, and measures of the Big overt integrity tests, personality-based integrity tests, . and measures of the Big 8,000 correlation coefficients. The Five dimensions. This resulted in more than Five dimensions. This resulted in more than 8,000 correlation coefficients. The personality scales scales were were assigned assigned to to Big Five dimensions dimensions using using the the classifications classifications personality Big Five by Barrick Barrick and and Mount Mount (1991) (1991)and andHough Houghet eta!. al.(1990). (1990).These Thesecorrelational correlational developed by developed data were were analyzed analyzed using using meta-analysis meta-analysis procedures procedures developed developed by by Hunter Hunter and data and Schmidt (1990a). Schmidt (1990a). data. Ones found that the true correlations Using results from the primary data, .85, and andconfirmatory confirmatoryfactor factoranalysis analysisindicated indicated between overt integrity tests averaged .85, that test intercorrelations were due to the presence of a single factor. Similarly, the .75, which, which,after after true correlations between personality-based integrity tests averaged .75, confirmatory factor analysis, also indicated that only a single factor explained the matrix. Ones formed a composite of the three overt integrity tests and correlated it with a composite of the four personality-based integrity measures; measures; she found a .61. Confirmatory factor analysis of the intercorrelation true score correlation of .61. matrix of the seven integrity tests indicated a shared general factor, with loadings ranging from .63 for the PDI Employment Inventory to .87 for the London House hierarchiPersonnel Selection Inventory. Further analysis suggested evidence for a hierarchi cal factor structure of integrity tests, with a general factor across tests and two factors—one for overt tests and the other for personality-based tests, specific group factors-one to test type. To determine the relation between integrity tests and Big Five dimendimen linear composite composite of of the the seven integrity integrity tests tests and and linear linear composites composites for for scales scales sions, a linear classified in Big Five dimensions were formed. True score correlations between the integrity composite and the Conscientiousness, Agreeableness, and Emotional .91, .61, .61, and and .50, .50, respectively. respectively. This This pattern pattern of of correlations correlations Stability composites composites were were .91, Stability were repeated for separate separate integrity integrity composites composites of of overt overt tests tests and and personality personalitywere repeated for based measures. based measures. Ones performed performed aameta-analysis meta-analysis ofthe of thesecondary secondarydata datato totest testthe thegeneralizabil generalizability of the results from the primary analyses. Of interest are the correlations between integintegrity tests and the Big Five personality dimensions. For this analysis, 423 integ rity-Conscientiousness correlations across 91,360 data points resulted in a true correlation of .42. True score correlations of integrity tests with Agreeableness and

CHAPTER CHAPTER 32 32

CONSCIEN11 0USNESS CONSOENTIOUSNESS

863 863

Emotional Stability measures were .40 and .33, respectively. The correlations with Extraversion and Openness to Experience Experience were -.08 - . 0 8 and .12. Meta-analysis was also used to determine the operational predictive validity of of integrity tests and to determine whether their validity comes from the Conscien Conscientiousness dimension. Ones et al. (1993), in the most comprehensive meta-analysis ever reported, based on 665 validity coefficients coefficients across 576,460 data points, estimated the mean true score validity of integrity tests for predicting supervisory supervisory ratings of of job performance to be .46. Using this correlation along with the true score correlation of of .23 between individual scales of Conscientiousness Conscientiousness and job performance performance (Bar (Barof .42 between integrity and rick & Mount, 1991) and the true score correlation of Conscientiousness (Ones, 1993), Ones (1993) partialed Conscientiousness from the integrity-job performance relation, which reduced the true score correlation from person.46 to .41. She determined that Conscientiousness as measured by individual person ality scales only partially explains the validity of of integrity tests for job performance. Conversely, when integrity test scores are partialed from from the Conscientiousness-job Conscientiousness-job performance performance relation, the true score correlation is reduced from from .23 to .05. Ones concluded that measures of Conscientiousness, as assessed by Big Five personality inventories, are part of the broader construct measured by integrity tests. She suggests that integrity tests tap a higher order factor that includes Conscientiousness, Agreeableness, and Emotional Stability. It is interesting to note that these three Big Five are also those that Digman (1990) finds to be important for predicting delinquency and grades at school.

V. BREADTH AND AND THEORETICAL DYNAMICS V, MEASUREMENT BREADm The Big Five structure emerged from the study of of natural language, specifically adjectival peer descriptions. We view this taxonomy not as a theory but as a useful Conscientiousness dimension of of the starting point for technical discussions. The Conscientiousness Big Five structure concerns social conformity conformity and impulse control. Questionnaire of these themes (John, 1990). evaluations of this construct reflect one or both of Conscientiousness, as well as the other Big Five dimensions, has enormous band bandwidth, and, of course, this is a source of criticism (Briggs, 1989). However, as broad as Conscientiousness descriptors are, it appears that the integrity construct is even broader. This is the point of earlier literature (Murphy, of our analysis and review. The earUer high-fidelity Con 1993) suggested that integrity assessments are narrow bandwidth, high-fidelity Conscientiousness measures. That is, in the hierarchical representation of of Conscientious Conscientiousness, integrity would appear at a low level and as such should be capable of of predicting specific behaviors. Ones's (1993) analysis suggests that integrity is even broader (and psychologically more complex) than Conscientiousness. How do we account for this? First, consider the derivation of of the Big Five constructs. The natural language of personality description relied historically on rational and factor analytic methods. The lexical approach for identifying personality characteristics begins by listing relevant terms from dictionaries. Attributes were

864 864

HOGAN HoGAN AND AND ONES ONES

classified on conceptual grounds and later subjected to cluster and factor analytic techniques (Cattell, 1943). Confirmatory factor analysis is now the method of choice for Big Five researchers aligning personality scales with the five-factor five-factor taxonomy. scientiousness is interpreted though its network of other Conscientiousness From this analysis, Con . inventory correlates. However, if we use an empirical approach to scale construction, either alone or in conjunction with rational and factor analytic methods, the results lead to a conbroader interpretation of the construct. Many integrity measures have been con structed empirically by comparing the item responses of persons known to be low in integrity with the responses of persons thought to be high in integrity. The criterion of interest (e.g., criminal versus noncriminal behavior (is definedfirst, first, and then inventory items are identified; those that discriminate between subjects on the criterion characteristic characteristic are are retained retained because because they they are are related related to to the the behavior behavior of of criterion interest. If If the the criterion criterion is is broad, broad, then then the the empirical empirical predictor predictor also also will will be be broad. interest. broad. is not not altogether altogether clear clear why whysimple, simple,factorially factorially pure pureconstructs constructsshould should Furthermore, it it is Furthermore, predict complex complex phenomena phenomena such such as as job job performance performance end and criminal criminal behavior. behavior. predict The CPI Socialization scale is a good example of the way in which broad criteria defined broad predictor measures. The criterion characteristic that Gough estab had in mind was the continuum of Socialization-Asocialization. As a way of establishing the validity of the scale, Gough tested groups that most people would agree in Socialization. Socialization. Mean Mean Socialization scale scale scores, scores, which which he he views views as as a differed in citizens,'' bank officers, and sociological continuum, range from high school "best citizens," school superintendents down to unmarried mothers, county jail inmates, and prison inmates (Gough, 1975). Similarly, Similarly, in in developing developing the the HPI HPI Reliability ReUability scale, scale, the the criteinmates (Gough, 1975). crite rion of of choice choice was was antisocial antisocial behavior behavior (J. (J. Hogan Hogan & & Hogan, Hogan, 1989). 1989). The The initial initial HPI HPI rion theircorrespondence correspondence item composites for the Reliability scale were chosen based on their with the structure of deviancy as revealed in earlier factor analytic studies of the CPI Socialization scale. In the case of the HPI, item composites that came from EmopersonaUty scales of Conscientiousness, Agreeableness, Extraversion, and Emo the personality tional Stability empirically distinguished felons from nonfelons and delinquents from nondelinquents. This multifaceted measure of integrity, which is is keyed against the delinquency criterion, has a broad range of antisocial behavioral correlates in the workplace-insubordination, sabo workplace—insubordination, excessive absences, tardiness, equipment sabotage, and negative supervisory ratings. To summarize, summarize. Conscientiousness as assessed using rational or factor analyti analytically derived measures necessarily will focus on a single well-defined construct. Conversely, if integrity measures are developed empirically and if the criterion characteristic to be predicted is broad (and reliable), then the resultant scale will be multifaceted and complex. These explanations are posed to account for the difference between measures of Conscientiousness and integrity. difference How do we interpret Conscientiousness or integrity scores in terms of personperson ality theory? Conscientiousness, defined as social conformity and impulse control, is the degree to which a person makes an effortless adaptation to authority. For person cannot cannot adapt adapt fully fully to to authority authority because because there there is is always always ambivalence ambivalence Freud, a person

CHAPTERS! CHAPTER 32

CONSOENTIOUSNESS CONSCIENTIOUSNESS

86S 865

in the relationship. From this developmental perspective, normally Conscientious guilt associ associbehavior is a strategy designed to accommodate authority and to avoid gUilt with critical criticalsuperiors superiorsor ortheir theirsymbols. symbols.Individual Individualdifferences differencesin inConscientious Conscientiousated with ness form a continuum, from those who lack Conscientiousness-who Conscientiousness—who are unable authority—to those who are excessively Conscientious Conscientious— to resolve conflicts with authority-to who are compulsive, stingy, dependent, and stubborn. However, recall the evolutionary features that characterize people. We evolved in groups, we live in groups, and we participate in a status hierarchy within our groups. To To some some degree, degree, we weare aremotivated motivatedto toengage engagein insocial socialinteraction, interaction,which which is the process by which we achieve social status and social acceptance. People seek both social status, even if it is in the form of trying to avoid losing it, and social acceptance, even if it is in the form of trying to avoid criticism. (1983) socioanalytic perspective, the vehicle for for participating participating From R. Hogan's (1983) in the group process is one's identity. Identity is a repertoire of self-presentations self-presentations that develop during youth and adolescence, and identity is the basis on which social granted—or withdrawn. We might develop the status and social acceptance are granted-or identity or vocational role of a scholar, counselor, minister, or person of integrity. When we we interact interact and and others others observe observe us, us, their their reactions reactions to to us us become become our our reputaWhen reputa tions. Reputations Reputations are are coded coded in in trait trait words words that that reflect reflect others' others' descriptions descriptions of of the tions. the status and and acceptance afforded our our identities. identities. And And so so we we come come full full circle circle to to the the status acceptance afforded lexicon—the content base for the Big Five. derivation of the trait lexicon-the Over time, a person who is described by others as conscientious develops the identity of a "person of integrity." Developmentally, conformity engenders social acceptance, and up to a point conformity also will facilitate social status. The Conscientious child probably enjoyed positive relations with parents, caretakers, and others of authority because of his or her tendency to conform and desire to get along. The Conscientious adolescent is reliable, gets things done for the group, goes along with the group, and is comfortable with adult authority. He or she (Consciendevelops the reputation for being dependable, responsible, and careful (Conscien tious). Respect for others also will lead to a reputation for being kind and trusting (Agreeableness), as well as being consistently calm and content (Emotional StabilStabil adult ity). This identity is reinforced through social acceptance and approval. In adulthood, the identity and the processes which support it normally will be outside of conscious awareness. As for measurement, reputation is assessed through others' standardized apap praisals. This Thiscan cantake takethe theform formof ofpeer, peer,spouse, spouse,or orsupervisors supervisorsratings ratingsusing usingadjective adjective checklists checkHsts and observer reports. Although many applied appHed psychologists attempt to have observers evaluate the particular behavior of actors, actors, observers are rarely able to do this. this. Instead, they construct impressions of an actor's characteristics (Bartlett, 1932), and these impressions are what drive the evaluations (Murphy, Martin, & & Garcia, 1982). 1982). On the other hand, we also consider a person's responses to a form of of social social interaction. interaction. The The person endorses endorses inven invenpersonality inventory to be a form tory items in the way that he or she would like to be regarded in social interactions. The respondent thinks about the question, considers the impression he or she would

866 866

HOGAN HoGAN AND AND ONES ONES

make with a particular particular response, response, and and endorses endorses the the item item to to convey convey the the desired desired image. image. R. Hogan ((1983) 1 983) points out that this process is not as conscious or deliberate as it is in this this description. In In fact, fact, he hecontends contendsthat thatbecause because identities identitiesare areso sowell well solidified solidified in adulthood, little conscious effort effort goes into self-presentations. Moreover, this moot the the issue issue of of "faking" "faking" theoretical perspective on item response dynamics makes moot al, 1990). or item distortion (also see Hough et aI., The ubiquitous issue of faking also can be interpreted from a socioanalytic perspective, and this leads us to our final point. Persons who endorse such faking items as "I have never told a lie" lie" or "I have never hated anyone" show excessive virtuousness in their interpersonal style. Conscientiousness is one of the normal personaUty, measures of which lack item content at the extreme dimensions of personality, ends. Items, which may resemble faking items, at the low end of conscientiousness enjoy are absent because test test takers takers find find the the content content offensive offensive or or invasive invasive (e.g., (e.g., **I are absent because "I enjoy using illegal drugs"). drugs"). Similarly, Similarly, there there are are few few items items at at the the high high end. If If such such items items using high Conscientious Conscientious people. people. Such were used, used, we we might might have have aa very very different different view view of of high were Such persons to have have reputations as being inflexible, self-righteous, self-righteous, perfection perfectionpersons are are likely likely to reputations as being inflexible, istic, judgmental, and and evangelistic. Their private identities are are likely to be be that that of of evangelistic. Their private identities likely to istic, judgmental, morally scrupulous, scrupulous, virtuous, virtuous, upright, upright, and and characterologically characterologically superior superior person. person. aa morally Context is is important important for for interpretation, interpretation, so so this this person person will will also also show show relatively relatively high high Context scores for for Emotional Emotional Stability Stability and and elevations elevations for for the the other other Big Big Five Five measures measures scores (Johnson, 11990). Persons who who score score low low on on Conscientiousness Conscientiousness measures measures are are and 990). Persons and (Johnson, are seen seen as as deviant-and deviant—and this this is is their their identity, identity, too. are too. We are optimistic about the use of measures and integrity integrity measures of Conscientiousness and in applied psychology. In some ways, the Big Five has had the effect effect of pulling applied psychologists away from the retarding influences of behaviorism. Through meta-analyses, we begin to see the true potential of Conscientiousness and integrity measures for predicting important real world criteria such as job job performance performance and job. As we learn more about this construct, we counterproductivity on and off off the job. will better understand its nuances and measurement applications. REFERENCES Allport, G. W., W., & AUport, G. & Odbert, Odbert,H. H.S.S.(1936). (1936).Trait-names: Trait-names:AApsycho-lexical psycho-lexicalstudy. study.Psychological Psychological Monographs, (No. 211). 4747 (No. 211) . Monographs, Barrick, Barrick, M. M. R., R., & & Mount, Mount, M. M. K. K. (1991). (1991). The The Big Big Five Five personality personality dimensions dimensions and and job job 44, 1-26. performance: performance: A A meta-analysis. meta-analysis. Personnel PersonnelPsychology, Psychology, 44, 1-26. Barrick, R., & (1993). Autonomy Barrick, M. M. R., & Mount, Mount, M. M. K. K. (1993). Autonomy as as aa moderator moderator of of the the relationships relationships ofApplied between the between the Big-Five Big-Five personality personality dimensions dimensionsand andjob jobperformance. performance.Journal Journal of Applied 1 1 1-118. Psychology, Psychology,78,78, 111-118. Barrick, & Strauss, Strauss,J.J.P.P.(1993). (1993).Conscientiousness Conscientiousnessand andperformance performance Barrick, M. M. R., R., Mount, Mount, M. M. K., K., & of sales representatives: Journal ofApplied representatives:Test Testofofthe themediating mediatingeffects effectsofofgoal goalsetting. setting. Journal ofApplied of sales

Psychology,78,78, nS-lll, Psychology, 715-722.

Bartlett, A study in experimental and social Cam Bartlett, F. F. C. C. (1932). (1932). Remembering: Remembering: A study in experimental andpsychology. social psychology, England: Cambridge Cambridge University University Press. Press. bridge, England: bridge,

Cam-

CHAPTER 32

CoNsaEimousNESS CONSCIEN110USNESS

867

Block, J. (1993). A contrarian contrarianview viewof ofthe thefive-factor five-factorapproach approachtotopersonality personalitydescription. description. Unpublished manuscript, University of California, Berkeley. Bond, M. M. H., H., Nakazato, Nakazato, H., H., & &Shiraishi, Shiraishi,D. D.(1975). (1975).Universality Universalityand anddistinctiveness distinctivenessinindimen dimenPsychology, 6, 346-357. sions of Japanese person perception. Journal Journalof ofCross-Cultural Cross-Cultural Psychology, 6,346-357. Borgatta, E. E. F. F. (1964). (1964).The Thestructure structureofofpersonality personalitycharacteristics. characteristics.Behavioral Behavioral Science, 9,8-17. Science, 9, 8-17. Borkenau, P., & & Ostendorf, F. (1989). Descriptive consistency and social desirability in self selfEuropeanJournal JournalofofPersonality, Personality, 31-45. and peer reports. European 3, 3, 31-45. Briggs, S. R. (1989). The optimum level of measurement for personality constructs. In & N. Cantor (Eds.), Personality Personalitypsychology: psychology:Recent Recenttrends trends and emerging D. M. Buss & and emerging (pp.246-260). 246-260).New NewYork: York:Springer. Springer. directions (pp. directions Cattell, R. Journal R. B. B.(1943). (1943).The Thedescription descriptionof ofpersonality: personality:Basic Basictraits traitsresolved resolvedinto intoclusters. clusters. Journal of of Abnormal and Social Social Psychology, Psychology,38, 38,476-506. 476-506. ofof personality. Yonkers-on-Hudson, NY: Cattell, R. B. (1946). (1946). Description Descriptionand andmeasurement measurement personality, Yonkers-on-Hudson, NY: Worid Book. World Cattell, R. R. B. B.(1947). (1947).Confirmation Confirmationand andclarification clarificationof ofprimary primarypersonality personalityfactors. factors.Psychome Psychometrika, 12, 12,197-220. 197-220. trika, & Schmidt, F. L. (1993). Personality, integrity, and white-collar crime: crime: A Collins, J. M., & PersonnelPsychology Psychology46, 46,295-311. 295-311. construct validity study. Personnel Conley, J. J. (1985). Longitudinal stability of of personality traits: traits: A multitrait-multimethod multitrait-multimethodmultioccasion analysis. Journal 49,49, 1266-1282. Journalof ofPersonality Personalityand andSocial SocialPsychology Psychology 1266-1282. & McCrae, McCrae, R. R. R. R. (1988). (1988).From Fromcatalog catalog to to classification: classification: Murray's Murray's needs needs and and Costa, P. T., Jr., & Psychology, 55,55, 258-265. the five-factor five-factor model. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 258-265. FL: Costa, P. T., Jr., & (1992). NEO-PI-R & McCrae, R. R. (1992). NEO-Pl-R professional professional manual. manual Odessa, FL: Psychological Assessment Resources. ofof psychological testing (2nd ed.). New York: Harper & Row. Cronbach, L. L.J.J.(1960). (1960).Essentials Essentials psychological testing (2nd ed.). New York: Harper & Row. structure: Emergence of the Five Factor model. Annual Annual Digman, J. M. (1990). Personality structure: Review Review of of Psychology, Psychology, 41, 41,417-440. 417-440. J. M., M.,& &Inouye, Inouye,J.J.(1986). (1986).Further Furtherspecification specificationofofthe thefive five robust factors personalDigman, J. robust factors ofof personal Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, 116-123. ity. Journal 50,50, 116-123. J. M., M.,&&Takemoto-Chock, Takemoto-Chock,N. N.K. K.(1981). (1981).Factors Factorsininthe thenatural naturallanguage languageofofpersonal personalDigman, J. ity: Re-analysis Re-analysis and andcomparison comparison of of six sixmajor major studies. studies.Multivariate MultivariateBehavioral Behavioral Research, ity: Research, 16, 149-170. 16, Fancher, R. E. (1990). Pioneers Pioneersof ofpsychology. psychology.New NewYork: York:Norton. Norton. Fiske, D. W. (1949). Consistency of the factorial structures of personality ratings from Journalof ofAbnormal Abnormaland andSocial SocialPsychology, Psychology,44,44, 329-344. 329-344. different sources. Journal & Barthol, Barthol, R. R. P. P. (1953). (1953).The Thevalidity validityof ofpersonality personalityinventories inventoriesininthe theselection selection Ghiselli, E. E., & of employees. Journal Journalof ofApplied AppliedPsychology, Psychology,37, 37,18-20. 18-20. Goldberg, L. R. (1981). Language and individual differences: The search for universals in personality lexicons. In L. Wheeler (Ed.), Review Reviewof ofpersonality personalityand andsocial socialpsychology psychology NJ: Erlbaum. (Vol. 1, pp. 203-234). Hillsdale, NJ: Goldberg, L. R. (1990). An alternative "description of personality": The Big-Five factor Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, 1216-1229. 59,59, 1216-1229. structure. Journal The development of markers markersof ofthe theBig-Five Big-Fivefactor factorstructure. structure.Psycho PsychoGoldberg, L. R. (1992). The logicalAssessment, Assessment,4,4,26-42. 26-42. logical Goldberg, L. R. R. (1993). (1993). The Thestructure structureof ofphenotypic phenotypicpersonality personalitytraits. traits,American AmericanPsychologist, Psychologist, 48, 26-34. 48, AmericanJournal Journalof ofSociology, Sociology, Gough, H. G. (1948). A sociological theory of psychopathy. American 53, 359-366. 53,

868 868

HoGAN AND AND ONES ONES HOGAN

Gough, Inventory. Palo Alto, CA:CA: Gough, H. H. G. G. (1957). (1957). Manual Manualfor for the theCalifornia CaliforniaPsychological Psychological Inventory, Palo Alto, Consulting Consulting Psychologists Psychologists Press. Gough, H. H. G. G. (1960). (I960). Theory Theory and and measurement measurement of of socialization. socialization. Journal Journalof ofConsulting ConsultingPsy PsyGough,

chology, 24, 24,23-30. 23-30. chology, Gough, H. H. G. G. (1972). (1972). Manual Manualfor for the thePersonal PersonalReaction ReactionBlank. Blank. Palo Alto, CA: Consulting Palo Alto, CA: Consulting Gough, Psychologists Press. Psychologists Press.

Manualfor for the theCalif California Psychological Inventory. Palo Alto, Gough, H. H. G. G. (1975). (1975). Manual ornia Psychological Inventory. Palo Alto, CA:CA: Gough, Consulting Consulting Psychologists Psychologists Press.

Gough, H. G. California PsychologicalInventory Inventory administrator's guide. Palo Alto, Gough, H. G. (1987). (1987). Calif ornia Psychological administrator's guide. Palo Alto, CA: CA: Consulting Psychologists Psychologists Press. Gough, H. H. G., G., & & Heilbrun, Heilbrun, A. A. B. Jr. Jr. (1983). (1983). The The Adjective Adjective Checklist Checklistmanual: manual: 1983 edition. 1983 edition. CA: Consulting Palo Alto, Alto, CA: Consulting Psychologists Psychologists Press. Gough, H. G., & Peterson, Peterson, D. D. R R. (1952). (1952). The The identification identification and and measurement measurement of of predisposi predisposiGough, H. G., & tional factors factors in in crime crime and anddelinquency. delinquency.Journal Journalof ofConsulting Consulting Psychology, tional Psychology, 16, 16,207-212. 207-212. Guion, Guion, R R. M., M., & &Gottier, Gottier,R R.F. F.(1965). (1965).Validity Validityof ofpersonality personalitymeasures measuresininpersonnel personnelselection. selection. Personnel PersonnelPsychology, Psychology,18, 18,135-164. 135-164. Hogan, J., & Hogan, Hogan, R R. (1989). (1989). How How to to measure measure employee employee reliability. reliability. Journal Journalof ofApplied Applied Hogan, J., & Psychology, Psychology, 74, 74, 273-279. 273-279. Hogan, J., Hogan, R., & Briggs, Briggs, S. S. R R. (1984). (1984). Development Development and validation validation of of employee Hogan, J., Hogan, R, & selection drivers. Tulsa, OK: University of Tulsa. selection tests testsfor forcombination combination drivers. Tulsa, OK: University of Tulsa. M. M. M. Page Page (Ed.), (Ed.), 1982 1982Nebraska Nebraska Hogan, R R. (1983). (1983). A A socioanalytic socioanalytic theory theory of of personality. personality. In In M. Hogan, Symposium Symposium on on Motivation Motivation(pp. (pp.55-89). 55-89).Lincoln: Lincoln:University UniversityofofNebraska NebraskaPress. Press. Hogan, R R. (1991). (1991). Personality Personality and and personality personality measurement. measurement. In In M. M. D. D. Dunnette Dunnette & & L. M. Hogan, Hough Hough (Eds.), (Eds.), Handbook of of industrial industrialand andorganizational organizationalpsychology psychology( Vol. (Vol.2,2,pp. pp. 327-398). 327-398). Palo Palo Alto, Alto, CA: CA: Consulting Consulting Psychologists Psychologists Press. Hogan, & Hogan, Hogan, J. J. (1992). (1992). Hogan Hogan Personality PersonalityInventory Inventorymanual. manual.Tulsa, Tulsa,OK: OK:Hogan Hogan Hogan, R., R, & Assessment Assessment Systems. Psychol Hogan, R, Hogan, R., & &Nicholson, Nicholson,R R.(1988). (1988).The Themeaning meaningofofpersonality personalitytest testscores. scores.American American Psychologist, ogist, 43, 43,621-626. 621-626. R, Jacobson, Hogan, validation Hogan, R., Jacobson, G., Hogan, Hogan, J., & Thompson, Thompson, B. (1987). (1987). Development Development and and validation Tulsa, OK: Hogan Assessment of service operations inventory. of a service operations dispatcher dispatcherselection selection inventory. Tulsa, OK: Hogan Assessment Systems. Systems. McCloy, RR. A.A. (1990). Criterion Hough, L. L. M., M.,Eaton, Eaton,N. N.K., K.,Dunnette, Dunnette,M. M.D., D.,Kamp, Kamp,J.J.D., D.,&& McCloy, (1990). Criterionrelated related validities of of personality personality constructs constructs and and the the effect effect of of response response distortion distortion on on those validities. validities. Journal Journalof ofApplied AppliedPsychology, Psychology,75,75,581-589. 581-589. Hunter, (1990a). Methods error andand biasbias Hunter, J. J. E., & & Schmidt, Schmidt, F. F. L. L. (1990a). Methodsof ofmeta-analysis: meta-analysis:Correcting Correcting error researchfindings. findings.Newbury NewburyPark, Park,CA: CA: Sage. in research Sage. Hunter, J. J. E., & Schmidt, Schmidt, F. F. L. L. (1990b). (1990b). Dichotomization Dichotomization of of continuous continuous variables: variables: The Hunter, E., & The implications for for meta-analysis. meta-analysis. Journal ofApplied Applied Psychology, Psychology,75,75,334-339. 334-339. implications Journal of John, O. O. P. P. (1990). (1990). The The Big Big Five Five factor factor taxonomy: taxonomy: Dimensions Dimensions of of personality personality in in the the natural natural John, language and and in in questionnaires. questionnaires.In InL. L.A. A.Pervin Pervin(Ed.), (Ed.),Handbook Handbookof ofpersonality: personality: Theory language Theory and research research(pp. (pp.66-100). 66-100).New NewYork: York:Guilford GuilfordPress. Press. Johnson, J. J.A. A.(1990, (1990,June). June).Unlikely Unlikelyvirtues virtues provide multivariate substantive information Johnson, provide multivariate substantive information aboutabout personality.Paper Paperpresented presentedatatthe the2nd 2ndannual annualmeeting meetingof ofthe theAmerican AmericanPsychological Psychological personality. TX. Society, Dallas, TX. Johnson, J. J. A., A., & &Ostendorf, Ostendorf,F.F.(1993). (1993).Clarification Clarificationof ofthe thefive fivefactor factormodel modelwith withthe theabridged abridged Johnson, Big-Five circumplex. Journal Journalof of Personality Personalityand andSocial SocialPsychology, Psychology, 65,65, Big-Five dimensional dimensional circumplex. 563-576. 563-576.

CHAPTER 32 32 CONSCIENTIOUSNESS CoNsciEimousNESS

869

Landy, F. J. (1986). Stamp collecting versus science: Validation as hypothesis testing. Ameri American Psychologist, Psychologist,41, 41,1183-1192. 1183-1192. Lorr, Lorr, M., & Youniss, Youniss, R R. P. P. (1973). (1973).An Aninventory inventory of ofinterpersonal interpersonal style. style.Journal Journalof ofPersonality Personality Assessment, Assessment,37, 37,165-173. 165-173. McCrae, R R. R, R., & Costa, P. T., Jr. (1987). Validation of the five-factor five-factor model of personality across instruments and observers. Journal Journal of of Personality Personalityand andSocial SocialPsychology, Psychology,52,52, 81-90. McCrae, R R. R, R., & Costa, P. T., Jr. (1989). Reinterpreting the Myers-Briggs Type Indicator Personality, from the perspective of the five-factor model of personality. Journal of of Personality, 57, 57, 17-40. R, & Costa, P. T., Jr. (1991, August). Does Lorr's lSI ISI measure measure the thefive-factor five-factor McCrae, R R. R., model? Paper Paper presented presented at at the the 99th 99th annual annual Convention Convention of of the the American American Psychological Psychological Association, San Francisco. McCrae, R R. R., R, & Costa, P. T., Jr. (1992). Discriminate vahdity validity of NEO-PIR facet scales. Educationaland andPsychological Psychological Measurement, 229-237. Educational Measurement, 52, 52, 229-237. R. R, R., Costa, P. T., Jr., & Piedmont, R R. L. (1993). Folk concepts, natural language, McCrae, R California Psychological Inventory and the fiveand psychological constructs: The California five Journalof ofPersonality, Personality,6161, 1-26. factor model. Journal , 1-26. Personalityand andassessment. assessment.New NewYork: York:Wiley. Wiley. Mischel, W. (1968). Personality workplace. Pacific Pacific Grove, Grove,CA: CA:Brooks/Cole. Brooks/Cole. Murphy, K. (1993). Honesty in the workplace. Murphy, K., Martin, c., C, & & Garcia, Garcia, M. M. (1982). (1982). Do Do behavioral behavioral observations observations scales scales measure measure Murphy, K., Martin, Journalof ofApplied AppliedPsychology, Psychology,67, 67,562-567. 562-567. observation? Journal The relationship relationshipbetween betweensocially socially desirable responding personnel Nolan, K. A. (1991). The desirable responding andand personnel integritytesting: testing:The The unique effect ofself-deception and impression management. integrity unique effect ofself-deception and impression management. UnpubUnpublished master's thesis, Georgia State University, Atlanta. Identification Nolan, Y., Johnson, J. A., & Pincus, A. (1994). Personality and drunk driving: Identification PsychologicalAssessment, Assessment,6,6,33-40. 33-40. of DUI types using the HPI. Psychological Noller, P., Law, H., & Comrey, A. L. (1987). Cattell, Comrey, and Eysenck personality NoUer, P., Personality factors compared: More evidence for the five robust factors? factors? Journal of of Personality and Social Social Psychology, Psychology,53, 53,77-782. 77-782. Norman, W. T. (1963). Toward an adequate taxonomy of personality attributes: Replicated factor Journalof ofAbnormal Abnormal and Social Social factor structure in peer nomination personality ratings. Journal Psychology, Psychology,66, 66,574-583. 574-583. Normative operating characteristics Norman, W. T. (1967). 2,800 2,800personality personalitytrait traitdescriptors: descriptors: Normative operating characteristics for a university universitypopulation. population.Ann Ann Arbor: Arbor: University University of of Michigan, Michigan, Department Department of of Psy Psyfor chology. R. (1966). (1966). Raters, ratees, and randomness in personality Norman, W. T., & Goldberg, L. R Psychology, 4, 44-49. structure. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 4, 44-49. O'Bannon, R M., Goldinger, L. A., & Appleby, G. S. (1989). Honesty and integrity R. M., integritytesting. testing. Atlanta, GA: Applied AppHed Information Information Resources. Ones, D. S. (1993). The construct constructvalidity validityofofintegrity integrity tests. Unpublished doctoral dissertation, tests. Unpublished doctoral dissertation, University of Iowa, Iowa City. Ones, D. S., Mount, M. K. Barrick, M. R., R, & Hunter, J. E. (1994). Personality and job performance: A critique of Tett, Jackson, & Rothstein (1991) meta-analysis. Personnel Personnel Psychology,47, 47,147-156. 147-156. Psychology, Convergent and divergent divergent validity validity Ones, D. S., Schmidt, F. L., & Viswesvaran, C. (1994). Convergent evidence ffor theBig-Five Big-Fivedimensions dimensionsofof personality: A meta-analysis. Unpublished evidence or the personality: A meta-analysis. Unpublished manuscript.

870 870

HOGAN HoGAN AND AND ONES ONES

& Schmidt, F. L. (1993). Comprehensive meta-analysis of c., & Ones, D. S., Viswesvaran, C, integrity test validities: Findings and implications for personnel selection and theories of job performance. Journal Journalof ofApplied AppliedPsychology, Psychology,78, 78,679-703. 679-703. Paajanen, G. E. (1985). PDI PDl Employment Employment Inventory. Inventory. Minneapolis, MN: MN: Personnel Deci Decisions, Inc. & Costa, P. T., Jr. Jr. (1992). (1992). An assessment of the Edwards Piedmont, R. L., McCrae, R. R., & Personal Preference Schedule from the perspective of thefive-factor model. Journal Journal five-factor model. of Personality PersonalityAssessment, Assessment,58,58,67-78. 67-78. of & Hogan, J. (1986). Development Development and and validation validationofof Raza, S., Metz, D., Dyer, P., Coan, T., & personnel selection personnel. Tulsa, OK:OK: University selectionprocedures proceduresfor forhospital hospitalservice service personnel. Tulsa, University of Tulsa. Rosen, A. S. S. (1977). (1977).On Onthe thedimensionality dimensionalityofofthe theCalifornia CaliforniaPsychological PsychologicalInventory Inventorysocializa socializaJournalof of Consulting Consultingand andClinical Clinical Psychology, 583-591. tion scale. Journal Psychology, 45, 45, 583-591. & Callahan, C. (1989). Integrity testing for personnel selection: Sackett, P. R., Burris, L. R., & An update. Personnel PersonnelPsychology, Psychology,42,42,491-529. 491-529. Schmidt, F. L., & & Hunter, Hunter,J.J.E. E.(1992). (1992).Development Developmentofofaacausal causalmodel modelofofprocesses processesdetermin determinCurrentDirections DirectionsininPsychological Psychological Science, 1, 89-92. ing job performance. Current Science, 1, 89-92. & Hunter, J. E. (1992). Annual Review Reviewof of Schmidt, F. L., Ones, D. S., & (1992). Personnel selection. Annual 43, 627-670. Psychology, Psychology, 43, Schmitt, N., Gooding, R. Z., Noe, R. A., & & Kirsch, M. M. (1984). Meta-analyses of validity studies published between 1964 and and 1982 1982and andthe theinvestigation investigationof ofstudy studycharacteristics. characteristics. Personnel Personnel Psychology, Psychology,37, 37,407-422. 407-422. Schuessler, K. E., & & Cressey, Cressey, D. D. R. R. (1950). (1950).Personality Personality characteristics characteristics of of criminals. criminals.American American Journal Journalof of Sociology, Sociology,55, 55,476-484. 476-484. Science Research Associates. (1983). Personal PersonalOutlook OutlookInventory. Inventory.Park ParkRidge, Ridge,IL: IL:Author. Author. Shannon, C, C., & Urbana: & Weaver, W. (1949). The The mathematical mathematicaltheory theoryofofcommunication. communication. Urbana: University of Illinois Press. Skinner, H. A., Jackson, D. N., & & Rampton, Rampton,G. G. M. M.(1976). (1976).The ThePersonality PersonalityResearch ResearchForm Formin in ofof Behavioral a Canadian context: context: Does Doeslanguage languagemake makeaadifference? difference?Canadian CanadianJoumal Journal Behavioral Science, Science,8,8,156-168. 156-168. Sutherland, E. H. (1951). Critique of Sheldon's 'Variety of Delinquent DeUnquent Youth.' American American Sociological Review, Sociological Review,16, 16,10-23. 10-23. erential Personality Questionnaire. Unpublished Tellegen, A. (1982). Briefmanual theDiff Differential Personality Questionnaire. Unpublished manualfor forthe manuscript, manuscript. University of Minnesota, Minneapolis. Tett, R. P., Jackson, D. N., & & Rothstein, M. (1991). Personality measures as predictors of job performance: A meta-analytic review. Personnel PersonnelPsychology, Psychology,44, 44,703-742. 703-742. Recurrentpersonality personalityfactors factorsbased basedonontrait traitratings ratings & Christal, R. C. (1961). Recurrent Tupes, E. C C.,, & TX: U.S. Air Force. (Tech. Rep. No. ASD-TR-61-97). Lackland Air Force Base, TX: Waller, N. G., & & Ben-Porath, Y. S. (1987). Is it time for clinical psychologists to embrace personaUty? American AmericanPsychologist, Psychologist,42,42, 887-889. 887-889. thefive-factor five-factor model of personality? Watson, D. (1989). Strangers' ratings of the five five robust personality factors: Evidence of Psychology, surprising convergence with self report. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 57, 57, 120-128. ). Manual Adjective Scales. Odessa, FL: Psychological J. S. S. (1991 (1991). Manualfor forthe theInterpersonal Interpersonal Adjective Scales. Odessa, FL: Psychological Wiggins, J. Assessment Resources.

PART VIII PARTVffl

APPLIED ApPLIED PSYCHOLOGY

This Page Intentionally Left Blank

CHAPTER 33 33

PERSONALITY, INTERACTIVE RELATIONS, AND AND APPLIED PSYCHOLOGY ApPLIED F . CHAPLIN CHAPLIN WILLIAM F. UNIVERSITY UNIVERSITY OF OF ALABAMA ALABAMA

informaThe purpose of personality assessment is "the obtaining and evaluating of informa reation regarding individual differences" (Cronbach, 1956, p. 173). Although the rea & Pearson, Pearson, sons for obtaining and evaluating such information are diverse (Fiske & the prediction of useful criteria is is aa major one. In to the the most 1970), the In the preface to decades. Person Personinfluential graduate text on personality assessment for the past two decades, and Prediction: Prediction:Principles PrinciplesofofPersonality Personality Assessment, Wiggins (1973) writes, ality and Assessment, Wiggins (1973) writes, on the prediction of socially socially relevant criteria criteria has prompted overriding emphasis on "An overriding to adopt Donald W. Fiske's suggestion that the the principle title .. . .. is is the the most me to of the subject matter of this text" (p. iii). Certainly, appropriate characterization appropriate characterization of the subject matter of this text" (p. iii). Certainly, for much much of of the the work work in in the the field field of of personality personality assessment assessment has has been been the impetus impetus for the for large large predictive predictive validity vaUdity coefficients coefficients (e.g., (e.g., Ghiselli, Ghiselli, 1956; Hase & the quest quest for & the 1956; Hase Goldberg, 1967; Jackson, 1971; Sines, 1964). Goldberg, 1967; Jackson, 1971; Sines, 1964). L PREDICTION PREDICTION MODELS MODELS I.

Contrary to popular belief, scientific predictions are nearly always inaccurate. "The government people didn't have aa good sense of how science works. They couldn't to grips with the the fact that there are are no no absolute truths .. .. .. that our our data come to H A.NDBOOK OF HANDBOOK OF PERSONA.LITY PERSONAinrYPSYCHOLOGY PSYCHOLOGY

873 873

COPYRIGHT 1997 BY ACADEMIC PlESS PRESS.. COPYII GHT 0 C 1997 ALL ALLliGHTS RIGHTSOFOFREPIODUCI1ON REPRODUCHONININANY ANYFORM FORMRESERVED. RESERVED.

874 874

WiLUAM F.CHAPLIN WILUAM F. CHAPLIN

reflect reflect closer and closer approximations of what might tum turn out to be true" (Richard Axel, quoted in Booth, 1989). Inaccuracy is a consequence of our necessary reliance on models of the phenomena we wish to predict-models predict—models that are generally fictitious fictitious perfection in the sense that they are simplifications. "Astronomers did not achieve perfection and never would, not in a solar system tugged by the gravities of nine planets, scores of moons and thousands of asteroids, but calculations of planetary motion forgot they were forecasts" (Gleick, 1987, p. p. 14). 14).Thus, Thus, were so accurate that people forgot model—an especially unless the phenomenon we wish to predict is as simple as the model-an unlikely circumstance when the phenomenon is human behavior—the behavior-the predictions will be "approximate" (i.e., inaccurate).

A. The The Bivariate Bivariate Linear Linear Model Model A. from another is the well-known The most basic model for predicting one variable from in Equation 1. 1. In this equation, y is the criterion simple linear relation shown in variable, x is the predictor variable, and b and a are constants reflecting the slope and intercept of the regression line, respectively. Typically, b and a are empirically derived to minimize the squared errors of prediction ("least-squares criterion"). The Pearson product-moment product-moment correlation coefficient coefficient provides an index of how well, again in a least-squares sense, the predictions match the observations. (1)

y ^bxbx ^+ a. =

simpUfication of of nearly all psychological psychological phenomena. phenomena. This model is an extreme simplification m the absence of measurement error, few criteria can be perfectly predicted Even in from a linear model involving one predictor. One apparent tactic for improving from predictions is to increase the complexity of the prediction models so that they more closely approximate the complexity of the criterion. B. B.

Multivariate Extension Extension Multivariate

tly used to increase the complexity of the prediction frequently The tactic that is most frequen model is the introduction of more predictor variables into the equation. The ratio ratiomultifaceted and multiply muhiply nale for this approach is that psychological criteria are multifaceted determined. Thus, combining variables that capture different different aspects of the criterion should more completely model the criterion and, presumably, more accurately predict it. The general form of this multivariate model is y = b\Xi 4- 62-^2 + • • • +fcjjcy+ • • • + bkXk +

fl,

(2) (2)

Xyisisthe theith fthofofkkpredictor predictorvariables, variables,bifttisisthe theith ithofofkkweights, weights, where y is the criterion, XI and a is the intercept. The output of Equation 2 is a set of predicted values of y. The correlation between these predicted values and the actual values is called a "multiple correla correlation"; tion "; it indexes how well the model predicts (in a least-squares sense) the criterion.

CHAPTER CHAPTER 33 33

875 875

iNrERACDVE INTERACTIVE RELATIONS

This model has has been the the mainstay of statistical prediction for most of the the constants history of personality assessment (Wiggins, 1973). Moreover, when the (i.e.,fe/S and aa) in the the model are are optimally (least-squares criteria) criteria) derived, this (Le., bis and ) in powerful features (Goldberg, 1991). In particular, it adjusts adjusts model has a number of powerful predictfor (a) differences differences in the units of measurement, (b) redundancy among the predict for ors, (c (c)) differential differential validities among the the predictors, and and (d (d)) all all regression effects. effects. ors, However, even in the unlikely event that this model included a comprehensive set of predictors, it would still be quite simplified. Specifically, this model is limited to Psychological phenomena phenomena a linear (additive) combination of the predictor variables. Psychological nonlinear function function of a set set of variables. are almost certainly a more complicated, nonlinear One type of nonlinear relation that has been repeatedly proposed for personality is an interaction or moderator effect. effect. predictions is c. C. Interactive Models

An interaction effect effect is usually discussed in the context of the analysis of variance. It is the main effects is defined as as occurring when the effects of one independent variable are different at different different levels of a second independent variable. One can then refer refer different to the second variable as a "moderator variable" because it moderates the relation the first independent variable and and the the dependent variable. However, the between the the as a moderator variable is statisti statistidesignation of one of the independent variables as not conceptually, arbitrary because interaction interaction effects effects are cally, though perhaps not is equally legitimate (statistically) to refer refer to to the the first first independent symmetrical. It is as a moderator moderator of the relation between the the second independent variable variable as the dependent variable. and the demonstrated some time ago, interaction effects effects can can also As Cohen (1978) demonstrated The general form form of the the two two be evaluated using a multiple regression paradigm. The predictor interactive model is y = biXi -f &2^2 + b3XiX2 + a,

(3)

the terms are are as as defined defined in Equation 2. where the the cross product of two independent variables is is related to Specifically, if the after the the effects effects of the the two two independent independent variables have the dependent variable, after the cross product, then an an interaction effect effect is from the been removed (partialed) from independent variables are qualitative, the value of F for present. Indeed, when the independent the interaction in the analysis of variance and the value of F for the partialed cross product will be identical. However, when one or both of the independent variables are quantitative, the multiple regression procedure will have greater statistical power not require that the the independent variables be rendered, usually because it does not the cross product is is computed (Cohen, 1983). arbitrarily, into categories before the has been been most most widely widely The multiple regression approach to modeling interactions has used within the context of prediction models. In this context it is typical to designate on an an aptitutde aptitutde test) test) and and one independent independent variable as the the "predictor" "predictor" (e.g., (e.g., scores scores on one variable as

876 876

WILUAM ClwUN WILLIAM F. F.CHAPUN

the other other as the "moderator" (e.g., an index of of test anxiety). However, just as is the case in the analysis of statistically arbitrary. of variance, this designation designation is statistically One important characteristic of moderator effects of interaction or moderator effects is that they are orthogonal or predictor orthogonal to the main effects effects of of the independent independent or predictor variables. Thus, of variance variance the interaction effect effect is defined residual effect effect after after in the analysis of defined as a residual effects of of the independent independent variables have been removed. Indeed, the sum the main effects of squares used to assess the interaction effect effect can be obtained of obtained by subtracting the of the two main effects effects from from the cell sum of sum of squares of of squares. Likewise, in not simple, correlation correlation between the multiple regression paradigm it is the partial, not dependent variable that indexes the moderator moderator effect effect the cross product and the dependent independence of of moderator moderator effects (Cohen, 1978). Failure to appreciate the independence effects has led Mansfield to some erroneous proposals by Bobko (1986) and Morris, Sherman, and Mansfield for increasing the power of of tests to detect interaction interaction effects effects (Cronbach, (Cronbach, (1986) for 1987). Summary D. Summ ary There is little doubt that multivariate and interactive equations are more faithful faithful of human behavior behavior than either the bivariate or models of or multivariate multivariate linear models. of further further complexities, such as power Moreover, the addition of power polynomials (e.g., transformations (e.g., logarithmic or arcsine), quadratic or cubic trends), nonlinear transformations higher order order interactions of of all these terms, might might well further further increase the and the higher correspondence between the equations and the phenomena phenomena they correspondence they model. However, before beginning the unending task of of considering more and more equabefore more complex equa tions, let us consider how the increase in complexity entailed by interactive models power to predict human behavior. has contributed to their power n. DESCRImON DESCMPWON AND AND EVALUATION OF SOME INTERACTIVE MODELS D.

In this section I will begin by briefly different contexts-personality briefly describing describing three different contexts—personality education—in which it was widely believed beUeved consistency, personnel selection, and education-in that prediction would substantially prediction models containing containing interaction interaction terms would substantially increase simpler models. However, I will note predictive accuracy over that achieved with simpler received dismal empirical support. that in all three contexts the interactive models received for their poor poor showing-reasons showing—reasons that that have Finally, I will consider the reasons for for the use of of interactive models in applied settings. important implications for A. Personality Consistency During the 1960s and 1970s the field of of personality personality assessment experienced experienced a series of attacks on one of of its most fundamental fundamental propositions, the consistency of personality consistency of of perpetrators of of this assault (cf. (cf. Mischel, 1968) marshalled marshalled evi evicharacteristics. The perpetrators dence that human behaviors, even highly similar behaviors, varied varied substantially

CHAPTER CHAPTER 33 33

INrERACI1VE INTERACTIVE RELATIONS

877 877

across different different situations. Such variability, they argued, was incompatible with the existence of the stable dispositions (traits) that were the primary object object of personal personality assessors' attention. In hindsight, this argument argument seems inconsequential inconsequential because the variability of specific specific behaviors implies little about the consistency of the con constructs indicated by those behaviors (Block, 1977; 1977; Golding, 1978; 1978; G. A. Kelly, 1955; Ozer, 1986). However, for some (e.g., Bem & Allen, 1974; 1974; Cheek, 1982; Kenrick Kenrick & & Stringfield, 1980; 1978), the only way to resurrect the concept of traits and 1980; Turner, 1978), reestablish the legitimacy of personality measurement appeared to be through the invocation of moderator variables. In one of the most influential influential papers in that decade, Bem and Allen (1974) differences in personality consistency; for for any proposed that there are individual differences one trait, some individuals will be consistent whereas others will not be. That is, these individual differences differences in consistency moderate the relation between two is, measures of the same trait. To support this proposition, Bem and Allen studied five personality personality traits traits and and reported reported the the results results from from two two of of them, them, friendliness friendliness and and five conscientiousness. For each trait, they divided a group of subjects into a high and low consistent group on the basis of a median split on two measures of crosscross situational consistency. The high consistent group generally exhibited higher corre correlations among different different measures of the target trait than the low consistent group, although this was true for for each of the measures of cross-situational consistency for only one of the traits. Bem and Allen concluded that personality traits were suffi sufficiently stable to warrant their continued study and measurement, but only within a subgroup subgroup of the popUlation. population.

B. B.

Differential Predictability Predictability Differential

In the 1950s, the business of personnel prediction and selection was in desperate The results results of of several several large-scale large-scale prediction prediction need of "new ideas" (McNemar, 1952). The studies (e.g., E. L. Kelly & Fiske, 1951; 1951; Office Office of Strategic Services Assessment Staff, 1948) were in, and the optimism with which these studies had been begun had been tempered tempered by the general finding that the state-of-the-art, state-of-the-art, cross-validated, optimal, multiple predictor assessment model was not doing much better than chance at forecasting forecasting performance. We now know some of the reasons for for the model's poor showing (Wiggins, 1973). These These include include (a) an an inadequate inadequate specification specification of the criterion (the "criterion problem"), (b) the reliance on a single index, rather than a more generalizable aggregate index, of the predictor variables, and (c) the reduction in the power of a multiple regression equation when it is applied to new samples (the problem of "capitalization on chance"). However, one of the first predictability(Ghiselli, (GhiselU, responses was the introduction of the concept of differential differentialpredictability 1956; 1956; 1960a, 1960b, 1963). The underlying premise of differential differential predictability, which explicitly violates one of the tenets of classical test theory, is is that some individuals are more predictable differences between predictable and unpredict unpredictthan others. If there are systematic differences differences should in principle be identifiable with a able individuals, then those differences

878 878

WILUAM F. CHAPIJN WILUAMF.CHAPUN

third (moderator) variable. Predictive efforts efforts can then be restricted restricted to those people for for whom those efforts efforts will be rewarded. To develop an index for for differentiating differentiating predictable individuals from from unpredictable ones, Ghiselli (1956, 1960b) suggested computing computing correlations between items (usually from from the predictor scales) and the absolute difference difference between the actual and predicted criterion scores of of a group of of individuals. Items that correlated correlated with this discrepancy could then be combined into a scale that would index unpredictability. In a series of of studies Ghiselli GhiselU demon demonstrated that this approach significantly significantly increased his ability abiUty to predict the perfor performance of of taxicab drivers (1956), sociability of college students (1960b), and the occupational level of of adult men (1960a). Thus predictability appeared appeared to moderate moderate occupational predictor-criterion relations. predictor-criterion Tecliniques C. Student Characteristics and Instructional Techniques of education is that there are some Perhaps the most pervasive belief belief in the field of for teaching students that are more effective effective than others. This belief belief is techniques for illustrated by the large amount amount of of time and effort effort that has been devoted devoted to develop developillustrated instructional methods (e.g., (e.g.. Gage, 1963; ing and evaluating instructional 1963; Haines & McKeachie, 1983; Skinner, 1968). It is also illustrated 1967; Husband, 1954; Levine & Wang, 1983; of us go through when it comes time to prepare prepare by the soul-searching that many of for the coming year. However, there is actually little evidence to suggest our courses for effort, and soul-searching have been well spent (Dubin & Taveggia, that this time, effort, after reviewing the empirical literature Uterature on this 1968). As Goldberg (1972) noted after issue, "the overwhelming overwhelming finding finding that has emerged from from hundreds of of studies . . . differing college instructional procedures do not not appear appear to produce any is that differing differences in average course achievement" (p. 154). consistent differences explanation for for this surprising finding finding is that that learning is more One popular explanation complicated than was first thought. Specifically, it was suggested that there was an interaction interaction between between student student characteristics characteristics and instructional techniques such that of student work poorly with other types, techniques that work well with one type of Tiffin, & Seibert, and vice versa (Cronbach, 1957, 1967; McKeachie, 1968; Snow, Tiffin, If this explanation explanation is true, not only are past and future future efforts efforts to develop 1965). If better teaching methods justified, but these efforts efforts should be redoubled. better

D. Empirical Evaluation For each of of these moderator of personality For moderator variable solutions to the problems of performance prediction, and educational effectiveness, effectiveness, the initial em emconsistency, performance pirical support for for the interactive models seemed promising. However, following following of optimism in which many investigators began to incorporate incorporate the an initial period of disquieting results began interactive models into their research and practice, some disquieting Specifically, researchers who tried to replicate the initial findings findings to be reported. Specifically, failed to do so. Following an extensive study of of the moderating moderating effects effects generally failed

CHAPTER CHAPTER 33 33

INTERACDVE INTERACTIVE RELATIONS RELATIONS

879 879

of cross-situational consistency, Chaplin and Goldberg (1984) concluded,

pubUcation of the Bern and Allen Alien study now The excitement that followed the publication needs to be tempered. Not only did we fail to find find any generalization of their traits, butweweeven evenfailed failed findingstotoa larger a largerand andmore morerepresentative representativesample sampleofoftraits, findings but to replicate theirfindings findings for one of the two traits they investigated. investigated, (p. (p. 1089) & Peake, 1982; This conclusion was echoed by other investigators (e.g., Mischel & Paunonen & 1985). Even a more recent confirmatory & Jackson, 1985). confirmatory report (Zuckerman moderator effects. Rather et al, aI., 1988) was not based on the discovery of substantial moderator effects it was based on demonstrating the statistical significance of small moderator effects with very large samples (Chaplin, 1991). In the case of of differential differential predictability, the ephemeral quality of of the putative 1966,1967; moderators was so widely documented (e.g., Brown & Scott, 1966, 1967; Goldberg, 1969; Kellogg, 1968; Strieker, 1967; 1972; Wallach & & Leggett, 1972) that 1968; Stricker, 1967; Velicer, 1972; 1969; himself concluded, "Furthermore, since the indications are that moderators Ghiselli himself are rather specific it might be that they, like suppressors, do not hold up well treatment" from sample to sample" (1963, p. 86). from 86). Finally, the search for "trait by treatment" 1964; Lub Lubinteractions in education led to similar disappointments (e.g., Goldberg, 1964; lin, 1965; 1965; McKeachie, 1963; 1963; Tallmadge, 1968). These poignant poignantfindings, whencoupled coupledwith withthose thoseconcerning concerningthe thefate fateofofattempts attempts These findings, when new empirical interaction scales, suggest suggest that that the the significant significant interac interacto construct new tions discovered . . . are discovered in in this thisproject project... are unlikely unUkelytotolead leadtotodifferential differentialpredictions predictions which are more valid than those achievable by general predictors alone. (Goldberg,1972, 1972,p. p. 200) (Goldberg, 200) Recently, a similar conclusion was reached by Dance and Neufeld (1988) about aptitude-by-treatment interaction research in clinical psychology. aptitude-by-treatment What Went Went Wrong? Wrong? E. What effects These examples all illustrate situations in which the existence of interaction effects was a foregone conclusion. Why, then, did the addition of interactive terms to prediction models fail to increase predictive validity? In answering this question I consider some logical, statistical, psychometric, and conceptual issues that must will consider confronted by those who would venture into the moderator domain. be confronted 1. Logical Issues 1. The complexity of human behavior that makes the existence of interaction effects effects makes itit likely likely that the size size of of any anygiven giveneffect effect will willbe besmall. small.In Inwriting writing a certainty also makes about the discouraging failure of many studies to detect aptitude-by-treatment aptitude-by-treatment interactions, Cronbach (1975) noted, "Once we attend to interactions, we enter a hall of mirrors that extends to infinity. However far far we carry our analysis-to analysis—to third order or fifth order, or any other order—untested interactions of a still higher order order-untested 119). Likewise, in discussing the cross-situational consistency can be envisioned" (p. 119).

880

WILUAM F. CitAPUN WILUAMF.CHAPUN

of behavior, Epstein (1983) wrote, Theoretically, any single instance of behavior can be predicted if all the right moderator variables are included. This is no more than to say that behavior is determined, and that if we knew everything that determined it, we could predict it. However, to do so might require the addition of of so many moderator moderator variables variables that they would generate interactions of such complexity as to make the proceproce dure unfeasible and the results uninterpretable. 377) 0.. ^77) (p

These statements lead to three related conclusions. First, the number of possipossi ble interaction interaction effects for for any dependent variable is very large, perhaps infinite. effect Thus, within the context of any uncontrolled "naturalistic" system, any one effect effects are roughly equal, the proportional effect effect is almost certain to be small. If the effects of any one interaction is 11k \lk (where k /: is a very large number). number). Second, any interaction interaction effect is itself itself going to be moderated by a higher order interaction effect. Third, as effect unspecified Nisbett (1977) noted, this renders the hypothesis that there exists some unspecified effect untestable (and also uninteresting), because any failure failure to detect moderator effect effect can be attributed to an as yet undiscovered higher order interaction. an effect 2. Statistical Statistical Issues Ismes

Even in the more limited domain of a specific study with a limited set of variables, effects will be statistically more difficult difficult to detect than main effects effects interaction effects 1977; Keppel, Keppel, 1982). 1982).First, First, statistical statistical power power is, is, in in part, part, aa function function of of sample sample (Cohen, 1977; size. In a standard analysis of variance the number of observations on which each cell mean is based will be less than the number of of observations on which each combination of of cell marginal mean is based, because the marginal means are a combination effects are based on cell means, whereas main effects effects are based means. Interaction effects on marginal means. function of effect effect size. The larger larger the effect, effect, the more Second, power is also a function likely it is to be detected. The general linear model always expresses as much of of the systematic variance in any experiment in terms of of main effects. Interaction effects effects are, by definition, based on the residual systematic variance. Thus, even ifif an interaction effect effect is present, especially if it is an ordinal interaction, it may produce observations that are captured by the additive, linear terms of the model. effect As always, a failure to reject the (null) hypothesis that there is no interaction effect does not mean that an interactive process is not operating. Third, the overall power of any analysis is generally decreased as one adds for testing interactions terms to the model being tested. In their recommendations for framework of multiple regression analysis, Cohen and Cohen (1983) within the framework suggested that investigators should be very selective about which variables and their possible interactions they include in their analysis. For example, with 10 10 research factors (a not unusually large number outside 45 two-way, 120 three-way interaction the experimental laboratory), there are 45 how many many individual individual IVs IVs [independent variables] variables].... sets, and heaven knows how . . . . By omitting possible possible interaction interactionsets setscarrying carryingchance chancevariance variancefrom fromthe theequation, equation,

881 881

CHAPTER CHAPTER33 33 INrERAC11VE INTERACTIVERELAnONS RELATIONS

as many df ([degrees degrees of freedom] as there are omitted IVs become available to the error term, thereby increasing . .. . . the statistical power of the tests that are performed. (Cohen Cohen, 1983, p. 348) Cohen, 1983, p, 348) (Cohen &<Sc 3. Psychometric Psychometric Issues Issues

coefficients. Specifically, Specifically, Measurement error has a deleterious impact on regression coefficients. measurement error generally attenuates correlations between variables, and this in the regression coefficients coefficients associated associated with unreliably introduces a negative bias in measured variables (Kenny, 1979). Because it is the product of of variables, it might moderator effect effect would be a seem that the reliability of the term that carries the moderator Unfortunately, the reliability reliability of of direct function of the reliability of its components. Unfortunately, of its component the cross product is, among other things, dependent on the scaling of variables. Thus, a cross product's reliability may not be easily discerned from the reliability of its components. between the reliability of Bohrnstedt and Marwell (1977) derived the relation between of the components. This relation is the product of two variables and the reliabilities of 2{MJs,) ({MllSDryy Mylsy ) rxy ++r�y + rxxryy Mxlsx) ({MylSy)r,y M�/S�) rxx + 2( M�/Snryy + ({MySl)r,, ( ) rxyxy ( ) '''^'^ "" ((MllSl) {MJISj) 2{MJs,) {Mylsy)r,y rly++ 1.0 1.0 ' ' ^^^ 2( 4 + ) ( + ) / Mxlsx Mylsy ) rxy ++r�y M� S� M�/S� =

r^yxyisisthe thereliability reliabilityofofthe thexyxyproduct; product;Mx Mxand andMy Myare arethe themeans meansofofx xand and where rxyxy y, y, respectively; Sx and Sy 5^, are the respective standard deviations; rxx rxxand andryy Vyy are are thethe respective reliabilities; and rxy Vxyisisthe thecorrelation correlationbetween betweenxxand andy.y. As can be seen in Equation 4, the reliability of a product depends not only of on the reliability of the component variables, but also on the correlation between respective means to standard devia deviathe component variables and the ratios of their respective transformations. The ratio of means to standard deviations will be changed by any transforma tion that affects the origins of the variables (i.e., that contains an additive constant). The scale dependence of cross-product reliabilities is certainly germane to transformations on the evaluation of recent discussions about the impact of scale transformations moderator effects (e.g., Cronbach, 1987; Dunlap & Kemery, 1987; Morris et al., aI., 1986). Moreover, as Bohrnstedt and Marwell (1977) have shown, it leads to certain anomalies in the estimation of cross-product reliabilities. For example, even if one of of the component variables is completely completely unreliable, the reliability of of the cross product will generally not be zero, and, under certain conditions, it will approach the reliability of the more reliable component variable. Also, when the correlation between the component variables is negative, the reliability of of the cross product will be lower than if the correlation is positive, even if the component reliabilities are held constant. More recently, Busemeyer and Jones (1983) have shown that these anomalies are eliminated if the component variables are converted to deviation scores. In this case. Equation 4 simplifies to case, r

— ^-yy "^ ^xxl'yy

"^'''^

rly + 1.0 '

where the terms are as defined in Equation 4.

.-v

(5)

^^^

882 882

WILUAM F. CHAPLIN WILUAMF.CHAPUN

As is clear from Equation 5, the reliability of a cross product cannot exceed the reliability of of its most reliable component. Typically, its reliability will be less. For example, consider two variables (in deviation score form), each with a reliability of of .80. If those variables are uncorrelated, the reliability of their cross product will be .64. Even if the correlation between the two variables rises to the maximum allowed by their reliability (.80), the reliability of of the cross product will be maximized at .78. As Busemeyer and Jones (1983) noted, "The presence of measurement error in the predictor variables will drastically reduce the power to detect a significant contribution from the product term" (p. 559). In addition to the other handicaps, detecting interaction effects must generally be done in the presence of more measurement error than is present for main effects. Moreover, the extent of of this measurement error cannot be directly estimated from reliabiUty of the component variables when the component variables are mea meathe reliability affected in sured on scales that are not centered around zero. Indeed, it may be affected anomalous ways by arbitrary scaling factors. 4. Conceptual Conceptual Issues Issues 4.

In the three examples of of personality consistency, personnel selection, and educa educational effectiveness, the primary hypothesis was that some moderator effect effect existed. of these examples, little effort effort was made to establish the construct However, in each of validity of of the proposed moderators, nor was any compelling theory for the specific moderator variable developed. Thus, Bem and Allen (1974) proposed that individu individucross-situationally als who reported being cross-situationally consistent would be cross-situationally consistent; Ghiselli (1956) proposed that people who were more predictable would be more predictable; and numerous educators proposed that techniques that were better suited to a student would improve his or her academic performance. That these predictions sound trivial reflects the fact that they all are limited to asserting the mere existence (cf. Meehl, 1978) of a moderator variable; the nature of of that variable and its operation are generally unspecified. unspecified. Standing in contrast to my pessimistic appraisal of the likelihood of of a successful successful search for moderator variables are the many studies that have reported significant significant and replicable interactions. These include the moderating influence of self-efficacy self-efficacy on behavior change (Bandura, 1977), the interaction of personality variables with stressful situations (Friedman & & Rosenman, 1974) or with arousal (Humphreys & & Revelle, 1984), the divergent perspectives of actors and observers (Jones & & Nisbett, 1972), the situational factors that influence bystander intervention (Darly & & Latane, Latan6, 1970), and the influence of of family family configurations on intelligence (Zajonc, 1976). However, unlike the studies reviewed at the beginning of this paper, these successful successful investigations all concerned specific moderator variables, and the conditions under which their effects effects would be observed had been predicted. Thus, the investigators were able to focus the full power of their design on those effects. In other words, they knew where to look. efforts to detect By comparison, consider Goldberg's (1972) appraisal of the efforts an interaction between teaching methods and student characteristics:

33 CHAPTER 33

883 883

INTERACTTVE RELATIONS INTERACDVE

While the directors of both research programs might argue that the personality they utilize utilizeare are "theory-based" "theory-based".. . . . .ititisisdoubtful doubtfulwhether whethereither eithertheory theory measures they For,at at the themoment, moment,we we have havefew few actually dictated these measurement decisions. For, theories in psychology-and psychology—and none none in in college college instruction-which instruction—which specify specify the the theories number and and nature nature of of those those personality personality characteristics characteristics predisposing predisposing students students to to number achieve differentially differentially in in different different college college courses courses . . .. .. . achieve (p. (p. 157) 157)

Tellegen, Kamp, and Watson (1982) made a similar point about the problem with moderators of cross-situational consistency, as did Wiggins (1973) with regard to differential differential predictability. "At present theories of personality structure do not ap apsufficiently articulated to guide selection of appropriate mathematical pear to be sufficiently from the myriad possibilities that exist" (Wiggins, 1973, prediction models from 1973, p. 79).

m in. IMPLICATIONS FOR ApPLIED APPLIED PSYCHOLOGY

fared well in applied settings. Moderator variables Interactive models have not fared difficult to discover and are more difficult difficult to replicate. In hindsight, these are difficult disappointing conclusions are not surprising: given the vast number of possible interaction effects, any one of them will generally be minuscule; the detection of of effects is further further hampered by measurement unreliability and low statistical such effects effects that are detected will be difficult difficult in the absence power; the replication of any effects of a strong theory that predicted the effect. What, then, are the implications of this analysis for for applied prediction? A. The The Importance Importance of of Theory Theory A. for finding moderator moderator variables is to start with a clear specification specification My prescription for of what the variables are and how they should be measured, as well as the conditions conditions under, and the manner in which, their effects effects will be manifested. The paper by Humphreys and Revelle (1984) on the influence of individual differences differences and situa situational factors on information information processing provides a sample script for for these actions. In summarizing their program of research, they write, "We ** Wenow nowcan canrelate relatepersonality personality dimensions to situations and tasks and make specific predictions about the conditions under which people who differ differ in impulsivity, achievement motivation, or anxiety differ in their performance differ performance in a variety of situations" (p. 180). Once this has been done the usual procedures for for increasing statistical power, including using large samples, reliable measures, and designs that maximize the size of the interaction effect, effect, can be employed. However, although good experimental form form may lead to the reliable detection of moderator effects, it will do so despite the despite the relatively relatively small small moderator size of those effects. Many efforts to find moderator variables have been motivated efforts moderator not by a theoretical interest, but by the desire to find variables that would substan substantially and "usefully" affect Bern affect the relation between a predictor and a criterion. Bem coefficients" (Mischel, and Allen (1974) sought to demonstrate that "personality coefficients"

884 884

WILUAM F. CHAPLIN WILUAMF.CHAPUN

1968) could greatly greatly exceed exceed the .30 barrier; Ghiselli (1956) intended intended his index of of differential differential predictability to be of of practical use; and and the same same goals are are evident evident in the aptitude-by-treatment aptitude-by-treatment interaction literature. Given Given that that interaction interaction effects effects are in the "real their practical impact. **real world" generally small, I remain pessimistic about about their B. Explanation versus Prediction evaluating a theoretical Prediction is the fundamental fundamental standard standard for for scientifically scientifically evaluating model. However, predictive accuracy accuracy does not imply a true true model; rather, predictive inaccuracy implies a false false model (Popper, 1962). inferential asymmetry asymmetry results inaccuracy 1962). This inferential because accurate accurate predictions can be based on very inaccurate inaccurate models (e.g., the model of of the sun as a fiery chariot on a circular circular track track did an excellent job of of an excellent rising of of the sun in the east). Thus, if efficient prediction, predicting the rising if the goal is efficient rather than explanation, the quest for for highly accurate models models may be unnecessary. rather Corrigan published pubUshed a review of of linear linear models In 1974 Dawes and Corrigan of the use of decision making. In this review review they concluded that that under under a broad reasonable in decision broad and reasonable of conditions conditions a linear combination combination of of unit-weighted unit-weighted variables variables outperformed outperformed the set of of optimally optimally (least-squares) weighted weighted variables variables in predicting same set of predicting a criterion. "robust beauty" of of these unit-weighted unit-weighted models was further further documented documented by The "robust Neither Dawes nor Corrigan Corrigan nor nor any other other decision scientists has Dawes in 1979. 1979. Neither contended that these linear models, unit weighted weighted or otherwise, are realistic ever contended of the phenomena phenomena they are predicting (e.g., Goldberg, 1968; Hoffman, Hoffman, portrayals of Hoffman introduced introduced the expression expression "paramorphic em1960). Indeed, Hoffman "paramorphic models" to em "The true true relationships need need not linear for linear models phasize that distinction. "The not be linear for linear approximated by linear to work; they must merely be approximated linear models" (Dawes, 1979, p. behavior can only be be understood understood and explained by 5573). 73) . Thus, although human behavior and explained often be be better better predicted appealing to very complex models, it can often predicted by relatively simple ones. C. The Virtues of of SimpUclty Simplicity c. development of of increasingly complex prediction prediction models has been guided by The development the seemingly straightforward straightforward assumption assumption that complex complex phenomena phenomena will be better predicted predicted by complex complex models. However, exactly the opposite opposite conclusion is consistent with the evidence.

In the derivation sample, it was clear that increases in the complexity of the prediction equations were associated with corresponding increases increasesininpredictive predictive accuracy. However, .. .. . . in a cross-validation sample, increases in complexity associated with corresponding decreases decreasesininpredictive predictiveaccuracy. accuracy.With With . .. .. were associated outperformed the more few exceptions, the usual multiple-regression equation outperformed complex models. (Wiggins, 1973, 1973,pp. pp,78-79) 78-79) (Wiggins,

Although Although complex relations can be modeled quite accurately, post post hoc, in a given sample, the cost of of accuracy is a sharp decrease in generality. generaUty. There are at least two reasons for for this.

88S 885

CHAnER CHAPTER 33 33 INn:RACI1VE INTERACTIVE RELATIONS RELATIONS

/. Chance ChanceDisguised Disguised as as Complexity Complexity 1. In any set of data it is likely that there will appear some intriguing patterns that are purely chance fluctuations. It is out of an appreciation for for the clever clever way in which chance can disguise itself itself that we have developed statistical methods to evaluate sample results against chance distributions. However, the standard standard multiple multiple regression equation is more easily seduced. Specifically, multiple regression equa equations rely heavily on any and all patterns of relations (chance or otherwise) among a set of variables in their single-minded pursuit of minimizing squared squared errors errors of of prediction. Thus, multiple regression equations "capitalize" on chance fluctuations to derive optimal regression coefficients, coefficients, and they do this all the more as terms are added to the equation (Wherry, 1931). Within a sample of of data it is not possible to distinguish random patterns from from ones. Thus, some of the complex terms that serve to characterize a nonrandom ones. given sample of data may be modeling "error." To separate chance patterns from (repUcate) (Mosier, 1956) the model with systematic ones, one must cross validate (replicate) a new set of data. In a new sample of data, the random patterns will be different different and the terms that modeled these patterns in the original sample will no longer be accurate. Moreover, they will generally be counterproductive. One reason that simple models outperform outperform complex ones in new samples is that the simple ones have been less influenced by chance patterns. Indeed, the basic bivariate linear model is largely impervious to capitalization on chance, and the unit-weighted multivariate linear Unear model does not rely on a sample to derive its regression parame parameters. It is precisely these models that are typically the most accurate when they are applied to new samples (Dawes, 1979). 2. f Simple Models 2. The The Nomothetic Power o of

Certainly not all complexity is the result of random processes. However, even when complex terms reflect systematic relations, simple models may still outpredict outpredict com complex ones. As As we we develop develop aa more detailed and complex complex characterization of aa phenom phenomena, we place it. These Theseconditions conditions and and qualifica qualificaplace more more conditions conditions and and qualifications qualifications on on it. tions necessarily restrict the generality of the model because in many situations all of the conditions may not be known or their operation may be further further qualified by other variables. Moreover, each additional measure of aa qualifying variable will will bring bring with it more measurement error and a reduction in statistical power. Simple models overcome these problems by aggregating the complexities into a smaller number of more reliable variables. variables. For example, it is difficult difficult to improve on the simple linear equation that uses high school grade-point average (GPA) to performance in college (APC). Certainly, an individual's academic predict academic performance function of ability, motivation, social skill, discipline, performance is a complex function peer support, and a host of other factors. However, the simple model of APC = /(GPA) will typically outperform outperform the more complicated model APC = !(ability, /(ability, f(GPA) motivation, etc.). The reason is that GPA is an aggregate that reflects, assuming some consistency between high school and college, the personality, ability, and =

=

886 886

WILUAM F. CuAPUN WILUAMF.CHAPUN

situational variables, combined in whatever complex way they combine. No one believes that the relation APC = = GPA is, even remotely, an adequate explanation explanation of of academic performance in college. To explain performance we need a complex model, but to predict performance the simple model that sidesteps the complexity by aggregation works better. We do not need to know the idiographic variables and their relations for each student; we get that information for free, so to speak, GPA. when we obtain GP A.

IV. rv. CONCLUSION The conclusion that follows from this review is an old one that can best be expressed of moderators and other by those who invested considerable energy in the pursuit of effort required to develop complex relations. "It is quite possible that the time and effort fruitfully spent in seeking improvements in reliability moderators might be more fruitfully of the sort that follow follow from classical psychometric psychometric theory" (Ghiselli, and validity of 1963, p. 86). "It still seems safe to assert that new predictions made on the basis of the most significant significant interaction effects are unlikely to be more valid than those of of general predictors alone" (Goldberg, 1972, p. 207). "The made on the basis of whole trick is to decide what variables to look at and then to know how to add" & Corrigan, 1974, p. (Dawes & p. 105). of psychological phe pheModels that are designed to increase our understanding of nomena need to reflect the conditional and nonlinear relations that are certainly nomena psychological variables. However, models that serve the more operating among psychological efficiently predicting important criteria are always handicapped pragmatic goal of efficiently by the inclusion of complex relations. Our efforts to improve predictions from personality variables should be devoted to the development of simple models that rely on higher order variables that automatically include the complexity that exists. Such models and variables will be of of little scientific interest because they mask the processes we wish to understand. They may. may, however, be of great practical utility.

ACKNOWLEDGMENTS Some of the ideas presented in this chapter were formulated formulated while while the author was was aa postdoc postdoctoral fellow at Stanford Stanford University, University, supported supported by byGrant GrantMH-15728 MH-15728from fromthe theNational NationalInstitute Institute of Mental Health. Additional support was provided by Grant MH-39077 from from the National Institute of Mental Health to the Oregon Research Institute. I thank Lewis R. Goldberg, John, Steven StevenPrentice-Dunn, Prentice-Dunn,Stephen StephenG. G.West, West,and andJennifer JenniferWilson Wilsonfor fortheir theirhelpful helpful Oliver P. John, reactions to initial drafts of this chapter.

REFERENCES Bandura, A. (1977). (1977). Self-efficacy: Self-efficacy: Toward Towardaaunifying unifyingtheory theoryof ofbehavioral behavioralchange. change.Psychologi Psychological Review, Review, 84, 84,191-215. 191-215.

CHAPTER 33 33 CuAPn:R

INTERACITVE RELAll0NS RELATIONS iNn:RACI1VE

887 887

Bern, D. J., & Allen, A. (1974). On predicting some of the people some of the time: The search for cross-situational consistencies in behavior. Psychological Psychological Review, Review, 81, 506-520. Block, J. (1977). Advancing the psychology of personality: Paradigmatic shift or improving the quality of research? In D. Magnusson & N. S. Endler (Eds.), Personality Personality at the crossroads: crossroads: Current issues in interactional interactional psychology psychology (pp. 37-63). Hillsdale, NJ: Erlbaum. Bobko, P. (1986). A solution of some dilemmas when testing hypotheses about ordinal interactions. Journal Journal of of Applied Applied Psychology, Psychology, 71, 323-326. Bohrnstedt, G. W., & Marwell, G. (1977). The reliability of products of two random variables. In K. F. Schuessler (Ed.), Sociological Sociological methodology: methodology: 1978 (pp. 254-273). San Francisco: JJossey-Bass. ossey-Bass. A clash of of cultures at meeting on misconduct. Science, 243, 598. Booth, W. (1989). A of predictability. Journal Journal of of Brown, F. G., & Scott, D. A. (1966). The unpredictability of Educational Educational Measurement, Measurement, 3, 297-301. Brown, F. G., & Scott, D. A. (1967). Differential Differential predictability in college admissions testing. Journal of of Educational Educational Measurement, 4, 163-166. Journal Busemeyer, J. R., & Jones, L. E. (1983). Analysis of multiplicative combination rules when causal variables are measured with error. Psychological Psychological Bulletin, 93, 549-562. Chaplin, W. F. (1991). The next generation of moderator research in personality psychology. Journal Journal of of Personality, Personality, 59, 143-178. Chaplin, W. F., & Goldberg, L. R. (1984). A A failure to replicate the Bern Bem and Allen study differences in cross-situational consistencies. Journal Journal of of Personality Personality and on individual differences Social Social Psychology, Psychology, 47, 1074-1090. Cheek, J. M. (1982). Aggregation, moderator variables, and the validity of personality tests: A peer rating study. Journal Journal of of Personality Personality and and Social Psychology, Psychology, 43, 1254-1269. Cohen, J. (1977). Statistical power power analyses for for the behavioral behavioral sciences (Rev. ed.). New York: Academic Press. Cohen, J. (1978). Partialed products are interactions; partialed powers are curve components. Psychological Bulletin, 85, 858-866. Psychological of dichotomization. Applied Applied Psychological Psychological Measurement, Measurement, 7, Cohen, J. (1983). The cost of 249-253. Cohen, J., & Cohen, P. (1983). Applied Applied mUltiple multiple regression/correlation regression/correlation analysis for for the behav behavioral ioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum. Cronbach, L. J. (1956). Assessment of individual differences. Annual Annual Review Review of of Psychology, Psychology, 7, 7, 173-196. Cronbach, L. J. (1957). The two disciplines of scientific psychology. American American Psychologist, Psychologist, 12, 671-684. Cronbach, L. J. (1967). How can instruction be adapted to individual differences? In R. M. Gagne Gagn6 (Ed.), Learning Learning and individual individual differences differences (pp. 23-43). Columbus, OH: Merrill. Cronbach, L. J. (1975). Beyond the two disciplines of scientific psychology. American American Psychol Psychologist, 30, 116-127. Cronbach, L. J. (1987). Statistical tests for moderator variables: Flaws in analyses recently Psychological Bulletin, 102, 414-417. proposed. Psychological Dance, K. A., & Neufeld, R. W. J. (1988). Aptitude-treatment interaction research in the clinical setting: A A review of of attempts to dispel the "patient uniformity" myth. Psycho Psychological Bulletin, 104, 192-213. Latan^, B. (1970). Bystander Bystander intervention. New York: Appleton-Century-Croft. Appleton-Century-Croft. Darley, J., & Latane,

888

WILLIAM CHAPLIN WILLIAMF. F.CHAPUN

Dawes, R. M. (1979). The robust beauty of improper linear models in decision making. American 571-582. American Psychologist, Psychologist,34,34, 571-582. Dawes, R. M., M., & & Corrigan, B. (1974). Linear models in decision making. Psychological Psychological Bulletin, Bulletin,81, 81,95-106. 95-106. paradox: A comparative analysis Dubin, R., & & Taveggia, Taveggia, T. T. C. C. (1968). (1968). The Theteaching-learning teaching-learning paradox: A comparative analysis of college college teaching teachingmethods. methods.Eugene: Eugene: University Oregon Press. University ofof Oregon Press. of Dunlap, W. W.P., P.,&&Kemery, Kemery,E.E.R.R.(1987). (1987).Failure Failuretotodetect detectmoderating moderatingeffects: effects:IsIs multicoUinearmulticollinear 418-420. ity the problem? Psychological PsychologicalBulletin, Bulletin,102, 102, 418-420. Epstein, S. (1983). Aggregation and beyond: beyond:Some Somebasic basicissues issueson onthe theprediction predictionof ofbehavior. behavior. Journal Journalof ofPersonality, Personality,51,51,360-392. 360-392. & Pearson, P. H. (1970). Theory and techniques of personality measurement. Fiske, D. W., & Annual Psychology,21, 21,49-86. 49-86. Annual Review of of Psychology, & Rosenman, R. (1974). Type TypeA A behavior behaviorand andyour yourheart. heart.New NewYork: York: Knopf. Knopf. Friedman, M., & Gage, N. L. (1963). Handbook of of research researchon onteaching. teaching.Chicago: Chicago: Rand McNally. Rand McNally. Journal Ghiselli, E. E. (1956). Differentiation Differentiation of individuals in terms of their predictability. Journal of of Applied Psychology, Psychology, 440 THA-yil. 0 374-377. GhiselU, E. E. (1960a). The prediction of predictability. predictabiUty. Educational Educationaland andPsychological Psychological MeaMea Ghiselli, surement,20, 20,3-8. 3-8. surement, Ghiselli, E. E. (1960b). Differentiation of tests in terms of the accuracy with which they 20, 20, 675-684. predict a given individual. Educational Educationaland andPsychological PsychologicalMeasurement, Measurement, 675-684. Ghiselli, E. E. (1963). Moderating effects and differential differential reliability and validity. Journal Journalof of Applied Psychology, Psychology, 47, 47,81-86. 81-86. Chaos:Making Makinga anew newscience. science. New York: Viking. New York: Viking. Gleick, J. (1987). Chaos: Goldberg, L. R. (1964). The effects of six teaching conditions on learning and satisfaction Schools,2,2,366-375. 366-375. Psychology in in the the Schools, in a televised college course. Psychology Goldberg, L. R. (1968). Simple models or simple processes? Some research on clinical 483-496. judgments. American American Psychologist, Psychologist,23,23, 483-496. personahty assessment: Goldberg, L. R. (1969). The search for configural relationships in personality MultivariateBehavioral Behavioral The diagnosis of psychosis and neurosis from the MMPI. Multivariate Research, Research,4,4,523-536. 523-536. Goldberg, L. R. (1972). Student personality characteristics and optimal college learning conditions: An Instructional An extensive extensive search search for for trait-by-treatment trait-by-treatment interaction interaction effects. effects. Instructional Science, Science,I,1,153-210. 153-210. Goldberg, L. (1991). Human mind versus regression equation: Five contrasts. In D. L. R. (1991). & W. Grove (Eds.), Thinking Thinkingclearly clearlyabout aboutpsychology: psychology: Essays in honor Essays in honor of of Cicchetti & E. Meehl Meehl (Volume {Volume1:1:Matters Matters public interest, 173-184). Minneapolis, MN: Paul E. ofof public interest, pp. pp. 173-184). Minneapolis, MN: University of Minnesota Press. Golding, S. S. L. L.(1978). (1978).Toward Towardaamore moreadequate adequatetheory theoryofofpersonality: personality:Psychological Psychologicalorganizing organizing (pp. 69-95). principles. In H. London (Ed.), Personality: Personality:AAnew newlook lookatatmetatheories metatheories (pp. 69-95). Washington, DC: DC: Hemisphere. Haines, D. D. B., & &McKeachie, McKeachie,W. W.J.J.(1967). (1967).Cooperative Cooperativeversus versus competitivediscussion discussion methods competitive methods 58,58, 386-390. in teaching introductory psychology. Journal Journalof ofEducational EducationalPsychology, Psychology, 386-390. of Hase, H. D., D., & & Goldberg, L. R. (1967). The comparative validity of different strategies of 231-248. deriving personality inventory scales. Psychological PsychologicalBulletin, Bulletin,67,67, 231-248. Psychological Hoffman, P. J. (1960). the paramorphic representation of clinical judgement. Psychological Bulletin, Bulletin,57, 57,116-131. 116-131. & Revelle, W. (1984). Personality, motivation, and performance: A Humphreys, M. S., & theory of the relationship between individual differences and information processing. Psychological PsychologicalReview, Review,91,91,153-184. 153-184.

CuAnER CHAPTER 33 33

lNn:RACI1VE INTERACTIVERELAnoNs RELATIONS

889 889

AmeriHusband, R. W. (1954). Television versus classroom for learning general psychology. Ameri Psychologist, 9, 181-183. can Psychologist, Jackson, D. N. (1971). The dynamics of structured personality tests: 1971. 1971. Psychological Psychological Review, 78, 229-248. Jones, E. E., & Nisbett, R. E. (1972). The actor and the observer: Divergent perceptions of the causes of of behavior. In E. E. Jones, D. E. Kanouse, H. H. Kelly, R. E. Nisbett, Attribution: Perceiving Perceiving the causes of of behavior behavior S. Valins, & B. Weiner (Eds.), Attribution: (pp. 79-94). Morristown, NJ: General Learning Press. differential predictor of Kellogg, R. L. (1968). The Strong Vocational Interest Blank as a differential Educational and and Psychological Psychological Measurement, Measurement, 28, 1213-1217. engineering grades. Educational prediction of of performance performance in clinical psychology. clinical psychology. Kelly, E. L., & Fiske, D. W. (1951). The prediction Ann Arbor: University of Michigan Press. psychology of of personal personal constructs (Vol. 1). New York: Norton. Kelly, G. A. (1955). The psychology Correlation and and causality. New York: Wiley. Kenny, D. A. (1979). Correlation of the beholder: Kenrick, D. T., & Stringfield, D. O. (1980). Personality traits and the eye of Crossing some traditional boundaries in the search for consistency in all people. Psycho PsychoReview, 87, 88-104. logical Review, researcher*shandbook handbook (2nd (2nded.). ed.).Englewood Englewood Cliffs, Cliffs, Design and and analysis: A researcher's Keppel, G. (1982). Design NJ: Prentice-Hall. perceptions: Implications Levine, J. M., & Wang, M. C. (1983). Teacher and student student perceptions: Implications for for learning. learning, Hillsdale, NJ: Erlbaum. Reinforcement schedules, scholastic aptitude, autonomy need, and Lublin, S. C. (1965). Reinforcement achievement in a programmed course. Journal Journal of of Educational Educational Psychology, 56,295-302. Psychology, 56, 295-302. achievement McKeachie, W. J. (1963). Research on teaching at the college and university level. In N. L. Gage (Ed.), Handbook Handbook of of research on teaching (pp. 1118-1172). Chicago: Rand McNally. McKeachie, W. J. (1968). Psychology at age 75: 75: The psychology teacher comes into his own. American Psychologist, Psychologist, 23, 551-557. American prediction of of performance performance McNemar, Q. (1952). Review of E. L. Kelly and D. W. Fiske, The prediction in clinical psychology. Psychology, 47, 857-860. psychology. Journal Journal of of Abnormal Abnormal and Social Social Psychology, Meehl, P. E. (1978). Theoretical risks and tabular asterisks: Sir Karl, Sir Ronald, and the slow Psychology, 46, 806-834. progress of soft psychology. Journal Journal of of Consulting Consulting and and Clinical Clinical Psychology, 46,806-834. Mischel, W. (1968). Personality Personality and and assessment. New York: Wiley. Mischel, W., & Peake, Peake, P. P. (1982). (1982). Beyond Beyond deja d6}hvu vuin inthe thesearch searchfor forcross-situational cross-situational consistency. consistency. Psychological Review, Review, 89, 730-755. Psychological Morris, J. H., Sherman, J. D., & Mansfield, E. R. (1986). Failures to detect moderating squares—moderated multiple regression: Some reasons and effects with ordinary least squares-moderated a remedy. Psychological Psychological Bulletin, 99, 99, 282-288. 282-288. Mosier, C. I. (1951). Problems and designs of cross-validation. Educational Educational and and Psychological Psychological Measurement, 11, 5-11. Nisbett, R. E. (1977). Interaction versus main effects effects as goals of of personality research. In Personality at at the crossroads: crossroads: Current issues in D. Magnusson & N. S. Endler (Eds.), Personality interactional psychology psychology (pp. 235-241) New York: Wiley. interactional Assessment of of men. New York: Rinehart. Rinchart. Assessment Staff. (1948). Assessment Office of Strategic Services Assessment Ozer, D. J. (1986). Consistency Consistency in personality: personality: A A methodological methodological framework. framework. New York: Springer-Verlag. Paunonen, S. V., & Jackson, D. J. (1985). Idiographic measurement strategies for personality promissory notes. Psychological Psychological Review, Review, 92, and prediction: Some unredeemed promissory 486-511.

890 890

WILUAM F. CHAPLIN WILUAMF.CHAPUN

Popper, K. R. (1962). Conjectures Conjecturesand andrefutations. refutations.New NewYork: York:Basic BasicBooks. Books. Sines, J. J. O. O. (1964). (1964).Actuarial Actuarialmethods methodsas asappropriate appropriatestrategy strategyfor forthe thevalidation validationof ofpsychologi psychological tests. Psychological PsychologicalReview, Review,71, 71,517-523. 517-523. Skinner, B B.. F. (1968). The technology technology of of teaching. teaching.New NewYork: York:Appleton-Century-Crofts. Appleton-Century-Crofts. film Snow, R. E., Tiffin, Tiffin, J., & Seibert, W. F. (1965). Individual differences differences and instructional film 56,56, 315-326. effects. Journal Journalof ofEducational EducatiormlPsychology, Psychology, 315-326. Stricker, Strieker, L. J. (1967). Compulsivity as a moderator variable: A replication and extension.

Journal Journalof ofApplied Applied Psychology, Psychology,50, 50,331-335. 331-335.

Relationships between between training training methods methods and and learner learner characteristics. characteristics. Tallmadge, G. K. (1968). Relationships Journal 59,59, 32-36. Journal of of Educational EducationalPsychology, Psychology, 32-36.

J., & & Watson, Watson, D. D. (1982). (1982).Recognizing Recognizingindividual individualdifferences differences in inpredictive predictive Tellegen, A., Kamp, 1., structure. Psychological PsychologicalReview, Review,89, 89,95-105. 95-105. Turner, R. G. (1978). Consistency, self-consciousness, and predictive validity of typical and of Research Research in inPersonality, Personality,12,12,117-132. 117-132. maximal personality measures. Journal of Comment on on the the general general inapplicability inapplicability of of Ghiselli's Ghiselli's moderator moderator system system Velicer, W. F. (1972). Comment for two two predictors. predictors. Journal Journalof ofApplied AppliedPsychology, Psychology,56, 56,262-265. Idl-ldS, for M. A., A.,& &Leggett, Leggett,M. M.I.I.(1972). (1972).Testing Testingthe thehypothesis hypothesisthat thataaperson personwill willbebeconsistent: consistent: Wallach, M. Journal Stylistic consistency versus situational specificity in size of children's drawings. Journal Personality,40, 40,309-330. 309-330. of Personality, of

Wherry, R. J. (1931). A new formula formula for for predicting the shrinkage of the multiple correlation Wherry, coefficient. Annals Annals of of Mathematical MathematicalStatistics, Statistics, 440-457. 2, 2, 440-457. coefficient. Wiggins, J. J. S. S.(1973). (1973).Personality Personality and prediction: Principles ofpersonality assessment, Reading, and prediction: Principles ofpersonality assessment. Reading, Wiggins, MA: Addison-Wesley. configuration and intelligence. Science, Science, 192, 192,227-236. 227-236. Zajonc, R. B. (1976). Family configuration C, & & Sartoris, 1. J. M. M. (1988). Zuckerman, M., Koestner, R., DeBoy, T., Garcia, T., Maresca, B. c., To predict some of the people some of the time: A reexamination of the moderator Social Psychology, variable approach in personality theory. Journal Journalof ofPersonality Personalityand and Social Psychology, variable 54, 1006-1019. 54,

CHAPTER 34 CHAPTER 34

PERSONALITY AND HEALTH PROGRESS AND PROBLEMS PROGRESS AND PROBLEMS IN IN PSYCHOSOMATICS PSYCHOSOMATICS

J, WIEBE AND TIMOTHY W. W . SMITH DEBORAH 1. UNIVERSITY OF OF UTAH UTAH UNIVERSITY

The past few decades have witnessed a remarkable resurgence of interest in the influence of personality on health (Suls & Rittenhouse, 1987). 1987). Sparked potential influence (M. Friedman & Rosenman, by theory and research on the Type A behavior pattern (M. 1974) and psychological hardiness (Kobasa, 1979), the study of personality and health has has been been an an integral integral component component of of the the developing developing fields of of behavioral behavioral medi medihealth cine and and health health psychology. psychology. Compared Compared to to their their psychoanalytic psychoanalytic predecessor, predecessor, current current cine approaches have been more closely tied to an expanding empirical base. As a result, significant, growing impact on psychology and medicine. such research has had a significant, of Yet despite these advances, work on personality and health has been the target of from rejection rejection (Angell, 1985) to careful careful noteworthy criticism. Varying in degree from discussion of methodological limitations (e.g., Holroyd & Coyne, 1987), the central theme of these critiques is that support for the basic psychosomatic hypothesis is tentative at best. The purpose of this chapter is to provide: (1) an overview of conceptual brief summaries of several of the more models guiding research in this area, (2) brief limitadeveloped areas of research, and (3) a discussion of common methodological limita We focus almost almost exclusively exclusively on on the role role of of personality personality processes processes tions and challenges. We of physical illness. However, personality is relevant in the development and course of issue—psychological responses to physical illness and its treat treatto another central issue-psychological ment. For For example, example, personality personality variables variables may may moderate moderate the the emotional emotional and and behav behavment. ioral effects effects of of acute acute medical medical crises, crises, as as well well as as the the adaptive adaptive demands demands of of chronic chronic ioral physical illness. illness. We We focus focus on on the the contribution contribution of of personality personality to to the the development development physical JlANDBOOK HANDBOOK OF OF PERSONALTl'Y PERSONALTTY PSYCHOLOGY PSYCHOLOGY

891 891

CoPYII GHT C COPYRIGHT © 1997 BY ACADEMIC ACADEMIC PRESS. PRESS. ALL lIGHTS OF REPRODUcnON IN ANY roRM RESERVED.

ALL RIGHR OF REPRODUCHON IN ANY FORM RESERVED.

WIEBE WiEBE AND ANDSMI1H SMITH

892

and course course of ofphysical physicalillness, illness,asasititrepresents representsthe theoldest oldestand andmost mostdeveloped developedapplica application of personality psychology to understanding disease.

I. MODELS OF ASSOCIATIONS BETWEEN PERSONALITY AND HEALm HEALTH This section provides an overview of four models that have been developed to explain the processes by which personality may influence health. Although the compremodels are discussed separately, they are not clearly mutually exclusive or compre hensive explanations of the personality-health association. Multiple models are often utilized to explain how a specific personality dimension may be related to health. health. A. Stress-Moderation Model Many approaches to personality and health assign a central role to the concept of of see Cohen, Cohen, 1979; 1979;Contrada, Contrada,Leventhal, Leventhal,&&O'Leary, O'Leary,1990; 1990;Hous Housstress (for reviews, see ton, 1989; Suls & & Rittenhouse, Rittenhouse, 1990). 1990).Although Although stress stress has has been been widely widely implicated implicated ton, 1989; Suls in the the development development of of disease, disease, demonstrated demonstrated stress-illness stress-illness relations relations are are quite quite weak weak in (Rabkin & & Struening, Struening, 1976). 1976). Such Such weak weak links links have have been been hypothesized hypothesized to to reflect reflect (Rabkin individual differences differences in in susceptibility susceptibility to to stress. stress. Thus, Thus, the the stress-moderation stress-moderation model individual model assumes that that stress stress causes causes illness illness and and that that dispositional dispositional factors factors make make one one more more or assumes or less vulnerable vulnerable to to its its pathogenic pathogenic effects. effects. less is generally generally hypothesized hypothesized to toaffect affect health healthvia viaseveral severalphysiological physiologicalpath pathStress is ways. ways. Specifically, Specifically, stress stressisisbelieved believedtotoactivate activatethe thesympathetic sympatheticand andneuroendocrine neuroendocrine systems (i.e., the the sympathetic sympathetic adrenal-medullary adrenal-medullary and and pituitary-adrenocortical pituitary-adrenocortical sys systems), resulting resulting in in increased physiological arousal. arousal. Frequent Frequent and andprolonged prolonged periods periods of arousal are presumed to place excessive strain on body organs and systems, eventually leading to illness (Krantz & aI., 1989; 1989; Selye, & Manuck, 1984; 1984; Menkes et al, Selye, 1956). 1956). Stress-induced neuroendocrine activation has also been reported to impair the effectiveness effectiveness of the immune system, thereby increasing the risk of disease otherwise inhibited by byimmunological immunologicalprocesses processes(e.g., (e.g.,infectious infectiousdiseases diseasesand andcancer; cancer; for reviews, see Herbert & 1993; O'Leary, 1990). & Cohen, 1993; The ways in inwhich whichstress stressmay mayinfluence influencethese theseunderlying underlyingpathogenic pathogenicprocesses processes are illustrated in Figure 1. 1. Although Although objective objective events eventsvary varyin intheir theirpotential potential to to elicit elicit arousal, it is generally the subjective appraisal of these events that activates the sympathetic and neuroendocrine systems. con systems. Hence, psychological stress and the concomitant physiological arousal have been theorized to occur when an event is appraised as threatening and presenting demands that cannot be met (Lazarus & Folkman, 1984). 1984). Cognitive and behavioral efforts for coping with stress also influ influence physiological arousal. Coping responses may exert effects by changing the intensity or duration of the current stressor, or by influencing influencing the likelihood that similar events will occur or be appraised as stressful in the future.

CHAPTER 34 34 CHAnER

PERSONALITY AND AND HEALlH HEALTH i'ERSONAUTY

893 893

Physiological Arousal

.::: eol.

• •

Illness

/ ! Objective, SStressful —•Appraisalof—•Coping Responses—•Subsequent Events Events .. g RespDnses_Subsequent """ ...sIul _ Objective, Events ^ and Outcomes Event as Events Threatening

Personality Personality

FlGURE FIGURE11 Interactional Interactionalstress-moderation stress-moderationmodel. model

Personality has been hypothesized to influence health at each point in this and coping process process (see Figure 1). 1). First, First, the thecognitive cognitive and and motivational motivational aspects aspects stress and of personality are beUeved believed to exacerbate, attenuate, or prevent the appraisal of stimuU as as threatening, threatening,thereby therebyaffecting affectingfluctuations in physiological physiological arousal. arousal. various stimuli fluctuations in Personality has also been theorized to influence pathogenic physiological arousal at the point of the coping response. If one believes that certain coping strategies influence the degree, duration, or frequency of experienced stress, then personality may impact impact health health by by determining determining whether whether more more or or less less adaptive adaptive coping coping responses responses may are utilized. utilized. are This basic stress-moderation model is limited in some respects. Although the straightforward approach is appealing, researchers have generally been unable to identify stable categories of coping responses that are consistently adaptive or is growing growing recognition recognition that thatcoping copingand and personal personalmaladaptive (Lazarus, 1990). There is ity constructs must be conceptualized in process-oriented terms that consider inter interactions between personality, coping, and situational factors over time (Contrada et aI., al., 1990; 1990; Houston, 1989; 1989; Lazarus, 1990; 1990; Lazarus & & Folkman, 1984). More recent stress-moderation models of personality and health move move beyond this adherence to older, static interactional approaches to personality, and explicitly acknowledge the reciprocal relations between persons and situations (Smith, 1989; transactionalviews viewsofofpersonality personalityassert assertthat that & Sanders, 1989). The various transactional Suls & people do not simply respond to situations-they situations—they also create the situations they encounter through their choices choices and and actions actions (Buss, (Buss, 1987; 1987; Cantor, Cantor,1990). 1990).For For example, example, antagonistic individuals may may create interpersonal conflicts through their argumenta argumentative social behavior, while agreeable persons might avoid such interpersonal strain. As depicted in Figure 2, the transactional stress-moderation approach identi identifies fies three pathways through which personality may influence the pathophysiology of disease. In In addition to impacting appraisal and and coping responses to a given extent, personality is likely to influence the objective events themselves. More frequent, severe, and enduring exposure to stressful events is likely to contribute to disease. The reciprocal nature of these processes is reflected in the acknowledgment that personality dispositions may be strengthened and maintained by objective events

7� T Th'"S: /

894

WIEBE AND SMI11I WKBEANDSMIIH

�---Physiological Arousal- ^ -I·� lIIness Illness

Appraisal of Coping --.. Subsequent Events Objective, Stressful --. —•Appraisal o f--' —•Co p i n g Responses Responses—•Subsequent ObJective, Event as as ^ and Outcomes Events Threatening

^Personality Personality

FlGUU FIGURE 22 Transactional Transactional stress-moderation stress-moderation model. model.

and situational contexts, creating the short- and long-term cycles central to many transactional approaches to personality (e.g., Carson, 1969; 1969; Wachtel, 1977).

B. Health Health Behavior Behavior Models Models B. contrast to the direct physiological linkages of the stress-moderation model, the In contrast health behavior behavior model model proposes proposes that that personality personality affects affects health health via via the the quality quality of health one's health practices (Cohen, (Cohen, 1979; 1979; Contrada et a1., al, 1990). 1990), It It is is well well accepted accepted that that one's health practices Contrada et poor health practices (e.g., (e.g., smoking, lack of of exercise, exercise, high-fat high-fat diet) diet) increase increase the the poor health practices smoking, lack risk of of developing developing aa variety variety of of illnesses illnesses (Blair (Blair et et al, Coyne & Holroyd, Holroyd, risk a1., 1989; 1989; Coyne 1982; Paffenbarger & Hale, Hale, 1975). 1975). Further, Further, personality personality variables variables such such as as hardiness hardiness 1982; Paffenbarger (Wiebe McCallum, 1986), 1986), neuroticism neuroticism (Costa (Costa & McCrae, McCrae, 1987b; McCrae, 1987b; McCrae, (Wiebe & McCallum, Costa, & Bosse, Bosse, 1978), and hostility hostility (Leiker (Leiker & Hailey, Hailey, 1988; Rask, 1990) 1990) are are related related Costa, 1978), and 1988; Rask, to the the tendency tendency to to engage engage in in various various healthy healthy and and unhealthy unhealthy behaviors. behaviors. Thus, Thus, accord accordto ing to to this this model, personality is is linked linked to to illness illness because because it it affects affects one's one's choice choice of model, personality of ing health behaviors. behaviors. This This basic model is is depicted in the the upper upper portion portion of of Figure Figure 3. health basic model depicted in This is is not not aa classic classic psychosomatic psychosomatic model model since since the the direct direct physiological physiological correlates correlates This of personality personality are are not not the the identified identified mediators. mediators. However, However, health health behaviors behaviors are are of increasingly recognized recognized as as contributing contributing to to the the association association between between personality personality increasingly and disease. disease. and Although personality dimensions have been associated with health behaviors, Although the causes causes of of such such associations associations are are unclear. unclear. It It is is possible possible that that psychological psychological variables variables the theorized to to influence influence health health practices practices are are inherent inherent aspects aspects of of some some personality personality theorized I.

11.

Personal!

Stressful Events — • Appraisal — • Coping

FlGultE fiGURE 33 Health Health behavior behavior model. model.

^ Illness

Physlofogical Arousal

34 PERSONAUTY PERSONALITYAND ANDREALm HEALTH CHAPTER 34

895 895

self-efficacy (Bandura, 1989) and internal locus of of control constructs. For example, self-efficacy hypothesized to mediate various (Lau, 1988; Strickland, 1978) may be reasonably hypothesized personality-health personality-health behavior relationships. It is also possible that that the relationship additional form between personality and health behaviors represents an additional form of of stress often deteriorate during times of (Horowitz et moderation. Health practices often of stress (Horowitz aI., al., 1979; Langlie, LangUe, 1977; Shachter, Silverstein, Kozlowski, Kozlowski, Herman, & & Liebling, 1977). 1977). affect the frequency frequency and and duration duration of of appraised appraised stress, then If dispositional traits affect personality dimensions may impact impact health by influencing influencing stress-induced stress-induced health these personality smoking and alcohol and behavior departures. Further, health behaviors such as smoking characterized as regressive or avoidant avoidant coping behaviors. substance use have been characterized influence the likelihood that one will cope with appraised appraised Personality may, thus, influence health-damaging behaviors because one has fewer fewer alternative coping stress through health-damaging in the lower portion resources. This more complex health behavior model is depicted in of Figure 3. of of the health behavior behavior model are likely Ukely to be hampered Tests of hampered by several inherent difficulties difficulties (Contrada (Contrada et aI., al., 1990). For example, personaUty personality appears to influence both health-enhancing health-enhancing and health-damaging health-damaging behaviors. Because these influence behaviors are only weakly correlated other (Harris correlated with each other (Harris & Guten, 1979; . Leventhal, Prohaska, & Hirschman, 1985) and fluctuate fluctuate over over time (Mechanic, 1979), it is unlikely unUkely that that their their aggregate will be strongly strongly correlated correlated with with personality. It is, thus, improbable that global lifestyle factors factors will prove prove to be powerful powerful mediators of the effects of personality on health. of effects of c. C. Constitutional Predisposition Model

Some researchers have suggested that statistical associations between between personality and health reflect rather than direct reflect the operation of of a third variable rather direct causal pro prothat individuals may be genetically cesses. This model, depicted in Figure 4, posits that pathophysiological processes (e.g., enhanced sympathetic predisposed to certain pathophysiological enhanced sympathetic reactivity or reduced reduced parasympathetic dampening dampening of of stress responses) which influ influreactivity development of of subsequent subsequent illness and and the behavioral, cognitive, and ence both the development of measured personality. Thus, although although prospective emotional aspects of prospective studies may identify robust associations between personality personaUty and and subsequent subsequent disease, the consti constiidentify predisposition model characterizes this association association as epiphenomenal. tutional predisposition studied across personality personality variables, While this model has not been systematically studied of personality (Bouchard, Lylkken, McGue, Segal, & & evidence of the heritability of

� Personality � Predisposition p. Physiological Physiological Predisposition----I.� Responsiveness Responsiveness� �' lIIness Illness FIGURE FIGURE 44 Constitutional Constitutional predisposition predisposition model. model.

896 896

WIEBE WiEBE AND AND SMI11I SMIIH

Tellegen, 1990) suggests that it is important to understand understand the influence influence of of genetic factors on personality-health associations. D. Dlness Illness Behavior Model

In this model, personality is hypothesized to affect affect illness behavior rather than actual illness. Illness can be defined as the presence of objectively objectively measured patho defined of pathophysiological processes such as high blood pressure, documented documented organic disease, or death. Illness behavior, in contrast, refers to actions people take when they of symptoms, work absen absenperceive themselves as ill. Such behaviors include reports of Although illness behavior is teeism, medical care utilization, and self-medication. Although & Lemke, 1990; Kaplan & & clearly related to actual health status (Idler, Kasl, & & Kotler, 1985; Maddox & & Douglas, 1973), this correlation Camacho, 1983; Kaplan & far from from perfect. Stoical individuals may display less illness behavior than their is far actual health health warrants, warrants, while while hypochondriacs hypochondriacs typically typically display excessive illness illness be beactual havior. Illness Illness behavior behavior is is heavily heavily influenced influenced by by psychological factors such such as as the the havior. psychological factors tendency to to perceive perceive physiological physiological sensations sensations and and to them as Cioffi, tendency to label label them as illness illness (see (see Cioffi, Pennebaker, 1982; Watson & & Pennebaker, 1989, for reviews). reviews). 1979; Pennebaker, 1982; Watson 1989, for 1991; Cohen, 1979; Thus, from from the the illness illness behavior behavior perspective, perspective, personality-health personality-health relationships are Thus, relationships are potentially artifactual. artifactual. potentially Figure 5 illustrates how psychological factors might influence influence illness behavior Research on a commonly used illness in the absence of underlying pathophysiology. Research measure—self-reported physical symptoms-suggests symptoms—suggests that symptom re rebehavior measure-self-reported ports are strongly affected differences in focus of of affected by situational and dispositional differences attention and by health beliefs and related cognitive processes (Pennebaker, 1982). As discussed later, the personality dimension of neuroticism includes a high level of somatic concern, which in tum turn increases symptom reporting. Actions such as staying home from from work and visiting a physician are also likely to be affected affected by the manner in which an individual perceives and labels ongoing physiological sensations. Thus, these seemingly more objective health measures may be heavily influenced influenced by psychological factors.

j �j /

^ Reports of of Symptoms

.. r--------r-----1.

Heightened I Labeling as al --. Normal Physiological ^ Perception of � ^Labeling as Normal Physiologic ll»ness llness and A Sensations ttention to Attention Sen Sensations

I

Personality

FIGURE FIGURE 55

Illness Illness behavior behavior model. model.

Health Health ^^^^ Care Utilization Utilization

^ •

CHAPTER CHAPTER 34 34

PERSONALITY PERSONALITY AND AND REALm H E ALIH

897 897

II. n. PERSONALITY CONSTRUCTS IN CURRENT RESEARCH A great variety of personality dimensions have been examined in the most recent cycle of research on personality and health. Although we cannot provide the depth and range of discussion needed for for a comprehensive comprehensive review, this section summarizes research concerning the most important personality constructs examined to date. In frameworks, assessment procedures, each case, we address four issues—conceptual issues-conceptual frameworks, evidence of associations with health, and methodological limitations. A. A. Type Type A A Behavior Behavior and and Hostility Hostility of the Type A pattern M. Friedman and Rosenman's (1959) description of pattern— competitiveness, hostility, impatience, achievement striving, job involvement, and style—represents the beginning of the current resurgence a loud, explosive speech style-represents actionof personality and health research. They described the Type A pattern as an action emotion complex and a style of response to environmental challenges and demands. Subsequent conceptual approaches to the Type A pattern have extended Subsequent descriptions and explanations of this behavioral style. Glass (1977) suggested that responding to perceived perceived threats to one's the Type A pattern represents a style of responding control over the environment. Compared to the more relaxed, easygoing Type B's, Type A's are more concerned with control, have a lower threshold for perceiving threats to control, and respond with more vigorous attempts to exert control. Price (1982) described the Type A pattern as the overt manifestation manifestation of an underlying tentative sense of self-esteem of self-esteem and beliefs beUefs concerning the necessity and difficulty difficulty of demonstrating self-worth through success and achievement. Several variations of of these basic conceptual approaches have been offered. offered. The major major conceptual descriptions of Type A behavior and its association with coronary disease identify identify physiological reactivity as the underlying pathophysi pathophysiological mechanism. Consistent with an interactional interactionsd stress-moderation model, Type A's are hypothesized to display larger stress-induced stress-induced increases in cardiovascular and neuroendocrine parameters (e.g., blood pressure, heart rate, circulating catecate cholamines) than their Type B counterparts. Quantitative and qualitative reviews have supported the hypothesis that Type A's exhibit more pronounced reactivity (Harbin, 1989; 1989; Houston, 1988). Other approaches to the issue of mechanisms linking Type A behavior behavior and health have adopted stress-moderationmodel. model.From Fromthis thisperspective, perspective. adopted a transactional transactionalstress-moderation Type A's not only respond to challenges with more pronounced reactivity, they also create more severe, frequent, and enduring stressors through their thoughts and actions (Smith, 1989; 1989; Smith & & Anderson, 1986). The The physiological physiological consequences consequences of this greater exposure to stressors contribute, in turn, to the development of of disease. Finally, some authors have proposed a constitutional predisposition model (Krantz & & Durel, 1983). 1983). From this perspective, the overt behavioral manifestations manifestations an underlying constitutional predisposition, specifically of the Type A pattern reflect an a more responsive or reactive sympathetic nervous system. This autonomic respon-

898

WIEBE WiEBE AND AND SMI'I1I SMITH

sivity not only produces the Type A behavioral style, but also contributes to the development of cardiovascular disease. Three assessment devices have been used in the bulk of the research on the Type A pattern and are considered primary by virtue of their inclusion in large, prospective studies of coronary risk: (SI; Rosen risk: the Type A Structured Interview (SI; Rosenman, 1978), the Jenkins Activity Survey (Jenkins, Rosenman, & & Zyzanski, 1974), and the Framingham Type A Scale (Haynes, Feinleib, & & Kannel, 1980). Although initially intended to assess the same construct, these measures are quite modestly & Rosenman, Rosenman, 1981; 1981;Matthews, Matthews,Krantz, Krantz, intercorrelated (Chesney, Black, Chadwick, & & MacDougall, MacDougall, 1982) 1982) and and exhibit exhibit distinct distinct patterns patterns of of correlations correlations with with Dembroski, & Smith, O'Keefe, O'Keefe, & &Allred, AUred,1989). 1989). al., 1981; Smith, other personality traits (e.g., Chesney et aI., Thus, the convergent and discriminant validity of measures of Type A behavior is questionable. This inequivalence of measures intended to assess the same construct has contributed to confusion and controversy surrounding the Type A concept. Research on the association between the Type A pattern and subsequent produced inconsistent inconsistent findings. findings. After After two two decades decades of of generally generally coronary disease has produced supportive data, a panel of experts convened by the American Heart Association concluded that the Type A pattern was a robust coronary risk factor (Cooper, & Weiss, Weiss, 1981). 1981).Several Several widely widely published published failures failures to to replicate replicate this this association association Detre, & (e.g.. Case, Heller, have recently raised doubts about the previous conclusion (e.g., «&Norusis, Norusis,1985). 1985). & Moss, 1985; 1985; Ragland & & Brand, 1988; 1988; Shekelle, Gale, & Case, & This controversy has been reduced to some extent by recent meta-analyses of the Miller, Turner, Turner, Tindale, Tindale, Posovac, Posovac, & & available prospective studies (Matthews, 1988; 1988; Miller, Dugoni, 1991). In previously healthy populations, the Type A pattern is associated with increased risk of initial, premature development of of coronary coronary disease. disease. However, However, this is true only of the Type A pattern when quantified by the SI, and is not true in high-risk groups (e.g., high levels of traditional coronary risk factors, or patients with preexisting disease). The study of high-risk populations appears to account for much of the recent inconsistencies. One result of the inconsistency of findings findings in this area has been increased attention on the health risks associated associated with with the the individual individual elements elements within within the the array of Type A characteristics. Prospective studies using the SI and scoring components separately have identified hostility as the the "toxic" "toxic" element elementwithin withinthe theType TypeA Apattern pattern (Dembroski, MacDougall, Costa, & & Grandits, 1989; 1989; Hecker, Chesney, Black, & Frautchi, 1988; 1988; Matthews, Matthews, Glass, Glass, Rosenman, Rosenman, & &Bortner, Bortner,1977). 1977).These Thesefindings findingshave have prompted considerable interest in the health consequences of hostility (see Smith, 1992, for review). Because of its use in an early cross-sectional study of coronary disease (R. B. Williams aI., 1980) and its availability in archival MMPI data sets, WilUams et al., the Cook and Medley (1954) Hostility (Ho) Scale has been pressed into service in several longitudinal studies. Although an equal number of studies have supported and failed to support an association between Ho scale scores and subsequent health 1989; Hearn, Heam, Murray, & (e.g.. Barefoot, Dodge, Peterson, Dahlstr6m, Dahlstrttm, & Williams, 1989; (e.g., 1989; Leon, Finn, Murray, & & Bailey, 1988; 1988; Shekelle, Gale, Ostfeld, & Leupker, 1989; prospective studies studiesusing usingdifferent differentself-report self-reportmeasures measuresof ofhostility hostility Paul, 1983), other prospective

CuAPl'ER CHAPTER 34 34

PERSONAUTY PERSONALITY AND AND HEALm HEALTH

899 899

have found this trait to predict subsequent illness 1987; Koskenvuo ilhiess (Barefoot (Barefoot et al., al, 1987; et al, al., 1988). A recent meta-analysis of this literature indicates that both behavioral ratings of hostility and the Ho scale are reliable predictors of coronary heart disease (CHD), although the latter latter association is smaller than the former former (Miller, Smith, Turner, Guijarro, & Hallett, 1993). Further, hostility as measured by the Ho scale and other self-report self-report instruments predicts all-cause mortality. That is, hostile persons are likely to die at an earlier age from from both CHD and other illnesses. As in the case of the broader Type A pattern, models of the mechanisms linking hostility and health have focused primarily on physiological reactivity to 1994; R. B. environmental stressors (Smith, 1994; B. Williams, Barefoot, & Shekelle, 1985). Several studies have found that hostile persons respond to social stressors such as interpersonal conflict conflict and provocation with larger increases in blood pressure than is exhibited by comparatively friendly, agreeable subjects (Christensen & Smith, is 1993; Hardy & Smith, 1988; Smith & Allred, AUred, 1989; Suarez & Williams, Williams, 1989). In 1993; Hardy Smith, 1988; Smith 1989; Suarez 1989). In addition to these stress-moderation studies, several studies suggest that hostile addition to these stress-moderation studies, several studies suggest that hostile persons experience experience more more frequent frequent and and severe severe interpersonal and have have interpersonal stressors stressors and persons fewer and less satisfactory social supports (Blumenthal, Burg, & Williams, 1987; fewer and less satisfactory social supports (Blumenthal, Burg, & Williams, 1987; Houston & & Kelley, Kelley, 1989; 1989; Smith Smith & & Frohm, Frohm, 1985; 1985; Smith, Smith, Pope, Sanders, Allred, AUred, & Pope, Sanders, & Houston O'Keefe, 1988). This pattern of psychosocial correlates of hostility may reflect the O'Keefe, 1988). This pattern of psychosocial correlates of hostility may reflect the impact of of hostile hostile stress-engendering stress-engendering behavior behavior on on the environment. impact the interpersonal interpersonal environment. Thus, transactional stress-moderation models may also be useful in explicating the Thus, transactional stress-moderation models may also be useful in explicating the health consequences of hostility (Smith & Pope, 1990). health consequences of hostility (Smith & Pope, 1990). One recent component analysis suggests that hostility may not be the only coronary-prone aspect of the Type A pattern. Houston, Chesney, Black, Cates, Gates, & H Hecker ecker (1992) found that both hostility and social dominance are independent GHD. Interestingly, the act of exerting social influence predictors of subsequent CHD. and control elicits heightened cardiovascular reactivity, and social dominance is GHD in animal models (see Smith & Ghristensen, 1992, for for review). associated with CHD Christensen, 1992, Thus, both socially controlling and hostile persons might be at increased coro coronary risk.

B. Neuroticism Neuroticism B. Neuroticism is a broad dimension of normal personality characterized by a tendency to experience chronic negative emotions and to display related behavioral and (Gosta & McCrae, McGrae, 1980,1987b; McGrae & Costa, Gosta, 1984). cognitive characteristics (Costa 1980, 1987b; McCrae Highly neurotic individuals thus have generally negative views of themselves and (McGrae & the world regardless of the objective reality. Neuroticism is quite stable (McCrae Gosta, 1984) and there is growing evidence that it is heritable heritable (Pedersen, Plomin, Costa, & Friberg, 1988; 1988; McGleam, & 1988; Rose, Koskenvuo, Kaprio, Sama, Sarna, & Langinvainio, 1988; McClearn, Tellegen et al., 1988). (McCrae, 1982). Familiar Many instruments exist for measuring neuroticism (McGrae, scales include the Eysenck Personality Inventory Neuroticism Scale (Eysenck &

900 900

WIEBE WiEBE AND AND SMI11I SMITH

Eysenck, 1968} 1968) and the Neuroticism factor of the NEO Personality Inventory (Costa & & McCrae, 1985). Watson and Clark (1984), however, have argued that many scales developed to measure seemingly different dimensions of negative affects (e.g., depression, trait anxiety, and neuroticism) correlate so strongly that they they can be interpreted as measuring the same latent construct. Neuroticism correlates significantly with various measures of illness (see Costa & & McCrae, 1987b; 1987b; H. S. Friedman & & Boothe-Kewley, 1987; 1987; Watson, 1988; Watson & & Pennebaker, Pennebaker, 1989, 1989,for forreviews). reviews).Because Becausenegative negativeemotions emotionsare areassociated associated and hazardous hazardous health health behaviors behaviors (e.g., (e.g., smoking, smoking, with both both autonomic autonomicfluctuations with fluctuations and alcohol alcohol or or substance substance use; use; Costa Costa & & McCrae, McCrae, 1987b), 1987b), such such associations associations suggest suggest that that neuroticism neuroticism may may contribute contribute to to the the development development of of disease. There There is is reason reason to to be be skeptical, however, since since the the majority majority of of studies studies have illness skeptical, however, have utilized utilized subjective subjective illness behavior behavior indices indices (e.g., (e.g., self-report self-report symptom symptom checklists; checklists; Stone Stone & & Costa, Costa, 1990). 1990). As As already argued, argued, such such measures measures may may reflect reflect tendencies tendencies to to perceive perceive or or report report physical physical already sensations as illness symptomology rather than any pathophysiological process. Recent research research has, has, thus, thus, attempted attempted to to distinguish distinguish between between the the illness illness behavior behavior Recent and the the psychosomatic psychosomatic effects effects of of neuroticism. neuroticism. and There pro There is is evidence evidence that that neuroticism neuroticism is is associated associated with with the psychological psychological proin the theillness illnessbehavior behaviormodel model(e.g., (e.g.,exaggerated exaggeratedencoding encodingand andrecall recall cesses articulated articulated in cesses of symptoms; 1992). Further, symptoms; Larsen, Larsen, 1992). Further, although although there there are are inconsistencies inconsistencies (Persky, (Persky, Kempthorne-Rawson, & at., & Shekelle, 1987; 1987; Shekelle et at., al, 1981; 1981; Somervell et al., 1989), 1989), data data generally generally indicate indicate that that neuroticism neuroticism does does not not predict predict the objective objective health health measure of mortality (e.g., Almada et ai., 1991; Costa & McCrae, 1987b; al., 1991; 1987b; Dattore, Shontz, & & Coyne, 1980; 1980; Kaplan & & Reynolds, 1988; 1988; Keehn, Goldberg, & & Beebe, 1974; Zonderman, Costa, & McCrae, 1989). One exception to this pattern is that depression may hasten death in high-risk populations, such as patients with chronic disease (Burton, Kline, Lindsay, & 1986; Carney, Carney, Rich, Rich, & & Freedland, Freedland, & Heidenheim, 1986; 1988; Frasure-Smith, Lesperance, & 1988; & Talajic, 1993). Although neuroticism may be associated with increased increased behaviorally behaviorally mediated mediated mortality mortality (e.g., (e.g., suicide suicide or oraccidents; accidents; Keehn et al, at., 1974), it does not appear to be a robust risk factor for disease in the general population. In support of the illness behavior hypothesis, patients complaining of chest pain who have been objectively shown to have healthy arteries score higher on neuroticism than do patients with actual disease or noncomplaining disease-free controls (Costa, Fleg, McCrae, & & Lakatta, 1982). Further, some researchers have reported a negative negative association association between between indices indicesof ofneuroticism neuroticismand andarterial arterialocclusion occlusion & Williams, Williams, 1992; 1992;Bass Bass& &Wade, Wade,1984; 1984;Elias, Elias, (Barefoot, Beckham, Peterson, Haney, & Robbins, Blow, Rice, & & Edgecomb, 1982). Prospective studies indicate that the dispositional tendency to report physical symptoms is associated with increased risk of developing angina-like chest pain syndromes, but not with objective evidence of of CHD such as documented myocardial infarction (Shekelle, Vernon, & & Ostfeld, 1991). Thus, neuroticism and related traits appear to be associated with cardiac 1991). symptom symptom reporting, reporting, but but not not with with the the development development of of actual actual cardiac cardiac disease. In summary, data suggest suggest that that neuroticism neuroticism increases increases the the tendency tendencyto toperceive, perceive. summary, available available data

CHAPTER 34

HEALIH PERSONALITY AND REALm

901

report, or act upon general physical sensations rather than the development of of physical illness itself. The relationship between neuroticism and illness behavior has specific implica implications for personality and health research. As mentioned mentioned above, above, this area of of research research has often often utilized subjective subjective health measures that are clearly contaminated with neuroticism. Although this approach may provide useful useful information information in the begin beginning stages of research, investigators must progress beyond this and clearly distin distinguish the psychological causes of illness behavior from from those of actual disease. Second, many personality personaUty variables that are commonly studied in the personality and health literature are either conceptually similar to neuroticism or have been 1987; Smith, Pope, found (e.g.. Funk & Houston, 1987; found to be associated with neuroticism (e.g., Rhodewalt, & Poulton, 1989). Any association between such personality variables and subjective health measures may, of may, therefore, reflect reflect the common influence of neuroticism. Personality and health researchers would benefit benefit from from more precisely assessing the personality variables of interest and demonstrating their distinction from broad constructs such as neuroticism. Finally, at a more complex level, many from of the variables hypothesized to mediate the effects effects of personality personaHty on health (e.g., stressful life events, social support, and health behaviors) are also correlated with stressful & Costa, 1984). 1984). This potential 1986; Schroeder & neuroticism (Depue & Monroe, 1986; theoretically different different constructs impedes accurate explication explication of person personoverlap of theoretically ality-health relationships. ality-health c. C. Hardiness The construct of hardiness represents the aggregate of beliefs that life is meaningful, meaningful, 1979). Such beliefs controllable, and challenging rather than threatening (Kobasa, 1979). hkelihood that an encountered life event is appraised are theorized to reduce the likelihood autoas stressful, thereby diminishing the potential of that event to induce adverse auto 1985). Thus, hardiness theory nomic arousal (Kobasa, Maddi, Puccetti, & Zola, 1985). rehes heavily on the interactional stress-moderation model. relies A variety of hardiness measures have appeared in the literature (see Funk, 1992; Orr Orr & Westman, 1990, 1990, for reviews). As discussed below, each has been 1992; criticized and there is currently no accepted standard standard hardiness scale. The initial assessment devices were developed by selecting a battery of existing personality scales that discriminated between high-stress-high-illness and high-stress-Iow-illness high-stress-low-illness 1982). Several shortened male executives (Kobasa, 1979; 1979; Kobasa, Maddi, & Kahn, 1982). versions of this battery have appeared in the literature, and a revised 50-item scale is currently being marketed (Personal Views Survey, The Hardiness Institute). selfHardiness has been primarily studied by examining its relationship with self reported physical symptoms. Most studies indicate that high-hardy subjects report & Gannon, 1988; 1988; Kobasa et aI., al., 1982; 1982; lower levels of concurrent symptoms (Banks & Kobasa & Puccetti, 1983; 1983; Rhodewalt & Zone, 1989; 1989; Roth, Wiebe, Fillingim, & 1989; P. W. W. Williams, Williams, Wiebe, Wiebe, & &Smith, Smith, 1992) 1992) and and subsequent subsequent symptoms symptoms than than Shay, 1989;

902 902

WIEBE AND SMI1H WIEBEANDSMTIH

1985; Wiebe & do low-hardy subjects (Banks & Gannon, 1988; Kobasa et aI., al, 1982, 1982,1985; McCallum, 1986). While these main effects effects appear consistent with hardiness theory, this interpre interpretation may be misleading for at least two reasons. First, hardiness theory is based on the stress-moderation model, which implies the presence of a statistical interaction between hardiness and stress. Some studies have reported this necessary interaction aI., 1982; Rhodewalt & Zone, 1989), but others (Banks & Gannon, 1988; Kobasa et al., have not (Funk & Houston, 1987; Roth et aI., al., 1989; Schmied & Lawler, 1986; Wiebe & McCallum, 1986). Second, these data are based upon subjective measures of of physical symptoms and are open to the interpretation that hardiness influences influences illness behavior rather than actual illness. This possibility takes on added significance significance when one considers that the hardiness scales appear heavily contaminated with neuroticism (Allred (AUred & Smith, 1989; 1989; Funk & Houston, 1987). In a test of of the convergent and discriminant validity of of the hardiness scales, Wiebe, Williams, and Smith (1990) found that separate measures of hardiness correlated more strongly with each other than they did with measures of of neuroticism. Despite such discrimination, however, hardiness and neuroticism were strongly correlated. This overlap makes it possible that demonstrated associations between hardiness and illness behavior occur as a function of the common influence of neuroticism rather than of of the portion of of hardiness that is distinct from neuroticism. of this possibility indic�te indicate that this does occur to some extent (Funk & Direct tests of Rhodewalt & Zone, 1989; P. W. Williams et aI., al., 1992). Houston, 1987; Rhodewalt The physiological mechanisms hypothesized to link hardiness and health have also been studied. Several studies have reported that, in response to laboratory stressors, high-hardy subjects display lower cardiovascular arousal than low-hardy subjects (Contrada, 1989; 1989; Wiebe, 1991; Wiebe & Williams, 1992). Allred AUred and Smith (1989), however, reported that high hardiness was associated with increased, rather than decreased, systolic blood pressure responses to evaluative threat. In addition to such broad tests of of hardiness and stress moderation, attempts have been made to identify identify the specific cognitive mechanisms by which hardiness is theorized to attenuate responses to stress. These studies have primarily focused focused on the mediating effects of stress appraisal. In support of hardiness theory, low effects of lowhardy subjects have been found to make fewer positive self-statements self-statements after an evaluative threat task than do high-hardy subjects (Allred (AUred & Smith, 1989). These individuals have also been reported to rate the same life experiences (Rhodewalt & Rhodewalt & Zone, 1989) and laboratory stressors (Wiebe, Agustsdottir, 1984; Rhodewalt 1991; Wiebe & Williams, WiUiams, 1992) as less positive and controllable than do high-hardy individuals. Finally, some data suggest that such appraisals mediate the relationship al., 1989; Roth et aI., between hardiness and health outcomes (Rhodewalt & Zone, 1989; 1989; Wiebe, 1991). Subjectively appraised stress has also been reported to mediate the relationship between hardiness and health behaviors (Wiebe & McCallum, 1986), suggesting suggesting that that adaptive adaptive stress stress appraisals appraisals may may impact impact health health by by reducing reducing stress-induced health behavior departures. stress-induced health behavior departures.

CHAPTER CHAPTER 34 34

PERSONAUTY PERSONALITYAND ANDREALm HEALTH

903 903

Although research on the cognitive aspects of of hardiness is fairly supportive, Although complete body of of literature literature is highly inconsistent. This may reflect reflect problems problems the complete measurement of of hardiness. In addition to the documented documented overlap with in the measurement neuroticism, the more commonly commonly used hardiness scales have been criticized because because different component component scores (commitment, control, and chal chalthey combine three different lenge) into one composite index. However, the accuracy of viewing hardiness as a unidimensional construct construct has been questioned questioned due to inconsistent findings · from from unidimensional factor of the hardiness scale and to reports that the commitment commitment and factor analyses of control components predict health outcomes while challenge does not (see Carver, Carver, Funk & Houston, 1987; 1991; Hull, VanTreuren, & 1989; Funk 1987; Hull, Lehn, & Tedlie, 1991; for reviews). Progress on the health consequences of of hardiness is Virnelli, 1987, for likely to be hampered until such psychometric issues are resolved.

D. Optimism and Explanatory Style Scheier and Carver (1985, 1992) propose that the personality trait of optimism effects on health. They define define dispositional optimism as a stable, has beneficial effects generalized expectation that good things will happen. In the context context of of their Control generalized expectation Theory approach approach to adaptation and self-regulation self-regulation (Carver & Scheier, 1982), these authors argue that individual differences differences in optimism are central to the process of of adjustment. When individuals recognize a discrepancy between their behavioral behavioral adjustment. When goals or standards and their present present situation, individual differences differences in optimism optimisminfluence subsequent subsequent actions. Optimists, anticipating positive outcomes, pessimism influence present will persist in attempts to reduce the discrepancy between goals and the present situation. That is, optimists should strive to solve problems and cope actively. In contrast, pessimists, anticipating bad outcomes, are likely to exhibit passive or fatalistic responses. Further, the more adaptive coping of of optimists should lessen fatalistic effects of of stressors on emotional adjustment adjustment and physical the otherwise deleterious effects health. Thus, this approach approach clearly conforms conforms to the general interactional stress stressmoderation model, with an emphasis on individual differences differences in secondary ap apmoderation subsequent coping. praisal and subsequent differences in optimism-pessimism, Scheier and Carver To assess individual differences developed an eight-item Life Life Orientation Test (LOT). Consistent Consistent with pre pre(1985) developed dictions, high LOT LOT scores have been found found to be associated with reduced reports of of physical illness (Scheier & Carver, 1985), higher levels of problem-focused problem-focused coping, of passive coping strategies such as avoidance (Scheier, Weintraub, & and less use of other interesting interesting tests of predictions derived from from the model, Carver, 1986). In other LOT scores are less likely to experience postpartum postpartum expectant mothers with high LOT depression (Carver (Carver & Gaines, 1987), and alcoholics with high LOT scores are more more depression likely to complete complete treatment treatment (Strack, Carver, & Blaney, 1987). Scheier et al. (1989) found that, compared compared to their pessimistic counterparts, optimistic cardiac surgery found patients exhibited more rapid postoperative postoperative recoveries and less likeHhood of an likelihood of intraoperative myocardial infarction. Optimism has also been found found to predict

904 904

WiEBE AND AND SMI11I SMTIH WIEBE

improved emotional adjustment to breast cancer, an effect that is apparently medimedi ated by adaptive coping responses (Carver et a1., al., 1993). To evaluate the overlap of neuroticism and optimism, Smith, Pope, Rhode Rhodewalt, wait, & & Poulton (1989) conducted a convergent-discriminant validity analysis. In three independent samples, the LOT was as closely correlated with two measures of neuroticism as it was with a second measure of optimism-the optimism—the Generalized & Hale, 1978). 1978). As a result, the LOT could be Expectancy for Success Scale (Fibel & construed as a measure of neuroticism (i.e., scored in the opposite direction), and at the very least is clearly contaminated with this trait. Such contamination raises the question of whether or not shared variance with neuroticism underlies the correlations of optimism with various indices of coping and adjustment. In two independent samples, statistical control of neuroticism scores eliminated the other othersignificant correlation between optimism and physical symptoms and coping wise significant &Poulton, Poulton,1989, 1989,Studies Studies11and and2). 2).These Theseresults results behaviors (Smith, Pope, Rhodewalt, & & Bosse, 1993) suggest that some and others (e.g., Mroczek, Spiro, Aldwin, Ozer, & reflecting of thefindings findings consistent with the optimism model can be interpreted as reflecting the more basic, established dimension of neuroticism. From a different conceptual background, Peterson and Seligman (1987) have described a construct construct similar to optimism. optimism. Explanatory Explanatory style style refers refers to to stable patterns of causal attributions individuals make for positive and negative events in their lives. Derived from the learned helplessness framework, an optimistic explanatory style refers to the characteristic tendency to make internal, stable, and global attributions for positive events, and external, unstable, and specific attributions for negative events. The pessimistic explanatory style is characterized by the opposite pattern of causal attributions. Explanatory style can be assessed with either structured questionnaires or a rating technique using written or verbal descriptions of events. The questionnaire presents a series of hypothetical positive and negative events, and requires causal attribution in response to an open-ended question as well as ratings of the stated cause on the three dimensions. The text-based system is an independent raterrater based approach, though it relies reUes on the same attribution dimensions. Although this approach has not been used extensively in health research, some results suggest that explanatory style may be related to subsequent health (for a review, see Peterson & & Seligman, 1987). For example, a pessimistic explana explanatory style is prospectively related to increased reports of illness, as well as visits to a physician. physician. The interpretive ambiguities of research research using illness illness behavior as an outcome apply to thesefindings. findings. However, Peterson, Seligman, SeUgman, and Valliant (1988) found that a pessimistic explanatory style was associated with physicians' ratings of subjects as less healthy over a 35-year follow-up. Given the somewhat more is more compelling. objective nature of this health index, this finding finding is Umitation of these provocativefindings One limitation findings concerns assessment procedures. The convergence of the questionnaire and rating approaches is not established, and independent evidence of construct validity is lacking. Thus, something assessed by these techniques may be useful in predicting health outcomes, but just what that

34 PERSONALITY PERSONALITYAND ANDREALm HEALTH CHAPTER 34

90S 905

potential mechanisms linking dimension is has not been clearly identified. Further, potential explanatory style and health have not been extensively explored. Preliminary find findings, however, suggest that pessimists are less likely to take active steps to treat treat their their illnesses than are optimists (Lin & Peterson, 1990). Thus, maladaptive, passive health behavior in response to initial illness may exacerbate exacerbate disease, contributing contributing observed in some studies. to the associations between explanatory style and health observed

E. Inhibited Power Motivation Motiyation

Power motivation is defined defined as the desire to have an impact on others by controlling, them (McClelland, 1975,1985), The influencing, aggressing against, or even helping them 1975, 1985). The conceptual model states that when inhibited inhibited or frustrated frustrated by psychological general conceptual processes such as self-restraint self-restraint or situational factors, power motivation contributes contributes power motivation to the development development of of physical illness. Thus, this is another interactional stressanother stress moderation approach: a motivationally based personality trait interacts with situamoderation motivationally personality situa tional factors to influence health. assessed through through the use of of projective projective tech techInhibited power motivation is assessed specifically the Thematic Thematic Apperception Apperception Test. Ratings Ratings of of responses responses can be niques, specifically fairly stable over time (see Jemmott, 1987, for for review). made reliably and are fairly Independent, thorough evaluations of of construct validity, vahdity, however, have not been Independent, reported to date. Inhibited health outcomes Inhibited power motivation has been linked to a variety of of health and pathophysiological pathophysiological mechanisms. In two cross-sectional cross-sectional studies, individuals high in inhibited power motivation were more likely Ukely to have high blood blood pressure pressure (McClel (McClelA prospective study found found that inhibited inhibited power associland, 1979). A power motivation motivation was associ of developing hypertension hypertension over a 20-year follow-up. Indi Indiated with increased risk of found to report report more physical viduals with this motivational dynamic have also been found power motivation has illness (McClelland & Jemmott, 1980). Further, inhibited power been linked to a reduced immunocompetence immunocompetence that could place such individuals at greater risk for for a variety of infections and neoplastic diseases (Jemmott et aI., al., 1983, 1990; McClelland, Alexander, & Marks, 1982; McClelland, Floor, Davidson, & Saron, 1980). Associations between this personality personality trait and high blood blood pressure pressure and sup suppressed pressed immune functioning functioning suggest plausible pathophysiological pathophysiological mechanisms. Lacking from from the available evidence, however, are larger, prospective studies evalu evaluating the utility of of this trait in predicting death or objective objective indications of of serious illness. Nevertheless, there are interesting parallels between this model and interac interactional stress-moderation stress-moderation research with both humans and animals. In humans, the attempt attempt to assert influence or control over others is associated with heightened heightened cardiovascular reactivity (Brown & Smith, 1992; Smith, Baldwin, & Christensen, Christensen, 1990). Thus, the social behaviors likely Ukely to be associated with power power and dominance hypothesized to increase disease risk. are related to physiological stress responses hypothesized In a series of studies by Kaplan, Manuck, and their colleagues (see Manuck, Kaplan, Adams, & Clarkson, 1988, for for review), chronic stress induced induced by repeated repeated reorgani-

WiEBE AND ANDSMlI1I SMITH WIEBE

906

zation produced accelerated zation of of social social groups groups produced accelerated atherosclerosis atherosclerosis in in dominant-but dominant—but not not subordinate-male monkeys. The stable individual difference domi subordinate—male monkeys. The stable individual difference variable variable of of dominance-submission nance-submission in in this this animal animal model model isis clearly clearlysimilar similar to tothe thehuman humanmotivational motivational variable described described by by McClelland, McClelland, and and the the social social reorganization reorganization stressor stressor is is similar similar variable to the the situational situational factors factors in in the the power power motivation motivation model. model. Although Although highly highly specula speculato tive, the the elegant elegant and and compelling compellingfindings from these these animal animal studies studies suggest suggest that that findings from tive, social dominance dominance and and power power motivation motivation may may be be useful useful traits traits in in future future personality personality social and health health research. research. and m . LIMITATIONS LIMITATIONS AND AND CHALLENGES CHALLENGESIN INPERSONALITY PERSONALITYAND AND m. REALm HEALTH RESEARCH

The preceding preceding overview overview of of selected selected topics topics in in personality personality and and health health research research clearly clearly The documents of recent recent activity. activity. Although Although several several areas areashave haveproduced produced poten potenflurry of documents aaflurry tially important importantfindings, recurring limitations limitations are are readily readily apparent. apparent. They They can can be tially findings, recurring be grouped into into three three categories-personality categories—personality assessment, assessment, model model testing, testing, and and descrip descripgrouped tion versus versus explanation. explanation. tion A. Personality A. Personality Assessment Assessment In summaries, the was In each each of of the the previous previous summaries, the issue issue of of valid valid personality personality assessment assessment was raised. In raised. In all all cases, cases, there therewas waseither eitheraalack lackofofsufficient sufficientevidence evidenceofofconstruct constructvalidity validity or or clear clear evidence evidence of of questionable questionable validity. validity. If If the the only only intent intent of of personality personality and and health research health research was was the the improved improved prediction prediction of of health health outcomes, outcomes, construct construct validity validity would example, it would be be irrelevant. irrelevant. For For example, it would would not not matter matter if if aa scale scale that that was was intended intended to to assess assess Type Type A A behavior behavior actually actually measured measured intelligence, intelligence, because becausethe the sole soleconcern concern regarding regarding the the personality personality measures measures would would be be their their incremental incremental predictive predictive utility. utility. Research field, however, Research in in this thisfield, however, is is clearly clearly more more ambitious ambitious than than improved improved statistical statistical prediction alone. The efforts efforts are intended to test specific conceptual hypotheses about the contribution of personality constructs constructstotosubsequent subsequenthealth. health.Yet Yetthe thetask task of of establishing the construct validity of personality assessments is all alltoo toooften often over overlooked. A related problem is the failure to evaluate the relation of newly proposed scales and andconstructs constructsto toestablished establisheddimensions dimensionsofofpersonality. personality.Overlap Overlapwith withneuroti neuroticism was aa common common concern, concern, despite despite the the fact fact that that this this dimension dimension has has aa long long history history in in personality personality taxonomies. taxonomies. Failure Failure to to address address this this concern concern creates creates the the possibility possibility of of reinventing old old traits traits under under new new labels labels (Holroyd (Holroyd & & Coyne, Coyne, 1987). 1987). The The potential potential reinventing result is is an an ever-expanding ever-expanding set set of of unintegrated unintegrated and and misleading misleading literatures. literatures. result Both by Both of of these these assessment assessment problems problems could could be be rectified rectified to to aa large large extent extent by greater use use of of existing existingtools toolsininthe thebroader broaderfield fieldof ofpersonality. personality.Rich Richand andcontinually continually greater refined approaches approaches to to measurement are aa keystone keystone of of personality personality research, research, and and the the refined measurement are emerging consensus consensus regarding regarding the thefive-factor model as as an an adequate adequate trait trait taxonomy taxonomy emerging five-factor model (Digman, 1990) would provide provide an an important important anchor anchor for for integrative integrative efforts. efforts. As As 1990) would (Digman,

34 CHAPTER 34

PERSONAUTY PERSONALITY AND REALm HEALTH

907 907

several authors have argued (Costa & McCrae, 1987a; Smith & Williams, WiUiams, 1992), examining the association between measures used in the personality personaUty and health literature and inventories based on the five-factor model of personality structure may be quite useful in this regard. B. B.

Testing Models Models of of Personality Personality and and Health Health Testing

Even with careful careful validation of measures, other problems may limit the correspon correspondence between the statistical hypothesis that is tested and the conceptual model that guides it. it. As has been described, the operational definitions of health and illness have enormous implications of impUcations for the interpretation interpretation of research results. Because of illness. Given ease and low cost, symptom reports are commonly used to measure illness. the interpretive ambiguity resulting from from this strategy, this approach is unlikely to produce important increments over existing knowledge. vaUd personality assessments are utilized and objective health out outEven if valid comes are assessed, it is important to recognize which features of the guiding conceptual models have been tested and which have not. For example, despite the fact that the predominant model of the cardiovascular effects effects of hostility is a person fact major prospective studies of this trait by situation, stress-moderation approach, the major have only examined the main effect effect of hostility. The predicted interactive effects effects of individual differences differences and situational parameters cannot be tested, because no assessments of situational factors are included (Matthews, 1983; 1983; Smith, 1989). If If the model is correct, the predictive power of the interaction is diluted by collapsing across situation factors. Similarly, most of these models specify some sort of physiological mechanism linking personality traits and illness. To To date, date,no nostudies studies have have evaluated evaluated the the complete complete path model in which the association between personality and actual physical disease is mediated by a physiological or health behavior mechanism (Krantz & Hedges, 1987). Intermediate strategies of associating personality with psychophysiological reactivity and immune functioning functioning are becoming more common and sophisticated. However, the association between these processes and subsequent illness is far from established. is, the overall significance of the short-term immunological, estabUshed. That is, neuroendocrine, or cardiovascular cardiovascular correlates of personality personaHty in experimental studies is simply not known (Blascovich & Katkin, 1993; 1993; Herbert & Cohen, 1993). Thus, the final common pathway in these models is not an established strand in the surrounding nomological net. In addition to not testing the complete model of personality-health personality-health associa associations, research designs often do not rule out competing explanations of documented relationships. For example, one common approach approach to testing associations between personality traits and health is the cross-sectional, group comparison design where personality personaUty characteristics are contrasted in currently currently ill and healthy groups. Al Although useful in preliminary hypothesis testing, significant useful significant group differences differences may actually reflect reflect somatopsychic processes processes rather rather than than psychosomatic effects. effects. That That is, is, personality differences may be the result, rather than the cause, of physical illness. differences

908 908

WiEBE AND AND SMI11I SMITH WIEBE

C Description Description versus versus Explanation Explanation C. It is clear that many of the most important and formidable challenges in this field involve description. Which measures are reliably related to substantive health outcomes, and which traits do these measures assess? Once such associations are documented, knowledge of how personality influences health becomes important. This more explanatory information information is likely to be useful in guiding the development of primary and secondary interventions. In cases where descriptive research supports the pursuit of explanation, devel developments in personality research could again prove useful. Cognitive and social approaches to personality provide conceptual and methodological tools for under understanding dynamic personality processes through which people interpret life tasks and their social environment, regulate and evaluate their behavior, and devise and implement problem-solving strategies. This more recent focus on the "doing" underaspects of personality is an important complement to the recent advances in under standing taxonomies and structure, or the "having" side of personality (Cantor, (Cantor, 1990). Thus, once again a more complete use of the larger field of personality research could lead to important important progress in understanding psychological influences influences on health. health. on

IV. rv. CONCLUDING THOUGHTS The study of personality and health currently confronts both daunting and exciting The advances advances in in several several areas areas are are sufficient sufficient to to support support continued continued efforts, efforts, challenges. The but simple methodological approaches and incomplete tests of the underlying mod models threaten the longevity of this psychosomatic renaissance. Evolving medical technologies now provide opportunities for more compelling evaluations of health outcomes and pathophysiological mechanisms than were previously available. Cur Current personality theory, research, and methods contain avenues for addressing equacritical limitations in research on the psychological side of the psychosomatic equa tion. Thus, lasting advances in the long-standing issue of the impact of personality on physical heahh health may now be possible through the creative application of recent developments in in personality personality psychology psychology and and medicine. medicine. We We eagerly eagerly await await such such ad addevelopments vances. vances.

REFERENCES Allred, Allred,

& Smith, Smith, T. T. W. W. (1989). (1989). The The hardy hardy personality: personality: Cognitive Cognitive and and physiological physiological K. D., D., &

responses to to evaluative evaluative threat. threat. Journal JournalofofPersonality Personality Social Psychology, andand Social Psychology, 56, responses 257-266. 257-266. Almada, S. S. G., Zonderman, A. A. B., Shekelle, R R. B., Dyer, A. A. R, R., Daviglus, Davigius, M. M. L., Costa, Costa, Almada, G., Zonderman, B., Shekelle, B., Dyer, P. T., Jr., & & Stamler, Stamler, J. J. (1991). (1991). Neuroticism Neuroticism and and cynicism cynicism and and risk risk of of death death in in middle middleP. T., Jr., aged men: men: The The Western Western Electric Electric Study. Study. Psychosomatic Psychosomatic Medicine, 53, 165-175. Medicine, 53, 165-175. aged

56,

CuArn:R CHAPTER 34 34

PERSONAUTY PERSONALITYAND ANDREALm HEALTH

909 909

Angell, M. (1985). Disease as a reflection of the psyche. New New England England Journal Journal of of Medicine, Medicine, 312, 1570-1572. Bandura, A. (1989). Human agency in social cognitive theory. American American Psychologist, Psychologist, 44, 1175-1184. Banks, J. K., & Gannon, L. R. (1988). The influence of hardiness on the relationship between between stressors and psychosomatic symptomatology. American American Journal Journal of of Community Community Psy Psychology, 16, 16, 25-37. Barefoot, J. C., C , Beckham, J. c., C , Peterson, B. L., Haney, T. L., & Williams, R. B. (1992). Measures of of neuroticism and disease status in coronary angiography patients. Journal Journal 0, 127-132. of of Consulting Consulting and and Clinical Psychology, Psychology, 6 60, Barefoot, J. C., C , Dodge, K. A., Peterson, B. L., Dahlstrom, Dahlstr5m, W. G., & Williams, R. R, B., Jr. (1989). The Cook-Medley Cook-Medley Hostility Hostility Scale: Item content and ability to predict survival. Psychosomatic Psychosomatic Medicine, 51, 46-57. Barefoot, c.,, Siegler, I. C c.,, Nowlin, J. B., Peterson, B., Haney, T. L., & Williams, Barefoot, 1. J. C R. B., Jr. (1987). Suspiciousness, health, and mortality: A follow-up study of 500 older Psychosomatic Medicine, 49, 450-457. adults. Psychosomatic C , & Wade, C. (1984). Chest pain with normal coronary arteries: A comparative study Bass, c., 4, 51-61. of of psychiatric and social morbidity. Psyc/tological Psychological Medicine, 114, Blair, S. H., Kohl, H. W., Paffenbarger, R. S., Clark, M. S., Cooper, K. H., & Gibbons, L. W. (1989). Physical fitness and all-cause mortality: A prospective study of healthy men and women. JAMA, 2395-2401. JAMA, Journal Journal of of the American American Medical Medical Association, Association, 262, 262,2395-2401. Blascovich, J., & Katkin, E. S. (1993). Cardiovascular Cardiovascular reactivity reactivity to psychological psychological stress and DC: American Psychological Association. disease. Washington, DC: c.,, Burg, M. M., & Williams, R. B. (1987). Psychological Blumenthal, J. A., Barefoot, J. C correlates of hostility among patients undergoing coronary angiography. British Journal of Medical Medical Psychology, Psychology, 60, 349-355. of Bouchard, T. J., Jr., Lylkken, D. T., McGue, M., Segal, B. L., & Tellegen, A. (1990). Sources of human psychological differences: The Minnesota study of twins reared apart. Science, 250, 223-228. Brown, P. C c.,, & Smith, T. W. (1992). Social influence, marriage, and the heart: heart: Cardiovascular of interpersonal control in husbands and wives. Health Psychology, Psychology, consequences of 11, 88-96. Burton, H. J., Kline, S. A., Lindsay, R. M., & Heidenheim, A. P. (1986). The relationship of 261-269. of depression to survival in chronic renal failure. Psychosomatic Psychosomatic Medicine, 48, 48,261-269. Buss, D. M. (1987). Selection, evocation, and manipulation. Journal Journal of of Personality Personality and and Social Psychology, 53, 1214-1221. Psychology, of personal personalCantor, N. (1990). From thought to behavior: "Having" and "doing" in the study of American Psychologist, Psychologist, 45, 735-750. ity and cognition. American Carney, R. M., Rich, M. W., & Freedland, K. E. (1988). Major depressive disorder predicts Psychosomatic Medicine, 50, cardiac events in patients with coronary-artery disease. Psychosomatic 627-633. Carson, R. C. (1969). Interaction Interaction concepts concepts in personality. personality, Chicago: Aldine. Carver, C. S. (1989). How should multifaceted personality constructs be tested? Issues Journal of of Personality Personality illustrated by self-monitoring, attributional style, and hardiness. Journal and Social Social Psychology, Psychology, 56, 577-585. and Carver, C. S., & Gaines, J. G. (1987). Optimism, pessimism, and post-partum depression. Cognitive Therapy Therapy and and Research, 11, 449-462. Cognitive

910 910

WIEBE WiEBE AND AND SSMITH MI1H

Carver, C. C. S., Pozo, Pozo, C., C, Harris, Harris, S. S. D., Noriega, Noriega, V., Scheier, Scheier, M. M. F., F., Robinson, Robinson, D. D. S., S., Ketcham, Ketcham, Carver, A. S., Moffat, F. L., & & Clark, Clark, K. K. L. L. (1993). (1993).How Howcoping copingmediates mediates the theeffect effect of of optimism optimism on distress: Journalof ofPersonality Personality distress: A study of women with early stage breast cancer. Journal and Social Social Psychology, Psychology,65, 65,375-390. 375-390. Carver, C. S., & & Scheier, M. F. (1982). Control theory: theory: A useful conceptual framework for Bulletin, 92, personality-social, personality—social, clinical, and health psychology. Psychological Psychological Bulletin, 1111-135. 1 1-135. Case, R R. B., Heller, S. S., Case, N. B., & & Moss, A. J. (1985). Type A behavior and survival 312, 737-741. after acute myocardial infarction. New Eng/and EnglandJournal JournalofofMedicine, Medicine, 312, 737-741. Chesney, M. A., Black, G. W., Chadwick, J. H., & & Rosenman, R R. H. (1981). Psychological 4, 217-229. correlates of the Type A behavior pattern. pattern. Journal Journalof ofBehavioral BehavioralMedicine, Medicine, 4,217-229. & Smith, T. W. (1993). Cynical hostility and cardiovascular response Christensen, A. J., & during self-disclosure. Psychosomatic PsychosomaticMedicine, Medicine,55,55,193-202. 193-202. Cioffi, D. D. (1991). (1991). Beyond Beyond attentional attentional strategies: strategies: A A cognitive-perceptual cognitive-perceptual model model of of somatic somatic Cioffi, 25-41. interpretation. Psychological PsychologicalBulletin, Bulletin,109, 109, 25-41. Cohen, F. (1979). Personality, stress, and the development of physical illness. In G. C. Stone, dbook (pp. F. Cohen, & Health psychology: psychology:AAhan handbook (pp.77-111). 77-111).San San & N. E. Adler (Eds.), Health Francisco: Jossey-Bass. Contrada, R R. J. (1989). Type Type A A behavior, behavior, personality personahty hardiness, hardiness, and andcardiovascular cardiovascular response response 57,57, 895-903. to stress. Journal Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, 895-903. Contrada, R R. J., Leventhal, Leventhal, H., & & O'Leary, O'Leary, A. A. (1990). (1990). Personality Personality and and health. health. In In L. L. A. A. Contrada, (pp.(pp. 638-669). New York: Pervin (Ed.), Handbook Handbookof ofpersonality: personality:Theory Theoryand andresearch research 638-669). New York: Guilford Press. Cook, W., & & Medley, Medley, D. D. (1954). (1954). Proposed Proposed hostility hostility and andpharisaic pharisaicvirtue virtue scales scalesfor for the the MMPI. MMPI. Journal Journal of ofApplied AppliedPsychology, Psychology,38, 38,414-418. 414-418. Cooper, T., Detre, T., & & Weiss, S. M. (1981). Coronary-prone behavior and coronary heart disease: A critical review. Circulation, Circulation,63, 63,1199-1215. 1199-1215. Costa, P. T., Jr., Fleg, J. L., McCrae, R R. R, R., & & Lakatta, E. G. (1982). Neuroticism, coronary artery disease, and chest pain complaints: complaints: Cross-sectional and longitudinal studies. Experimental 37-44. ExperimentalAging AgingResearch, Research,8, 8, 37-44. Costa, P. T., Jr., & & McCrae, R R. R R. (1980). Somatic complaints in males as a function of age and neuroticism: A longitudinal analysis. Journal 3, 3, 245-257. Journalof ofBehavioral BehavioralMedicine, Medicine, 245-257. manual. Odessa, Costa, P. T., Jr., & & McCrae, R R. R R. (1985). The The NEO NEO Personality PersonalityInventory Inventory manual, Odessa, FL: FL: Psychological Assessment Resources. Costa, P. P. T., T., Jr., Jr.,&&McCrae, McCrae,RR.RR.(1987a). (1987a).Personality Personalityassessment assessmentininpsychosomatic psychosomaticmedicine. medicine. Costa, In T. T. M. M. Wise Wise (Ed.), (Ed.), Advances in inpsychosomatic psychosomaticmedicine medicine (pp. 71-82). Basel: Karger. (pp. 71-82). Basel: Karger. In Costa, P. P. T., T., Jr., Jr., & & McCrae, McCrae, R R. R R. (1987b). (1987b). Neuroticism, Neuroticism, somatic somatic complaints, complaints, and and disease: disease: Costa, Is the the bark bark worse worse than than the the bite? bite? Journal Journalof ofPersonality, Personality,55,55,299-316. 299-316. Is Coyne, J., J., & & Holroyd, Holroyd, K. K. (1982). (1982). Stress, Stress, coping, coping, and and illness: illness: A A transactional transactional perspective. perspective. In In Coyne, T. Millon, C. Green, & &R R. Meagbee (Eds.), Handbook of of clinical clinicalhealth healthpsychology psychology T. (pp. 105-127). 105-127). New York: York: Plenum Plenum Press. (pp. Dattore, P. P. J., Shontz, F. F. C, & Coyne, Coyne, L. L. (1980). (1980). Premorbid Premorbid personality personality differentiation differentiation of c., & cancer and non-cancer groups: groups: A test of the hypothesis of cancer proneness. Journal Journal cancer of Consulting Consultingand andClinical ClinicalPsychology, Psychology, 388-394. 388-394. of 48,48, Dembroski, T. T. M., M.,MacDougall, MacDougall,J.J.M., M.,Costa, Costa,P.P.T.,T.,Jr., Jr., Grandits, (1989). Components Dembroski, && Grandits, G.G. A.A. (1989). Components of hostility as predictors of sudden death and myocardial infarction in the Multiple Risk Factor Factor Intervention Intervention Trial. Trial. Psychosomatic PsychosomaticMedicine, Medicine,51,51, 514-522. 514-522. Risk

34 PERSONAUTY PERSONALITYAND ANDIlEum HEALTH CHAPTER 34

911 911

Depue, R. R. A., A., & & Monroe, Monroe, S. S. M. M. (1986). Conceptualization Conceptualization and and measurement measurement of of human Depue, disorder in in life life stress stress research: research: The The problem problem of of chronic chronic disturbance. disturbance. Psychological Psychological disorder Bulletin, . Bulletin,99, 99,36-51 36-51.

Digman, J. M M.. (1990). Personality structure: Emergence ooff the five-factor five-factor model. Annual Annual

Review Review of Psychology, Psychology, 41, 41,417-440. 417-440. c., Rice, A. P., & Elias, M. F., Robbins, M. A., Blow, F. C, & Edgecomb, T. L. L. (1982). Symptom reporting anxiety, and depression in arteriographically arteriographicaUy classified middle-aged chest pain patients. Experimental 45-51. ExperimentalAging AgingResearch, Research,8, 8, 45-51. & Eysenck, S. B. (1968). Eysenck Eysenck Personality Personality Questionnaire Questionnairemanual. manualSan San Eysenck, H. J., & Diego, CA: CA: Educational and Industrial Testing Service. Fibel, B., & & Hale, W. D. (1978). The Generalized Expectancy for Success Scale-A Scale—A new Psychology, 46, 46, 924-931. measure. Journal Journal of of Consulting Consultingand andClinical Clinical Psychology, 924-931. Frasure-Smith, N., Lesperance, Lesperance, F., F., & & Talajic, Talajic, M. M. (1993). Depression Depression following following myocardial myocardial Frasure-Smith, infarction: Impact on six-month survival. JAMA, Journal Journal of of the theAmerican American Medical Medical Association, Association,270, 270,1819-1825. 1819-1825. Friedman, H. H. S., S., & & Booth-Kewley, Booth-Kewley, S. S. (1987). The The "disease-prone "disease-prone personality": personality": A A metaFriedman, meta analytic view view of of the the construct. construct. American American Psychologist, Psychologist,42,42, 539-555. 539-555. analytic Friedman, M., & Rosenman, Rosenman, R. R. H. H. (1959). Association Association of of aa specific specific overt overt behavior behavior pattern pattern Friedman, M., & with increases increases in in blood blood cholesterol, cholesterol, blood blood clotting clotting time, time, incidence incidence of of arcus arcus senilis senilis and with clinical coronary coronary artery artery disease. disease. JAMA, Journal Journalof ofthe theAmerican AmericanMedical Medical Association, clinical Association, 169, 1286-1296. 1286-1296. 169, Friedman, M., M.,& &Rosenman, Rosenman,R. R.H. H.(1974). (1974).Type TypeAAbehavior behavior and your heart. New York: Knopf. Friedman, and your heart. New York: Knopf. Funk, S. S. C. C. (1992). Hardiness: Hardiness: A A review review of of theory theory and and research. research. Health Health Psychology, Psychology, 11, 11, Funk, 335-345. Funk, S. c., (1987). A critical analysis of the Hardiness Scale's validity C, & & Houston, B. K. (1987). 53,53, 572-578. and utility. Journal Journal of ofPersonality Personalityand andSocial SocialPsychology, Psychology, 572-578. C. (1977). Behavior Behaviorpatterns, patterns,stress, stress,and and coronary disease, Hillsdale, Erlbaum. Glass, D. C. coronary disease. Hillsdale, NJ: NJ: Erlbaum. Harbin, T. J. (1989). The relationship between Type A behavior pattern and physiological responsivity: A quantitative review. Psychophysiology, Psychophysiology,26, 26,110-119. 110-119. Hardy, J. D., D., & & Smith, Smith, T. T. W. W. (1988). Cynical Cynical hostility hostility and and vulnerability vulnerabiUty to to disease: disease: Social Hardy, support, life stress, and physiological response to conflict. Health Health Psychology, Psychology, 7, 447-459. Harris, D. D. M., M., & &Guten, Guten,G. G.(1979). (1979).Health Healthprotective protectivebehavior: behavior:An Anexploratory exploratorystudy. study.Journal Journal of 20,20, 17-29. of Health Healthand andSocial SocialBehavior, Behavior, 17-29. Haynes, S. G., Feinleib, M., & & Kannel, W. B. (1980). The relationship of psychosocial factors to coronary heart disease in the Framingham Study: III. Eight-year incidence Study: III. of coronary heart disease. American 37-58. AmericanJournal Journalof ofEpidemiology, Epidemiology,111, 111, 37-58. Hearn, M. M. D., D.,Murray, Murray,D. D.M., M.,&&Leupker, Leupker,R.R.V.V.(1989). (1989).Hostility, Hostility,coronary coronaryheart heartdisease, disease,and and Hearn, total mortality: mortality: A A 33-year 33-yearfollow-up follow-upstudy studyof of university universitystudents. students.Journal JournalofofBehavioral Behavioral total Medicine,12, 12,105-121. 105-121. Medicine, Hecker, M. M. H. H. L., L., Chesney, Chesney, M. M. A., A., Black, Black, G. G. W., W., & & Frautchi, Frautchi, N. (1988). Coronary-prone Coronary-prone Hecker, behaviors in the Western Collaborative Group Study. Psychosomatic PsychosomaticMedicine, Medicine,50,50, 153-164. Herbert, T. B., B., & & Cohen, Cohen, S. S. (1993). (1993). Stress Stress and and immunity: immunity: A A meta-analytic meta-analytic review. review. Psychoso PsychosoHerbert, T. matic Medicine, Medicine,55,55,364-379. 364-379. matic Holroyd, K. K. A., & Coyne, Coyne, J. J. C. C. (1987). (1987). Personality Personality and and health health in in the the 1980s: Psychosomatic Holroyd, A., & 19805: Psychosomatic medicine revisited? revisited? Journal Journalof ofPersonality, Personality,55,55,359-375. 359-375. medicine

AND SMI11I WIEBE WlEBEANDSMnn

9U 912

Horowitz, M. J., Berfari, R, R., Hulley, S., Blair, S., Alvarez, W., Borhani, N., Reynolds, A., & risk factors, and coronary disease. Psychosomatics, Psychosomatics, Simon, N. (1979). Life events, risk

20, 20, 586-592.

Houston, B. K. (1988). Cardiovascular and neuroendocrine reactivity, global Type A, and components of Type A behavior. In B. K. Houston & & c. C. R R. Snyder (Eds.), Type A A behavior andand intervention (pp.(pp. 212-253). NewNew York: Wiley. behaviorpattern: pattern:Research, Research,theory, theory, intervention 212-253). York: Wiley. (1989). Personality dimensions in reactivity and cardiovascular disease. In Houston, B. K. (1989). research N. Schneiderman, S. M., Weiss, & & P. G. Kaufman (Eds.), (Eds.), Handbook of of research methods incardiovascular cardiovascular behavioral medicine 495-509). New York: Plenum Press. (pp.(pp. 495-509). New York: Plenum Press. methods in behavioral medicine Houston, B. K., Chesney, M. A, A., Black, G. W., Cates, Gates, D. S., & Hecker, M. H. L. (1992). Behavioral clusters and coronary heart disease risk. Psychosomatic Psychosomatic Medicine, Medicine, 54,

447-461.

& Kelley, K. E. (1989). Hostility in employed women: Relation to work Houston, B. K., & and marital experiences, social support, stress, and anger expression, expression. Personality Personalityand and Social Psychology PsychologyBulletin, Bulletin,15,15, 175-182. Social 175-182.

Hull, 1. J. G., G., Lehn, Lehn, D. D. A, A., & & Tedlie, Tedlie, J. J. C. C. (1991). A A general general approach approach to to testing testing multifaceted multifaceted Hull, personality constructs. constructs. Journal Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, 932-945. personality 61,61, 932-945. Hull, J. J. G., VanTreuren, VanTreuren, R R. R, R., & & Vimelli, Vimelli, S. S. (1987). Hardiness Hardiness and and health: health: A A critique critique and Hull, alternative approach. approach. Journal Journalof of Personality Personalityand andSocial SocialPsychology, Psychology, 518-530. alternative 53,53, 518-530. Idler, E. E. L., L., Kasl, Kasl, S. S. V., V., & & Lemke, Lemke, J. J. H. H. (1990). Self-evaluated health health and and mortality mortality among among Idler, (1990). Self-evaluated the elderly elderly in in New New Haven, Haven, Connecticut Connecticut and and Iowa Iowa and and Washington Counties, Counties, Iowa, the AmericanJournal JournalofofEpidemiology, Epidemiology,131, 131, 91-103. 1982-1986. American 91-103.

Jemmott, J. J. B., III. III. (1987). Social Social motives motives and and susceptibility susceptibility to to disease: disease: Stalking Stalking individual individual Jemmott, differences in health risks. risks. Journal Journalof ofPersonality, Personality,55,55,267-298. 267-298. Jemmott, J. J. B., III, Ill, Borysenko, J. J. Z., Z., Boresenko, M., McClelland, McClelland, D. C, R., Jemmott, c., Chapman, R, & Benson, Benson, H. H. (1983). (1983). Academic Academic stress, stress, power power motivation, motivation, and and decrease decrease in in Meyer, D., & Meyer, salivary secretory immunoglobulin A secretion rate. Lancet, Lancet, 1,1,1400-1402. 1400-1402. Jemmott, J. B., III, Ill, Hellman, C., C, McClelland, D. C., C, Locke, S. E., Kraus, L., Williams, R R. M., & & Valeri, C. R R. (1990). Motivational syndromes associated with natural killer cell activity. Journal 13,13, 53-73. Journalof ofBehavioral BehavioralMedicine, Medicine, 53-73. Jenkins, C. C. D., Rosenman, R R. H., & & Zyzanski, S. J. (1974). Prediction of clinical coronary heart disease by a test for the coronary-prone behavior pattern. New England EnglandJournal Journal of of Medicine, Medicine,23, 23,1271-1275. 1271-1275.

Kaplan, G. A, A., & & Camacho, T. (1983). Perceived health and mortality: mortaUty: A nine-year follow followup of the Human Population Laboratory cohort. American AmericanJournal JournalofofEpidemiology, Epidemiology,

117, 117,292-304. 292-304.

A., & & Kotler, P. L. (1985). Self-reports predictive of mortality from ischemic Kaplan, G. A, heart disease: A nine-year follow-up of the Human Population Laboratory cohort. heart Journal 38,38, 195-201. Journalof of Chronic ChronicDiseases, Diseases, 195-201.

Kaplan, G. G. A, A., & & Reynolds, Reynolds, P. P. (1988). Depression Depression and and cancer cancer mortality mortality and and morbidity: Kaplan, Prospective evidence evidence from from the the Mameda Mameda County County Study. Study.Journal Journalof ofBehavioral BehavioralMedicine, Medicine, Prospective 11, 11, 1-13.

Keehn, R R. J., Goldberg, I. D., & & Beebe, Beebe, G. G. W. W. (1974). (1974). Twenty-four-year Twenty-four-year mortality mortality follow-up follow-up Keehn, of army veterans veterans with with disability disability separations separations for for psychoneurosis psychoneurosis in in 1944. 1944. Psychosomatic Psychosomatic Medicine, Medicine,36, 36,27-46. 27-46.

Kobasa, S. S. C. C. (1979). Stressful Stressful life life events, events, personality personality and and health: health: An An inquiry inquiry into into hardiness. Kobasa, hardiness. Journalof ofPersonality Personalityand andSocial Social Psychology, 1-11. Journal Psychology, 37, 37, 1-11.

CHArn:R CHAPTER 34 34

PERSONAUTY PERSONALITYAND ANDHEALnI HEALTH

913 913

Kobasa, S. C, c., Maddi, S. R, & Kahn, S. (1982). S. R., (1982). Hardiness and health. A prospective study. Journal Psychology, 42, 42, 168-177. Journalof ofPersonality Personalityand andSocial Social Psychology, 168-177. Kobasa, S. S. c., C, Maddi, Maddi, S. S.R, R.,Puccetti, Puccetti,M. M.c., C,&&Zola, Zola,M. M.A A.(1985). (1985).Effectiveness Effectivenessofofhardiness, hardiness, of Psychosomatic exercise, and social support as resources against illness. Journal of Research, Research,29, 29,525-533. 525-533. Kobasa, S. c., 983). Personality and social resources in stress resistance. C, & Puccetti, M. C. (1 (1983). Journal Psychology, 45,45, 839-850. Journalof ofPersonality Personalityand andSocial Social Psychology, 839-850. Koskenvuo, M., M., Kapiro, Kapiro,J., J.,Rose, Rose,R R.J., J.,Kesaniemi, Kesaniemi,A, A.,Sarna, Sama,S., S.,Heikkila, Heikkila,K., K.,&&Langinvainio, Langinvainio, H. (1988). Hostility as a risk factor factor for mortality and ischemic heart disease in men. PsychosomaticMedicine, Medicine,50,50, 330-340. Psychosomatic 330-340. A. (1983). Psychobiological substrates of the Type A behavior Krantz, D. S., & Durel, L. A pattern. Health Health Psychology, Psychology,2,2,393-411. 393-411. (1987). Krantz, D. S., & Hedges, S. M. (1 987) . Some cautions for research on personality and health. Journalof ofPersonality, Personality,55,55,351-357. 351-357. Journal Krantz, D. S., & & Manuck, S. B. (1984). Acute psychophysiologic reactivity and risk of of PsychologicalBulletin, Bulletin, cardiovascular disease: A review and methodological critique. critique. Psychological 96, 435-464. 435-464. 96, K. (1977). (1977). Social Social networks, networks,health health beliefs, beliefs,and andpreventive preventive health health behavior. behavior.Journal Journal Langlie, J. K. of Health Healthand andSocial SocialBehaviors, Behaviors,18,18, 244-260. of 244-260. R. J. (1992). Neuroticism and selective encoding and recall of symptoms: Evidence Larsen, R from a combined concurrent-retrospective study. Journal of of Personality Personality and and Social Social from Psychology,62, 62,480-488. 480-488. Psychology, Lau, R Health R. R R. (1988). Beliefs Beliefs about about control control and and health behavior. behavior. In In D. D. S. S. Oochman Gochman (Ed.), (Ed.),Health behavior:Emerging Emergingresearch research perspectives 43-63). York: Plenum Press. (pp.(pp. 43-63). NewNew York: Plenum Press. behavior: perspectives Personalityand and R. S. (1990). Stress, coping, and illness. In H. S. S. Friedman (Ed.), Personality Lazarus, R disease 7 120). New disease(pp. (pp.997-120). NewYork: York:Wiley. Wiley. R. S., & Folkman, S. (1 (1984). Stress, appraisal, appraisal,and andcoping. coping. New York: Springer. Lazarus, R 984). Stress, New York: Springer. & Hailey, B. Leiker, M., & B. J. (1988). A link between hostility and disease: Poor health habits? BehavioralMedicine, Medicine,3, 3,129-133. 129-133. Behavioral G. R, R., Finn, S. E., Murray, D., & & Bailey, J. M. (19 (1988). Leon, O. 88). The inability to predict from hostility scores or MMPI items related to Type A behavior. cardiovascular disease from Journalof ofConsulting Consultingand andClinical Clinical Psychology, 597-600. Journal Psychology, 56, 56, 597-600. R. S. (1985). Preventive Preventive health health behavior behavior across across Leventhal, H., Prohaska, T. R., R, & Hirschman, R Prevention ininhealth healthPsychology Psychology the life span. In J. C. Rosen & L. J. Solomon (Eds.), Prevention (Vol. 8, pp. 1199-235). 99-235 ). Hanover, NH: NH: University Press of New England. Lin, E. H., & Peterson, C. (1990). Pessimistic explanatory style and response to illness. BehaviourResearch Researchand andTherapy, Therapy, 243-248. Behaviour 28,28, 243-248. G. L., & & Douglas, E. B. (1973). Self-assessment Self-assessment and health: A longitudinal study Maddox, O. 87-93. of elderly subjects. Journal 14, 14, Journalof ofHealth Healthand andSocial SocialBehavior, Behavior, 87-93. R., Adams, M. R, R., & & Clarkson, T. B. (1 (1988). Effects of stress and Manuck, S. B., B., Kaplan, J. R, 988) . Effects the sympathetic nervous system on coronary arterydisease in the cynomolgus macaque. American Heart HeartJournal, Journal,116, 116, 328-333. American 328-333. A. (1983). Assessment issues in coronary-prone behavior. In T. M. Dembroski, Matthews, K. A T. H. Schmidt, & heart disease & O. G. Blumchen (Eds.), Biobehavioral Biobehavioralbases basesofofcoronary coronary heart disease (pp. 62-78). Basel: Karger. Matthews, K. A A. (1988). CHD and Type A behavior: Update on and alternative to the Booth-Kewley and Friedman quantitative review. Psychological 373-380. PsychologicalBulletin, Bulletin,104, 104,373-380. -

914

WIEBE AND SMI11f WiEBEANDSMnH

Matthews, K. A., A, Glass, D. C, C., Rosenman, R R. H., & & Bortner, R R. W. W. (1977). Competitive drive, Pattern A, and coronary disease: A further analysis of some data from the 30,30, 489-498. Western Collaborative Group Study. Journal Journalof of Chronic ChronicDiseases, Diseases, 489-498. M., & Matthews, K. A., Krantz, D. S., Dembroski, T. M., & MacDougall, J. J. M. M. (1982). Unique and common conmion variance in structured interview and Jenkins Jenkins Activity Activity Survey Survey measures measures of of Social Psychology, 42,42,303-313. 303-313. the Type A A behavior behavior pattern. pattern.Journal Journalof ofPersonality Personalityand and Social Psychology, New York: Irvington-Halsted. McClelland, D. C. C. (1975). Power: Power: The The inner innerexperience. experience. New York: Irvington-Halsted. McClelland, (1979). Inhibited McClelland, D. D. C. C. (1979). Inhibited power power motivation motivation and and high high blood blood pressure pressure in in men. Journal Journal of ofAbnormaL Abnormal Psychology, Psychology,88, 88,182-190. 182-190. McClelland, Human motivation. motivation.Glenview, Glenview,IL: IL:Scott-Foresman. Scott-Foresman. McClelland, D. D. C. C. (1985). (1985). Human McClelland, D. C, Alexander, C, & Marks, Marks, E. E. (1982). (1982). The The need need for for power, power, stress, stress, immune immune McClelland, C., Alexander, C., & function, and illness 91,91,61 61-70. illness among amongmale maleprisoners. prisoners.Journal JournalofofAbnormal AbnormalPsychology, Psychology, -70. McClelland, D. C, Floor, E., Davidson, R R. J., J., & & Saron, Saron,C. C.(1980). (1980).Stressed Stressedpower powermotivation, motivation, McClelland, C., Floor, sympathetic activation, activation, immune function, function, and and illness. illness. Journal Journalof ofHuman HumanStress, Stress,6(2), 6(2), 111-19. 1 -19. McClelland, McClelland, D. D. C., C, & & Jemmott, Jenunott, J. J. B., B., III. III. (1980). (1980). Power Power motivation, motivation, stress, stress, and and physical physical Journalof ofHuman HumanStress, Stress,6(4), (5(4), 6-15. illness. Journal 6-15. McCrae, McCrae, R R. R R. (1982). (1982). Consensual Consensual validation validation of of personality personality traits: traits: Evidence Evidencefrom fromself-reports self-reports Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, 293-303. and ratings. ratings. Journal 43,43, 293-303. McCrae, R R. R., & Costa, Costa, P. P. T., T., Jr. Jr. (1984). (1984). Emerging Emerginglives, lives,enduring enduring dispositions: Personality McCrae, R, & dispositions: Personality adulthood. Boston: Boston:Little, Little,Brown. Brown. in adulthood. McCrae, R R. R., Costa, P. P. R., & Bosse, Bosse, R R. (1978). (1978). Anxiety, Anxiety, extraversion, extraversion, and and smoking. smoking. R, Costa, R, Jr., & McCrae, British Journal JournalofofSocial Socialand andClinical Clinical Psychology, 269-273. British Psychology, 17, 17, 269-273. Mechanic, D. D. (1979). The stability stability of of health health and and illness illness behavior: behavior: Results Results from from aa 16-year 16-year Mechanic, (1979). The follow-up. American AmericanJournal JournalofofPublic PublicHealth, Health, 1142-1145. 69,69, 1142-1145. follow-up. Menkes, M. M. S., S., Matthews, Matthews, K. K. A, A., Krantz, Krantz, D. D. S., S., Lundberg, Lundberg, U., U., Mead, Mead, L. L. A, A., Qaqish, Qaqish, B., B., Menkes, Liang, K., K., Thomas, Thomas, C. C. B., B., & & Pearson, Pearson, T. T. A A. (1989). (1989). Cardiovascular Cardiovascular reactivity reactivity to to the the Liang, cold pressor test as a predictor of hypertension. Hypertension, Hypertension,14, 14,524-530. 524-530. cold Miller, T. Q., Smith, T. W., Turner, C. ScHallett, Hallett,A A.J.J.(1996). (1996).AAmeta metaC. W., Guijarro, M. M. L., &

analytiC analytic review review of of research research on onhostility hostilityand andphysical physicalhealth. health.Psychological PsychologicalBulletin, Bulletin, 119, 119, 322-348. 322-348. Miller, T. Q., Turner, C. W., Tindale, R R. S., Posovac, E. J., & & Dugoni, B. B. (1991). Reasons for the trend toward null findings in research on Type A behavior. Psychological Psychological Bulletin, Bulletin,110, 110,469-485. 469-485. Mroczek, D. K., Spiro, Bosse, RR. (1993). Construct validation Spiro, A, A.,Aldwin, Aldwin,C. C.M., M.,Ozer, Ozer,D.D.J.,J.,&& Bosse, (1993). Construct validation of optimism and pessimism in older men: men: Findings from the Normative Aging Study. Health Health Psychology, Psychology,12, 12,406-409. 406-409. O'Leary, A O'Leary, A. (1990). (1990). Stress, Stress, emotion, and and human human immune immune function. function. Psychological PsychologicalBulletin, Bulletin, 108, 108, 363-382. 363-382. Orr, E., E., & & Westman, Westman, M. M. (1990). (1990). Does Does hardiness hardiness moderate moderate stress, stress, and and how?: how?: A A review. review. In Orr, and M. Rosenbaum Rosenbaum (Ed.), (Ed.), Learned Learned resourcefulness: resourcefulness:OnOncoping copingskills, skills,self selfcontrol, control, and M. adaptive behavior behavior(pp. (pp.64-94). 64-94).New NewYork: York:Springer. Springer. adaptive Paffenbarger, R R. S., & & Hale, W. W. E. E. (1975). (1975). Work Work activity activity and and coronary coronary heart heart mortality. mortality. New New Paffenbarger, 545-550. England 292, EnglandJournal JournalofofMedicine, Medicine, 292, 545-550. Pedersen, N. N. L., L.,Plomin, Plomin,R, R.,McClearn, McCleam,G. G.E., E.,&&Friberg, Friberg,L.L.(1988). (1988).Neuroticism, Neuroticism,extraversion, extraversion, Pedersen, ofof Personality and and related related traits traitsin inadult adulttwins twinsreared rearedapart apartand andreared rearedtogether. together.Journal Journal Personality and Social Social Psychology, Psychology,55, 55,950-957. 950-957. Pennebaker, J. J. W. W.(1982). (1982).The Thepsychology psychologyofofphysical physicalsymptoms. symptoms.New New York: Springer-Verlag. York: Springer-Verlag. Pennebaker,

34 CHAPTER 34

PERSONALITYAND ANDIlEALm HEALTH PERSONALITY

915 915

Persky, V. W., Kempthorne-Rawson, J., & Shekelle, R. B. B. (1987). of (1987). Personality and risk of cancer: 20-year follow-up of the Western Electric Study. Psychosomatic Psychosomatic Medicine, Medicine, 49, 435-449. 49, Seligman, M. M.E. E.P.P.(1987). (1987).Explanatory Explanatorystyle styleand andillness. illness.Journal Journal Personality, C, & Seligman, ofofPersonality, Peterson, c., 55, 55, 237-265. Peterson, c., C, Seligman, M. E. P., & Vaillant, G. E. (1988). Pessimistic explanatory style is a risk factor illness: A A thirty-five-year thirty-five-year longitudinal longitudinalstudy. study.Journal Journalof ofPerson Personfactor for physical illness: ality and Social Social Psychology, Psychology,55, 55,23-27. 23-27. Price, V. A. (1982). Type A and practice. NewNew York: A behavior behaviorpattern: pattern:AAmodel modelfor forresearch research and practice. York: Academic Press. Science,194, 194,10131013Rabkin, J. G., & Struening, E. H. (1976). Life events, stress, and illness. Science, 1020. Ragland, D. R. & Brand, R. J. (1988). Type A behavior and mortality from from coronary heart 318, 65-69. disease. New England EnglandJournal JournalofofMedicine, Medicine, 318, 65-69. Rask, M. K. in in thethe cynical hostility and and health relationship. Unpub K.(1990). (1990).Mediating Mediatingvariables variables cynical hostility health relationship. Unpublished doctoral dissertation, University of Utah, Salt Lake City. stsdottir, S. (1984). On the relationship of hardiness to the Type A Rhodewalt, Rhodewah, F., & Agu Agustsdottir, behavior pattern: Perception of life events versus coping with life events. Journal of of ResearchininPersonality, Personality,18,18, 212-223. 212-223. Research Rhodewalt, F., & Zone, J. B. (1989). Appraisal of life change, depression, and illness in Journalof ofPersonality Personalityand andSocial Social Psychology, 81-88. hardy and nonhardy women. Journal Psychology, 56, 56, 81-88. Rose, R. J., Koskenuvo, M., Kaprio, J., Sarna, S., & Langinvainio, H. (1988). Shared genes, personaUty: Data from 14,288 adult Finnish 10shared experiences, and similarity of personality: twins. Journal Journalof of Personality Personalityand andSocial Social Psychology, 161-171. Psychology, 54,54, 161-171. The interview interview method method of of assessment assessment ofthe of the coronary-prone coronary-prone behavior behavior Rosenman, R. H. (1978). The pattern. In T. M. Dembroski, S. M. Weiss, J. L. Shields, S. G. Haynes, & M. Feinleib (Eds.), Coronary-prone Coronary-pronebehavior behavior(pp. (pp.55-70). 55-70).New NewYork: York:Springer-Verlag. Springer-Verlag. hardiness, Wiebe, D. D. J., J., Fillingim, Fillingim, R. R. B., B.,& &Shay, Shay,K. K.A. A.(1989). (1989).Life Lifeevents, events,fitness, fitness, hardiness, Roth, D. L., Wiebe, of and health: A simultaneous analysis of proposed stress-resistance effects. Journal of Personalityand andSocial SocialPsychology, Psychology, 136-142. 57,57, 136-142. Personality Schachter, S., Silverstein, B., Kozlowski, E. T., Herman, L. P., & Liebling, B. (1977). Effects Effects of Experimental Experimental Psychology: Psychology: of stress on cigarette smoking and urinary pH. Journal of General,106, 106,24-30. 24-30. General, Scheier, M. F., & Carver, C. S. (1985). Optimism, coping, and health: Assessment and implications of generalized outcome expectancies. Health Health Psychology, Psychology, 4,4,219-247. 219-247. Scheier, M. F., F., & & Carver, Carver,C. C.S.S.(1992). (1992).Effects Effectsofofoptimism optimismon onpsychological psychologicaland andphysical physicalwell well16,16, being: Theoretical Theoretical overview overview and and empirical empirical update. update. Cognitive CognitiveTherapy Therapyand andResearch, Research, 201-228. Scheier, M. F., Matthews, K. A., Owens, J. F., F., Magovern, Magovem, G. J., Lefebvre, R. c., C, Abbott, R. A., & Carver, C. S. (1989). Dispositional optimism and recovery from coronary artery bypass surgery: The beneficial effects effects on physical and psychological well-being. Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, 1024-1040. Journal 57,57, 1024-1040. stress: Divergent Scheier, M. F., Weintraub, J. K., & Carver, C. S. (1986). Coping with stress: Journalof ofPersonality Personalityand andSocial SocialPsychology, Psychology, strategies of optimists and pessimists. Journal 51, 1257-1264. stressSchmied, L. A., & Lawler, K. A. (1986). Hardiness, Type A behavior, and the stress Journalof ofPersonality Personalityand andSocial Social Psychology, illness relationship in working women. Journal Psychology, 51, 1218-1223.

916 916

WIEBE WiEBE AND AND SMI'I1f SMITH

Schroeder, D. D. H., & & Costa, Costa, P. P. T., Jr. Jr. (1984). (1984). Influence Influence of of life life event stress stress on on physical physical Journal of flaws? Journal illness: illness: Substantive Substantive effects effects on on methodological methodologicalflaws? ofPersonality Personalityand andSocial Social Psychology, Psychology,46, 46,853-863. 853-863. Selye, H. (1956). The The stress stressof oflife. life.New NewYork: York:McGraw-Hill. McGraw-Hill. Shekelle, Shekelle, R. R. B., B., Gale, Gale, M., M., & & Norusis, Norusis, M. M. (1985). (1985). Type Type A A score score (Jenkins (Jenkins Activity Activity Survey) and risk of recurrent coronary heart disease in the Aspirin Myocardial Infarction Study. American 221-225. AmericanJournal JournalofofCardiology, Cardiology,56,56, 221-225. Shekelle, Shekelle, R. R. B., Gale, Gale, M., M., Ostfeld, Ostfeld, A. A. M., M., & & Paul, Paul, O. O. (1983). (1983). Hostility, Hostility, risk of of coronary coronary heart disease, PsychosomaticMedicine, Medicine,45,45,109-114. 109-114. disease, and and mortality. mortality. Psychosomatic Shekelle, Shekelle, R. R. B., B., Raynor, Raynor, W. J., J., Ostfeld, Ostfeld, A. A. M., M., Garron, Garron, D. D. c., C, Bieliauskas, Bieliauskas, L. L. A., A., Liu, Liu, S. S. C., C, Maliza, Maliza, c., C, & & Paul, Paul, O. O. (1981). (1981). Psychological Psychological depression depression and and 17-year 17-year risk risk death death from from 1 17-125. cancer. PsychosomaticMedicine, Medicine,43,43, 117-125. cancer. Psychosomatic Shekelle, Shekelle, R. R. B., B., Vernon, Vernon, S. S. W., & & Ostfeld, Ostfeld, A. A. M. M. (1991). (1991). Personality Personality and and coronary coronary heart disease. Psychosomatic PsychosomaticMedicine, Medicine,53,53,176-184. 176-184. disease. pattern: Additional Additional avenues avenues Smith, T. T. W. (1989). (1989). Interactions, Interactions, transactions, transactions, and and the the Type Type A A pattern: Smith, in the the search search for for coronary-prone coronary-prone behavior. behavior. In In A. A. W. Siegman Siegman & & T. T. M. M. Dembroski Dembroski in or coronary-prone (Eds.), (Eds.), In search searchffor coronary-pronebehavior behavior(pp. (pp.91-116). 91-116).Hillsdale, Hillsdale,NJ: NJ:Erlbaum. Erlbaum. Smith, T. T. W. (1992). (1992). Hostility Hostility and and health: health: Current Current status status of of aa psychosomatic psychosomatic hypothesis. hypothesis. Smith, Health Psychology, Psychology,11,11,88-96. 88-96. Health Smith, T. T. W. (1994). (1994). Concepts Concepts and and methods methods in in the the study study of of anger, anger, hostility, hostility, and and health. heahh. In Smith, In A. W. Siegman Siegman & & T. T. W. Smith Smith (Eds.), (Eds.), Anger, Anger, hostility, hostility, and andthe theheart heart(pp. (pp.23-42). 23-42). A. Hillsdale, NJ: NJ: Erlbaum. Hillsdale, Erlbaum. Smith, Smith, T. T. W., & & Allred, AUred, K. K. A. A. (1989). (1989). Blood Blood pressure pressure responses responses during during social social interaction in Journalof ofBehavioral BehavioralMedicine, Medicine, 135-143. high and and low low cynically cynically hostile hostile males. 12,12, 135-143. high males. Journal T. W., W., & & Anderson, Anderson,N. N.B. B.(1986). (1986).Models Modelsof ofpersonality personalityand anddisease: disease:An Aninteractional interactional Smith, T. Smith, approach and Social approach to to Type Type A A behavior behavior and andcardiovascular cardiovascularrisk. risk.Journal JournalofofPersonality Personality and Social Psychology,50, 50,11166-1173. Psychology, 166-1 173. Smith, Smith, T. T. W., Baldwin, Baldwin, M., M., & & Christensen, Christensen, A. A. J. J. (1990). (1990). Interpersonal Interpersonal Influence Influence as as active active Psychophysiology, 27, coping: coping: Effects Effects of of task task difficulty difficulty on on cardiovascular cardiovascular reactivity. reactivity. Psychophysiology, 27, 429-437. 429-437. rela Smith, Smith, T. T. W., & & Christensen, Christensen, A. A. J. J. (1992). (1992). Cardiovascular Cardiovascular reactivity reactivity and and interpersonal interpersonal relations: Clinical Psychol tions: Psychosomatic Psychosomatic processes processes in insocial socialcontext. context.Journal JournalofofSocial Socialand and Clinical Psychology, ogy, 11, 11, 279-301. 279-301. Smith, Smith, T. T. W., W., & &Frohm, Frohm,K. K.D. D.(1985). (1985).What's What'sso sounhealthy unhealthyabout abouthostility? hostility?Construct Constructvalidity validity and psychosocial psychosocial correlates correlates of of the the Cook Cook and and Medley Medley Ho Scale. Health Health Psychology, Psychology, and Ho Scale. 4, 503-520. 503-520. 4, T. W., W., O'Keeffe, O'Keeffe, J. J. L., L., Allred, AUred, K. K. D. D. (1989). (1989). Neuroticism, Neuroticism, symptom symptom reports, reports, and and Type Type Smith, T. Smith, A behavior: behavior: Interpretive Interpretive cautions cautions for for the the Framingham Framingham scale. Journalof of Behavioral Behavioral scale. Journal A Medicine, Medicine,12, 12,1-11. 1-11. Smith, T. T. W., & & Pope, Pope, M. M. K. K. (1990). (1990). Cynical Cynical hostility hostility as as aa health health risk: risk: Current Current status status and Smith, and future directions. directions. Journal Journalof ofSocial SocialBehavior Behaviorand andPersonality, Personality, 5, 77-88. 5, 77-88. future Smith, T. T. W., Pope, Pope, M. M. K., K., Rhodewalt, Rhodewalt, F., F., & & Poulton, Poulton, J. J. L. L. (1989). (1989). Optimism, Optimism, neuroticism, neuroticism, Smith, coping, coping, and and symptom symptom reports: reports: An An alternative alternative interpretation interpretation of of the the Life Life Orientation Orientation Test. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 640-648. Psychology, 56,56, 640-648. Test. Smith, T. T. W., Pope, Pope, M. M. K., K., Sanders, Sanders, J. J. D., Allred, K. K. D., & O'Keeffe, O'Keeffe, J. J. L. L. (1988). (1988). Cynical Cynical Smith, D., Allred, D., & hostility at at home home and and work: work: Psychosocial Psychosocial vulnerability vulnerability across across domains. Journal Journal of of hostility ResearchininPersonality, Personality,22,22, 525-548. Research 525-548.

CHAPTER 34 34 CHArTER

PERSONALITY AND AND IIEALnt HEALTH PERSONALITY

917 917

Smith, T. W., & Williams, P. G. (1992). Personality and health: Advantages and Limitations of the Five Factor Model. Journal of of Personality, Personality,60, 60,395-423. 395-423. Somervell, P. D., Kaplan, B. H., Heiss, G., Tyroler, H. A., Kleinbaum, D. G., & Obrist, of P. A. (1989). Psychologic distress as a predictor of mortality. American Journal of Epidemiology, Epidemiology, 130, 1013-1023. Stone, S. S, V., v., & Costa, P. T., Jr. (1990). Disease-prone Disease-prone personality personality or or distress-prone distress-prone personal personality? The role of neuroticism in coronary heart disease. In H. S. S. Friedman (Ed.), Personality Personalityand anddisease disease(pp. (pp.178-200). 178-200).New NewYork: York:Wiley. Wiley. Strack, S., Carver, C. S., & Blaney, P. H. (1987). Predicting successful successful completion of an aftercare aftercare program following treatment treatment for for alcoholism: The role of dispositional opti optiof Personality Personalityand andSocial SocialPsychology, Psychology,53, 53,579-584. 579-584. mism. Journal of Internal-external expectancies expectancies and and health-related health-related behaviors. behaviors.Journal Journal Strickland, B. R. (1978). Internal-external of . of Consulting Consulting and and Clinical ClinicalPsychology, Psychology,46,46,1 192-1211 1192-1211. E.. c., C, & Williams, R R.. B., Jr. (1989). Situational determinants of cardiovascular and Suarez, E men. Psychosomatic PsychosomaticMedicine, Medicine,51, 51,404-418. 404-418. emotional reactivity in high and low hostile men. of Suls, J., & Rittenhouse, J. D. (1987). Personality and health: An introduction. Journal of Personality,55, 55,155-167. 155-167. Personality, Suls, J., & Rittenhouse, J. D. (1990). Models of linkages between personality and disease. Personalityand anddisease disease(pp. (pp.38-64). 38-64).New NewYork: York:Wiley. Wiley. In H. S. Friedman (Ed.), Personality Suls, J., & Sanders, G. S. (1989). Why do some behavioral styles place people at coronary search of of coronary-prone coronary-prone risk? In A. W. Siegman & T. M. Dembroski (Eds.), In search behavior behavior (pp. (pp. 1-20). Hillsdale, NJ: Erlbaum. Tellegen, A., Lykken, D. T., Bouchard, T. J., Jr., Wilcox, K. J., Segal, N. L., & Rich, S. Personality similarity similarity in in twins twins reared reared apart apart and and together. together. Journal Journalof of Personality Personality (1988). Personality Psychology,54, 54,1031-1039. 1031-1039. and Social Psychology, Psychoanalysisand andbehavior behaviortherapy. therapy.New NewYork: York:Basic BasicBooks. Books. Wachtel, P. L. (1977). Psychoanalysis Watson, D. (1988). Intraindividual and interindividual analyses of positive and negative activities.Journal Journal affect: Their relation to health complaints, perceived stress, and daily activities. of Personality Personalityand andSocial SocialPsychology, Psychology,54, 54,1020-1030. 1020-1030. of Watson, D., & Clark, L. A. (1984). Negative Negative affectivity: affectivity: The Thedisposition disposition to toexperience experience aversive aversive emotional states. states. Psychological PsychologicalBulletin, Bulletin,96, 96,465-490. 465-490. Watson, D., & Pennebaker, J. W. (1989). Health complaints, stress, and distress: Exploring the central role of negative affectivity. affectivity. Psychological PsychologicalReview, Review,96, 96,234-254. 234-254. Wiebe, D. J. (1991). Hardiness and stress moderation: A test of proposed mechanisms. of Personality Personality and andSocial SocialPsychology, Psychology,60, 60,89-99. 89-99. Journal of Wiebe, D. J., & McCallum, D. M. (1986). Health practices and hardiness as mediators in Health Psychology, Psychology, 5,5,425-438. 425-438. the stress-illness relationship. Health Wiebe, D. J., & Williams, P. G. (1992). Hardiness and health: A social psychophysiological perspective on stress and adaptation. Journal of of Social Social and Clinical Clinical Psychology, Psychology,11, 11, 238-262. Williams, P. W., W., & & Smith, Smith,T. T. W. W.(1990). (1990).Hardiness Hardinessand andneuroticism: neuroticism: Overlapping Overlapping Wiebe, D. J., Williams, constructs?Paper Paperpresented presentedatatthe the1 11th annualmeeting meetingofofthe theSociety SocietyofofBehavioral Behavioral 1th annual constructs? Medicine, Chicago. (1992). Coping processes as mediators of the Williams, P. W., Wiebe, D. J., & Smith, T. W. (1992). relationship between hardiness and health. Journal of of Behavioral Medicine, Medicine, 15, 237-255. Barefoot, J. C, B. (1985). The health consequences of Williams, R. B., Jr., Barefoot, c., & Shekelle, R. B. of hostility. In M. A. Chesney & R. H. Rosenman (Eds.), Anger and and hostility hostilityin incardiovas cardiovascular cular and and behavioral behavioraldisorders disorders(pp. (pp.173-185). 173-185).New NewYork: York:Hemisphere. Hemisphere.

918 918

WiEBE AND ANDSSMITH MI'DI WIEBE

& Whalen, Whalen, R. R. (1980). (1980). Williams, R. B., Jr., Haney, T. L., Lee, K. L., Kong, Y., Blumenthal, J., & Type A behavior, hostility, and coronary atherosclerosis. Psychosomatic Psychosomatic Medicine, Medicine, 42, 42, 539-549. Zonderman, A. B., Costa, P. T., & McCrae, R. R. (1989). Depression as a risk for cancer nationallyrepresentative representativesample. sample.JAMA, JAMA,Journal Journalo/the of theAmerican American Medical mortality in aanationally Medical Association, Association, 262, 262,1191-1195. 1191-1195.

CHAPTER 35 CHAPTER 35

PERSONALITY DIAGNOSIS AND PERSONALITY DISORDERS LESLIE C. MOREY VANDERBILT VANDERBILT UNIVERSITY UNIVERSITY

L INTRODUCTION L INTRODUCTION

The conceptualization of personality disorder represents a particular challenge to research in the areas of personality and abnormal psychology. In part, part, this this difficulty difficulty from the the puzzling puzzling and and inconsistent inconsistent behavior behaviorof of individuals individualsbearing bearingthis thisdiagno diagnostems from but it also reflects the the vague and and ill-defined nature of the the very concept of sis, but of the very definition of personality personality disorder. As a further complication, the disorder has undergone changes in recent years. years: of competing competing approaches approaches for for the the understanding understanding At present, there are a number of and representation of personality disorder. Some of these models are tied to specific otherare are more moredescriptive descriptivein innature. nature.This Thischapter chapterprovides provides etiological theories, while other conceptuaUzing an overview of the major approaches that have been proposed for conceptualizing In addition, important controversies in in the the area will be explored, these disorders. In including issues pertaining to the diagnosis, temporal and cross-situational stability, and validity of these concepts. n. THE THE IIISTORY HISTORY OF OF PERSONALITY PERSONALITY DISORDER DISORDER ll.

The concept of personality disorder likely has a history as lengthy as the conceptual The ancient appreciation of undesirable and and maladaptive history of personality. The character traits was apparent in the works of many classical dramatists and authors H ANDBOOK OF HANDBOOK OF PERSONA.LITY PERSONALITY PSYCHOLOGY PSYCHOLOGY

919 919

COPYRIGHT C © 1997 BY BY ACADEMIC PlESS PRESS.. COmlGHT ALL liGHTS OF IEPRODUCllON IN ANY FORM RESUVED.

ALL RIGHTS OF REPRODUCHON IN ANY FORM RESERVED.

920 920

LESLIE C. CMoREY LESUE MOREY

as well well as as in inthe thedoctrine doctrineof ofthe thefour fourhumors humorsand andtheir theircorresponding correspondingtemperaments. temperaments. Modern psychiatric concepts of personality disorder probably had their roots in Pinel's concept of manie manie sans delire, delire, or madness without loss of reason. These individuals were thought to typically present with disordered behavior and emoemo tional instability, but with no corresponding loss of contact with reality. Prichard further expanded upon this concept by elaborating the features of what he called moral insanity insanity (Mack, (Mack, 1975). 1975).Prichard Prichardemphasized emphasizedthat thatthese thesecases casesof of insanity insanitydid did deterioration of the the intellectual faculties, faculties, but but rather rather were were impaired impaired not suffer from a deterioration habits." Furthermore, these impairments were seen as in "feelings, temper, or habits." "natural"behaviors, behaviors,rather ratherthan thanasasqualitatively qualitativelydistinct distinctoror"unnatural" "unnatural" variations of "natural" behaviors. In proposing this idea. Prichard greatly expand the boundaries of the concept of insanity, which under his writings included current notions such as personaUty disorder and substance dependence. Prior to this time, conduct of this personality sort would have largely been considered a moral shortcoming, to be dealt with by religious and/or legal authorities. 19th century, century,Koch Koch(1891) (1891)introduced introducedthe theterm term"psycho "psychoToward the end of the 19th affliction pathic" as a personality descriptor. This personality style tended to be an affliction of the community, typically criminal or other antisocial behavior. This label became popular within European psychiatric schools, and many writers of the time specuspecu lated that the psychopathic character resulted from some abnormal hereditary disposition. Emil Kraepelin, the preeminent nosographer of his day, described different varieties of psychopathic personalities (e.g., "born criminals") that he felt represented individuals in undeveloped stages of psychosis, which were largely constitutional in origin (Kraepelin, 1902). Similarly, Birnbaum and Gruhle also proposed genetic mechanisms underlying such character defects (Schneider, 1923/1958). Schneider's approach to personality disorder was noteworthy in that his Kurt Schneider'S concept was both broader and more precisely demarcated from other forms of of abnormalpersonal personalmental disorder. Schneider (1923/1958) described two concepts, abnormal psychopathicpersonality. personality.Abnormal Abnormalpersonality personalitywas wasdefined definedasas ity and psychopathic aa variation conceived range of average personality. variationupon uponan anaccepted acceptedyet yetbroadly broadly conceived range of average personality. The variation variation may may be be expressed expressed as as an an excess excess or or deficiency deficiency of of certain certain personal The personal and whether whether this this isisjudged judgedgood goodor orbad badisisimmaterial immaterialtotothe theissue. issue.The The qualities and qualities saint and and the the poet poet are are equally equally abnormal abnormal as as the the criminal. criminal. All All three three of of them them fall fall saint as we weconceive conceive itit so sothat that all all persons personsof of outside the the range range of of average average personality personality as outside note may may be be classed classed as as abnormal abnormal personalities. note personalities.

(Schneider, 1923/1958, pp. 2-3) (Schneider, 192311958, pp. 2-3)

Schneider's psychopathic psychopathicpersonality personalityconcept concept was was analogous analogous to to what what isis cur curSchneider'S inreference reference to to was used used in rently called personality disorder (the term "psychopathic" was "psychopathological," rather than referring to antisocial behavior). For Schneider, abnormal personality becomes personality disorder when it either (1) causes a suffer because of his or her abnormality abnormaUty or (2) causes the community to person to suffer suffer because of one's abnormality. Thus, Schneider's definition of disordered

CHAFfER CHAPTER 3S 35

PERSONALITY DIAGNOSIS AND DISORDERS PERSONALITY

921

personality was based upon two criteria, one being statistical rarity and the other being the impact of the personality personaUty upon the person and upon society. In doing so, Schneider sought to remove the "moral judgment" implicit in definitions such as that offered offered by Prichard. Furthermore, Schneider was was dissatisfied dissatisfied with the biological offered by investigators such as Kraepelin. According to Schneider, definition offered Abnormal (psychopathic) personality may apparently be based upon some spe specific bodily condition the antecedents antecedents are are not not morbid morbid organic organicprocesses. processes.They They condition but but the are morphological and functional variations which it would strain reasonable conjecture to suppose are not in some way formative of the condition. The associated psychic abnormalities as morbid morbid abnormalities cannot cannot therefore therefore be be regarded regarded rightly rightly as phenomena and there seems to be no logical ground for calling the resultant personalities sick ones. (Schneider, pp. 9-10) (Schneider,192311958, 1923/1958, pp. 9-10)

Schneider identified identified a number of distinct types of personality which were precursors to many contemporary concepts. His classification classification of 10 psychopathic personaUty types provided the first widely used descriptive scheme for personality personality disorder in psychiatry (Vaillant, 1987). Schneider's personality types have been found to have significant significant relationships to current personality disorder concepts (Standage, 1986). In fact, certain writers have proposed that the few Schneiderian types which do not have direct contemporary equivalents, such as hyperthymic personality, should also also be be included included due due to to their their theoretical theoretical importance importance and and clini clinipersonality, should cal utility. A. A. Early Early Psychoanalytic Psychoanalytic Models Models of of Personality Personality Disorder Disorder

Freud's 1908 paper entitled "Character and Anal Eroticism" stimulated thinking about the development of personality traits and development, in contrast to the development of symptomatic features with which psychoanalysis had been primarily concerned prior to that time. However, Freud's interests along this line seemed limited to identifying identifying derivatives of instinctual drives, and his writings on broader functioning were not extensive. In contrast, JJung forms of personality functioning ung paid particular attention to personality and individual differences, most notably in his well-known personality typology (lung, (Jung, 1923). Jung identified identified certain dichotomous attitude personality function types which accounted for important individual differences, with types and function the most fundamental introverted and and fundamental of these involving the difference difference between the introverted extroverted attitude types. Interestingly, despite Jung's close association with the extroverted fundamental differences differences were the developmental theory of Freud, he felt that such fundamental likely to be inborn, writing, "two children of the same mother may exhibit contrary attitudes at an early age, though no change in the mother's attitude can be demon demonimporstrated. Although nothing would induce me to underrate the incalculable impor tance of parental influence, this familar experience compels me to conclude that the decisive factor factor must be looked for in the disposition of the child" (Jung, 1923, p. 415).

LESLIE MOREY LESLIE C. CMoREY

922 922

Of Freud's followers, Franz Alexander of particular particular Of Alexander and Wilhelm Reich were of importance importance in the development development of of psychoanalytic concepts of of personality personality disorder. disorder. Alexander Alexander drew the distinction distinction between symptomatic symptomatic neuroses neuroses and character character neuro neuroses (i.e., personality personaUty disorder) on the basis of of the primary primary mode of of instinct gratifica gratification and the type of of conflicts conflicts typically noted noted with these types of of patients. According According to Alexander, instinctual gratification could be differentiated gratification differentiated into the autoplastic autoplastic mode, where the locus of of gratification gratification was internal to the patient patient and impulses were were gratified gratified symbolically, symboUcally, and the alloplastic alloplastic mode, where gratification gratification results primarily from external objects and impulses were acted out. The former former mode was character characterfrom of the symptomatic neuroses, while the latter defined defined personality disorder. istic of Among Among the psychoanalytic theorists, Wilhelm Reich also was one of of the most important important writers in the area of of personality and character. Reich extended extended Freud's Freud's area of of character character formation formation and delineated delineated a number number of of character character writings in the area disorders, including among others the impulsive, hysterical, and masochistic types. Reich's discussions discussions were central in the development development of of the concept concept of of character character opposed to ego dystonic) disturbances. In early psycho psychodisorders as ego syntonic (as opposed psychoanalysis were ego-dystonic ego-dystonic in analytic writings, those disorders amenable to psychoanalysis represented manifestations manifestations of of impulses unacceptable to the ego. that the symptoms represented ego-syntonic experiences were those phenomena phenomena consistent with the In contrast, ego-syntonic of the ego. For For Reich, this distinction could could be used to identify identify the difference difference goals of between neurotic symptoms and character character disorders. In his words, words. between The neurotic symptom is experienced as a foreign body and creates a feeling of The neurotic neurotic character character trait, on on the the other hand, such such as as the the exaggerated exaggerated being ill. The orderliness of of the the compulsive compulsive character character or or the the anxious anxious shyness shyness of of the the hysterical hysterical orderliness character, are are organically organically built built into into the the personality. personality. One One may may complain complain about about character, being shy shy but but does does not not feel feel ill ill for for this this reason. reason. being 1949, p. (W, Reich, Reich, 1949, p. 42). 42). (W.

pointed out that character character represents represents a specific specific way of of being for for an individual, Reich pointed an expression expression of of his or her total past, while a symptomatic disorder disorder generally corresponds to a single experience.

m. APPROACHES TO PERSONALITY DISORDER i n . CONTEMPORARY ApPROACHES A. Standard Nomenclature: The Diagnostic and Statistical Manual of of Mental Disorder

As can be seen from from the preceding history, there have been many widely differing differing conceptual models for for personality personality disorders. Historically, this has led to a bewilder bewilderconceptual of different different classifications classifications of of mental disorder (Menninger, 1963), resulting ing array of Tower of of Babel where mental health workers use different different words words in a conceptual Tower concept and identical words to connote different different concepts to designate the same concept (the term "borderline" is a good example of the latter phenomenon). During the period following following the end of of World World War II, there were at least four four major major competing

CHAPTER CHAPTER3S 35

PERSONALITY PERSONALITYDIAGNOSIS DIAGNOSISAND ANDDISORDERS DISORDERS

923

standardclassification systems for mental disorder (Raines, 1952). In an attempt to standard ize mental health terminology, the American Psychiatric Association published the first edition of of the Diagnostic Diagnostic and Statistical Manual of of Mental Mental Disorder Disorder (DSM-I) in 1952. The second edition of of this manual (the DSM-II) was published pubHshed in 1968 (American and was relatively similar in content and format to its predecessor (American Psychiatric Association, 1968). In 1980, the American Psychiatric Association published the DSM-III, which of the earlier classification systems. These revisions came was a substantial revision of DSMin response to a period of intense criticism of psychiatric diagnosis. The DSM III (American Psychiatric Association, 1980) contained a number of innovations difficulties associated designed to improve the utility of diagnosis. To address the difficulties specific with diagnostic reliability, the DSM-III incorporated the use of relatively specific criteria which a person must meet in order to receive a diagnosis, which led to different diagnosticians than had been noted in the greater agreement between different past (Spitzer, Forman, & Nee, 1979). However, it should be noted that the criteria used in the DSM will not guaran guarantee that all persons with a particular diagnosis will share critical features. Many of the definitions provided in the DSM (including all of of the personality disorder definitions) are set up in a format format where an individual must have a subset of several features in order to be diagnosed. This approach to diagnostic classification classification is called polythetic, meaning that there are no "necessary and sufficient" conditions which polythetic, all persons in a particular diagnostic category must share. Thus, for the nine DSM DSMIV (American Psychiatric Association, 1994) diagnostic features of Borderline Borderiine Per Personality, at least five of of which are required for diagnosis, there are 256 different different criteria combinations that can result in this diagnosis. Such an approach to classifica classification acknowledges the heterogeneity of of personality disorder manifestations, but at the same time makes the development of diagnostic instrumentation quite difficult. difficult. The DSM-IV, consistent with previous major medically oriented nosologies, concepturepresents a categorical approach to taxonomy. In such a system, a person conceptu ally either has a disorder or does not. However, the use of a categorical approach in DSM-IV did not necessarily mandate that all category members would be strictly homogeneous, as the use of of polythetic criteria illustrates. This aspect suggests the adoption of a prototype prototype model of categorization (e.g., Smith & Medin, 1981). Proto Prototype classes may be contrasted to classically defined categories, where class member membership is defined by the presence of necessary and sufficient sufficient conditions; for example, members of the classically defined concept "even numbers" are those numbers evenly divisible by two. Under the assumptions of the prototype model, the set of defining features represents an "ideal type" or best example of class members (i.e., the prototype), and decisions regarding class membership are made on the basis of similarity comparison to this prototype (Smith & Medin, 1981). As a result, some class members are clearly better exemplars of the concept than others (Rosch, 1973); for example, a robin is a better exemplar of the class "bird" than is a penguin, of this discrete category. Among the categories of mental yet each are members of disorders represented in DSM-IV, the personality disorders have been singled out

924 924

LESLIE C. CMoREY LaUE MOREY

as concepts which are best modeled as prototypes (Blashfield, Sprock, Pinkston, & Hodgin, 1985). In fact, some revisions to the DSM were intended to make the personality disorders more representative of a prototype model by removing any single necessary features from the definitions definitions (Widiger, (Widiger,Frances, Frances,Spitzer, Spitzer,&&Williams, Williams, 1988). Thus, the use of prototype categorization allows an introduction of some dimensionality into an essentially categorical personality diagnostic system. For patients may may both bothshare sharethe the DSM-IV DSM-IV diagnosis diagnosisof of Borderline Borderline example, although two patients Personality Disorder, this does not imply that they are equally good examples of the category, nor that they will present with the same number or severity of borderline personality features. Another important aspect of the DSM-IV system is the use of a multiaxial approach to diagnosis. Rather than receiving one diagnosis (as was the case prior to the DSM-III), five separate evaluations are to be made. Axis I, which composes most of the instrument, describes the clinical clinical syndromes, syndromes,while whileAxis AxisIIIIisisused usedtoto personality disorders disorders and certain developmental disorders. Axis III reflects code personality conditionsthat thatare arejudged judgedtotobe beininsome someway wayrelated relatedtotoaabehavior behaviordisorder. disorder. physical conditions stressorswhich whichmay mayhave havebeen beenrelated relatedtotothe theonset onset Axis IV includes psychosocial stressors of current difficulties. Finally, Axis V codes the current or recent level of the adaptive functioning. functioning. person's adaptive The multiaxial approach introduced in DSM-III represented what was wasperhaps perhaps one of the most important developments in the recent history of the personality Thisclassification classification system systemattempted attemptedto tomake makeexplicit explicitthe thequalitative qualitativediffer differdisorders. This ences between personality disorders and traditional psychiatric diagnoses such as schizophrenia. In particular, particular,the thepersonality personalitydisorders disorderswere wereseen seenas astrait traitdisturbances; disturbances; personality disorder reflects reflects aa long-standing long-standing personality personality trait trait that is, in this system a personality whichcame cameto tobe beinflexible, inflexible,maladaptive, maladaptive,and andcausally causallyrelated relatedtotosubjective subjective (or traits) which first distress or impairment in functioning. These disturbances were thought to be first evident in childhood or adolescence, continuing without periods of remission or exacerbation throughout adulthood. By implication, the clinical syndromes such as schizophrenia and depression could be differentiated dififerentiated from personality disorders in state (i.e., more temporary) manifestations. A total of 10 trait that they involve state disturbances are defined using operational criteria in DSM-IV. These disorders were arranged into three "clusters" "clusters" or orsuperordinate superordinategroupings groupingsof ofdisorders: disorders:Cluster Cluster A (including Schizotypal, Schizoid, and Paranoid Personality Disorders), Cluster C (Dependent, B (Histrionic, Narcissistic, Antisocial, and Borderline), and Cluster C Avoidant, and Obsessive-Compulsive). However, in practice the conceptual conceptual approach approach taken taken in inDSM-IV DSM-IVfor forseparating separating personality disorders from clinical syndromes continues to be unclear. The distincdistinc diagnosis are areinconsistent inconsistent and andoften oftenpuzzling. puzzling. tions drawn between these two axes of diagnosis For example, dysthymic disorder, characterized by a long-standing, relatively mild contradisdepression, has been thought by some to represent a personality style, in contradis tinction to the mood disorder it is conceptualized as in the DSM-IV. Furthermore, the DSM approach to personality disorder has been criticized for many reasons selection of of the theparticular particularpersonality personalitydisorder disorder aside from conceptual ambiguity. The selection

CuAPl'ER CHAPTER 3S 35 PERSONAUTY PERSONALITYDIAGNOSIS DUGNOSISAND ANDDISORDERS DISORDERS

925 925

constructs has been criticized for for both lack of theoretical coherence (Millon, 1981) and and clinical verisimilitude (Gunderson, 1983). DSM-based DSM-based personality disorder disorder diag diagfound to have relatively low reliability (Mellsop, Vargh Varghnoses invariably have been found ese, Joshua, & Hicks, 1982; Spitzer et aI., al., 1979), among the lowest of any of of the overiap extensively extensively categories in the entire diagnostic manual. The categories also overlap (Morey, 1988b; Pfohl, Coryell, Zimmerman, & Stangl, 1986), such that an individual individual particular personality personality disorder diagnosis has at least a 50% chance of of receiving any particular for another another personality disorder. Finally, the coverage of of personality personality meeting criteria for disorder has been criticized criticized (Peele, 1986), suggesting that there are many patients patients disorder personality disorder disorder who do not meet criteria criteria for for any of of the Axis II disorders. with a personality DSM-III-R revisions indicate that these changes had limited The data addressing the DSM-III-R addressing such shortcomings (Morey, 1988b), and the DSM-IV modifica modificasuccess in addressing tions also seem to have had limited impact on some of of these issues. On the positive side, the approach to personality personality disorder disorder taken in the DSM On of tremendous heuristic importance. For example, the greater greater specificity specificity has been of of personality personality disorder disorder definitions definitions beginning with the DSM-III fostered fostered the develop developof ment of of several assessment assessment devices with which to ascertain ascertain these diagnoses with ment far geater reliability reliability than had been previously previously possible. These efforts efforts made possible far unprecedented amount amount of of research research on the personality disorders (Blashfield & (Blashfield & an unprecedented classification of personality personality disorders is a McElroy, 1987). 1987). Furthermore, the DSM classification reasonable reflection reflection of of the empirical covariation of commonly commonly encountered encountered person personreasonable under the assumptions upon which the DSMs were founded founded ality problems, under (Morey, 1988a). Despite these important aspects, there are shortcomings in the DSM taxonomy of personality disorder disorder that create difficulties difficulties for for all scientific scientific efforts efforts directed at of treatment of of these conditions. Such shortcomings have investigating the etiology or treatment number of of investigators to propose alternative approaches to personality personality led a number disorder conceptualization, conceptualization, many of of them dimensional ones. The following following sections disorder few of of the most promising promising models that have garnered attention in re rereview a few cent years. B. Millon's Biosocial Learning Typological Model

Theodore Millon (1969, 1981) has developed a personality model that roughly parallels the typological organization of of the DSM-III personality disorders, and Millon (1983) was active in the development of of the DSM-III conceptualization conceptualization of of some of of the personality personality disorders. However, Millon's organization reflects a conceptually based approach not mirrored in DSM-III. Millon views personality personality as a habitual coping pattern which represents a particular way of achieving achieving positive positive reinforcements and avoiding punishment. In Millon's system, reinforcement reinforcement lies at reinforcements of the personality model; personality styles are seen as differing differing primarily the core of reinforcement is pursued and obtained. obtained. in the ways in which reinforcement Millon (1969) originally originally identified identified personality coping patterns based upon upon of instrumental behavior. The first dichotomy dichotomy involves the activeactivethree dichotomies of

926

Lam; C. MOREY LESUECMOREY

the initiative in passive dimension, which refers to whether the the individual takes the pursuing reinforcement or alternatively such behavior is largely reactive to environ environmental events. The second distinction involves the pleasure-pain dimension, refer referring to aa tendency to be primarily motivated either by by the the pursuit of pleasure or instead by the avoidance of aversive events. events. The The third third dimension dimensionrefers refersto toself-other, self-other, or the relative importance of subjective and objective modes of experience. the and Millon utilized these dichotomies in in creating aa lO-fold 10-fold (originally an an 8-fold, and revised to be consistent with DSM-III-R) typology of personality styles. These styles may be placed in a "reinforcement matrix" matrix" similar similar to to that that presented presented in in Table Table I. The rows of this matrix address what Millon calls the "instrumental polarity," The the polarity,'* The or the the manner in which a person seeks reinforcement (actively (actively or passively). The columns of this table refer to the "source polarity," or the sources from which the or the reinforcement will primarily be sought. With respect to sources, Millon outlines reinforcement types:independent, independent, seeking reinforcement from self; dependent, or seeking five types: oror seeking reinforcement from thethe self; dependent, or seeking five ambivalent,or oruncertainty uncertaintyand andvacillation vacillationininreinforce reinforcereinforcement from others; ambivalent, detached,aaseeming seemingfailure failuretotoseek seekororatatleast leastexperience experiencereinforce reinforcement seeking; detached, any source; and and discordant, discordant, or oraaseeming seemingreversal reversalofofthe theexperience experienceof of ment by any reinforcement and punishment. According to Millon, the thepersonality personalitydisorders disordersrepresent representthe themaladaptive maladaptivepoles poles of these 10 basic personality styles that may have self-destructive consequences of how the individual is unaware. Under persistent environmental adversity, howwhich the ever, Millon proposes that the the personality functioning will decompensate toward one of three pathological personality syndromes, syndromes, depending upon the person's person's basic sources of reinforcement. According to Millon (1987), those with an an independent style of reinforcement deteriorate to aa Paranoid personality personaUty syndrome, those with TABLE TABLE II Millon's Typological Model of Personality and and Personality Disorders Preferred Preferred source of reinforcement reinforcement Instrumental msirumeniai pattern pattern

Passive'* Passivea

Normal

Mild Severe

Activea Active*

Normal Mild Severe

Dependent

Cooperative

Dependent

Borderline Sociable

Histrionic Borderline

Independent

Ambivalent

Confident

Respectful

Paranoid

Paranoid

Forceful

Sensitive Passive-aggressive

Narcissistic

Antisocial Paranoid

Compulsive

Borderline

Discordant

Self-abasing

Self-defensive Self-defensive Borderline

Antagonistic

Sadistic Paranoid

Detached

Introversive Schizoid

Schizotypal

Inhibited

Avoidant Schizotypal

Note. Adapted from from Millon (1987). Note. The first first type type mentioned mentioned indicates indicates normal normal personality personality pattern; pattern; the thesecond second indicates indicates mild mild personality personality a" The and the the third indicates severe personality personaUty disorder pattern. disorder pattern; and

CHAPTER 3S 35 CHAPTER

AND DISORDERS PERSONALITY DIAGNOSIS AND

927 927

a dependent style will become Borderline, and those with a detached style develop personaUty, while the ambivalent and discordant types may become a Schizotypal personality, coneither Paranoid or Borderline. Millon's typology incorporates a number of con offer a number structs central to personality theory, and its parallels to the DSM offer cUnical advantages. advantages. However, However, itit should should be be noted noted that that the the relationship relationship of research and clinical 1985, 1986; of Millon's ideas to DSM constructs has been controversial (Millon, 1985, Widiger, Williams, Spitzer, & Frances, 1986). C. Interpersonal Interpersonal Models Models of of Personality Personality Disorder Disorder c. Within the broad rubric of psychodynamic theories, a school of thought emerged which viewed personality as shaped by social and cultural circumstances to a greater extent than by biological or instinctual factors. Harry Stack Sullivan proposed a systematic interpersonal interpersonal theory that viewed personality as "the relatively enduring pattern of recurring interpersonal (Sulli interpersonal situations which characterize a human life" (SuUivan, 1953, 1953, p. 1111). 11). Sullivan's ideas were futher futher systematized by Timothy Leary, a young academic psychologist who later pursued other interests. Leary, in appraising the impact of Sullivan, felt that Sullivan "convincingly buried the much-berated remains of descriptive, Kraepelinian, and negatively-value-toned psychiatry, but provided no substitute classification classification system" (Leary, 1957, 1957, p. 10). 10). Leary attempted attempted to provide a classification classification scheme which would allow an interpersonally based appraisal of personality function Interfunction and dysfunction. In his inftuential influential 1957 book book Inter of personal Diagnosis Diagnosis of of Personality, Personality, Leary presented a two-dimensional model of interpersonal behavior in which the variables were arrayed in a circular manner. This circular array is divided into quadrants through the use of two orthogonal affiliation(the (thehorizontal horizontalaxis), axis),and andthe theother otherinvolving involving dimensions, one representing affiliation interpersonal dominance As aa result, result, 16 sections sections of of the the circumplex circumplex dommance (the vertical axis). As could be described that represented different different blends of affiliation affiliation and dominance. Leary's book stimulated a great deal of interest and research in an interper interpersonal circumplex model of personality. Some of this research pointed to various shortcomings of the Leary model (Wiggins, 1982). For example, there seemed to be a noticeable gap between octants PA (Autocratic) and NO (Hypernormal) suggesting that persons in these adjoining octants were more different different than por portrayed by the model. Furthermore, certain variables hypothetically at opposing ends of the circle did not seem to be inversely related, as would be expected under the model. As a result, a number of investigators have proposed revisions to the original Leary formulation. Many of these revisions, such as those proposed by Schutz (1958), Lorr Bishop, and McNair (1965), Wiggins Wiggins (1979), (1979), Kiesler Kiesler (1983), (1983),and and Benja Benjamin (1974), have attracted significant research attention in their own right. A number of authors have argued that an interpersonal taxonomy is crucial for an adequate conceptualization of personality disorder (Benjamin, 1993). McLemore potenand Benjamin (1979) proposed that the interpersonal approach represents a poten tial improvement upon traditional psychiatric classification systems. In their view, the primary advantage of the interpersonal approach to diagnosis is that it is pre-

928 928

LESLIE C. CMoREY MOREY

scriptive as well well as asdescriptive; descriptive;ininother otherwords, words,ititspecifies specifiessteps stepstototake takeininimplement implementing a successful treatment. This system has been seen as particularly relevant as a descriptive scheme for personality disorder (McLemore & & Brokaw, Brokaw, 1987). 1987).Accord According to to McLemore McLemore and and Brokaw, Brokaw, there there are are four four basic basic assumptions assumptions underlying underlying an ing interpersonal approach approach to to the the conceptualization conceptualization of of personality personaUty disorder: disorder: interpersonal 1. signifi1. Deeply ingrained behavioral styles are much more psychologically signifi cant than discrete "symptoms." *'symptoms." This Thisassumption assumptionimplies imphesthat thatthe theinterpersonal interpersonalstyle style tends to be consistent over time, potentially giving rise to varying symptomatic patterns. Thus, Thus, the theinterpersonal interpersonalstyle stylemust mustbe beaddressed addressedififthe thesymptomatic symptomaticpresen presentation is to be permanently ameliorated. 2. Personality disorders reflect disordered relationships with other people. McLemore and Brokaw note that these relationships need not be with people in the immediate social milieu, but may also include relations with individuals no longer present, or even deceased. 3. Effective of Effective treatment of personality disorder requires the interruption of self-defeating self-defeating interpersonal cycles. cycles. As such, the interpersonal diagnostic approach (a) what other people have done to the patient; leads to an understanding of: (a) (b) (6) how others' behavior pulls maladaptive behavior from the patient; (c) how the patient may engineer interpersonal circumstances which may maintain the disor disordered pattern; and (d) what the patient must do to break out of this cyclical maladaptive pattern (Strupp & & Binder, 1984). 4. The establishment of an interpersonal relationship is a prerequisite to personality disorder treatment. It is through through this this relationship relationship that that the the self-defeating self-defeating cycles described described above abovemay maybe beexperientially experientiallyobserved observedby byboth boththerapist therapistand andpatient, patient, allowing for some intervention. It has been noted noted that there there are are several several conceptual and and empirical relationships relationships between DSM personality disorder categories and the interpersonal circumplex model (Frances, 1982; 1982; Kiesler, Kiesler, 1986; 1986;Morey, Morey,1985; 1985;Widiger Widiger&&Kelso, Kelso,1983; 1983;Wiggins, Wiggins, 1982; 1982; Wiggins Wiggins& &Pincus, Pincus,1994). 1994).However, However,the theinterpersonal interpersonaldomains domainsalso alsotaps tapsmany many aspects of behavior not well represented in traditional personality diagnostic syssys tems. In fact, a study conducted by Morey (1985) and replicated by Dejong, DeJong, tems. van den Brink, Jansen, and Schippers (1989) suggested that the DSM personality personaHty disorders maynot not adequately adequately tap tapmany manyfacets facetsof ofinterpersonal interpersonalbehavior, behavior,particularly particularly affiUative components. The interpersonal model of personality is clearly pertithe affiliative perti nent to personality disorder classification and should be considered as a promisnent to personality disorder classification and should be considered as a promis ing alternative. alternative. ing

D. Dimensional Dimensional Conceptualizations Conceptualizations of of Personal Personal Disorder D. Disorder In contrast contrast to to the the categorical categorical models models historically historically proposed proposed within within psychiatry, psychiatry, a In different approach to the conceptualization of personality disorder has somewhat different been most prominent within academic psychology. psychology. This Thisapproach approachviews viewspersonality personality functiondisorder as the extreme of a continuous dimension of normal personality function-

CHAPTER DISORDERS CHAPTER35 35 PERSONALITY PERSONALITYDIAGNOSIS DIAGNOSISAND AND DISORDERS

929

ing. The dimensional perspective has gained popularity in psychology for a number of reasons (Widiger, 1993; Widiger & Frances, 1985). For example, it has been suggested that dimensional dimensional personality models are more precise (Frances, 1982). Furthermore, dimensional dimensional clinical judgments of personality personality disorder are far more reliable than categorical judgments (Heumann & Morey, 1990). Categorical classifi classifications with high overlap such as Axis II of of the DSM (Morey, 1998b) make the identification identification of "pure" diagnostic groups very difficult, difficult, with the resulting subjects probably not representative of the majority of patients presenting in routine clinical practice. Also, in general, psychometric research on personality traits has failed to identify sharp distinctions between adaptive and maladaptive maladaptive personality personality character characteristics (Eysenck, Wakefield, & Friedman, 1983). The dimensional psychometric tradition in modem psychology dates back to the research in individual differences performed in Great Britain by Sir Francis Galton. The British tradition produced a number of important scholars in the area of personality and individual differences. Perhaps the two best known are Hans Eysenck (e.g., Eysenck & Eysenck, 1985) and Raymond Cattell (e.g., Cattell, 1965), each of of whom developed models of personality structure utilizing dimensional statistical methods of factor analysis. Cattell originally identified 16 "source" traits of observed behavior; subse subsewhich he believed operated as the underlying source of quent studies from Cattell's laboratory discovered as many as 9 additional source different maladaptive personality variants could be represented traits. For Cattell, different as different constellations of these source personality traits, each of of which tend to share relatively low scores on a trait described as "ego strength." Eysenck has proposed that personality is best modeled as consisting of three independent dimensions, which he has label neuroticism, psychoticism, psychoticism, and extrover extroversion-introversion. sion-introversion. The neuroticism trait refers to a tendency to be emotionally reactive; the psychoticism trait involves emotional independence (i.e., impersonal, unempathic); and the extroversion trait refers to sociability sociability and venturesomeness. In contrast to Cattell's personality dimensions, Eysenck maintains that these major dimensions are uncorrelated in the general population, such that each reflects an independent aspect of personality style. In his theoretical writings, Eysenck has proposed that these traits reflect heritable biological mechanisms, with neuroticism representing a highly reactive autonomic nervous system, extroversion involving low resting cortical activation, and psychoticism involving hormonal mechanisms (Eysenck & Eysenck, 1985). For Eysenck, personality disorders are viewed as extreme positions along these personality dimensions. In his words, "the concept of of personality disorders is not seen as a categorical diagnosis, but as behavior characterized by the confluence of three major dimensions of personality, determined in its precise operation by of personaUty, the predominance of one or another of these, and shading gradually and impercepti imperceptibly into more normal types of behavior" (Eysenck, 1987, p. 215). Because of of this emphasis upon the dimensionality of of behavior, Eysenck has been a leading critic of the traditionally categorical psychiatric approach to personality disorder (Eysenck et aI., al, 1983).

930 930

LESLIE MOREY LESLIE C. CMoREY

Another dimensional approach to personality conceptualization which has recently attracted research attention is the the so-called so-called "five-factor" "five-factor" model model of personal personality. The five-factor model, proposed initially by Tupes and Christal (1961) and refined by Norman (1963), has a number of elements in common other popular refined dimensional approaches; in fact, it resembles an integration of the Eysenck model, the higher order factors of Cattell's theory, and the two interpersonal dimensions factors described by Leary (1957) and his successors as described earlier. The five factors 1984): Neuroticism, Neuroticism, characterized may be described as follows (McCrae & Costa, 1984): self-satisfied nature; by worry, insecurity, and self-pity, as opposed to a calm and self-satisfied Extroversion, referring referring to a sociable and affectionate affectionate person in contrast to a sober, Extroversion, Openness, implying an imaginative, independent personality reserved individual; Openness, Agreeableness,characterized characterizedby byaatrust trustcontrasted to a conforming, orderly nature; Agreeableness, helpful attitude in contrast to a suspicious, exploitative orientation; and Consciing, helpful Consci entiousness, denoting denoting aawell-organized, well-organized, careful, careful, disciplined disciplined person person as as opposed opposed to to aa entiousness, careless, weak-willed individual. As pointed out by Costa and McCrae (1986), (1986), the utility and robust nature of of the five-factor model have been supported in a number of research studies. In addition, there is substantial evidence to suggest that these five factors reflect enduring characteristics which persist throughout much of adult life (McCrae & Costa, 1984). It has been suggested by McCrae and Costa that the personality personaUty disorders may be well represented by thefive factors, with all such disorders tending five to be high on Neuroticism and specific disorders reflecting reflecting particular constellations of the remaining four four factors. Widiger, Trull, Clarkin, Sanderson, and Costa (1994) specify hypothetical patterns for each of the DSM personality disorders on the five factors and their subfacets, with many of the disorders displaying facets of five of Neuroticism. The robustness of thesefive five factors of personality across many diverse perempirical studies highlights their potential as a useful model of the structure of per sonality. E. Assessment Assessment and and Diagnosis Diagnosis of of Personality Personality Disorder Disorder E. As mentioned earlier, the introduction of the DSM-III did not ameliorate problems personaUty disorder diagnosis, which continued to be among the associated with personality al., 1982). 1982). Some authors have least reliable of all mental disorders (Mellsop et a!., speculated that possible explanations for this relative lack of reliability may in part be attributed to idiosyncratic interviewing styles by clinicians. Other studies suggested that systematic biases, perhaps related to demographic features such as the gender or race of the client, may be occurring in the assignment of certain Ochoa, 1989; 1989;Warner, Warner, 1978). 1978).For Fornumerous numerous personality disorder diagnoses (Morey & Ochoa, reasons, it soon became clear that advances in diagnostic instrumentation for the personality disorders were required if substantive research was to be performed. performed. As noted by several reviews, this need led to the development of several new diagnostic techniques for for use with personality disorders. In addition, a number of of the more time-honored personality assessment instruments have been adapted

CHAPTER3S 35 PERSONAUfY PERSONALITYDIAGNOSIS DIAGNOSISAND ANDDISORDERS DISORDERS CHAPTER

931 931

for use with this aim. aim. The reader is encouraged to peruse the several reviews of instrumentation in this area, such as those by Widiger and and Frances (1987), J. H. Reich (1987), or Zimmerman (1994). There are a number of thorny issues to resolve in evaluating personality personaUty disorder assessment instruments. The definition of criterion groups for research is a perennial problem, since there are no well-validated markers for these concepts. For the most part, the described instruments follow follow some approximation of the DSM format since it is currently the standard nomenclature in mental health. identified. However, the DSM concepts themselves are ill defined and unreliably identified. difficult to interpret: As such, negative results in assessment validity studies are difficult personaUty disorder constructs, Are the instruments not useful for diagnosing DSM personality the constructs themselves not not useful? useful? As As noted noted by by Stangl, Stangl, Pfohl, Pfohl, Zimmerman, Zimmerman, or are the Bowers, and Corenthal, (1985), this creates a "catch-22" for diagnostic researchers that is not easily resolved. difficulty besetting diagnostic research with the personality personaUty disorders Another difficulty involves diagnostic overlap. It has been well established that there is a great deal of overlap among the personality personaUty disorders (Morey, 1988b; Pfohl et al., of aI., 1986); consequently, it is typical to find patients presenting with features of a number of of different personaUty disorders. As a result, identifying a relatively pure "criterion different personality identifying group" in this area becomes exceedingly difficult. difficult. Furthermore, even if such groups are obtained, they may be somewhat artificial in that they are not representative of usual patient populations. This high diagnostic overlap suggests that differentiating differentiating difficult for any instrument. among these disorders will be very difficult A final consideration in assessing these conditions involves modes of data A Self-report data collection coUection is subject to distortion arising from several gathering. Self-report individsources, such as impression management efforts, and personality disordered individ uals are notorious in this regard. For example, when "frequent lying" is one of the definitional criteria (as it is for antisocial personality), personaUty), it is safe to assume that self selfdefinitional report information provided by these individuals may be suspect. Futhermore, individuals with personality disorders often do not come into contact with mental health professionals in an entirely voluntary fashion; typically, there are situational elements that provide motivation for distorting their self-presentation. Hare (1985) gives the example of one highly psychopathic prisoner who had his own MMPI scoring keys and an MMPI research library that he used as the basis for a consulting service of "manufacturing" profiles for other inmates. Such concerns have led some reUance upon self-report self-report information in establishing researchers to caution against a reliance personaUty disorder diagnoses. personality selfHowever, the use of structured interviews, touted as an alternative to self judgreport data that may circumvent such distortions through the use of clinical judg ment, is also problematic. A A primary consideration is that many of the traits which must be identified to establish personality disorder diagnoses may defy direct inin quiry. For example, one of the DSM-III criteria for Paranoid Personality Disorder involved the lack of a true sense of humor; it is not easy to imagine a structured interview question which would yield meaningful information about this attribute.

932 932

LESLIE C. MOREY

In contrast to many of of the Axis I disorders, diagnosis of of personality disorder does not solely involve gathering information about the patient, but also entails developing information about developing a sense of the patient as a person. The constraints imposed by the comparatively of constraints imposed comparatively rigid format format of of structured structured interviews may in fact be a hindrance hindrance to such efforts. efforts. Another Another formidable formidable problem problem with such interviews is that they are based upon upon the assumption assumption that personality personality judgments can be made with uniform uniform ease by mental assumption that has long been suspect (e.g., Taft, Taft, 1955). health professionals, an assumption combinaIn general, it is advised that personality disorder assessment involve some combina of self-report self-report and interview interview contact contact in order to maximize available information. information. tion of IV. NATURE OF PERSONALITY DISORDER: DISORDER: rv. THE NATURE CURRENT CONTROVERSIES majority of of instances, the term personality personality disorder disorder is employed to refer refer to In the majority DSM a collection of conditions that may not seem to have much in common. The DSM attempt to provide explicit diagnostic criteria for for a number number of of manuals, which attempt unfortunately vague when called upon to provide specific personality disorders, is unfortunately between a definition of superordinate construct. Without such a definition, the line between personality disorder and other personalpersonality other clinical phenomena, as well as that between personal disorder and normal normal personality, becomes quite murky. The remainder remainder of of this ity disorder chapter is devoted to exploring facets of of such a definition. definition. chapter personality disorder disorder denotes a maladaptive maladaptive personality constel constelAs used here, a personality manifest across lation which: (1) is evident early in life; (2) is stable over time; (3) is manifest diverse situations; (4) is an ego-syntonic condition; (5) is particularly evidenced evidenced by interpersonal interpersonal disruption; (6) represents an extreme of of normal variation variation among people; and (7) may arise as a result result of of multiple influences. Each of these assumptions assumptions represents an area of of some controversy, and there is often often minimal evidence that can be offered of these claims. The following offered in support support of following sections briefly briefly examine examine the evidence and the implications of of each of of these assumptions for for the conceptualiza conceptualization of of personality disorder.

A. The Seven Assumptions 1. Are Evident i . Personality Personality Disorders Disorders Are Evident Early Early In in Lif Lifee

There There are a number number of of follow-back follow-back studies which demonstrate that adults with psychological disorders tend tend to have had childhoods childhoods characterized characterized by relatively poor adjustment adjustment (Parker (Parker & & Asher, 1987). Although much of of this literature focused focused poor & Wertheim, 1976), for certain personal personalupon psychotic patients (e.g., Rolf, Knight, & constructs the research research evidence supporting supporting the early appearance of of personality personality ity constructs fairly compelling. Perhaps the most thoroughly thoroughly examined area has problems is fairly personality characteristics. A A well-known study in this domain involved antisocial personality conducted by Robins (1966), who prospectively prospectively followed followed children who exhibited was conducted

35 CHAPTER 35

PERSONALITY PERSONAUTY DIAGNOSIS AND DISORDERS

933 933

antisocial behavior and found that such behavior was predictive of sociopathic actions during adulthood. Subsequent research has also shown that adult antisocial from childhood features, such as aggressiveness and behavior can be predicted from manifest as early eariy as ages 6 to 9 years (Loeber & Dishion, 1983). In fact, stealing, manifest the predictive utility of childhood behavior in predicting adult antisocial personality led to an incorporation of such material into the DSM-III and DSM-III-R diagnostic for this disorder. criteria for For other personality constellations, there is relatively little information information on personalhow early in life such features may be identified. One longitudinal study of personal difficulties were evident by age ity disordered males demonstrated that adaptive difficulties & Vaillant, 1985). There There is is isolated isolated evidence evidence to to suggest suggest that that adult adult deficits deficits 14 (Drake & observed in interpersonal skills, which are prominent in personality disorder, can be observed at an early age (Havighurst, Bowman, Liddle, Matthews, & Pierce, 1962; Robins, of little 1966). Although a childhood behavioral style of shyness and withdrawal is of for later psychopathology (Parker & Asher, 1987), most follow followpredictive value for characterback studies of schizophrenia-spectrum patients indicate that a childhood character (e.g.. Ricks & & Berry, 1970). ized by a withdrawn interpersonal style was common (e.g., such. Schizoid and Schizotypal Personality Disorders might be expected to As such, age. However, research on other demonstrate observable manifestations at an early age. personality disorders disorders constructs constructs is is either either lacking lacking or or yields yields unimpressive unimpressive results. results. For For personality example, the the Fels Fels Longitudinal Longitudinal Study Study (Kagan (Kagan & & Moss, Moss, 1962) 1962) reported reported only only minor minor example, associations between between dependency dependency behaviors behaviors identified identified during during childhood childhood and and associations adult dependency. dependency. adult A major study of personality personaUty stability within the normal range was conducted by Block (1971), who examined personality personality characteristics during early adolescence adolescence (junior high school), middle adolescence adolescence (senior high school), and adulthood (gener (generGunior for 171 ally in their thirties) for 171 subjects studied at the Institute of Human Development at University of California, Berkeley. The average correlation between personality configuration during junior high school and senior high school was .77 and .75, configuration while between senior high school and adulthood the average correlation was .56 1971). These data indicate that, and .54 for men and women, respectively (Block, 1971). on average, there appears to be appreciable personality consistency across the time intervals examined by Block, even for personality within the normal range. As a distinguished result, one would expect that most personality disorders, which are distinguished inflexibility, would begin to be evidenced by early adolescence ifif by rigidity and inflexibility, not earlier. However, there is clearly a significant amount of research which needs to be done in order to better substantiate this assumption. 2. Personality Disorders Are Reasonably Stable over Time 2.

In general, it is assumed that personality disorders tend to be fairly stable over the difficulties which these individuals experience adult years, although the nature of the difficulties may vary across time. The evidence cited above suggests that personality difficulties difficulties in adulthood tend to be evident fairly early in life, supporting the contention that such disorders are stable over time. Furthermore, there is considerable evidence

934 934

LESUE C. MOREY LESUECMOREY

which indicates that personality stability is evident well into older age (McCrae & Costa, 1984). 1984). As such, one should not expect that personality personality disorders will be confined confined to a relatively youthful youthful segment of of the population. population. Unfortunately, Unfortunately, the temporal persistence of of personality disorders has been been further further demonstrated demonstrated in a number number of of studies of treatment samples, which supports supports the clinical observation that these disorders are fairly fairly refractory refractory to treatment. For example. followed a cohort example, Pope, Jonas, Hudson, Cohen, and Gunderson (1983) followed of patients diagnosed as having Borderline Personality Disorder over periods rang rangof from 4 to 7 years; this period included at least one hospitalization and numerous numerous ing from found that 67% of of these patients continued to meet criteria criteria treatment contacts. They found for Borderline Personality Personality at follow-up follow-up despite treatment efforts. efforts. Similarly, McGlas McGlasfor conducted a long-term follow-up study of of roughly 100 inpatients diag diaghan (1983) conducted hospitalizanosed with Borderline Personality Disorder at admission to the index hospitaliza years following following tion. Outcome data on these patients were collected an average of 15 years from this hospitalization. Between 44 and 50% of these individuals the discharge from particuwere diagnosed as Borderline Personality at follow-up, depending upon the particu for diagnosis. Additionally, 46% of of Borderline patients were still lar criteria used for form of of treatment at the time of of follow-up follow-up contact. Such results support support in some form the contention contention that the problems associated associated with personality disorders tend to be be stable over lengthy time periods. Orie One widely believed bit of of clinical lore concerning the personality disorders involves the belief belief that they tend to "burn-out" **bum-out" during later years; that is, the personality personality difficulties difficulties are thought thought to abate sometime during ages 40 to 55. Although Although of patients with personality disorder are often often difficult difficult to con conlongitudinal studies of duct, there is evidence that these problems seem to decrease over time (Glueck & McGlashan, 1986; 1986; Robins, 1966). 1966). However, However, in in McGlashan's McGlashan's longitudi longitudiGlueck, 1968; McGlashan, nal study of of borderline borderline personality, a curvilinear curvilinear trend was observed observed with reemer reemergence of of difficulties difficulties noted after after age 60. According According to McClashan, "many middle middleaged borderline patients develop stable instrumental functioning functioning but not close social and personal relationships. The latter deficit deficit appears not to change with time and may 'haunt' these patients in a symptom-exacerbating symptom-exacerbating fashion fashion as they age and lose their their work capacities and opportunities with the attendant structure and gratifica gratifications" (McGlashan, 1986, p. tions" (McGlashan, 1986, p. 29). 29). In assessing personality personality disorder, the assumption of of temporal stability is an important important consideration. To firmly firmly establish establish these diagnoses, it is important important to make a careful careful determination of the person's life history without being unduly influenced influenced by immediate situational circumstances or concurrent concurrent Axis I psychopathology. With careful careful questioning, a life life pattern pattern will emerge which often often will be telling with respect to personality difficulty. As such, interviews should not neglect historical personality difficulty. historical information information which may be buried in discussions of current precipitating circumstances. buried of current 3. 3.

Personality Disorders Disorders Are Are Reasonably Reasonably Stable Stable across across Personality Diff erent Situations Different Situations

The assumption that people behave consistently across diverse situations is one generated a great deal of controversy within personality psychology (e.g., (e.g.. which has generated

CHAPTER35 35 CHAPTER

PERSONALITYDIAGNOSIS DIAGNOSISAND AND DISORDERS PERSONAUfY DISORDERS

935 935

Epstein & O'Brien, 1985; Kenrick & Funder, 1988; Mischel, 1968, 1983). Walter Mischel, for example, created a furor with his 1968 book on this topic, entitled Personality and and Assessment. Assessment In that volume Mischel pointed out that much psycho psychoPersonality differences in logical research has not supported the contention that individual differences behavior are very stable across situations. As a result, Mischel and others have himself is a much suggested that the situation in which the person finds her or himself stronger determinant of behavior than internal personality dispositions. However, according to the DSM, personality disorders reflect reflect traits which by definition are stable across different different situations, even though this "stability" may definition of ego ego well be maladaptive. Block and Block (1980) have described the construct of modify his or her behavior as resiliency, which refers to an individual's capacity to modify a function of the demands of the environment. Presumably, resilient individuals of different different have the capacity to vary their behavior according to the demands of powerful mediators of of behavior in situations; as such, situations may seem to be powerful resilrelatively resilient individuals. Block and Block have demonstrated that this resil differentiator of people; individuals with personality iency is a stable and consistent differentiator disorders would characterized in their scheme as nonresilient personalities. In other of the adaptation mechanism words, a personality disorder represents a failure of described by Block and Block, whereby the personality is stable to a problematic extent. However, this assumption must not be interpreted to mean that personality disordered individuals will always behave in a manner consistent with their diagno diagnopersonality across sis. In general, it is important to recognize that the stability of personality situations is evident when one considers aggregates of of behavior rather than single behavioral instances (Epstein & O'Brien, 1985). In other words, a trend suggestive of a personality disorder only becomes apparent over a number of of of observations of a particular individual. Even with the most rigid personality personaUty disorder that a clinician is likely to encounter, it is unlikely that we can predict precisely what that individual will do in a given situation with a high degree of accuracy. It is for this reason that judgments of relatively specific behavioral instances, such as immediate dangerous dangerousdifficult to make (Ewing, 1983). The concept of of personality disorder ness, are so difficult implies that such individuals act in a certain way with much greater frequency and in more situations than is expected of most people. However, it does not imply that these people will behave in this manner at all times and in all situations. An important implication of this assumption is that the diagnostician should not assume that the personality-relevant personality-relevant behavior a person demonstrates during an initial interview is necessarily representative of their typical functioning. Unfortu Unfortunately, many explanatory models of mental disorder include theoretical constructs such as transference (e.g., Zetzel, 1956) or stimulus generalization (e.g., Dollard DoUard & Miller, 1950), which imply that the examinee's behavior in such an interview interview situa situainconsistion should be typical of their personality style. However, this conclusion is inconsis tent with a large body of research evidence in the personality field. A A 1-hr interview with a mental health professional professional hardly constitutes a representative sample of of behavior, and a personality disorder diagnosis cannot be assigned with much confi-

936 936

LESLIE MOREY LESLIE C. CMoREY

dence following such a contact. It is essential that such judgments be based on information gathered across repeated observations. Although the gathering of these details is often not feasible for the diagnostician, it is frequently possible to gain such information from relatives or other peers who indeed have had repeated contacts with the examinee. 4. 4. Personality Disorders Are Ego-Syntonic Conditions

personality disor disorThis assumption addresses the experience of the individual with a personality the descriptions descriptions provided provided by by psychoanalytics psychoanalyticssuch suchas asAlexander Alexander der. Consistent with the and Reich, Reich, it it is is assumed assumed that that personality personality disorders disorders involve involve traits traits which which are are an an and essential part part of of the the personality, personality, rather rather than than symptoms symptoms which which are are experienced experienced by by essential the person person as as alien alien to to the the personality. personality. As As such, such, patients patients with with these these disorders disorders are the are often portrayed portrayed as as having having little little insight insight into into the the nature nature of of their their difficulties. difficulties. However, However, often this characterization characterization is is aa bit bit unfair unfair given given that that the the nature nature of of the the difficulties difficulties experi experithis enced by by individuals individuals with with personality personality disorders disorders tend tend to to be be somewhat somewhat different different than than enced those associated associated with with symptomatic symptomatic disorders. disorders. For For example, example, Reich Reich points points out out that that those personality problems problems are are more more readily readily rationalized rationalized than than symptoms: symptoms: *'A reason is is personality "A reason often given given for for neurotic neurotic character character traits traits which which would would be be immediately immediately rejected rejected as as often absurd if if it it were were given given for for symptoms: symptoms: 'he *he is is just just that that way'" way'" (W. (W. Reich, Reich, 1949, 1949, p. p. 43). absurd 43). In other other words, words, aa "way "way of of being" being" is is less less likely likely to to be be seen seen as as an an immediate immediate source source In of distress distress than than would would be be ego-alien ego-alien symptoms symptoms such such as as anxiety anxiety or or obsessions. obsessions. of The assumption assumption that that personality personality disorders disorders are are by by nature nature ego-syntonic ego-syntonic creates creates The host difficulties difficulties for for their their assessment. assessment. For For the the most most part, part, the the presenting presenting problems problems aa host with which which most most individuals individuals with with Axis Axis II disorders disorders present present involve involve experiences experiences with directly pertinent pertinent to to establishing establishing aa diagnosis. diagnosis. For For example, example, aa person person with with panic panic directly disorder is is likely likely to to rapidly rapidly identify identify his his or or her her problems problems as as associated associated with with transient transient disorder episodes of of severe, severe, debilitating debilitating anxiety; anxiety; his his or or her her experience experience is is directly directly pertinent pertinent episodes to the the criteria criteria required required for for diagnosis. diagnosis. On On the the other other hand, hand, the the presenting presenting problems problems to of the the personality personality disordered disordered individual individual are are rarely rarely so so straightforward. straightforward. They They are of are often seeking seeking evaluation evaluation at at someone someone else's else's request. request. Their Their presenting presenting complaints complaints often often involve involve the the behavior behavior of of other other people, rather than than their their own own experiences. experiences. often people, rather by definition, definition, they they do do not not see seethe thephenomena phenomenamost mostpertinent pertinentto tothe theestablish establishAlmost by Almost ment of of the the diagnosis diagnosis as as being being problematic; problematic; at at best, best, they they recognize recognize that that the the conse consement quences of of these these phenomena phenomena lead lead to to some some difficult difficult circumstances. circumstances. As As such, such, the the quences self-reported problems problems of of such such persons persons will will often often obscure obscure rather rather than than clarify clarify the the self-reported clinical picture. clinical picture. 5. 5.

Personality Disorders Are PrImarily Primarily Distinguished by Interpersonal Interpersonal Disruption Disruption

Over the past past several several decades, decades, aa number number of of writers writers have identified identified interpersonal interpersonal behavior as an important focus for the study of personality and psychopathology 1964; Homey, 1945; 1945; Kiesler, 1986; 1957; McLemore & (Adams, 1964; 1986; Leary, 1957; & Benjamin, 1979; Sullivan, 1953; 1953; Wiggins, 1982). 1982). One focus of such attention has concerned 1979; the utility utility of of the the interpersonal interpersonal approach approach as as aa foundation foundation for for the the diagnosis diagnosis and and the

CHAPTER CHAPTER 35 35

PERSONAUTY PERSONALITY DIAGNOSIS DIAGNOSIS AND AND DISORDERS DISORDERS

937 937

classification classification of all functional functional mental disorders. For example, Adams (1964) identi identiof psychopathology and fied a large degree of overlap between most measures of social behavior, and he concluded that what has traditionally been considered "mental illness" is actually more accurately represented as a maladaptive variant of interpersonal behavior. More recently, McLemore and Benjamin (1979) proposed a taxonomy of social behavior which they felt had the potential to serve as a "psychosocial alternative" to the DSM system. These authors claimed that "DSM-III shows near total neglect of of & Benjamin, Benjamin, social psychological variables and interpersonal behavior" (McLemore & 1979, 18). Other authors (i.e., Frances, 1980; sug 1979, p. 18). 1980; Widiger & Kelso, 1983) have suggested that interpersonal models, while not likely to replace the entire DSM-III taxonomy for for personality classification, have a great deal of potential to serve as a taxonomy functioning. Each of these writers has observed that many psychiatric diagnoses frequently based upon reports or observations of interpersonal behavior. are frequently The assumption made here is that personality disorders, unlike certain other dysfunctional primarily primarily through their expression in forms of mental disorder, are dysfunctional Even though though this this interpersonal interpersonal disruption disruption may well well be be aamanifesta manifestathe social milieu. Even tion of some intrapsychic difficulties, the maladaptiveness which qualifies personality disorders as mental disorders can only be evident in an interpersonal context. For example, traits like passive-aggressiveness or avoidant behavior are unlikely to be very maladaptive for a person stranded on a desert island without other persons present. In contrast, conditions such as schizophrenia or panic disorder will tend to cause problems for our hypothetically stranded person, even without another individual to whom such symptoms can be reported. As described by McLemore and Brokaw (1987), personality disorders are "disturbances" in the sense that the behavior of such people is disturbing to someone else. for the assessment assessment of This interpersonal assumption has unique implications for of personality disorder. Most significantly, it calls into question the utility of relying upon self-reported self-reported information information in establishing diagnoses. Traditionally, assessment in clinical psychology and psychiatry has relied reHed heavily upon information information provided directly by the patient, whether gathered in the form of of an interview or through the use of questionnaires. An assumption that personality personaUty disorders are inherently interpersonal problems requires that this interpersonal field should be sampled in some way. One manner in which important important information information concerning interpersonal interpersonal rela relationships may be obtained is through the use of peer ratings or other sociometric strategies. The use of peers in obtaining assessment information information has long been seen as a powerful powerful technique in research with children and adolescents (e.g., Ausubel, 1955), but it has been applied to adults with relative infrequency. This neglect is unfortunate unfortunate since, as Wiggins (1973) has stated, "in general, peer ratings have been shown to have substantial generalizability in terms of of inter-observer inter-observer agreement, generalizability of external structure across diverse rating groups and conditions, and generalizability generalizabiUty to criterion criterion situations of social importance" (Wiggins, 1973, p. 378). 378). For the personality disorders, an informant informant such as a close friend p. friend or relative

LESUE C. MOREY LESUECMOREY

938 938

may be an ideal source of behavior patterns of information information about the long-standing long-standing behavior of personality functioning functioning in patients of the examinee. Self-reported Self-reported information information about personality with these disorders is often often distorted distorted by situational situational circumstances; it should should be noted noted that that this distortion may be either in a positive/idealized positive/idealized (Tyrer et al., a l , 1984) or a negative/pathological negative/pathological (Edell, 1984) direction. Although Although the reports of of any particular information of multiple informants should information may contain some bias, the use of multiple informants portrayal of of the subject's pertinent provide a pattern that is likely to be an accurate portrayal greatest extent possible, information information from from peers, friends, and characteristics. To the greatest establishment of disorder diag diagrelatives should be considered in the establishment of personality disorder noses. 6. Jlect Extremes 6. Personality Personality Disorders Disorders Re Reflect Extremes o off Normal Normal

Personality Personality Variation Variation

of debate among contempo contempoThis assumption, which is a rather controversial point of classification researchers, holds that the difference difference between those individuals rary classification personalities is a difference difference of of with personality disorders and those with "normal" personalities rather than a difference difference of of kind. In other other words, it is assumed degree rather assumed that there is boundary or discontinuity discontinuity between the normal normal and the abnormal abnormal ranges no natural boundary of personality. Although the DSM calls for for a discrete categorical categorical decision whereby of disorder or they do not, there is little Httle evidence the person either has a personality disorder of considering personality personaUty disorder as a categorical categorical "disease entity." In in support of Uttle evidence to suggest that any functional functional psychiatric psychiatric disorder disorder has fact, there is little identified as such (Kendell, 1975), but those considerations are beyond beyond the been identified assumed to represent represent the extreme scope of this book. Here, a personality disorder is assumed of a normally distributed personality personaUty constellation. As a result, any boundary end of disorder is inherently inherently arbitrary. drawn between normal personality disorder "personality disorder" and "psychiatric disturbance" The distinction between "personality is also not one which is easily drawn. One interesting approach approach to drawing this distinction distinction was taken by Foulds (1971). Foulds separated separated what he called called personality deviance (i.e., personality disorder) from from personal illness (i.e., clinical syndrome). syndrome), and and he proposed proposed a model of of the relationship between between these conditions whereby they were viewed as overlapping but conceptually independent domains. In making conceptually independent this distinction. distinction, he focused focused upon quantitative aspects of of these these conditions. conditions, namely. namely, the distributions of symptoms (features and traits (features (features of personal personal illness) and (features of of personality deviance) in various populations. Expanding upon this approach. approach, Morey and Glutting Glutting (1994) identified identified four four quantitative features that could be used to distinguish normal personality traits from from features features of psychopathology. psychopathology. Normal and and Abnormal Abnormal Personality Personality Constructs Constructs Differ Differ in the Distribution Distribution a. Normal of Their Their Related Related Features in the General General Population. Population, Foulds hypothesized hypothesized that of abnormal symptoms should have distributions which have a marked marked positive skew (i.e., infrequently) in normal samples but but are roughly roughly normally normally distributed (i.e occur infrequently) contrast, normative personality traits should be distributed distributed in clinical samples. In contrast. .•

CHAPTER CHAPTER 35 35 PERSONALITY PERSONALITYDIAGNOSIS DIAGNOSISAND ANDDISORDERS DISORDERS

939 939

in a roughly Gaussian (i.e, bell-shaped) manner in the general population; a sample of of individuals with "deviant" personalities personaUties are distinguished by the personality trait being manifest to a degree rarely encountered in the general population. It should be noted that both types of constructs may be of clinical interest. Various regions of each type of construct may represent an area of concern; a person can be having difficulties because he or she manifests a particular normative trait to an extreme difficulties degree (e.g., introversion), or because he or she manifests an abnormal construct difference is in the to even a slight degree (e.g., suicidal ideation). The primary difference of the construct; the individual with a clinical trait (i.e., psychopathology) psychopathology) nature of different from normals, while individuals with an may be somehow qualitatively different ''abnormal amount" of a normative personality trait are quantitatively distinct; that "abnormal difference of degree rather than kind. is, a difference fe. Normal Normal and and Abnormal Abnormal Personality Personality Constructs Differ Differ Dramatically Dramatically in Their b. Social Desirability. Desirability. Assessment investigators have long recognized that self-report self-report personality tests can be vulnerable to efforts efforts at impression management. In particu particular, much concern has been expressed about the influence of efforts efforts to respond in a socially desirable desirable fashion on such tests. Various Various diverse diverse and creative creative efforts efforts have have been directed at resolving this dilemma, including the empirical keying strategy behind the development of of the original MMPI as well as the subsequent use of the "k-correction" and the forced forced choice matched item alternatives employed in the Edwards Personal Preference Preference Schedule. However, for self-report tests that focus on "abnormal" constructs, these strategies tend not to work very well. It is suggested that the reason for these problems is that abnormal constructs are inherently socially undesirable. As such, most measures of social desirability responding will correlate quite highly with measures of abnormal constructs. In contrast, the social desirability of normative personality personaUty features is more ambiguous, less evaluative, and more likely to be tied to aa specific specific context. context. For For example, example, the the trait trait adjective adjective "talkative" "talkative" likely to be tied to might be a socially desirable characteristic in a salesperson but not in librarian. might be a socially desirable characteristic in a salesperson but not in aa librarian. There is likely to be little consensus among people as to whether being "talkative" is aa desirable desirable or or undesirable undesirable characteristic, characteristic, whereas whereas characteristics characteristics such such as as "de "deis consensually as undesirable. pressed" or "delusional" will invariably be viewed consensually This implies that the social desirability of a construct may be useful as an indicator people, of its status in capturing normal or abnormal variation between people. Measures of of Abnormal Abnormal Personality Personality Constructs Differ Differ Dramatically Dramatically c. Scores on Measures between Clinical and Community Community Samples, While Scores of of Normal Normal Constructs Do Do between Not. This criterion is based upon the assumption that, in dealing with an abnormal personality construct, "more" is worse; that is, the more of the construct a person has, the greater the impairment the person manifests and the more likely the person is to come to the attention of mental health professionals. For example, when considering disordered thinking as a personal characteristic, greater amounts of considering thought disorder will be associated with greater impairment and need for for interven intervention. Thus, a clinical population should invariably obtain higher scores on measures

940 940

LESUE C. MOREY LESUECMOREY

of such constructs than a community sample. In contrast, for a normative personality of trait, the adaptive direction of of scores is less clear-cut. Given the assumption that such traits are normally distributed, then the traits are inherently bipolar, and of the trait may be maladaptive. Thus, even if if clinical extreme scores at either end of samples were restricted to persons with problems on a particular normative trait (e.g., extreme scores on Introversion-Extraversion), there would still be no reason differences between clinical and community subjects, as the extreme to suspect mean differences of the clinical subjects at either end of of the continuum would be expected to scores of balance out.

d. Measures of of Normative Normative Personality Personality Traits Should Should Demonstrate Demonstrate FactoriaV Factorial/ Correlational Correlational Invariance Invariance across across Clinical Clinical and Community Community Samples, While Measures Measures of Abormal Abormal Traits May May Not. The basic assumption behind this criterion is that of the correlation pattern that gives abnormal constructs their syndromal coherence coherence should only emerge in samples where there is adequate representation of individuals manifesting manifesting the syndrome (i.e., clinical samples). In community samples, which may association include relatively few individuals who have a clinical syndrome, the association of the same syndrome may be no greater than that between any between features of two features selected randomly. As an example, if depression were defined by five necessary and sufficient sufficient criteria, and these five criteria were intercorrelated in a community sample that contained no depressed subjects, the average correlation correlation of nondepressed individuals, between these features might well be zero. In a sample of self-esteem may only be associated at chance levels since sleep problems and low self-esteem associaindividuals who share the putative causal process that underlies the clinical associa of these features have been removed from the sample. It is the covariation of tion of these features in individuals considered to be depressed that lends a correlation pattern to these features. Thus, highly intercorrelated sets of features (i.e., syn syndromes) might emerge from a factor analysis of clinical subjects that would not be identified in a sample of subjects from the community. In contrast, those traits that describe normal variation in personality would be expected to capture this variability among clinical as well as normal subjects. Even though the clinical subjects may be, as a group, more extreme on normal personality traits, similar correlational patterns among elements of of the trait should be obtained. For example, the construct of of Extroversion-Introversion Extroversion-Introversion should iden identify tify meaningful differences differences among clinical subjects as well as normal subjects, and the intercorrelation of the behaviors that make up this construct should be similar predictable empirical results with respect in the two populations. This should yield predictable to the factor structure (for multifaceted scales/constructs) and the average item coefficient alpha, for unidimensional constructs); for a norma normaintercorrelation (i.e., coefficient tive trait, these results should be similar in clinical and non clinical samples. In nonclinical contrast, these values may well differ if an "abnormal" construct is being examined. Empirical work has supported this assumption with respect to many features of personality disorder. For example, Tyrer and Alexander (1979) examined a of personality disorder classification scheme similar to the system by Schneider that

CHAPTER CHAPTER 35 35

PERSONALITY PERSONALITY DIAGNOSIS DIAGNOSIS AND AND DISORDERS DISORDERS

941

was described earlier. Their factor factor analyses found no qualitative difference difference between the personalities of individuals with personality disorder and those of individuals without such disorders. In their words, "the results of the factor factor analysis reveal that the underlying structure of variables is similar in both those with and those without primary personality disorder and hence supports the concept of personality disorders as being at the extreme of a multidimensional continuum" (Tyrer & Alexander, 1979, p. 166). 7. 7. Personality Disorders Are Multiply Multiply Determined

This assumption proposes that there is no single specific etiology for the personality different disorders. Rather, personality traits are seen as evolving from a number of different environmental/situational in inorigins, including constitutional, developmental, and environmentaVsituational fluences. Futhermore, within each of these domains it is assumed that there are fluences. different routes to the same personological presentation. For example, it is not different difficult to imagine that dependency behaviors may arise out of constitutional constitutional difficult from limitations, or from early development (e.g., parental overinvolvement), or from prolonged situational pressures (e.g., an abusive spouse). V. CONCLUSION V.

It is important to note that all of the seven assumptions just described represent areas of considerable controversy within both psychology and psychiatry. The above discussions should not be interpreted as indicating that these issues are resolved; major conceptual issues in this area in rather, they may be viewed as a guide to major need of resolution. It is important to understand that the conceptualization of of personaUty disorder presented here is not necessarily consonant with the viewpoint personality expressed in the DSM. However, they do represent an attempt to make explicit, in a way which the American Psychiatric Association manuals do not, the domain of phenomena denoted by the concept of "personality disorder." REFERENCES B. (1964). (1964). 'Mental illness' or interpersonal behavior? American AmericanPsychologist, Psychologist, Adams, H. B. 19, 19, 191-197. Diagnostic andstatistical statistical manual of mental and manual of mental disor disorAmerican Psychiatric Association. (1952). (1952). Diagnostic der, Washington, DC: DC: Author. der. American Psychiatric Association. andand statistical manual ofmental disorder Association.(1968). (1968).Diagnostic Diagnostic statistical manual of mental disorder (2nd ed.). ed.). Washington, DC: DC: Author. Association.(1980). (1980).Diagnostic Diagnostic statistical manual of mental disorder American Psychiatric Association. andand statistical manual ofmental disorder DC: Author. (3rd ed.). ed.). Washington, DC: Association.(1987). (1987).Diagnostic Diagnostic statistical manual of mental disorder American Psychiatric Association. andand statistical manual ofmental disorder rev. ed.). DC: Author. (3rd rev. ed.). Washington, DC:

942 942

LESLIE MOREY LESLIE C. CMoREY

American Psychiatric Association. (1994). Diagnostic manual ofof mental disorder Diagnosticand andstatistical statistical manual mental disorder (4th ed.). Washington, DC: Author. Ausubel, D. P. (1955). Sociempathy Sociempathy as a function function of sociometric status in an adolescent group. Psychological PsychologicalReview, Review,62, 62,378-390. 378-390. group. Benjamin, L. L. S. S. (1974). Structural Structural analysis analysis of of social social behavior. behavior. Psychological Psychological Review, 81, Benjamin, 81,

392-425.

of personality: A re reBenjamin, L. S. (1993). Dimensional, categorical, or hybrid analyses of Psychological Inquiry, Inquiry, 4,4,91-134. 91-134. sponse to Widiger's proposal. Psychological Blashfield, R. K., & McElroy, McElroy, R. A A. (1987). The The 1985 joumal journal literature literature on on the the personality personality Blashfield, 536-546. disorders. Comprehensive ComprehensivePsychiatry, Psychiatry,28,28, 536-546. (1985). Exemplar prototypes of Blashfield, R. K., Sprock, J., Pinkston, K., & Hodgin, J. (1985). of personality disorder disorder diagnoses. Comprehensive Comprehensive Psychiatry, Psychiatry,26, 26,11-21. 11-21. time, Berkeley, CA: Bancroft. Bancroft. Block, J. (1971). Lives through time. J. H. H. (1980). The The role role of of ego-control ego-control and and ego-resiliency ego-resiliency in in the the organization organization Block, J. J. & Block, J. of behavior. In W.A. W.A Collins (Ed.), Minnesota Symposium on Child Psychology (Vol. 13) pp 39-101.) Hillsdale, NJ: Erlbaum. scientificanalysis analysisof ofpersonality. personality,Baltimore: Baltimore:Penguin PenguinBooks. Books. Cattell, R. B. (1965) The scientific for clinical Costa, P. T., & McCrae, R. R. (1986). Personality stability and its implications for 407-423. psychology. Clinical ClinicalPsychology PsychologyReview, Review,6, 6, 407-423. DeJong, C. A A. J., J.,van vanden denBrink, Brink,W., W.,Jansen, Jansen,J.J.A A.M., M.,&&Schippers, Schippers,G.G.M.M.(1989). (1989). Interpersonal Dejong, Interpersonal aspects of DSM-III axis II: Theoretical hypotheses and empirical findings. findings. Journal of of Personality PersonalityDisorders, Disorders,3,3,135-146. 135-146.

DoUard, J., & Miller, N. E. (1950). Personality Personality and and psychotherapy. psychotherapy. New NewYork: York: McGraw-Hill. McGraw-Hill. Dollard, Drake, R. E., & Vaillant, G. E. (1985). A validity study of Axis II of DSM-III. American Journal Journalof ofPsychiatry, Psychiatry,142, 142,553-558. 553-558.

Clinical validity and utility. Journal Edell, W. S. (1984). The Borderline Syndrome Index: Oinical of Nervous Nervous and and Mental MentalDisease, Disease,1 72, 172, 254-263. of 254-263.

Epstein, S., & O'Brien, E. J. (1985). The person-situation debate in historical and current 513-537. perspective. Psychological PsychologicalBulletin, Bulletin,98,98, 513-537. C. P. P.(1983). (1983)."Dr. "Dr.Death" Death"and andthe thecase casefor foran anethical ethicalban banon onpsychiatric psychiatricand andpsychologi psychologiEwing, C. cal predictions of dangerousness in capital sentencing proceedings. American Journal

of of Law and Medicine, Medicine,8,8,407-428. 407-428.

Eysenck, H. J. (1987). The definition definition of personality disorder and criteria appropriate for for their description. Journal 211-219. Journalof ofPersonality PersonalityDisorders, Disorders,1, 1, 211-219. Eysenck, H. J., & Eysenck, M. W. (1985). erences. New (1985). Personality Personalityand andindividual individualdiff differences. NewYork: York: Plenum Press. Friedman, A A. F. F. (1983). (1983).Diagnosis Diagnosisand andclinical cUnicalassessment: assessment: Eysenck, H. J., Wakefield, J. A., & Friedman, The DSM-III. DSM-III. Annual Annual Review Reviewof ofPsychology, Psychology,34,34,167-193. 167-193. The A. (1971). Foulds, G. A (1971). Personality deviance and personal symptomatology. Psychological Psychological

Medicine, Medicine,1,1,222-233. 222-233.

Frances, A A. (1980). The DSM-III personality disorders section: A commentary. conmientary. American

Journal 1050-1054. Journalof ofPsychiatry, Psychiatry,137, 137, 1050-1054.

A. (1982). Dimensional Dimensional and and categorical categorical systems systems of ofpersonality personality diagnosis: diagnosis:A Acompari compariFrances, A ComprehensivePsychiatry, Psychiatry, 516-527. son. Comprehensive 23,23, 516-527. Glueck, S., & Glueck, Glueck, E. E. T. T. (1%8). Non-delinquentsin inperspective. perspective,Cambridge, Cambridge,MA: MA: Harvard Glueck, (1968). Non-delinquents Harvard University Press. Gunderson, J. G. (1983). DSM-III diagnoses of personality disorders. In J. P. Frosch (Ed.), Current personality disorders (pp.(pp. 20-39). Washington, DC:DC: American Currentperspectives perspectivesonon personality disorders 20-39). Washington, American Psychiatric Press.

CHAmR CHAPTER 35 35 PERSONAUTY PERSONALITYDIAGNOSIS DIAGNOSISAND ANDDISORDERS DISORDERS

943 943

Hare, R. D. (1985). Comparison of procedures for the assessment of psychopathy. Journal of Consulting 53,53, 7-16. Consultingand andClinical ClinicalPsychology, Psychology, 7-16. Havighurst, R. J., Bowman, P. H., Liddle, G. P., Matthews, C. V., & Pierce, J. V. (1962). Growing Growing up upin inRiver RiverCity. City.New NewYork: York:Wiley. Wiley. Heumann, K. A., & Morey, Morey, L. L.C. C.(1990). (1990).Reliability Reliabilityof ofcategorical categoricaland anddimensional dimensionaljudgments judgments 147, 498-500. of personality disorder. American AmericanJournal JournalofofPsychiatry, Psychiatry, 147, 498-500. Horney, (1945). Our inner conflicts. conflicts.New NewYork: York:Norton. Norton. Homey, K. (1945). Jung, C. G. (1923). Psychological Psychologicaltypes. types.New NewYork: York:Harcourt, Harcourt,Brace. Brace. (1962). Birth Kagan, J., & Moss, H. A. (1962). Birth to maturity: maturity:AAstudy studyininpsychological psychologicaldevelopment. development. New York: Wiley. Kendell, R. E. (1975). The role of of diagnosis diagnosis ininpsychiatry. psychiatry.London: London:Blackwell BlackwellScientific. Scientific. Profiting from from controversy: controversy: Lessons Lessons from from the the person personKenrick, D. T., & Funder, D. C. (1988). Profiting situation debate. American 23-34. American Psychologist, Psychologist,43,43, 23-34. Kiesler, D. J. (1983). The 1982 interpersonal circle: A taxonomy for complementarity in PsychologicalReview, Review,90, 90,185-214. 185-214. human transactions. Psychological Kiesler, D. J. (1986). Interpersonal methods of diagnosis and treatment. In J. O. Cavender (Ed.), Psychiatry Psychiatry (Vol. (Vol. 1,1,No. No.4,4,pp. pp.1-23). 1-23).New NewYork: York:Lippincott. Lippincott. Koch, J. L. (1981). Die psychopathischen Ravensburg: Maier. psychopathischenminderwertigkeiten. minderwertigkeiten. Ravensburg: Maier. Kraepelin, E. (1902). Clinical physicians (A.(A.R.R. Clinicalpsychiatry: psychiatry:AAtextbook textbookfor forstudents studentsand and physicians Diefendorf, Trans.) (6th ed.). ed.). London: Macmillan. Interpersonaldiagnosis diagnosisofof personality. New York: Ronald. personality. New York: Ronald. Leary, T. (1957). Interpersonal Dishion, T. T.(1983). (1983).Early Earlypredictors predictorsof ofmale maledelinquency: delinquency:AAreview. review.Psychologi PsychologiLoeber, R., & Dishion, cal Bulletin, Bulletin,94, 94,8-99. 8-99. cal Lorr, M., Bishop, P. F., & McNair, D. M. (1965). (1965). Interpersonal types among psychiatric patients. Journal Journalof ofAbnormal AbnormalPsychology, Psychology,70,70,468-472. 468-472. Mack, J. J.E. E. (1975). (1975).Borderline Borderiinestates: states:An Anhistorical historicalperspective. perspective.InInJ.J.E.E.Mack Mack(Ed.), (Ed.),Borderline Borderline states states ininpsychiatry psychiatry(pp. (pp. 1-21). 1-21).New New York: York: Grone Grune &&Stratton. Stratton. dispositions: Personality in in McCrae, R. R., & & Costa, P. T. (1984). Emerging Emerginglives, lives,enduring enduring dispositions: Personality adulthood. adulthood. Boston: Boston:Little, Little,Brown. Brown. McGlashan, T. H. (1983). The borderline syndrome. I. Testing three diagnostic systems. Archives 1311-1318. Archives of of General GeneralPsychiatry, Psychiatry,40,40, 1311-1318. McGlashan, T. H. (1986). The Chestnut Lodge follow-up study: III. Long-term outcome of of borderline personalities. Archives Archives of of General GeneralPsychiatry, Psychiatry,43,43,20-30. 20-30. McLemore, C. W., W., & & Benjamin, Benjamin, L. L.S. S.(1979). (1979).Whatever Whateverhappened happenedto tointerpersonal interpersonaldiagnosis? diagnosis? A psychosocial alternative to DSM-III. American Psychologist, Psychologist,34, 34,17-34. 17-34. McLemore, C. W., & Brokaw, D. W. (1987). Personality disorders as dysfunctional dysfunctional interper interpersonal behavior. Journal 1, 1, 270-285. Journal of ofPersonality PersonalityDisorders, Disorders, 270-285. Mellsop, C., C, Varghese, F., Joshua, S., & Hicks, A. (1982). The reliability of Axis II of DSMAmericanJournal JournalofofPsychiatry, Psychiatry, 139, 1360-1361. III. American 139, 1360-1361. balance.New NewYork: York:Viking VikingPress. Press. Menninger, K. (1963). The vital balance. Millon, T. (1969). Modern Modem psychopathology. psychopathology. Philadelphia: Philadelphia: Saunders. Saunders. Disorders of ofpersonality: personality:DSM-IlI DSM-IIIAxis Axis/I.II. New York: Wiley. Millon, T. (1981). Disorders New York: Wiley. Millon, T. 38,38,804-814. 804-814. T. (1983). (1983). The TheDSM-III: DSM-III:An Aninsiders insidersperspective. perspective.American AmericanPsychologist, Psychologist, Millon, T. (1985). The The MCMI provides provides aa good good assessment assessment of of DSM-III DSM-III disorders: disorders:The TheMCMIMCMIJournalof ofPersonality PersonalityAssessment, Assessment, 379-392. II will prove even better. Journal 49,49, 379-392. Millon, T. (1986). (1986). The MCMI and DSM-III: Further Personality Further commentaries. Journal of of Personality Assessment,49, 49,205-207. 205-207. Assessment,

944 944

C. MOREY MOREY LESLIE C.

Millon, T. (1987). Millon clinical multiaxial inventory-II manual. Minneapolis, MN: MN: National National Computer Systems. Mischel, W. (1968). Personality Personality and and assessment. assessment. New York: Wiley. of predictability and consistency of of persons: Mischel, W. (1983). Alternatives in the pursuit of Stable data that yield unstable interpretations. Journal Journal of of Personality, Personality, 51, 578-604. Morey, L. C. (1985). An empirical comparison of DSM-I1I approaches to of interpersonal and DSM-III classification classification of of personality disorders. Psychiatry, Psychiatry, 48, 358-364. Morey, L. C. (1988a). The categorical representation representation of of personality personality disorder: A A cluster analysis of DSM-I1I-R personality features. Journal of DSM-III-R Journal of of Abnormal Abnormal Psychology, Psychology, 97, 314-321. Morey, L. C. (1988b). Personality disorders under DSM-I1I DSM-III and DSM-III-R: An examination examination of of convergence, coverage, and internal consistency. American American Journal Journal of of Psychiatry, Psychiatry, 145, 573-577. C , & Glutting, J. H. (1994). The Personality PersonaUty Assessment InventQry: Inventory: Correlates Morey, L. c., abnormal personality. In S. Strack & M. Lorr (Eds), Differentiating Differentiating with normal and abnormal normal and and abnormal abnormal personality personality (pp. 402-420). New York: Springer. normal of adherence to diagnostic criteria: criteria: Morey, L. C C.,, & Ochoa, E. S. (1989). An investigation of of the DSM-I1I DSM-III personality disorders. Journal Journal of of Personality Personality Disor DisorClinical diagnosis of ders, 3, 180-192. taxonomy of of personality attributes: Replicated Replicated Norman, W. T. (1963). Toward an adequate taxonomy nomination personality personahty ratings. Journal Journal of of Abnormal Abnormal and Social Social factor structure in peer nomination Psychology, 66, 66, 574-583. Psychology, R. (1987). Peer relations and later personal adjustment: Are low lowParker, J. G., & Asher, S. R accepted children at risk? Psychological Psychological Bulletin, 102, 357-389. R. (1986). Report of of the Speaker-Elect. American American Journal Journal of of Psychiatry, Psychiatry, 143, 1348Peele, R 1353. Zinmierman, M., & Stangl, D. (1986). DSM-III personality disorders: Pfohl, B., Coryell, W., Zimmerman, Diagnostic overlap and internal consistency of of individual DSM-III criteria. Comprehen Comprehensive sive Psychiatry, Psychiatry, 27, 21-34. Pope, H. G., Jonas, J. M., Hudson, J. I., Cohen, B. M., & Gunderson, J. G. (1983). The of DSM-I1I DSM-III borderline personality disorder: A A phenomenologic, family history, validity of follow-up study. Archives Archives of of General General Psychiatry, Psychiatry, treatment response, and long-term follow-up 40, 23-30. Raines, G. N. (1952). Foreword. In Diagnostic Diagnostic and and statistical manual manual of of mental mental disorder. disorder Washington, DC: DC: American Psychiatric Association. Reich, J. H. (1987). Instruments measuring DSM-III and DSM-III-R personality disorders. Journal Journal of of Personality Personality Disorders. Disorders, 1, 220-240. Reich, W. (1949). Character Character analysis. New York: Orgone Institute. Ricks, D. P., F., & Berry, J. C. (1970). Family and symptom patterns that precede schizophrenia. In M. Roff Roff & D. Ricks (Eds.), Life Life history history research in psychopathology, psychopathology, (pp. 31-39). Minneapolis: University of of Minnesota Press. Deviant children grown grown up: A sociological sociological and and psychiatric psychiatric study study of of Robins, L. N. (1966). Deviant sociopathic sociopathic personality. personality. Baltimore: Williams & Wilkins. Rolf, J. E., Knight, R, R., & Wertheim, E. (1976). Disturbed preschizophrenics: Childhood Journal of of Nervous Nervous and and Mental Mental Disease, Disease, symptoms in relation to adult outcomes. Journal 162, 274-279. Rosch, E. (1973). On the internal structure of perceptual and semantic categories. In T. E. Moore (Ed.), Cognitive Cognitive development development and and the acquisition acquisition of of language (pp. 27-48). New York: Academic Press.

CHAPTER 35 35 CHAPTER

PERSONALITYDIAGNOSIS DIAGNOSISAND ANDDISORDERS DISORDERS PERSONAUTY

945 945

Schneider, K. (1958). (1958). Psychopathic Psychopathic personalities. personalities. London: London: Cassell. Cassell. (Original (Original work work pub published 1923). Schutz, W. C. (1958). FlRO-B: theory of of interpersonal behavior. NewNew FIRO-B:AA three threedimensional dimensional theory interpersonal behavior. York: Rinehart. Smith, E. E., & Medin, D. L. (1981). (1981). Categories Categoriesand andconcepts. concepts,Cambridge, Cambridge,MA: MA:Harvard Harvard University Press. Spitzer, R. L., Forman, J. B., B., & Nee, J. (1979). (1979). DSM-I1I DSM-III field trials: I. Initial interviewer diagnostic reliability. American 36,36, 815-817. AmericanJournal JournalofofPsychiatry, Psychiatry, 815-817. Standage, K. (1986). A clinical and psychometric investigation comparing Schneider'S Schneider's and 35-46. the DSM-III typologies of personality disorders. Comprehensive ComprehensivePsychiatry, Psychiatry,27,27, 35-46. Stangl, D., Pfohl, B., Zimmerman, M., Bowers, W., & Corenthal, C. (1985). (1985). A structured interview for the DSM-III personality disorders. Archives of General Psychiatry, Psychiatry,42, 42, of General 591-596. (1984). Psychotherapy Psychotherapyininaanew newkey. key.New NewYork: York:Basic BasicBooks. Books. Strupp, H. H., & Binder, J. (1984). interpersonaltheory theoryof ofpsychiatry. psychiatry.New NewYork: York:Norton. Norton. (1953). The interpersonal Sullivan, H. S. (1953). Taft, R. (1955). The ability to judge people. Psychological PsychologicalBulletin, Bulletin,52, 52,1-23. 1-23. Taft, Recurrentpersonality personalityfactors factorsbased basedonontrait traitratings. ratings, C, & & Christal, R. E. (1961). Recurrent Tupes, E. C., (USAF ASD Tech. Rep. No. 61-97). Lockland Air Force Base, TX: U.S. Air Force. (1979). Classification Classification of personality disorder. British Journal of of Tyrer, P., & Alexander, J. (1979). Psychiatry,135, 135,163-167. 163-167. Psychiatry, Baiter, A., Giller, Ciller, E., & Tyrer, P., Cicchetti, D. V., Casey, P. R., Fitzpatrick, K., Oliver, R., Balter, for assessing person person(1984). Cross-national reliability study of a schedule for Harkness, L. (1984). Journalof ofNervous Nervousand andMental MentalDisease, Disease, 718-721. 172,172, 718-721. ality disorders. Journal Vaillant, G. E. (1987). A developmental view of old and new perspectives on personality Journalof ofPersonality PersonalityDisorders, Disorders, 146-156. 1, 1, 146-156. disorder. Journal Warner, R. (1978). The diagnosis of antisocial and hysterical personality disorders: An Journalof ofNervous Nervousand andMental MentalDisease, Disease, 166, 839-845. 166, 839-845. example of sex bias. Journal The DSM-I1I-R DSM-III-R categorical categorical personality personality disorder disorder diagnoses: diagnoses: A A critique critique Widiger, T. A. (1993). (1993). The and an alternative. Psychological PsychologicalInquiry, Inquiry,4,4,75-90. 75-90. (1985). The The DSM-I1I DSM-III personality personality disorders: disorders: Perspectives Perspectives from from Widiger, T. A., & Frances, A. (1985). psychology. Archives Archives of of General GeneralPsychiatry, Psychiatry,42,42,615-623. 615-623. Widiger, T. A., & Frances, A. (1987). Instruments and inventories for of for the measurement of personality disorders. Clinical ClinicalPsychology PsychologyReview, Review,7,7,49-76. 49-76. Widiger, T. A., Frances, A., Spitzer, R. L., & Williams, J. B. W. (1988). The DSM-I1I-R DSM-III-R personality disorders: An overview. American Journal of Psychiatry,145, 145,786-795. 786-795. of Psychiatry, Widiger, T. A., Trull, T. J., Clarkin, J. F., Sanderson, Sanderson, C., C, & & Costa, P. P. T. T. (1994). (1994). A A description description of of the DSM-III-R and DSM�IV DSM-IV personality disorders with the five-factor model of personality. personahty. In P. T. Costa T. A. Widiger (Eds.), Personality Personality disorders disordersand andthe thefive fivef actor model of personality, (pp. factor ofpersonality, (pp.41-56). 41-56).Washington, Washington,DC: DC:American American Psychological PsychologicalAs Association. Clinical Psychology PsychologyReview, Review, Widiger, T. A., Kelso, K. (1983). Psychodiagnosis of Axis II. Clinical 3, 491-510. B., Spitzer, R. L., & Frances, A. (1986). The MCMI and DSM DSMWidiger, T. A., Williams, J. B., brief rejoinder to Millon. Journal of of Personality PersonalityAssessment, Assessment,50, 50,198-204. 198-204. III: A brief J. S. (1973). (1973). Personality Personalityand andprediction: prediction:Principles Principles personality assessment. Wiggins, J. ofof personality assessment. NewNew York: Wiley. J. S. S. (1979). (1979). A A psychological psychological taxonomy taxonomyof of trait-descriptive trait-descriptive terms: terms:The The interpersonal interpersonal Wiggins, J. domain. Journal Journalof ofPersonality Personalityand andSocial Social Psychology, 395-412. Psychology, 37, 37, 395-412.

946 946

LESLIE C. CMoREY LaUE MOREY Wiggins, J. J. S. S. (1982). (1982). Circumplex models of interpersonal behavior in clinical psychology. In P. Kendall & J. psychology J. Butcher Butcher (Eds.), (Eds.), Handbook Handbook of of research researchmethods methodsininclinical clinical psychology (pp. 183-221). New York: Wiley. (pp. Wiggins, J. S., & Pincus, A. L. (1994). Personality structure and the structure of personality disorders. In P. T. Costa and T. A. Widiger (Eds.), Personality Personalitydisorders disordersand andthe thefive fivefactor ofpersonality (pp. 73-93). 73-93). Washington, Washington, DC: DC: American American Psychological Psychological As Asf actor model of personality (pp. sociation. sociation. Journal of Zetzel, E. (1956). Current Current concepts concepts of of transference. transference. International IntemationalJournal of Psycho-Analysis, Psycho-Analysis,

37, 369-375. 37,

Zimmerman, M. (1994). Diagnosing personality disorders: A review of issues and research General Psychiatry, Psychiatry,51, 51,225-245. 225-245. of General methods. Archives of

CHAPTER 36 CHAPTER 36

TRENDS AND PRACTICES IN PSYCHOTHERAPY PSYCHOTHERAPY OUTCOME ASSESSMENT AND THEIR IMPLICATIONS FOR PSYCHOTHERAPY PSYCHOTHERAPY AND APPLIED ApPLIED PERSONALITY LAMBERT MICHAEL J, J. LAMBERT

BRIGHAM BRIGHAM YOUNG YOUNG UNIVERSITY UNIVERSITY

EDWIN C. SUPPLEE TRIPLER CENTER TRIPLER REGIONAL REGIONAL MEDICAL MEDICAL CENTER

In this chapter we will focus on research in personality and psychotherapy. There are many ways in in which which personality has has been been discussed discussed and and analyzed analyzed in in psychother psychotherapy. These include studies on the effects of therapist personality on clients from hypothedifferent diagnostic groupings; for example Whitehorn and Betz (1954) hypothe sized that A and B therapist types had differential outcomes with schizophrenic and neurotic patients. Also present present are are studies studies of of patient patient personality personality and and its its relation relationand ship to both therapy outcome and the manner in which therapists and patients HANDBOOK HANDBOOK OF OF PERSONALITY PERSONALUY PSYCHOLOGY PSYCHOLOGY

947 947

COPYRIGHT COFYRIGHT C © 1997 1997 BY BY ACADEMIC ACADEMIC PRESS. PRESS. ALL liGHTS OF REPRODUcnON IN ANY FORM RESERVED.

ALL RIGHTS OF REPRODUCHON IN ANY FORM RESERVED.

948 948

LAMBERT LAMBERTAND ANDSUPPLEE SUPPLEE

participate in therapy sessions. This work is characterized by studies of of the relation relationego strength (as measured by the Rorschach, MMPI, or judges ship between patient patient strength measured judges ratings) and indices of of improvement improvement (Garfield, (Garfield, 1978). The effect effect of of similarities similarities between therapist therapist and client cUent personality, as well as other matching strategies based based upon numerous personality dimensions, has also been the subject numerous subject of of considerable considerable research (Berzins, 1977). Typical Typical of this research research is Carson's (1969) Interaction Interaction research Concepts Concepts of of Personality, Personality, in which it is argued that complementary complementary and incompatible incompatible personality styles can be identified identified and will lead to differential differential outcome. personality Although Although research on therapist and patient personality could form form the basis of of this chapter, we would be writing mostly about about the history history of of psychotherapy psychotherapy and contemporary research issues. To a large extent personality variables variables not about contemporary have not been used successfully successfully to ideally match therapists and patients, to select for treatments, or to predict predict therapy outcome. It has proved difficult difficult to patients for identify effectively measure those that identify salient personality characteristics and to effectively have been hypothesized to be important in the therapeutic process. While it would therapy through through be interesting to discuss past and present attempts to facilitate therapy for our failures, we have chosen a different different personality research and the reasons for for our analysis is the myriad ways in which focus. The particular vantage point for psychotherapy have been assessed, and changes in patients who have undergone psychotherapy assessment how assessment practices interact with the study of personality. First, the assessment of outcome will be put in an historical perspective. This will be followed followed with a of of current assessment assessment practices and their relation relation to personality. Finally, discussion of of current current practices for for personality personality and psychother psychotherwe will discuss the implications of for future future research. research. apy and suggest directions for construction, Byrne (1964) noted that in the early days of personality test construction, experiments experiments and psychometricians were advancing independent independent of each other, leav leaving "the former former short short on measurement measurement and the latter short short on theory." The two have since been wedded, with mutual advancement. In contrast, treatment treatment outcome outcome research and personality research continue to develop in completely separate separate spheres. Personality researchers regard the advancement advancement and usefulness usefulness of of personal personality measurement measurement more optimistically optimistically (Craik, 1986) than psychotherapy psychotherapy researchers researchers (Lambert, Bergin, 1986). As (Lambert, Shapiro, Shapiro, & Bergin, As we we shall shall point point out, out, there there is is an an urgent urgent need need for for psychotherapy psychotherapy researchers and those interested in measurement measurement and personality to join join forces forces if significant significant advances are to be made. I. HISTORICAL OVERVIEW effects of of psychotherapy psychotherapy has a relatively short history, dating Research into the effects 1930s. The psychotherapy of of the day was, of course, psychoanalytically psychoanalytically back to the 19308. oriented. And quite naturally, theoreticians and practitioners were ambitiously prominent disturb disturbengaged in changing the "structure of personality" as well as the prominent ing symptoms of of central concern to the client. Freud contended that the analyst for assessing the outcome of of therapy. was the prime, and perhaps the only, medium for

36 CHAPTER 36

PSYCHOTHERAPY O OUTCOME ASSESSMENT PSYCHOmERAPY UTCOME ASSESS MENT

949 949

This appears to be because of of the multifarious multifarious and complex complex dimensions dimensions present present in each case and in each session, and the traditional scientific research conducted in scientific that day did not account for all of those variables, at least to Freud's satisfaction. account for of those satisfaction. Most reports of of outcome outcome were of of a qualitative nature nature and single case studies, Most although there were also quantitative studies produced (Rachman, 1971; Lub Luborksy orksy & Spence, 1978). In those early early quantitative quantitative studies, outcome was reported in terms of of percent percentage of of those improving, without much much detail on the type or quality of of outcome outcome and without the use of of reliable rating scales based on patient reports or objective objective without observer ratings. In the Berlin Psychoanalytic Psychoanalytic Institute's report (ca. 1930), the criteria that Fenichel used to determine determine success success were an "analytically ''analytically acceptable acceptable personality personaUty Fenichel improvement of of symptoms (Bergin & Lambert, 1978). The interpre interprechange" and an improvement of the results have been contended contended since Eysenck's 1952 survey survey (Bergin, 1971; tations of the controversy controversy left largely unresolved unresolved (Garfield (Garfield & & Bergin & Lambert, 1978), with the depending on how how one views the data, the improvement improvement rate was was Bergin, 1986); depending either 31 or 91% of determining determining personality personality change change either 91 % (Bergin, 1971). Here, the means of reasonably assumed assumed to be the analyst's judgement, but the specifics specifics of of how how the is reasonably judgement is arrived at are not attended to. In this study, as well as many others others judgement of this period, personality personality functioning functioning was reported as global improvement improvement in gross of functioning as viewed by the therapist (Luborsky & Spence, 1978). This observation observation functioning for other schools in the early half half of of this century, including including appears to hold true for behaviorism (Eysenck (Eysenck & Beech, 1971) and humanism (Rogers, 1942). behaviorism Historically, one of of the most most important landmarks of of psychotherapy, psychotherapy, and of body of of studies studies produced by Rogers and his associates associates research on outcome, was the body starting in the early 1950s (Rogers & Dymond, 1954). This research continued to emphasize personality personality change but deviated in important ways from research on of research studies (Rogers & Dymond, 1954), psychoanalysis. In their early volume of stated that, given given the facilitative faciUtative therapeutic conditions, conditions, Rogers and his associates stated the client client will will reorganize reorganize himself himself at at both both the the conscious conscious and and the the deeper deeper levels levels of of the his personality personality in in such such aa manner manner as as to to cope cope with with life life more more constructively, constructively, more more his intelligently, and and in in aa more more socialized socialized as as well well as as aa more more satisfying satisfying way. way. More More intelligently, specifically specifically it is hypothesized hypothesized that the client will change in his perception perception of of self, will become become more more understanding understanding of of self self and and others, others, more more accepting accepting of of self self and and will others, more creative, more adaptive, more self-directing self-directing and autonomous, more more mature in in his his behavior, behavior, less less defensive, defensive, and and more more tolerant tolerant of of frustrations. frustrations. mature (p. 4)

The The outcome outcome measure measure did not rely solely on clinical observation, observation, but the clients' of successful successful therapy therapy were objectified objectified by the use of of the O-sort Q-sort routine. They They views of outlined from from their research research "necessary "necessary and sufficient" therapist therapist characteristics, client characteristics, and therapeutic outcome (Truax & Mitchell, 1971). the early 1950s most therapy outcome research on personality personality was was While in the proposed (or was striving to be) a nomothetic endeavor, there were some who proposed

9SO 950

LAMBERT AND SUPPLEE SuppLEE

research with the individual and with personality. George Kelly's (1955) personal personalconstruct psychology is an example. In order for the therapist to better understand a client's personal-construct system, Kelly devised the Role Construct Repertory Repertory (Rep) Test. The Rep Test could also be used as an outcome measure, to see if and how a person's core constructs may have changed via therapy. Gordon Allport AUport advocated quantitative methods in in single singlecase casestudies studiesas asimportant importantfor forunderstanding understanding the individual's personality structure, although ahhoughhis hismethods methodswere werenot notapplied appliedsystem systematically in behavior change research (Kiesler, 1971). With this growing popularity of applied behaviorism in the 1960s and 1970s, there was a drastic change in the methods used to assess outcome and in the status and importance of personality-based treatment. The behaviorally oriented theorists were interested in measuring readily observable change. Wolpe was primarily con concerned with anxiety as the antecedent of neurotic behavior and the control (reduc (reduction) of anxiety as the measure of outcome of desensitization (Eysenck & & Beech, 1971). The criterion for success of Lazarus's adaptation of desensitization, behavior rehearsal, was behavioral change primarily regarding assertiveness (Lazarus, 1971). Hans Eysenck postulated that personality could be divided into two major factors. Persons were seen as varying on the dimensions of instability-stability and on an extroversion-introversion continuum. These factors were hypothesized to be based on neurophysiological structures, with a predisposition to neurosis dependent upon both genetic and environmental influence. Understanding to what extent these factors exist within an individual should determine the therapeutic approach. He did not, however, expound on how these factors might or might not change as a result of treatment, and in fact seemed to view them as rather stable traits that therapy was not aimed at modifying (Eysenck, 1987; 1987; Eysenck & & Beech, 1971). Operant conditioning, as studied by Skinner and others, has been used to wide variety variety of of divergent divergent disorders: disorders: with with elicit changes in component behaviors of a wide schizophrenics, improvement in disorganized thinking, apathy, social withdrawal, and bizarre verbalizations; modification of delinquent behavior in children; and treatment of marital problems. Here again we have some strong positive findings findings for the efficacy efficacy of a technique, as with most behavioral techniques and as per the measurement criteria, but it is difficult to define changes in personality per se behav (Krasner, 1971). Of course, this is irrelevant for most behaviorists and even behavioral therapists. However, there are some theorists from the behavioral tradition who, early on, measured more than changes in the frequency of specific behaviors. Bandura (1971), using multiple outcome measures (i.e., a behavioral test of avoidance, a fear inventory, and a semantic differential technique to obtain attitudinal ratings), reported changes through modeling that lie not only in behavior but in affect and personalattitude. But even Bandura placed relatively little attention on integrated personal ity change, focusing more on limited aspects of personality. Despite the radical shift from personality-based therapies to behaviorally continued to tobe beaapocket pocketof ofresearch researchand andtheory theorythat thatespoused espoused based therapies, there continued an interest in, and in fact concentrated on, assessing more integrated and less

CHAPTER 36 36 CHAPTER

PSYCHOTIlERAPY PSYCHOTHERAPY OUTCOME OUTCOME ASSESSMENT ASSESSMENT

951

reductionistic aspects of people. For For example, example, Albert Albert Ellis Ellis (1973) (1973) described described rational rationalemotive therapy as a method method of of personality personality change change that that quickly quickly and and efficiently efficiently helps helps the the individual individual to to foster foster and and implement implement his his natural natural human humantendencies tendenciesto to gain gainmore moreindividuality, individuality, freedom freedom of of choice, choice, and and enjoyment enjoyment and and also also helps helps him him to to discipline discipline himself himself so that minimizes his tendencies to suggestible, that he he minimizes his natural natural human human tendencies to be be conforming, conforming, suggestible, and and unenjoying. unenjoying. It It actively actively and and didactically, didactically,as aswell wellasasemotively emotivelyand andbehaviorally, behaviorally, shows shows him him how how to toabet abetand andenhance enhanceone oneside sideofofhis his humanness humannesswhile whilesimultane simultaneously squelching) ously changing changing and and living living more more happily happily with with (and (and not not repressing repressing or or squelching) another side side of of his his humanity. humanity. another (p. 200) 200) (p.

Even more recently Robert Wallerstein (1989) summarized findings from from the Foundation and reiterated the Psychotherapy Research Project of the Menninger Foundation bepsychoanalytic position on change in psychotherapy by drawing a distinction be tween two basic types of change, structural change, change, which which is is based based on on the the interpretive interpretive resolution resolution of of unconscious unconscious structural intrapsychic conflicts, conflicts, and and behavioral behavioral change, change, or or change change in in manifest manifest behavior behavior intrapsychic patterns that that represents represents nothing nothing more more than than altered altered techniques techniques of of adjustment. adjustment. It patterns It is presumed presumed that that only only behavioral behavioral change changecan can result resultfrom fromsupportive supportivepsychothera psychotherais peutic techniques techniques and and implementations. implementations. Intrinsic Intrinsic to to this this dichtomizing dichtomizing between between peutic change has has been been the the assumption assumption that that only only structural structural change, change, as as brought brought kinds of of change kinds has aa guar guarabout through through conflict conflict resolution resolution and and appropriately appropriately achieved achieved insight, about insight, has antee of of stability stability and and durability. durabiUty. antee (p. 203) 203) (p. This longitudinal study was started in 1954 and still continues. Among the quantita quantitative measures used was Luborsky's Health-Sickness Rating Scale (Wallerstein, 1989). Qualitatively measured patients, treatment, and situational variables were taken as well. Positive "structural" change, which occurred in all three of the therapeutic modes, was defined, as "changes in specific intrapsychic configurations, configurations, in the patterning of defenses, in thought and affect affect organization, in anxiety tolerance, and in ego strength" (p. 203). 203). Projective testing provided provided support support for for the clinical evaluation of structural change; however, the reports have not discussed the nature of the projective tests again. A finding that was not predicted by the initial study group was that significant significant structural change was found found in the supportive psychother psychotherapy group, a group that received a mode of psychotherapy in which attempts to foster analytically relevant insight was minimal or nonexistent. The Menninger study is typical of early studies of of psychotherapy that applied devices developed from from Freudian dynamic psychology. Not at all uncommon was the use of projective methodologies, including the Rorschach Inkblot Test, the Thematic Apperception Test (TAT), drawing a person, and sentence-completion methods. Problems with the psychometric qualities of these tests, their reliance on from a theoretical position based on the unconscious have inference, and derivation from resulted in their waning use as indices of outcome. Rarely today does one hear the

952 952

LAMBERT AND SUPPLEE LAMBERT AND SuPFLEE

virtues of such tests for outcome measurement. Changes in the quality of fantasy material as produced on the TAT and other projective tests simply do not convince most researchers and observers that significant improvement has occurred in the actual lives of patients, let alone their personality integration. Although the espoused goal of many different different therapies is to reconstruct the personality (e.g., RET, analytic), the major effects reported by those researching outcome are in the areas of decreased anxiety and pain, increased freedom of action, and so forth. Currently, change in personality, that is, core personality, as a result of therapy therapy is is seldom measured measured and isiseven even considered considered irrelevant irrelevant to to the the goals goals of many treatment studies (Lambert et al., aI., 1986). Currently there is considerable skepticism about the value of personality assessment. This skepticism comes from including: (1) the popularity of behavioral approaches and the various quarters, including: corresponding lack of interest in standard assessment methods that elaborate on corresponding dynamics; (2) (2)the the humanistically humanistically derived derivedbelief belief that thatthe thetesting testingand anddiagnos diagnosinternal dynamics; tic enterprise is itself an unhelpful way of relating to persons seeking help; (3) the growing popularity of cognitive methods with their emphasis on self-talk, irrational beliefs, and similar limited constructs; (4) the belief belief that personality tests do not work very well and have unimpressive validity coefficients because they largely coefficients work measure personality personaUty traits to the exclusion of situational variables; (5) the excessive time required by measures such as the TAT, Rorschach, and MMPI; MMPI; and (6) the current emphasis on brief, crisis-oriented treatments that are short term and merely restore a person to prior levels of functioning or focus on other limited goals (Lambert et et al., (Lambert aI., 1986). 1986). Figure 11 may help the reader grasp the changes in outcome assessment that the past 50 SOyears, years,including includingthose thosejust justdiscussed. discussed.Figure Figure1 1suggests suggests have occurred over the

Change rated by therapist ---• -�. Multiple sources dominated by client ratings Ratings of gross change ---• --i�� Specific change/multiple technology Theory & & personality bound ----.� • Practically important! important/ symptom based symptom based

Change is unidirectional ------�� For better or worse

Change is unidirectional

• For better or worse

Change is unidimensional -----�� Change is Change is unidimensional • multidimensional Change is multidimensional

Changes are stable ---• --�� Changes are unstable

FIGURE 11 Developmental Developmental history history of ofoutcome outcome assessment. assessment. F1GUU

CHAPTER 36 36 CHAnER

PSYCHOTHERAPY OUTCOMEAss ASSESSMENT PSYCHOTIII RAPY OurCOME ESME S NT

953 953

several dimensions upon which assessments have varied since the focus of initial research research to the present. The field field has gradually moved moved from from complete complete reliance on improvement to the use of of outcome indices of therapist ratings of gross/general improvement specific specific symptoms that are quantified quantified from from a variety of viewpoints, including the patient, outside observers, relatives, physiological physiological indices, and environmental environmental data such as employment employment records. Assessment Assessment procedures procedures are becoming becoming more more complex complex and are also relying more heavily on standardized instruments that deal with specific specific kinds of change rather than theory-based measures such as projective tests. Re Researchers are more more sensitive to the possibility of both both positive and negative change change and construct outcome assessment devices that tap both kinds of change. There There growing awareness awareness that change is multidimensional multidimensional and that many facets facets has been a growing of the person change simultaneously and sometimes sometimes dyschronisticly. There of There is an awareness as well as empirical support for the ideas that changes come come intuitive awareness about on multiple multiple levels, that a small intervention intervention may generate generate multiple multiple changes, about singular aspect of human behavior behavior may be effected effected by multiple multiple approaches approaches and that a singular & Schimek, 1964). Changes are not permanent permanent and researchers are (Luborsky & making greater efforts efforts to understand the patterns of change change over over time time rather than making stable or final. final. It is difficult, difficult, however, to ferret out whether whether viewing change as stable changes test, situation (Kenny & & measured are due to the person, the or the situation (Kenny changes Campbell, 1989). ll. n. PERSONALITY CHANGE MEASUREMENT INSTRUMENTS: CURRENT STATUS OF KNOWLEDGE AND PATTERN OF USE itself from Although outcome research has divorced itself from the sole use of theoretically theoretically based, single measures of change, the result has been great divergence in the criteria after reviewing assessment outcome used. Froyd and Lambert (1989), after assessment practices in outcome studies published in 20 major journals between the years of of 1983 and 1988, 1988, found found that no less than 1,430 measures were applied in 348 outcome outcome studies, and of these, 840 were used only once! The type, number, and quality of measures varied greatly across journals, disorders, and treatment methods. Some Some data from from this literature review are presented in Table I. By far, the most frequently instrument was the frequently used instrument Beck Beck Depression Inventory (BDI), ( B D I ) , followed followed by the State-Trait State-Trait Anxiety Anxiety Inventory, weight, and the Hamilton Rating Scale for Depression (HRSD). ( H R S D ) . Of the multitude multitude of remaining diverse instruments, many yield indices of depression, anxiety, global of symptoms, maritaVfamily marital/family relations, cognitions, and self-concept, self-concept, while personality personality measurement, in the traditional sense, was rare. In the Froyd and Lambert Lambert (1989) review only two projective projective measures were readily recognizable, the Rorschach (seen twice) and a human figures drawing test (seen once). Similarly, in an earlier review of outcome outcome studies published in the Journal of of Clinical Clinical and and Consulting Consulting Psychology Psychology (JCCP) (JCCP) between between 1978 and 1982, Journal 1982, outcome measures (Lambert, (Lambert, projective techniques were not mentioned at all as outcome psychoanalytic and psychodynamic psychodynamic 1983). It appears that even those who espouse psychoanalytic

LAMBERT LAMBERT AND AND SUPPLEE SUPPLEE

954 954 TABLE TABU I

Frequency Frequency and and Percentage Percentage of of Measures Measures Used Used

in 348 348 Outcome Outcome Studies Studies According According to to in

Their Content'' ContentO

Content Content Intrapersonal Intrapersonal

Interpersonal

Frequency

% %

1,053

74 74

132 132

9

240

Social role performance Total Total

17 17

1,430

Most frequently used measures in a broad sample of outcome studiesO studies^ Outcome measure

Frequency

Beck Depression Inventory

42

Weight Weight

23 23

& & Trait Trait Forms) Forms) Hamilton Depression Hamihon Rating Rating Scale Scale for for Depression Symptom Symptom Checklist-90 Checklist-90 (R) (R) Self-Efficacy Self-Efficacy Rating Rating Scales Scales Locke-Wallace Locke-Wallace Marital Marital Adjustment Adjustment Scale Scale Blood Blood pressure pressure Heart Heart rate rate MMPI MMPI (all (all forms) forms)

State-Trait (State State-Trait Anxiety Anxiety Inventory Inventory (State

18 18 18 18 14 14 14 14 13 13 12 12 11 11 10 10

" Based Based on on aareview review of of all all outcome outcome studies studies published published in in 20 20 exemplary exemplary

°

journals over a 5-year period. Reprinted by permission of the authors Froyd and Lambert (1989).

orientations utilize more readily interpretable measures such as the HRSD, BDI, and Symptoms Checklist-90(R). As can be seenfrom from the Froyd and Lambert (1989) and Lambert (1983) reviews, many of the outcome devices used in recent times obserare atheoretical measures of symptoms, behavioral checklists, and the direct obser vation of target behaviors.^ behaviors. t The same general trend found in broad reviews of psychotherapy research is also reflected in measurement practices in reviews of single disorders. For example, & Payne (1990) studied the assessment practices used in Ogles, Lambert, Weight & Some controlled outcome studies of agoraphobia published between 1966 and 1988. Some of the results of this review are shown in Tables II and III. ^ItItisisinteresting interestingtotonote, note,however, however,that thatalthough althoughprojective projectivepersonality personalitymeasures measuresmay mayhave havefallen fallen

1

out psychotherapy outcome outcome researchers, researchers, the out of of favor favor with with psychotherapy the Rorschach, Rorschach, TAT, TAT, and and sentence-completion sentence-completion

tests are are still among the the assessment assessment instruments instruments most most widely widelyutilized utilizedby by psychotherapists psychotherapists(Rychlak, (Rychlak,1981). 1981).

CHAPTER CHAPTER 36 36

PSYCHOlll ERAPY OUTCOME PSYCHOTHERAPY OUTCOME ASSESSMENT ASSESSMENT

9SS 955

TABLE TABLE n n Number of Agoraphobia Instruments Used in 170 Outcome Studies by Category of Measure

Number of Agoraphobia Instruments Used in 170 Outcome Studies by Category of Measure

Number of specific scales/measures

Category

Category

Number of specific scales/measures

Fear and anxiety measures

27

Fear and anxiety measures Behavioral measures

27 10 9 10 9 8 39 8 39 6 6 14 22 14 22 135

Depression measures Behavioral Mental evaluation Depression

Unstandardized rating scales (symptom-based anxiety ratings) Mental evaluation Physiological Unstandardized

rating scales (symptom-based anxiety ratings)

Personality Physiologicaland general symptomsG

Personality and general symptoms'* Others Others Total Total

Specific scales and and their their frequency frequency of of use use are are listed listed in in Table Table III. III. Based Based on on Ogles, Ogles, Lambert, Lambert, Weight, Weight, & & c scales G" Specifi

Payne (1990). Payne

data shows showsthat thatover over135 135 separate separatemeasures measureswere wereemployed employed Inspection of these data in the 170 studies analyzed. This is an amazing diversity when one considers that the focus of treatment (agoraphobia), the goals of treatment (anxiety reduction and behavior change), and the interventions employed (mostly behavioral and TABLE TABLE HI ni Personality/General Symptom InstrumentsG

Personality/General Symptom Instruments" Instruments

Frequency of use in 170 studies

Personality Eysenck Personality Inventory 16 Personality FactorInventory Questionnaire Eysenck Personality Minnesota Multiphasic Personality Inventory 16 Personality Factor Questionnaire Maudsley MinnesotaPersonality MultiphasicInventory Personality Inventory Willoughby Personality Inventory Maudsley Personality Inventory Willoughby Personality Inventory Total

3 13 3 2 3 2 2 23

Instruments

Frequency of use in 170 studies

Personality

Symptoms

Hopkins Symptom Checklist-90

13

Wittenborn Psychiatric Rating Scale

17 12 3 3 2 2 2 2 2 2 1

Total

39

Middlesex Hospital Questionnaire

Brief Brief Psychiatric Rating Scale Cornell Medical Index

Tavistock Self-Assessment Self-Assessment

Treatment Emergent Signs and Symptoms

Based on on Ogles, Ogles, Lambert, Lambert, Weight, Weight, G" Based

«& Payne Payne (1990). (1990). &

956 956

LAMBERTAND ANDSUPPLEE SUPPLEE LAMBERT

cognitive therapies) were very limited in scope. It is also surprising that only five personality measures were used (Table III) in the 170 studies examined. There is little evidence for strong interest in personality and changes in personality dimen dimensions reflected in the traditional measures of the core personality traits of of persons who have the symptoms of agoraphobia. Of central interest to the psychotherapy researchers is the degree to which different types of measures reflect the size of different measures and different of changes that are occuring as a result of therapy. As part of the analysis undertaken by Ogles et effect size of various measures and classes of measures was estimated. al (1990), the effect assessed patients with the This analysis combined the results of several studies that assessed reflected the differences differences same measures (or classes of measures) so that comparisons reflected between measures rather than differences in treatment type, patient population, and procedural differences, such as the time lapse between pre- and post-testing. The results of this analysis are presented in Table IV. The data in Table IV suggest that the more that dependent measures assess the limited and actual targets of of treatment, the larger the treatment effects. Thus, measures that tap changes in anxiety and avoidance in high-stress agoraphobic agoraphobic situations resulted in larger treatment effects than measures that tap fear generally (such as the Fear Survey Schedule) or personality in the broad sense. This finding is not unique to agoraphobia and can be seen in other meta metaanalytic comparisons of various disorders. Tables V and VI suggest this trend, although they are not based on within-group comparisons and are thus less sound methodologically than the within-group data on agoraphobia. To the psychotherapy researcher, data like these further further reinforce the value of of general symptom of employing symptom-specific symptom-specific measures prior to or instead of TABLE IV IV

Overall Effect Effect Size (ES) Means and Standard Deviations by Scale

N" N°

MES

SDES

65 65

2.66 2.66

11.83 .83

Self-Rating Severity Self-Rating

31 31

52 52

2.30

1.14

Fear Questionnaire

56

1.93

1.30

Behavioral Approach Test (BAT)

54

11.15 . 15

1.07

Scale Scale

Phobic Anxiety Anxiety Phobic

& Avoidance Avoidance &

Global Assessment Scale

Anxiety Anxiety during during BAT BAT

Depression measures

Fear Survey Schedule Heart rate

48 48

60

26 21 21

2.12 2.12 1.36 1.36 11.11 .11

.99 .99 .44 .44

1.55 1.55

.85 .85

.72

.47

.56

G* N N represents representsthe thenumber numberof oftreatments treatmentswhose whoseeffects effects were weremeasured measured by each scale. scale. A study may contribute more than one treatment. Reprinted from from Ogles, Lambert, Weight, the American American Psychological Association. Association.

& of & Payne by permission of

PSYCHOTHERAPY OUTCOME AsSESSMENr ASSESSMENT PSYCHOTIlERAPY

CHAPTER CHAPTER 36 36

957 957

TABLE V V

The Relationship of Assessment Content to Psychotherapy Outcome:

The The Result Result of of Choosing Choosing aa Target Target Problem Problem on on Outcome Outcome across across Therapies Therapies

Number of of effect effect sizes sizes Number

Resulting Resulting effect effect size size

Fear/Anxiety Fear/Anxiety

719

11.06 .06

Social Behavior Behavior Social

391 391

.95

Target Target problem

Adjustment Adjustment

Self-Esteem

Personality traits Personality Achievement

.96

66 66

.95 .95

164

.52

30 30

.28 .28

139 139

Note. Note, Based on Shapiro and Shapiro (1982).

measures and personality inventories. If you want to demonstrate the power of a treatment, do not measure outcome with personality tests or inventories that tap underlying styles, attitudes, or the patterns that are usually sought in personality assessment. These results do suggest that the personality measures often applied in psychotherapy research may measure what they sometimes purport to measure: the relatively enduring attitudes, interests, needs, preferences, and patterns that make personality stable. Alternatively, traditional personality measures may not be sensitive to "stable" "stable" attributes that do change through psychotherapy.

TABLE VI VI Average, Measured Average, Standard Standard Deviation, Deviation, and and Number Number of of Effect Effect Sizes Sizes Classified Classified by by Type Type of of Outcome Measure* Outcome Outcome measure measure category category Fear-anxiety Fear-anxiety

Average effect effect Average

Standard Standard deviation deviation

Number Number of of effects effects

1.12 1.12

11.72 .72

647 647

0.84

0.56

0.86

70

383

1.13

59

Vocational or personal development

0.85

Measures and ratings of global

0.80

Addiction

0.77

0.52

55

Self-esteem

0.69

0.97

99 99

Work or school achievement Life indicators indicators of of adjustment adjustment Life Personality traits

0.49

0.76

215 215

0.31

0.56

18

Emotional-somatic complaint adjustment

Physiological Physiological stress stress Sociopathic Sociopathic behaviors behaviors

0.71

0.64 0.64 0.46 0.46

"Based Based on on Smith, Smith, Glass, Glass, and and Miller Miller (1980), (1980), p. p. 109. 109.

a

Note. Note. Reprinted Reprinted by by permission permission Johns Johns Hopkins Hopkins University University Press.

11.08 .08

0.82

0.43

50 50

94 94 35 35

958 958

LAMBERT AND SUPPLEE

m. IN i n . WHAT ASPECfS ASPECTS OF PERSONALITY ARE BEING MEASURED IN PSYCHOmERAPY OUTCOME RESEARCH? PSYCHOTHERAPY It is abundantly clear from from our review of psychotherapy psychotherapy outcome research that the field, field, in general, has moved a long way from interest in personality change per se. The common practice in contemporary research is to offer offer specific treatments for specific disorders and measure change with three to six scales rather closely related exto the targets (symptoms and symptom complexes) of treatment. As may be ex further into the abstract domain pected when researchers study variables that move further Hkely to see large changes, although the of personality functioning, they are less likely of theoretical interest. small changes that do occur in the personality domain may be of There remains some research interest in tapping the more general and abstract dimensions of persons that are often thought of as part, if not central aspects, of of turn now to a discussion of some of these personality dimensions personality. We tum inferences about personality that can be drawn from current psychother psychotherand to inferences apy research. A. Depression Depression A. Depression is a clear example of a disorder that has been popularly characterized from a symptomatic perspective. The DSM-IV and DSM-III-R have and diagnosed from for diagnosing provided relatively comprehensive, atheoretical, and concrete criteria for depression, and many of the methods used to define depression reflect that some 1983). The theoretical orientations from from which orientation (Moran & Lambert, 1983). foundation of these measures have been derived have also been the foundation of the most widely studied forms of therapy with depression, that is, the cognitive, behavioral, cognitive cognitivebehavioral, and pharmacological therapies (HoUon (Hollon & Beck, 1994; Shapiro & ShaSha piro, 1982). Depression has been one of the more popular topics of treatment outcome research and is measured both in studies of depression and as a variable of interest in other studies of psychotherapy such as marital distress ((Jacobson, Jacobson, Follete, & Pagel, 1986), 1986), anxiety (Borkovec & Mathews, 1988), and substance abuse (Alden, 1988). The Beck Depression Inventory was the most widely used instrument in Froyd and Lambert's (1989) survey. From the time of Lambert's 1976-1980 survey S-year survey undertaken in the late eighties, use of the instrument has risen to a 5-year from 5 times to over 25 times in JCCP alone. from Depression was also frequently frequently measured with the Hamilton Depression Depression Rat Rating Scale based on an assessment interview. Both instruments tap the cardinal signs of depression but do not attempt (an in fact avoid) measuring broader aspects of of meapersonality. Most of the depression measures have been described as "state" mea sures (Moran & Lambert, 1983), sampling signs, and symptoms from 2 weeks to a in the past. Scores Scores from from these thesemeasures measures can can certainly certainly be be interpreted interpreted couple of months in to be an indication of certain active personality variables as they tap physiological

CHAPTER 36 CHAPTER

PSYCHOTHERAPY OUTCOME ASSESSMENT PSYCHOTIfERAPY

959 959

functioning functioning in the form of sleep, gastrointestinal symptoms, and sexual interests, as well ll;ass thoughts and mood. Changes in any of these dimensions would indicate a change in some aspect of a person's personality. personaUty. But little is known about the degree to whiCh these scales tap deeper, more stable personality which personaUty dimensi9n�. dimension?. They correlate highly with other symptomatic measures ·of depression and anxiety, as of well as other measures of psychopathology (Moran & Lambert, 1983). There is a clear need for research on such personality correlates as dependency, passivity, and overreliance overreUance on certain psychological defense mechanisms or coping strategies. for instruments that have been devised These comments are equally true for directly from of depression. Cognitive con from cognitive theories and the treatment of constructs differ on a number of important aspects, including the level of inference differ inference and depth of awareness tapped by various measures. Psychologists interested in focused on such surface surface concepts as positive and nega negacognitive interventions have focused tive self-statements, irrational beliefs, and cognitive distortions. These concepts have been measured with tests such as the Automatic Thoughts Questionnaire reference to dee deeper personality dimensions. There is a high correlation without reference per personality between these cognitive scales and measures such as the BDI. But much less is known about changes on these dimensions and their relationship to changes in personality dimensions such as depressive schemata. The failure failure deeper cognitive personality to understand the deeper meanings (if any) of scores on scales like the HRSD for psychotherapy research and for and Beck Inventory have several implications for appropriate selection and treatment assignment. There is reason to believe that the assessment of personality dimensions in addition to symptoms could have a positive effect effect on clients who suffer suffer with depres depresclassifying some forms of mood disorders (e.g., sion. There is a movement toward classifying dysthymia) as a character disorder (Kocsis & Francis, 1987), drawing parallels between dysthymia and personality disorders. Akiskal (1983) has delineated several "character-spectrum dysphoric pattern" which has forms of depression, including a "character-spectrum pharmacotherapy and psychother psychothera slow, early onset and is particularly resistant to pharmacotherapy apy. In addition there appears to be growing consensus among clinicians that dede often mask underlying personality disorders that are not ad adpressive symptoms often dressed in many treatments. The presence of a personality disorder that goes unassessed and untreated probably leads to the high level of relapse in depressed patients. Thus the undue emphasis of symptoms, characteristic of contemporary research on depression, fails to deal with the complexity of persons with this disorder and often often leads to limited understanding of treatment-by-patient treatment-by-patient interactions. We see a similar pattern in the assessment of changes in anxiety. B. B.

Anxiety Anxiety

Related to and often coexisting with depression is anxiety. In recent years anxiety frequently measured symptom complex. The instrument has been the second most frequently most often used is AI; Spielbeger, Gorsuch, & is the State-Trait State-Trait Anxiety Inventory (ST (STAI; & Lushene, 1970), which attempts to deliver an index of both lasting (trait) and

LAMBERT AND SUPPLEE SuppLEE

960

transient (state) anxiety; anxiety; a composite score may also be obtained. Most of the studies have found change to occur in both the state and the trait anxiety scores of the ST AI. These changes are frequently STAI. frequently noted along with changes in other depression, abstinence abstinence from from drinking drinking (Alden, (Alden, 1988), 1988), psychological variables such as depression, Type A behavior pattern (K. R. Kelly & & Stone, 1987), and and specific specific phobic phobic anxiety anxiety & Mathews, Mathews, 1988). 1988).Maintenance Maintenanceof ofdecreased decreasedanxiety anxietyscores scoreson onfollow follow(Borkovec & up may indicate alteration in one's behavior pattern; however, it is questionable STAI, whether trait anxiety is actually being measured, particularly with the ST AI, as the construct validity of the Trait scale has been called into question (Chaplin, 1986). So even the Trait scale may be tapping more superficial and transient aspects of personality. Like depression, research on anxiety-based disorders suggests an emphasis oli oh the the most most symptomatic symptomatic aspects aspects of of the the person person with with little little regard regard for for the the emphasis personality dynamics dynamics or or patterns patterns that that bring bring anxiety anxiety into into play. play. As As with with depression, depression, personality emphasis in in future future research research might might profitably profitably be be placed placed on on these these patterns, patterns, dynamics, dynamics, emphasis or structures structures to to increase increase the the likelihood likelihood that that more more permanent permanent change change will will result result or from psychological psychological interventions. from interventions. There are numerous personality personaUty traits and patterns that can be identified as disorders. One area of contemporary concern that underlying anxiety and related disorders. can serve as an illustration is research research into into the the interaction interaction between between personality personality and (PMS). Interest Interest in in this this syndrome syndrome isisevident evident in in its its inclusion inclusion premenstrual syndrome (PMS). as a narrow classification, Premenstrual Dysphoric Disorder, under consideration diagnosisininthe theDiagnostic Diagnosticand andStatistical StatisticalManual Manual(American (AmericanPsychiatric PsychiatricAssoci Associas aadiagnosis 1994). Although pharmacological and psychological interventions focus on ation, 1994). the symptoms present during the acute phase of the disorder, there is considerable interest in the personality and life event factors that may predict and exacerbate (1986) as well as Palmer (1986) (1986) have this disorder. Keye, Hammond, and Strong (1986) outlined several personality traitc; traits of at least a portion of women with severe PMS. PMS. personality pattern pattern suggests suggests that that the the subtype subtype demonstrates demonstrates aastable stablepersonality personality This personality pattern, even in the postmenstrual phase, that that isis characterized characterized by byverbal verbalacting actingout, out, undercontrol with insufficient insufficient thinking and deliberation, observer-perceived drama and self-centeredness, and strong needs for affection affection but with conflicts over depen dependency. dency. They were also described as being demanding yet sensitive to the demands of others as well as crying easily and being depressed. Given this personality configuration, these women are especially vulnerable to the extremely distressing premenstrual episode as well as to environmental related to to their theirpattern pattern of ofinteracting interactingwith withfamily familymembers. members.Thus Thustreatment treatment stressors related aimed at and assessments targeted only toward the acute symptoms may have little lasting effect. Although a demanding request, psychotherapy and psychotherapy research may may be be most most productive, productive, in in the the long long run, run, by by attending attending to to personality personality research dimensions in in assessing assessing changes changes in in patients. dimensions

C. Self-Concept Self-Concept C. In Froyd and Lambert's (1989) review, there are several dimensions that were assessed and can be identified as reflecting aspects of traditional personality func-

CHAPTER 36 36 CHAPTER

PSYCHOTHERAPY OUTCOME ASSESSMENT PSYCHOlllE RAPY OurCOME

961

tioning that date back to the original client-centered research and social-learning theory. One such aspect is is people's views of themselves as measured by self-concept self-concept scales, measures of self-esteem, and self-efficacy self-efficacy ratings. Measures of self-concept self-concept have been used in such diverse research areas as eating disorders (Dworkin & Kerr, 1987; 1987; Leitenberg, Rosen, Gross, Nudelman, & Vara, 1988), depression (Graff, (Graff, Whitehead, & LeCompte, 1986), marital therapy (Snyder & Wills, 1989), and crisis intervention (Viney, Clarke, Bunn, & Benjaman, 1985). 1985). For example, Williams, performance desensitiza desensitizaTurner, and Peer (1985), comparing guided mastery and performance found that perceived self-efficacy self-efficacy "was a uniformly accurate predictor of tion, found of effects regardless of the treatment received." treatment effects Like depression and anxiety, self-concept appears to be an important dimen dimension in a wide variety of disorders and to more closely resemble traditional personal personalexample with with aa disorder of of contemporary contemporary concern concern might might ity/theoretical concerns. An example be bulimia nervosa. In this disorder, compensation for of for low self-esteem and fears of as reasons behind the bulimic's wishes and compulsive rejection have been proposed as attempts to achieve an idealized body. Dworkin and Kerr (1987), in a study of of women with body image problems, found self-cathexis (self-concept) to be related to body cathexis. Despite the fact that self-concept could be considered central in eating disorder research, it is more common for researchers to study weight and possibly eating attitudes than self-concept self-concept per se. In a study by Leitenberg et al. (1988) comparing exposure plus response prevention with cognitive-behavioral therapy for for bulimics, buUmics, all treatment groups were found found to experience increases in self-esteem. The authors, however, found no difference in weight increase or decrease for any of the groups. Similarly, significant difference in a review of anorexia nervosa therapy outcome research (reporting on more than studies). Qualls Quails and Berman (1988) also found a lack of relationship between 100 studies), weight gains and psychological improvement. The lack of concordance between these measures of outcome suggests the need to measure both and to remain cautious about the common-sense link between them. While the targets targets of treatment treatment often often reflect both symptoms and self-perception, treatment outcome more often often is limited to weight gain or eating attitudes. Again, changes in personality structure seem an important but neglected topic even in outcome studies that use measures self-concept. of self-concept. D. General General Issues Issues D. Beyond the three broad areas already outlined (depression, anxiety, and self selfis little cohesion between the remaining measures that is is particularly concept), there is meaningful physio meaningful for the study of personality. Direct observation of behaviors and physiological measures, such as weight loss, skin conductance, heart rate, and number of of 25% of the outcome measures used times behaviors occurred, represent about 25% 1983). These measures do not readily translate 1989; Lambert, 1983). (Froyd & Lambert, 1989; into personality indices, although one can infer that lasting change in weight loss, smoking, or even physiological measures such as skin conductance and blood cate-

962

LAMBERT AND AND SUPPLEE SuppLEE

cholamine levels indicates a change in in some pattern of psychological functioning (i.e., in personality). Linking changes in in these behaviors to changes in personality is a difficult task. Within the trend of measuring specific rather than global improvement is the effort to tailor change criteria to the individual client. In other words, if a client seeks help for severe depression and shows little evidence of pathological anxiety, the emphasis would be on changes in depression rather than changes in anxiety. Attempts to clarify the specific complaints and create specific therapeutic goals are reflected in measures such as the Goal Attainment Scaling, the Battle Target & Lambert, 1989; Lambert, 1989; Lambert, Complaints, and other "homemade" measures (Froyd & 1983). 1983). The problems with these measures are that as a means of structuring the patient's goals they leave open the possibility that the individualized goals will become or remain poorly defined subjective decisions created by the patient or the clinician. Though they undoubtedly reflect personality, they are attempts to elimielimi nate personality factors and focus on observable behaviors. Measures of relationships and social role performance have been occasionally used in outcome research. Several procedures have been found to be useful in significant measuring how the individual functions in the world, from the view of significant others (e.g., the Personal Adjustment and Role Skills, The Katz Ajustment Scale, and the Social Adjustment Scale), from self-report (e.g., the Social Adjustment Scale, SR version), and from trained observers (e.g., the Structural Analysis of Social Behavior Observations). These scales provide factors from which variables, helplessdirectly and indirectly tied to personality, can be derived, such as anxiety, helpless DeJulio, 1983; ness, suspiciousness, and negativism (Lambert, Christensen, and Dejulio, al., 1986). Lambert et aI., behavRather than inferring changes in personality from symptom scales and behav ioral observations, one might think that outcome researchers would employ multidi multidimensional personality inventories to make this task unnecessary. However, there is a decidedly clear lack of interest in including such scales in outcome studies. Although multidimensional personality inventories such as the MMPI, Edwards Personal Preference Schedule, and California Psychological Inventory continue to be used in clinical practice, they are not used in outcome research. The Edwards Personal Preference Schedule was used once over a 5-year span, utilizing only (and finding findingchanges changesin) in)the theAffiliation Affiliationand andIndependence Independencescales scales(Jacobson (JacobsonetetaI., al.,1986). 1986). While in times past the MMPI has been the most popular objective personality measure in outcome research (Beutler & & Crago, Crago, 1983), 1983),itit isis used used today today far far less less than than specific symptom indices (Lambert, 1983). 1983). In a 5-year review the MMPI was found only 10 times, and only then with selected scales (i.e., depression) or the sum of & Lambert, 1989). the clinical scales used (Froyd & The Eysenck Personality Inventory, while recommended as having a more reliable factorial structure and as being a better general measure of change in & Crago, 1983), was used once in a 5-year personality than the MMPI (Beutler & (Graff et al, review (Graff aI., 1986), with less neuroticism found as a result of therapy. The same can be said for the ipsative measures that came from Kelly and AUport Allport

CHAPTER 36 36 CuAPl'ER

PSYCHOTHERAPYOUTCOME OUTCOMEASSESSMENT ASSESSMENT PSYCHOTIlERAPY

963 963

(Kiesler, 1971), whose theories have been influential influential in American American psychology: the instruments they created created are rarely seen. Kelly's repertory grid is virtually nonexis nonexistent in outcome research literature in the United United States. The authors's experience experience interesting and often often enlighten enlightenhas been that, although subjects find the Rep grid an interesting computer programs ing instrument, it is quite arduous to complete. And although although computer programs eschew its use. continued to eschew are available for scoring, researchers have continued A A paradox paradox of psychotherapy psychotherapy research is that while theoretical theoretical interest in personality change change remains high, its measurement measurement is either not undertaken or, when personality changes in personality personality dimensions are small in comparison comparison to symptomatic symptomatic it is, changes more clinically meaningful meaningful and changes. These smaller changes may actually be more important for relapse prevention prevention and the suggest reorganizations that could be important maintenance of treatment gains, yet they are increasingly increasingly eschewed eschewed by psychother psychotherapy researchers researchers who w h o fear fear that broad personality personality measures may not reflect reflect some some of A t least for the practical outcomes of therapy interventions and patient change. At psychotherapy researchers in the United States are content to assess for the present, psychotherapy form or discrete symptoms symptoms and limited behaviors. the most part micro changes in the form Medard Boss, in a discussion discussion on reality, posited, "What makes a rock a rock?" answer to all of these Is it its shape? Is it its texture? Its color? Its weight? The answer questions is no, but take them away and there is no rock (Rychlak, 1981). What then is personality? Is it a reaction to stress in terms of anxiety, or anger, or B D I score? compliance? Is it weight change, or number of cigarettes smoked, or BDI Again we might answer, "No, "No, but take them away and there is no concept-no concept—no personality.'* The purpose here is not to resurrect resurrect the Kantian Kantian/Lockean personality." lLockean debate, exclude any but to emphasize the fact that there is no easy way to include or exclude psychological variable when discussing the nature of personality. But if anything is psychotherapy is measured, it is that clear from our review of the way change in psychotherapy different than the way persons were the way persons are studied today is highly different outcome research. It is more appears studied in the first therapy outcome more practical, yet it appears more fragmented fragmented and even chaotic, and it is based based only only peripherally peripherally on personality personality demands theory. Traditional personality measures have proved to be ill suited to the demands of psychotherapy psychotherapy research and, perhaps, even more ill ill suited to the social, political, of academic climate in which research is undertaken today. What implications implications and academic psychotherapy outcome outcome and personality? does this have for psychotherapy IV. SUMMARY, CONCLUSIONS, AND FUTURE DIRECTIONS IV.

Our review review of psychotherapy psychotherapy research research and personality personality change suggests suggests a number number recommendations: of conclusions conclusions and recommendations: of 1. Psychotherapy Psychotherapy research has changed a great deal from from its origins as inten intensive examination or individual cases through the lenses of psychoanalytically ori of psychoanalytically orifrom an integrated integrated ented therapists who both conceptualized and rated change from personality theory was central. framework in which personality framework

964

AND SUPPLEE LAMBERT AND

2. The movement from a theory-based analysis has been gradual and incomplete but owes its impetus to several identifiable forces—the forces-the emergence of behaviorism, the failure of traditional personality assessment devices to reflect patient change, changes in the Diagnostic and Statistical Manual with its greater emphasis on observable symptoms, the emergence of short-term therapies, the political and financial financial pressure to demonstrate practically important changes, and the proliferation of competing therapies, many of which are only loosely tied to personality theory. We live in a practical world-a world—a world less willing to tolerate academic endeavers unless they show some immediate benefit for the "consumer" of services. 3. The proliferation of theories, and the diversity of thought on behavior change, has resulted in an exponential rise in methods and devices for assessing changes in patient status. Very few of these measures can be accurately called measures of personality change, personality personaUty integration, or even personality. 4. In many ways this appears to be a positive development for the patient of receiving treatment. We can manage to identify and measure discrete aspects of personal functioning and we can test the effects of interventions on these discrete areas of functioning. The treatment situation is a competitive market place, and the diversity enhances and broadens the search for causal relationships and effec effective interventions. 5. Despite the positive side of the assess the current status of treatment outcome assessment, we have a sense that the field field is in chaos. With large numbers of new assessment instruments created every year and few being used more than once, it is hard to believe that the empirical side of psychotherapy will advance at a fast at all) as a whole. There will undoubtedly be pockets of rapid development, pace (if at but but as as a discipline psychotherapy and and behavior change seems to to lack lack orderly orderlyprogress progress toward the goal of empirically based practice with a clear understanding of mechamecha nisms of change. 6. As the crisis grows and becomes apparent, it seems possible that it will provide the opportunity, if not the necessity, for reorganization. This propitious moment cannot be far off. We cannot predict or imagine the theory of personality change that can unify even a small section of thefield. field. Eclectic approaches perhaps The Handbook of come the closest to forcing an integration. But as any reader of The of EclecticPsychotherapy Psychotherapy(Norcross, (Norcross,1986) 1986)ororthe theJournal Journal Integrative Eclectic Eclectic ofof Integrative andand Eclectic cansee, see,eclectic eclectictheories theoriesthemselves themselvesare arebeing beingdeveloped developedwith withlight lightPsychotherapycan Psychotherapy Uttle impact. ening quickness but little 7. Nevertheless, we remain optimistic about the future and call for the interinter ested theorist or dedicated student to lend a hand. It does seem that personality in dramatic imim theory combined with continued research may eventually result in provements in the way we work with people who are suffering and need our expertise. A beginning point may be to attempt a more thorough understanding of the meaning of symptomatic change for personality organization.

R 36 CHAPTER 36 CllArTE

PSYCHOTHERAPY OUTCOME OUTCOME Ass ASSESSMENT PSYCHOnlERAPY ESSMENT

965 965

REFERENCES Akiskal, H. S. (1983). Dysthymic disorder: disorder: Psychopathology of proposed chronic depressive subtypes. American 140, AmericanJournal JournalofofPsychiatry, Psychiatry, 140,11-20. 11-20. Alden, L. E. (1988). (1988). Behavior Behaviorself-management self-management and and controlled-drinking controUed-drinkingstrategies strategiesin in aacontext context of secondary prevention. Journal Clinical Psychology, 56, 56, 280-286. Journalof ofConsulting Consultingand and Clinical Psychology, 280-286. American Psychiatric Psychiatric Association. Association. (1994). (1994). Diagnostic Diagnosticand andstastical stastical manual of mental disorders manual ofmental disorders American DC: Author. (4th Rev. ed.) Washington, DC: Bandura, A. (1971). (1971). Behavior Behaviormodification. modification.New NewYork: York:Holt, Holt,Rinehart Rinehart&& Winston. Winston. Bandura, Bergin, A. A. E. E. (1971). (1971). The The evaluation evaluation of of psychotherapeutic psychotherapeutic outcomes. outcomes. In In S. S. L. L. Garfield Garfield & Bergin, A. E. Bergin (Eds.), The The handbook handbook of of psychotherapy psychotherapy and andbehavior behaviorchange change(1st (1sted., ed., pp. pp. 217-270). New York: Wiley. Bergin, A. E., & Lambert, M. M. J. (1978). (1978). the evaluation of therapeutic outcomes. In S. L. Garfield & A. E. Bergin (Eds.), The The handbook handbook of ofpsychotherapy psychotherapyand and behavior behaviorchange change pp. 139-190). New York: York: Wiley. (2nd ed., pp. Berzins, J. I. (1977). Matching therapists and patients in psychotherapy. In A. S. Gurman & & A. M. Razin (Eds.), Effective A Ahandbook ofofresearch (pp.(pp. 442-481). Effectivepsychotherapy: psychotherapy: handbook research 442-481). New York: Pergamon Press. Beutler, L. E., & Crago, M. (1983). Self-report measures of psychotherapy outcome. In M. J. Lambert, E. R. Christensen, R. Edwin, & assessment & S. S. Dejulio DeJulio (Eds.), (Eds.), The The assessment of of psychotherapy psychotherapyoutcome outcome(pp. (pp.453-497). 453-497).New NewYork: York:Wiley. Wiley. M. (1988). Treatment of nonphobic anxiety disorders: disorders: A Borkovec, T. D., & Mathews, A. M. comparison of nondirective, cognitive, and coping desensitization therapy. Journal Journalof of 877-884. Consulting 56,56, Consultingand andClinical ClinicalPsychology, Psychology, 877-884. Byem, D. (1964). (1964). Assessing Assessing personality personality variables variables and and their their alteration. alteration. In In P. P. Worchel Worchel & Byern, D. Byrne (Eds.), Personality Personalitychange change(pp. (pp.38-68). 38-68).New NewYork: York: Wiley. D. Wiley. Carson, R. R. C. C. (1969). Interaction Interactionconcepts conceptsofof personality. Chicago: Aldine. personality. Chicago: Aldine. Carson, Chaplin, W. W. F. F. (1986). (1986). Strait-trait Strait-trait Anxiety Anxiety Inventory. Inventory. Test Test Reviews, Reviews,1,1,626-632. 626-632. Chaplin, Craik, K. K. H. H. (1986). (1986). Personality Personality research research methods: methods: An An historical historical perspective, perspective. Journal Journal of of Craik, Personality, Personality,54, 54,19-51. 19-51. Dworkin, S. S. H., H., & Kerr, Kerr, B. B. A. A. (1987). (1987). Comparison Comparison of of interventions interventions for for women women experiencing experiencing Dworkin, body image problems. Journal 34,34, 136-140. Journalof ofCounseling CounselingPsychology, Psychology, 136-140. Ellis, A. (1973). Rational-emotive therapy. In R. Corsini (Ed.), Current Currentpsychotherapies psychotherapies(pp. (pp. 167-206). Itasca, IL: IL: F. E. Peacock. Eysenck, H. J. (1987). The role of heredity, environment, and "preparedness" "preparedness" in the genesis of neurosis. In H. J. Eysenck & I. Martin (Eds.), Theoretical Theoreticalfoundations foundationsofofbehavior behavior therapy therapy (pp. (pp.78-97). 78-97).New NewYork: York:Plenum PlenumPress. Press. Eysenck, H. J., & & Beech, R. (1971). Counterconditioning and related methods. In S. L. Garfield & A. E. Bergin (Eds.), The The handbook handbook of ofpsychotherapy psychotherapyand andbehavior behaviorchange change (lst pp. 543-611). New York: Wiley. (1st ed., pp. outcome Froyd, J. J. D., D., & Lambert, Lambert, M. J. J. (1989, (1989, April). April). A A survey surveyand and critique critiqueofofpsychotherapy psychotherapy outcome measurement. measurement.Poster Posterpresented presentedatatthe theWestern WesternPsychological PsychologicalAssociation, Association,Reno, Reno,NV. NV. Garfield, S. L. (1978). Research on client variables in psychotherapy. In S. L. L. Garfield & AnAn empirical of psychotherapy psychotherapyand andbehavior behaviorchange: change: empirical A. E. Bergin (Eds.), Handbook of analysis analysis (2nd (2nded., ed.,pp. pp.191-232). 191-232).New NewYork: York:Wiley. Wiley. (1986). The The handbook of ofpsychotherapy psychotherapyand and behavior behavior Garfield, S. L., & Bergin, A. E. (Eds.). (1986). change(3rd (3rded.). ed.).New NewYork: York:Wiley. Wiley. change

LAMBERT AND SUPPLEE SuppLEE

966

Graff, R R. W., Whitehead, G. I., L, III, & & LeCompte, LeCompte, L. L. (1986). (1986). Group Group treatment treatment with with divorced divorced women using cognitive-behavioral cognitive-behavioral and andsupportive-insight supportive-insight methods. methods.Journal Journalof ofCounsel Counsel-

ing Psychology, Psychology,33, 33,276-281. 276-281.

Hollon, HoUon, S., & & Beck, A. T. (1994). Research on cognitive therapies. In A. E. Bergin & S. L. Garfield (Eds.), Handbook of of psychotherapy psychotherapy and andbehavior behaviorchange change(4th (4thed., ed.,pp. pp.428428466). New York: Wiley. Jacobson, N. Predicting who will benefit from behavioral N. S., S.,Follete, FoUete,W. W.C., C,&&Pagel, Pagel,M.M.(1986). (1986). Predicting who will benefit from behavioral marital therapy. Journal Psychology, 54, 54, 518-522. Journal of of Consulting Consultingand andClinical Clinical Psychology, 518-522. Kelly, G. A. (1955). The ofof personality. Thepsychology psychologyof ofpersonal personalconstructs: constructs:Vol. Vol1. LA Atheory theory personality. New York: Norton. R, & Kelly, Kelly, K. K. R., & Stone, G. G. L. L. (1987). Effects Effects of of three three psychological psychological treatments treatments and and self selfmonitoring on the reduction of type A behavior. Journal Journalof of Counseling CounselingPsychology, Psychology,

34, 34, 46-54.

Kenny, Kenny, D. D. A., & & Campbell, Campbell, D. D. T. T. (1989). (1989). On On the the measurement measurement of ofstability stability in inover-time over-time data. data.

Journal Journal of of Personality, Personality,57,57,445-481. 445-481.

W. R, R., Hammond, Hammond,D. D.C., C,& &Strong, Strong,T.T.(1986). (1986).Medical Medicaland andpsychological psychologicalcharacteristics characteristics Keye, W. Keye, of of women women presenting presenting with with premenstrual premenstrual symptoms. symptoms. Obstetrics Obstetrics and and Gynecology, Gynecology,68, 68, 634-637.

Kiesler, D. D. J. J. (1971). Experimental Experimental designs designs in in psychotherapy psychotherapy research. research. In In A. A. E. E. Bergin Bergin & Kiesler, & The handbook handbookof ofpsychotherapy psychotherapyand andbehavior behaviorchange change(1st (1sted., ed., S. L. L. Garfield Garfield (Eds.), (Eds.), The S. pp. 36-74). New New York: York: Wiley. pp. Kocsis, J. J. H., & Francis, Francis, A. A. J. J. (1987). A A critical critical discussion discussion of of DSM-III dysthymic disorder. disorder. DSM-III dysthymic H., & Kocsis, American Journal Journalof ofPsychiatry, Psychiatry,144, 144, 1534-1542. American 1534-1542.

Krasner, L. L. (1971). The The operant operant approach approach in in behavior behavior therapy. therapy. In In A. A. E. E. Bergin Bergin & & S. S. L. Krasner, Garfield (Eds.), (Eds.), The The handbook of psychotherapy psychotherapy and andbehavior behaviorchange change(1st (1sted., ed.,pp. pp. handbook of 612-652). New New York: York: Wiley. Lambert, M. M. J. J. (1983). Introduction Introduction to to assessment assessment of of psychotherapy psychotherapy outcome: outcome: Historical Historical Lambert, perspective and current issues. In M. M. J. Lambert, E. R. Christensen, & S. S. S. Dejulio DeJulio outcome (pp. 3-32). New York: Wiley. (Eds.), The The assessment assessmentofofpsychotherapy psychotherapy outcome (pp. 3-32). New York: Wiley. Lambert, M. J., J., Christensen, E. R., R, & ofof psychotherapy & Dejulio, DeJulio, S. S. S. S. (1983). (1983). The Theassessment assessment psychotherapy outcome. outcome. New NewYork: York:Wiley. Wiley. Lambert, M. J., Shapiro, D. A., & & Bergin, Bergin, A. A. E. E. (1986). (1986).The The effectiveness effectiveness of of psychotherapy. psychotherapy. behavior In S. L. Garfield & & A. A. E. E. Bergin Bergin (Eds.), (Eds.), The Thehandbook handbookofofpsychotherapy psychotherapyand and behavior change change(3rd (3rded., ed.,pp. pp.157-212). 157-212).New NewYork: York:Wiley. Wiley. Lazarus, A. A. (1971). Behavior Behavior therapy therapyand andbeyond. beyond.New NewYork: York:McGraw-Hill. McGraw-Hill. Leitenberg, H. Rosen, J. c., C, Gross, J., Nudelman, S., & & Vara, L. L. S. (1988). Exposure plus response-prevention treatment of bulimia buhmia nervosa. Journal Journalof ofConsulting Consultingand andClinical Clinical Psychology, Psychology, 56, 56,535-541. 535-541.

Luborsky, L., L., & &Schimek, Schimek,J.J.(1964). (1964).Psychoanalytic Psychoanalytictheories theoriesofoftherapeutic therapeuticand anddevelopmental developmental change: Implications for assessment. In P. Worchel & & D. Byrne (Eds.), Personality Personality change change (pp. (pp.73-99). 73-99).New NewYork: York:Wiley. Wiley. Luborsky, L., & & Spence, D. P. (1978). Quantitative research on psychoanalytic therapy. In S. L. Garfield & & A. E. Bergin (Eds.), The The handbook handbookof ofpsychotherapy psychotherapyand andbehavior behavior change change (2nd (2nded., ed.,pp. pp.331-368). 331-368).New NewYork: York:Wiley. Wiley. Moran, P. W., & Lambert, M. J. (1983). A review of current assessment tools for monitoring Dejulio (Eds.), changes in depression. In M. J. Lambert, E. R R. Christensen, & & S. S. S. DeJulio The outcome (pp. 263-303). New York: Wiley. The assessment assessmentofofpsychotherapy psychotherapy outcome (pp. 263-303). New York: Wiley. York: Guilford Norcross, J. (1986). Handbook of of eclectic eclecticpsychotherapy. psychotherapy.New New York: GuilfordPress. Press.

CHAPTER CHAPTER 36 36

PSYCHOTHERAPY OUTCOME OUTCOME ASSESS ASSESSMENT PSYCH011lERAPY MENT

967 967

Ogles, B. M., & Lambert, M. M. J., Weight, D. G., & Payne, I. I. R R. (1990). Agoraphobia Agoraphobia outcome outcome

measurement: Psychological Assessment, 2, 317-325. measurement:AAreview reviewand andmeta-analysis. meta-analysis. Psychological Assessment, 2, 317-325.

Palmer, S. (1986). MMPI profiles of MMPI profiles of women women with withpremenstrual premenstrualsyndrome. syndrome.Unpublished Unpublisheddoc doctoral dissertation, Brigham Young University, Provo, UT. Qualls, R C., & Berman, J. S. (1988, August). Anorexia nervosa: nervosa: AA quantitative quantitativereview reviewofof Quails, R. C, annual the the treatment treatmentoutcome outcomeliterature. literature.Paper Paperpresented presentedat atthethesymposium symposiumofofthethe96th 96th annual Convention of the American Psychological Association in Atlanta, GA. Rachman, S. (1971). The effects effects of of psychotherapy. psychotherapy.Oxford, Oxford, England: England: Pergamon Pergamon Press. Press. Rogers, C. R R. (1942). Counseling Counselingand andpsychotherapy. psychotherapy.Boston: Boston:Houghton-Mifflin. Houghton-Mifflin. Rogers, C. R, personalitychange. change.Chicago: Chicago: R., & Dymond, R R. F. (1954). Psychotherapy Psychotherapy and and personality University of Chicago Press. Rychlak, J. F. (1981). Personality Personalityand andpsychotherapy. psychotherapy.Boston: Boston:Houghton-Mifflin. Houghton-Mifflin. Shapiro, D. A., & Shapiro, D. (1982). Meta-analysis Meta-analysis of of comparative comparative therapy therapy outcome outcome studies: studies: A replication PsychologicalBulletin, Bulletin,92, 92,605-640. 605-640. replication and refinement. Psychological ts of Smith, M. L., Glass, G. V., & Miller, T. I. (1980). (1980). The benefi benefits ofpsychotherapy. psychotherapy.Baltimore: Baltimore: Johns Hopkins University Press. R. M. (1989). Behavioral versus insight-oriented marital therapy: therapy: Snyder, D. E., & Wills, R Effects on individual and interspousal functioning. Journal of of Consulting Consultingand andClinical Clinical Effects

Psychology, Psychology, 57, 57,39-46. 39-46.

Spielberger, C. D., Gorsuch, R R. L., & Lushene, Lush^ne, R R. E. (1970). State-Trait State-TraitAnxiety AnxietyInventory. Inventory. Palo Alto, CA: CA: Consulting Psychologists' Press. Truax, C. C. B., B., & & Mitchell, K. K. M. M. (1971). Research Research on on certain certain therapist therapist interpersonal interpersonal skills skills in in Truax, relation to to process process and and outcome. outcome. In In S. S. L. L. Garfield Garfield & & A. A. E. E. Bergin Bergin (Eds.), (Eds.), The The handbook relation of psychotherapy psychotherapy and and behavior behaviorchange change(1(1st ed.,pp. pp.299-344). 299-344).New NewYork: York:Wiley. Wiley. st ed., of Viney, L. L., et a1. al. (1985). Crisis-intervention counseling: counseling: An evaluation evaluation of long- and short shortof Counseling Psychology,32, 32,29-39. 29-39. term effects. Journal of Counseling Psychology, R. S. (1989). The The Psychotherapy Psychotherapy Research Research Project Project of of the the Menninger Menninger Foundation: Foundation: Wallerstein, R of Consulting Consultingand and Clinical ClinicalPsychology, Psychology,57,57, 195-205. An overview. Journal of 195-205. Whitehorn, J. C., C, & Betz, B. (1954). (1954). A study of psychotherapeutic relationships between physicians and schizophrenic patients. American Journal of of Psychiatry, Psychiatry,3,3,321-331. 321-331. Williams, L. S., Turner, S. M., & Peer, D. F. (1985). Guided mastery and performance performance desensitization treatments for for severe acrophobia. Journal of of Consulting Consulting and and Clinical Clinical

Psychology, Psychology, 53, 53,237-247. 237-247.

This Page Intentionally Left Blank

INDEX INDEX

Ability, Ability, 523

Activity inhibition, inhibition, 487 Activity

Abnormal personality, personality, 920-921 920-921 Abnormal

Actor-observer effect, effect, 626-627 626-627 Actor-observer

Ability traits, traits, 11 Ability 11

Actor, 750-754, 750-754, 853-854, 853-854, 865 Actor,

Abnormal Abnormal psychology, psychology, 919-941 919-941

Adaptation, 354-360, 354-360, 837-838, 837-838, 840, 840, 903, 903, 950 Adaptation,

Absorption, 362, 362, 827, 827, 831, 838, 840 Absorption, 831, 838, 840

anatomical, 354, 359-360 anatomical, 354, 359-360

behavioral, 354-355

Academic psychology, 928

Acceptance, 26, 590, 663; 663; see also Group

Adaptive functioning, 348-349, 924; 924; see also also

Accountability, 664, 664, 669 669 Accountability,

Adaptive landscape, landscape, 333-335

acceptance; acceptance; Social Social acceptance acceptance

Axis V AxisV

Accuracy, 620-621, 620-621, 625-641, 625-641, 649-656, 649-656, 659-673 659-673 Accuracy,

Adaptiveness, 544, 544, 547, 547, 552, 552, 553, 553, 555, 555. 557, 557, 564, Adaptiveness, 567 567

consensus, consensus, 651-652 651-652 correspondence, 651

Adaptive Adaptive significance, significance, 651, 651, 653, 653, 799, 799, 808-809 808-809

functional/pragmatic, functional/pragmatic, 650-653 650-653

Adjustment, 455, 691, -754 691, 751 751-754

as distinct from agreement, 639-641

Adjectival Adjectival markers, markers, 746

information processing processing (cue (cue use), 650, 650, 652, 652, information

Adoption studies, 244-245, 249, 257-258,

654-655

internal consistency, 650, 652, 655, 661 normative models, 650, 652, 653-654, 668n operational, operational, 620-621, 620-621, 650, 652-653, 652-653, 670-671, 673 673 pragmatic utility, 651 stereotypic, 629-630, 629-630, 631-632 631-632 stereotypic, Achievement, 16-17, 46, 46, 56, 56, 486, 486, 489, 773, 773, 776, 776, Achievement, 897 Act-frequency Act-frequency approach, 26 Action-emotion complex, complex, 897 Action-emotion Action orientation, orientation, 491 Action 491 Action theory, theory, 489, 489, 491, 491, 504 504 Action Activation, 390, 390, 518-519, 518-519, 534, 534, 715-716, 715-716, 787 Activation, 787 Active-passive dimension, dimension, 925-926 925-926 Active-passive Activity, 346, 346, 375 Activity, 375 general, 770 social, social, 770, 770, 783 969 969

371-374, 375, 587 Adult development, 199-207, 275, 554 Aduh stages, stages, 275 in women, 291-309 Adult personality, 378 Adventure-seeking, 399, 400; 400; see also also Excitement-seeking; Sensation-seeking; Thrill and adventure seeking Aesthetics, 827, 830, 839 Affect, 26, 172-175, 656, 657, 658, 659, Affect, 25, 25,26,172-175, 659, 661, 661, 663, 666, 772-790, 832, 832, 950 negative, 492, 772, 772, 778, 778, 780, 783, 783, 805, 816, 816, 900 positive, 486, 772-790 primary, 25 Affect 174, 175 Affect congruence congruence model, model, 172, 172,174,175 Affective/motivational 816, 827 Affective/motivational model, model, 804, 804, 807, 807, 816, 827 Affective outcomes, 499 499 Affective Affective space, space, 517-518 517-518 Affective

970 970

INDEX

Affective tasks, 294 Affective Affiliation, 56, 486-488, 489, 774, 788, 927-928 Affiliation, Affiliative trust, 488 Affiliative Affiliative web, web, 418 Affiliative Affordances, 718 718 Affordances, effects, 281 Age effects, Age invariant invariant factors, 746 Age Agency, 24-25, 489, 748, 751, 801-802 Agentic power, 488 Agentic Age stereotypes, 275 Aggregation, 21 346, 352-353, 360, 543, 551, 558, Aggression, 45, 346, 558, 582, 588-592, 598, 600, 600, 608. 608, 804, 806, 933 in childhood, childhood, 804, 806 Agoraphobia, 954-957 Agreeableness, 27, 737, 748, 795-816, 858, 861, 861, 862, 930; 930; see also Nurturance Agreement/accuracy distinction, 639-641 Agreement/accuracy Alienation, 852-854 Allocation, 431 Allocentrism, 446, 446, 448 AUocentrism, Allport, 7-8, 10-12, 14, 18, 20, AUport, Gordon, 4-5, 6, 7-8,10-12,14,18. 73, 76, 84, 111, 28, 41, 43, 46, 48-49, 63, 73,76, 111, 123-125, 127-130, 166, 167, 417, 422, 671, 123-125,127-130,166,167, 671, 711, 739, 744-745, 754-755, 771. 771, 796, 950, 962-963 Altruism, 57, 796, 808-809 838, 840 Ambiguity, tolerance for, 444 444,, 832, 838, Ambition, 751-754, 775-788, 856 American Psychiatric Association, 923 American Psychological Association, 421 Amnesia, 556-557 Analysis of correlations, 621, 628-639 Analysis of mean differences, differences, 621-628 Analysis of of variance, 630-631 630-631 Anatomical adaptation, 354, 359-360 Anomaly-environment Anomaly-environment interactions, 250-253 Anomie, 602 602 Anti-Semitism, 46, 46, 57, 57, 58 Antisocial behavior, 864, 920, 933 Antisocial Antisocial personality, 931, 932 Anxiety, 16-17, 27, 27, 362, 362. 405-406, 455, 455, 549-551, 557, 564, 685, 836, 685, 691, 702, 702, 787, 787, 788-789, 788-789. 836. 900 900,, 936, 936, 950, 950. 952, 952, 955, 955. 956, 956. 959-960, 961, 961, 962 962 Applied psychology, 13, 13, 849-866, 947-964 631 Apportionment of variance, 631 631 conceptual, 631 statistical, 631 631 Appraisal model, model, 466, 466, 474 Appraisal Approach/avoidance, 595, 595, 605, 605, 609, 609, 787, 787, 805-806 805-806 Approach/avoidance, Approval, 750 750 Approval,

Archival approach, 43, 46 Arousability hypothesis, 388-408 Arousability Arousal Arousal emotions, 347-348 Arousal modulating modulating system, 392-393 Artists, 825-826, 842 Ascendance, 770-788 Ascending reticular reticular activating activating system (ARAS),

389, 393, 406 Assertiveness, 751-754, 814, 950 Assessment procedures, 953-957, 958 Association, 450-451, 454 Associative links, 715 Associative networks, 723 Attachment, 26, 57, 210, 211-212, 215-216, 347. 347, 352-353, 805-806 352-353. Attention. Attention, 562-563, 566, 750, 827, 838-839 Attention/self-regulation Attention/self-regulation theory, 561 Attitude, 950 Attribution Attribution theory, 563-564, 626-627, 657, 659 826. 836-837, 836-837. 840 Authoritarianism, 16-17, 418, 826, Authoritarian 46. 836-837 836-837 Authoritarian personality, personality, 46, Authority. 851-854, 851-854. 864-865 864-865 Authority, attitudes toward, 851, 864 864 853. 865 865 conflicts with, 851-852. 851-852, 853, ranking (AR), 448 Authority ranking Autonomic nervous nervous system, system. 347-348, 347-348. 804-805 Autonomic Autonomic response specificity, specificity. 392 Autonomous Stage, Stage, 204 Autonomous Autonomy, 295, 609, 827, 860 Autonomy, Autostereotypes, 455-456 455-456 Autostereotypes, 895, 903, 903, 956 956 Avoidance, 895, 404 Avoidance of monotony, 404 Avoidance of of negative negative affect, affect. 492 Avoidance Axes of the DSM 924, 932, 932, 934, 936; 936; see also also Clinical Axis I, 924, syndromes II, 924, 924, 929; 929; see also also Personality Axis II, disorders 924; see also also Physical conditions Axis III, 924; also Psychosocial stressors 924; see also Axis IV, 924; 924; see also also Adaptive functioning functioning Axis V, 924;

Bandwidth. 852, 852. 854, 854, 857, 857, 863 Bandwidth, Behavior. 345-363, 345-363. 712, 712, 832, 832, 853; 853; see see also also Behavior, behavior; Antisocial behavior; Interpersonal behavior; behavior Social behavior 345, 355-363 355-363 human, 345, mammalian. 345-349, 345-349, 354-355 354-355 mammalian, prescribed, prescribed. 59lf 591f primate. 345-355 345-355 primate, processes and classes of, of. 44, 44. 54, 54, 57 proscribed, 591 591

971 971

INDEX INDEX of, 521 social systems of, suppressed, 582, 591 Behavioral activation system, 390, 787 Behavioral Behavioral adaptation, 354-355 Behavioral change, 950-951, 955, 964 Behavioral Behavioral checklists, 954 Behavioral facilitation system, 787 787 Behavioral facilitation Behavioral genetics, 22, 43, 44, 244-246, 368, 371, 378, 380, 384, 587, 840 Behavioral inhibition system, 390 609, 891-908 Behavioral medicine, 609, Behavioral outcomes, 499 Behavioral prediction, 108, 640-641, 652, 935 935 in job performance, 850, 857, 860-861 Behavioral style, 928 Behavioral traits, 77-79 Behavioral verification, 619 Behaviorism, 43-44, 55, 74, 96, 97-98, 585, 590, 590, 592-593, 712, 714, 746, 866, 949-950, 952, 952, 958, 964 964 Behavior observability, 634-635 Beliefs, 78, 546-547, 838, 853, 952, 959 Betrayal, Betrayal, 466, 475-481 and individual differences, differences, 478-480 narratives, narratives, 476-478 and and social social network, network, 479 Between-cultures 441, 443, 443,454 Between-cultures variance, variance, 441, 454 Bias, 272, 649-673, Bias, 272, 649-673, 650n 65On Bidirectional Bidirectional influences, influences, 169,211,242 169, 211, 242 Big 28, 45, 87, 88, 88,127-128,144, Big Five, Five, 27, 27, 28, 45, 84, 84, 87, 127-128, 144, 206, 209, 259-260, 271, 280, 333, 378-380, 383, 452-459, 468, 522, 561, 718, 722, 722, 737-758, 768, 780, 796-816, 826-842, 849-866, 906-907, 930 hierarchical structure within, 144 Big Four; Four; see also Behaviorism; Humanistic psychology; Psychoanalytic approach; Psychometrics; Traits Big Four orientations or traditions, traditions, 43-45 Big Three Motives, 486-489; see also also Achievement; Mfiliation; Affiliation; Power Bin model, 726 Biobehavioral system, 786-788 Bioevolutionary approach, approach, 798, 808-809 Biographical correlates, correlates, 478 Biographical Biological-affect perspective, perspective, 803-804 Biological-affect Biological factors factors in in personality, personality, 387-408 Biological Biological potentials, 521, 521, 523, 523, 529, 529, 531 Biological potentials, 531 Biological systems systems of of behavior behavior (instincts), (instincts), Biological 520-521 Biophysical conceptions, conceptions, 711-712, 717 Biophysical Biosocial conceptions, conceptions, 711-712, 717 Biosocial

Biosocial Biosocial factors, 307 see Biosocial Biosocial learning typological typological model, 925; 925; see also Learning theory; Social learning theory Biosocial position, 112 Bipolar disorder, 787-788 Blind area, 80-83, 85 Blind spot, 836 Borderline personality disorder, 57, 58, 544, 922, 924, 927, 934 Broad-sense heritability, 370, 380 Bystander intervention, 57, 605, 608 Canalization, 246-249, 258 Capability, 584, 586, 685 Capacity, 521, 521, 523, 536-537 Cardiovascular activity, 389, 398-400 Cardiovascular Cardiovascular disease, disease, 898, 900 900 Case studies, 43, 48, 55, 918, 950, 963 Case studies, self, 218 Categorical Categorical self, 498; see also also Category breadth, breadth, 444-445, 444-445, 498; Category Construct breadth Category mistake, 520, 524 Category Category structure, structure, 720-721 Category Causal attribution, attribution, 622, 713-715, 719, 904 Causal 904 Central catecholemines, catecholemines, 402-403 Central Central system, 589, 589, 598-599, 607-608 607-608 Central nervous nervous system, Central 717, 721 Central traits, traits, 717, 721 Change criteria, Change criteria, 962 962 Change measurement measurement instruments, instruments, 953-957, 959 Character disorder, disorder, 959 Character neuroses, neuroses, 921 Child effect effect on on family, family, 254-255 Child Child effect effect on on mother, mother, 249 Child Childhood, elongation elongation in in humans, humans, 356-357 Childhood, Childhood aggression, aggression, 804, 806 Childhood Child minor physical anomalies, 249-253 and androgens, androgens, 250 and and and sex sex differentiation, differentiation, 250 241, 418, 442, 447-448, 456-459, Child-rearing, 241, 608 Child sexual abuse, 546; also Repressed Repressed 546; see see also memories Child temperament/personality, temperament/personality, 210, 210, 213, 214, Child 213, 214, 241, 374-378, 932-933 and behavior, behavior, 210, 210, 213, 213, 214, 214, 932-933 and Chronic disease, disease, 891, 891, 900 900 Chronic Circumplex model, model, see see Interpersonal Interpersonal circumplex circumplex Circumplex model model Classical test test theory, theory, 145-148 Classical Classification systems systems for for mental mental disorder, disorder, Classification 922-925 Class inclusion inclusion taxonomy, taxonomy, 516n Class

972 972

INDEX Clinical psychology, 937

Clinical samples, 938-940 938; see Clinical syndromes, 924, 932, 934, 936, 938; see also Axis I Clinical vs. vs, statistical prediction, 18, 18, 45 Cluster analysis, 492 Cluster model, 726 Cognition, 357, 564, 656, 661, 661, 712-728, 832; 832; see see also also Social cognition Cognitive aberration, 837 Cognitive ability, 826, 831-835 Cognitive algebra, 713-714, 717 Cognitive appraisal, appraisal, 524-526, 532 Cognitive-behavioral therapy, 958, 961 Cognitive complexity, 444, 841 Cognitive components, 827 Cognitive conservation, 564 Cognitive curiosity, 349, 357, 359-361 Cognitive deconstruction, 698-702 Cognitive defense mechanisms, 525 Cognitive 661, 665 Cognitive dissonance, 564, 661, Cognitive distortion, 959 Cognitive neuroscience, 716 Cognitive outcomes, 499 19, 23, 23, 43, 43, 44, 555, Cognitive psychology, 17, 17,19, 712-713, 740, 748, 840 Cognitive representational theories, 172, 173-174, 175 173-174,175 Cognitive social '-594; see social learning, 590, 593593^594; seealso also theory; Social learning Learning theory; Cognitive styles, 73, 486, 841 Cognitive tests, 86 Cognitive theories, 959 Cognitive therapy, 952, 955, 958-959 Cognitive traits, 77-79 Cohort effects, 432 Cohort studies, 269-285, 293 Collective self-statements, self-statements, 449-450 449-450 Commitment, 499-500 Communal sharing (CS), 448 Communion, Communion, 24-25, 489, 748, 751, 751, 801-802 Community samples, 939-940 Competence, 658, 688-689, 750-754, 773, 786, 786, 813, 814 Competency view, 750-755 Competition, 466 Complexes, 73 Complex hierarchical information processing system (CHIPS), 490 Complexity, 444 Component model, 466, 472-474 Component responses, 523-533

Conditioning, 591-594, 719, 719, 950 classic, 593-594, 719 operant, 594, 950 via punishment, 591-593

Conflicting goals, 716 Conflicting

Conflicting values, 602, 602, 604 Conflicts, 836-837 836-837 Conflicts,

Conflict theory, 295-296, 306 Conformist Stage, Stage, 203-204 203-204

Conformity, 451, 451, 585, 605, 608, 608, 750, 750, 752, 752, 796, Conformity, 585, 60S, 796, 798, 849, 856, 863-865

Confounding influences, 274 Confounding Confounding variables, 206, 832 Confounding

contributors to interactions, 249-253 Congenital contributors others* and and self-judgments, self-judgments, Congruence of of others' Congruence 617-641 Connectionist model, 716 Conscience, 585, 851 Conscientious/Conformist Stage, 204 Conscientious/Conformist

Conscientiousness, 27, 27, 206-207, 206-207, 444, 444, 588, 588, 737, 737, 801, 832, 832, 849-866, 930 Conscientious Stage, 204 Consciousness, 826, 835 Consensus accuracy, 651-652 also Consistency, 74, 74, 634, 634, 655, 655, 690, 690, 861; Consistency, 861; see also Internal consistency descriptive, 655 evaluative, 655 Consistency Seeker, 650, 656, 658, 660-662, 668 Constitutional predisposition model, 895-897 Construct breadth, 444-445, 498 Construct elaboration, 15 Constructivism (phenomenalism), 296, 618 Construct measurement, 15 Construct stability, 145 640, Construct validity, 15, 126, 150-159, 638, 15,126,150-159, 638, 640, 740, 856, 856, 861, 904, 905, 906, 960 960 Contextual factors, factors, 487, 487, 489 Contextualism, 420, 420, 423, 423, 488 488,, 495 Continuous Continuous variation, 200 Contrast effects, 376 Control, 700, 855, 897, 899, 903, 700, 827, 855, 903, 905 Control elements, 725 Control models, 392 Control processes, 555-556 Controls, 581�; 581-609; see also Impulse Impulse control external, 582, 582, 586 internal, 581-609 Control systems, 25, 25, 490 Control Control theory, 489-490, 497, 903 Convergence, 799-801, 799-801, 828 Convergent validity, 15, 154-155, 619, 15,154-155, 619, 641, 641, 745, 745, 750, 898, 902, 902, 904 904

973 973

INDEX Conversion experience, 684 Cooperation, 795, 795, 798 Coping, 57, 551, 551, 552-553, 556, 557, 560-561, 837, 892-893, 903, 903, 925-927, 959 Coping-defense-fragmentation Coping-defense-fragmentation triad, 552-553, 566-567 Coronary heart disease, 807, 816, 897-899 Correspondence accuracy, 651 Cortical activation, 389, 394-395 Cortical development, 589 Cortical functions, 598-599 Counterproductive strategies, 694-695 Counterproductive Coverage of of domain, domain, 151-152 151-152 Coverage Covertness, 450 450 Covertness, Creativity, 46, 825, 834, 837, 840 Criminals, 581-582, 581-582, 587, 607, 607, 610 Criminals, Criminology, 852 Criminology, Crisis-oriented treatments, 952, 952, 964 Criterion groups, groups, 930 Criterion-related validity, validity, 858, 858, 861-862 861-862 Criterion-related Cross-cultural comparisons, comparisons, 291-293, 367-368, Cross-cultural 291-293, 367-368, 439-459, 527-528, 607, 755-756 Cross-cultural psychology, 43, 44, 367-368, 439-459 Cross-sectional studies, studies, 272, 272, 292 Cross-sectional Cross-situational consistency, 166, 166,167n, Cross-situational 167n, 810, 812-813, 850, 919, 932, 934-936 Cross-species consistency, 528, 531-532 654 Cue, 11, 650, 654 Cultural beliefs, beliefs, 546-547 546-547 Cultural Cultural conflict, 602-603 602-603 Cultural conftict, Cultural relativism, 440, 585 Cultural relativism, 440, 585 Cultural rules, rules, 590-591 590-591 Cultural Cultural scripts, scripts, 443 443 Cultural Cuhural syndromes, syndromes, 443-450, 443-450, 451 Cultural Cultural values, 489 Culture, 20, 439, 442-450, 456-459, 826-828, 830-831 Curiosity, 825, 839, 842 Current concerns, 275, 485, 492, 494-495, 497, 500 "Curve of life," 282 Curvilinearity, 200 Cybernetics, 18-19, 392 Cyclical maladaptive maladaptive pattern, pattern, 928 928 Cyclical Cynical trust, trust, 488 488 Cynical

Charies, 317-321, 325-326, 367, Darwin, Charles, 374-375, 384, 719 Defense mechanisms/processes, 525, 543-545, 959 547, 551, 552, 560, 836, 959 Defense taxonomies, 566-568

Defensive maturity, 553-554, 566-567 Defensiveness, 80, 80, 836 Defensive profile, 552 Defensive Defensive style, 553-555 Definition Definition by exclusion, 583, 583, 608 Dehumanization of victim, 605-606 Delay of gratification, 46 Denial, 546, 562, 565, 565, 837 Dependability, 473, 473, 856, 856, 861 Dependency, 295, 295, 959 787-789, 861, 861, 900 Depression, 787-789, Depression, 900,, 903, 903, 924, 924, 940, 958-962 Depressive realism, 564, 564, 660 Description, 74, 76, 77-79, 86, 110-113, 121-123 86,110-113,121-123 Description vs. explanation, 434, 906, 908 of traits, 744-758 Descriptive Descriptive consistency, 655 Desires, 78 Deterrence, 591-592 Developmental analysis, 565 Developmental history, 559 Developmental psychology, 23, 23, 466, 471-472, 685, 695, 835 835 Deviancy, 585, 597, 602, 602, 853 Deviant self-concept, 603 Deviant values, 602-603 Diagnosis, see Psychodiagnosis Diagnostic and Statistical Manual of Mental Disorders, 45, 47, 554-555, 666 666,, 721, 721, 750, 750, 841, 922-441, 958; 958; see also also Axes of the DSM Diagnostic groupings, 947 Diagnostic overiap, overlap, 929, 931, 931, 937 937 Dialectical reasoning process, 132-135 Differential elevation, 629-630 Differentiation, 504-505 Dimensional approach, 928-930 Dimensions of cultural variation, 440-450, 454-459 Dimensions of personality variation, 440, 451, 451, 737 Dimensions of social behavior, 440, 450 Dimensions of social situations, 172 Direct effects, 425-426, 428-429 Directional fractionality, 392 Directional tendencies, 327-330 Direction of effect, 243, 257 Directive function, 490 Direct perception, 718 Direct/realist approach, 719 Discreteness/continuum in types and traits, 87-89

974 974

INDEX Discriminant Discriminant validity, validity,

15, 154-155, 750, 789, 898, 15,154-155, 902, 904 902,904 Disease, 487-488, 497, 504, 589, 599-600, 891-908 Disinhibition, 391, 391, 399-401, 399-401, 781-782 Displacement activities, 78 Dispositional Dispositional attribution, 626-627 Dispositional Dispositional conceptions, 232-233 Dispositional Dispositional constructs, constructs, 17

Dispositional Dispositional influences influences on on choice choice of of educational educational situations,

181

leisure leisure situations, situations,

181

occupational situations, situations, occupational

181-182

Dispositionally incongruent situations, situations, Dispositional ly incongruent

176,

182-184

Dispositional state, state, 497 Dispositional Dispositional strategy, 166,167n Dispositional strategy, 166, 167n Dispositional variables, variables, 520 520 Dispositional Dispositions, see Traits Traits Dispositions,

Dissembling, Dissembling,

86 86

Dissociation, 450-451, 450-451, 454, 454, 837 837 Dissociation, Distortion, 649, 649, 656, 656, 661, 661, 663, 663, 665-666, 665-666, Distortion,

668-669, 668-669, 959 671, 686, 931, 938, 959 Divergent thinking, thinking, 833 Divergent Dogmatism, Dogmatism, 838, 842 Domain scales, 746 746 Domain scales, Domain specificity, 323-324 Domain specificity, Dominance, 348-349, 352-353, 352-353, 360, 360, 748-750, 748-750, Dominance, 348-349, 802, 814, 816, 899, 905-906, 774, 776-790, 802, 927; see also also Extraversion, Extraversion, Surgency Surgency Dominance/agency/status, 748 Drives, 10, 485, 921-922 Drives, Dual-coding hypothesis, 723 Dyadic interactional interactional view, 748-750, 754 Dynamic lattice, lattice, 12 Dynamic Dynamic psychology, 835-837, 951, 951, 953 953 Dynamic traits, 11 Dyslexia, 57 Dysthymia, 924, 959 Eating disorders, disorders, 700-702, 961 Eating Ecological Ecological approach, approach, 718 Ecological fal fallacy, Ecological lacy, 428 Ecology, Ecology, 456-459, 495

Econometric Econometric techniques, techniques, 424 Education, 826, 830-831

Educational Educational psychology, psychology, 835, 835, 840 840 Educational Educational situations situations and and dispositions, dispositions,

Effectance Effectance motivation, motivation,

Efficacy, 688, 690, 950 Ego, 595 Ego actions, 551-552

773

181

Ego development, development, 23-24, 23-24,199-207 Ego 199-207 Ego dystonic condition, 922 Egoism, 685 Egoist, 650, 656, 658, 669 Ego processes, 552 Ego psychology,

55

Ego resilienCy, resiliency, 635, 827, 935 Ego strength, 929, 947

Ego syntonic syntonic condition, 922, 932, 936 685 Electrodermal Electrodermal activity, activity, 389, 397, 400, 563 Electroencephalograph (EEG), 393-394 Electroencephalograph Elementary information information processes, processes, 716 Elementary Elevation, 629-630 Elevation, Elongation of of childhood childhood in in humans, humans, 356-357 Elongation Emic constructs, constructs, 440-442, 448, 454 Emic Emotion, 347-348, 513-537, 543-568, 687, 691, Emotion, 701-702 arousal, 347-348 347-348 arousal, categories, 515-517, 519 dimensions, dimensions, 515, 517-519 evolutionary theory theory of, of, 559, 561 evolutionary and language, language, 513-517, 530-532, 534 and relationship, 347-348 taxonomies, 515-516 taxonomies, Emotionall ambivalence, ambivalence, 838 Emotiona Emotional capacities, capacities, 536-537 Emotional Emotional core, core, 767, 767, 772-790 772-790 Emotional Emotional creativity, creativity, 522, 522, 535 Emotional 535 Emotional expression, expression, 80S, 805, 807 Emotional 807 Emotional independence, independence, 929 Emotional Emotional intelligence, Emotional intelligence, 522, 537 Emotionality, 209, 209, 375, 375, 770, 770, 773-790, 773-790, 804, 804, 899; Emotionality, 899; also Affect Affect see also negative, 770, 899 negative, positive, 804 positive, 773-790, 804 Emotional Emotional potentials, potentials, 536-537 Emotional reactions, reactions, 519-520, 523-533 Emotional Emotional reactivity, 929 Emotional Emotional roles, roles, 536-536 536-536 Emotional Emotional rules, rules, 534-535, 534-535, 537 Emotional 537 920; see see Emotional stability, 858, 861-863, 861-863, 866, 866, 920; Emotional stability, 858, also Neuroticism Neuroticism also Emotional states, 519-521, 522-523, 533-534 Emotional Emotional syndromes, syndromes, 519-520, 522-523, 528, Emotional 533-536 Emotional traits, traits, 77-79 Emotional Emotion-focused strategies, strategies, 561, 561, 566 Emotion-focused 566 Empathy, 57, 605, 609, 609, 795, 795, 809, 809, 813; 813; see also also Empathy, 57, 60S, Nurturance Nurturance Encoding, 712, 712, 715, 715, 725 Endocepts, 531 Endocepts,

Egotism, Egotism,

97S 975

INDEX Enduring dispositional dispositional view, view,

745-747, 745-747, 754 10, 861 10, Enhancement Enhancement processes, processes, 548, 564, 566 Enjoyment, 492 Environment, Environment, nonshared, 244-245, 258, 371, 371, 379-381 Environment, Environment, shared, shared, 371, 371, 376, 379-381 Environmental 166, 217, Environmental influences, influences, 20, 20,166, 217, 368, 368, 371, 371, 376, 376, 381-383, 381-383, 456-459, 456-459, 586, 586, 589, 589, 596, 596, 814-815, 897, 897, 899, 899, 908, 908, 935, 935, 941, 941, 950, 950, 960 960 814-815, Environmentalism, Lockean, Lockean, 13 Environmentalism, Environmentally induced induced strategic strategic individual individual Environmentally differences, differences, 331-332 331-332 Environmental press, press, 489 489 Environmental Environmental stressors, stressors, 960 960 Environmental Environmental variation, variation, 371 371 Environmental Environment of of evolutionary evolutionary adaptedness adaptedness Environment (EEA), 322 322 (EEA), Episodic disposition, disposition, 523, 523, 533 533 Episodic Epistemological concerns, concerns, 47, 47, 62 62 Epistemological Epistemology of of subjective subjective personality personality Epistemology judgments, 131, 131,138 judgments, 138 Equality matching (EM), 448 448 Equilibrium, 10, 10,12 Equilibrium, 12 Erikson, Eric, 17, 17, 41, 41, 54-55, 54-55, 58, 58, 62, 62, 271, 271, 275, 275, 294-295, 426, 426, 468-469, 468-469, 683, 683, 685-686, 685-686, 695 695 294-295, Error paradigm, 652 652 Escapist behavior, 697-701, 697-701, 702 702 Escapist Esoteric thinking, 835 835 Esteem, 492 Esteem, 492 Ethical continuum, Ethical continuum, 604 604 Ethical judgment, judgment, 589, 589, 592, 592, 595-596, 595-596, 598, 598, 607 Ethical 607 Ethical prohibitions, prohibitions, 585, 585, 587, 587, 589, 602,, 604 604 589, 602 Ethical Ethical sensitivity, sensitivity, 589, 589, 599 599 Ethical Ethnocentrism, 441 441 Ethnocentrism, Ethnomethodologism, 423 423 Ethnomethodologism, Etic constructs, constructs, 440-442, 440-442, 448, 448, 450, 450, 454 Etic 454 Evaluation Evaluation (negative/positive), (negative/positive), 518-519, 518-519, 525, 529, 534 Evaluative connotations connotations of construct construct labels, 835 Evaluative consistency, consistency, 655 Evoked potentials, 401, 401, 563 Evoked responses (ER), (ER), 394-396, 404 404 Evolution, theory theory of, of, 317, 319-320, 465-467, 475, 852-854, 865 Evolutionary adaptation, 317-340 Evolutionary biology, 26, 318 Evolutionary heritage, 345, 363 Evolutionary hypotheses, 320, 325 Evolutionary models, 326-327 Evolutionary path, path, 319 Evolutionary psychology, 78, 317-340, 588, 797-798, 808-809 Energy,

Evolutionary theory of of emotion,

Evolutionary trends, 345, 363

Excitation, 805

Excitement-seeking,

559, 561

87; see also also Adventure

seeking; Sensation seeking; seeking; Thrill Thrill and seeking;

adventure adventure seeking

Executive Executive control control structure, structure, 716

Exemplar Exemplar view,

720-722 776-777 Existential self, 218 Existential self, Expectancy Expectancy for for success, success, 499-500 499-500 Experimental psychology, 13 110-113, 121-123; Explanation, 74, 76, 77-79, 77-79,110-113,121-123; also Description Description vs. vs. explanation explanation see also Explanatory Explanatory style, style, 903-905 903-905 Explicit motives, 505 Explicit 505 Exploration, 302-303, 302-303, 346, 346, 826, 826, 840 840 Exploration, Expressive Expressive behavior in in children, children, 222-223 222-223 Expressive reactions, reactions, 524, 524, 528-530, Expressive 528-530, 532 532 External controls, 582, 582, 586 External 586 External criteria, criteria, 153-154, 153-154, 622, 622, 652 652 External External traits, traits, 627-628 627-628 External Extraparental nepotistic nepotistic investment, 322 Extraparental 322 Extraversion/introversion, 12, 12, 27, 27, 85, 85, 87-88, Extraversion/introversion, 153-154, 361, 361, 378-379, 388-390, 393-399, 393-399, 153-154, 378-379, 388-390, 400-403, 406, 406. 635-637, 635-637, 737, 737, 746, 746, 748, 400-403, 748, 767-790, 800-801, 800-801, 814, 814, 816, 816, 827-828, 827-828, 841, 767-790, 841, 858, 863, 863, 921, 921, 930, 930, 939, 939, 940, 940, 950 950 858, Extrinsic instigation, 592 592 Extrinsic instigation, Eysenck, Hans, Hans, 7, 7,12,13, 27, 33-34, 33-34, 111, 111, 132, 132, Eysenck, 12, 13, 27, 176,179,183, 249, 270, 270, 336, 336, 353, 353, 361, 361, 176, 179, 183, 249, 387-391, 393, 393, 399, 399, 404-405, 404-405, 407, 407, 440, 440, 739, 387-391, 739, 747, 768-779, 768-779, 773, 773, 781, 781, 929 747, 929

Exhibitionism, Exhibitionism,

Facet scales, scales,

746-747 367-368, 520, 520, 528-529, 719 Factor analysis, 389, 492, 552, 561, 587-588, 738, 740, 742, 748-757, 769, 771, 828, 831, 831, 860, 860, 862-864, 903, 929, 940-941 Factor structure comparison, 280 Factor theory, 11, 206 206 False consensus bias, 627 Family as unit, 253 Family cohesion, 256 Family influences, 241-260 Family 516, 721 Family resemblances, resemblances, 514, 516, Fantasy, 827, 827, 832 Fantasy, 832 Favorability, 690, 690, 702 702 of traits, traits, 624-625 624-625 of Fearfulness, 346, 346, 353-353, 353-353, 360 Fearfulness, 360 Fears, 853 Fears, 853 Facial Facial expressions,

976 976

INDEX INDEX Feedback, 489-490, 525, 529, 654, 685-686, 688-692, 702 Feelings, 524, 524, 526, 532-533; 532-533; see also also Affect; Affect; Emotion; Emotionality Fidelity, 854, 854, 863 Field dependence-independence, 16-17, 362, 445, 841 445,841 Field theory, 10 Filtering, 562

748 Fit, 166, 213, 748 Five Factor Model,

Big Five Five see Big

Flexibility, 827, 832 Flexibility, Focus of of convenience, convenience, 744, 744, 748, 748, 750-751, Focus 750-751 , 753-754

Folk concepts, concepts, 73, 73, 851, Folk 851, 852

Genotype-environment correlation, 257 Gerontology, 269

Gestalt approach, 717 Getting along/getting ahead, ahead, 26, 26, 465-481 465-481,, 662, 853 Gilligan's three-stage model for women's growth, 295

Global lifestyle factors, 895

Goal-based motivational theories, 172-173, 175 Goal complexity, 504-505 Goal conflict, 491, 491, 501-504 Goal-directed action, 491

Goal-directed behavior, 485, 489 Goal-directedness, 490

Goal-directed units, units, 492 492

Forced choice choice rating, rating, 630 Forced

Goal instrumentality matrix, 500, 500, 504 Goal instrumentality matrix, 504

Free-response listing, listing, 498-499, 498-499, 506 Free-response 506

Goals, 78, 78, 485-506, 485-506, 686, 686, 690, 690, 694, 694, 701, 701, 716, 716, 717; Goals, 717; see see also Personal goals

Foreclosures, 695-696 695-696 Foreclosures, Formality, 450-451 450-451 Formality,

Freud, Sigmund, Sigmund, 5-6, 5-6, 9, 9,11,13, 18, 44-55, 44-55, 78, Freud, 11, 13, 18, 78, 218, 418-419, 503, 544, 545, 554, 556-557, 671, 851-852, 864, 921-922, 585, 594-596, 671, 948, 951 Functionalism, 294, 299, 304 Functional/pragmatic criteria for accuracy, FunctionaVpragmatic 650-653 Fundamental attribution error, 627 Fuzzy-set, 721

Gender differences, 252, 282, 291-309 Gender-related personality change, 304-307 Gender roles, 291-310, 427 Gender specialization, 294 General arousal model, 389-392, 406 General causality orientation, 362 Generalizability, 1 17-138, 148-149, 150, 862 GeneralizabiUty, 49, 49,117-138,148-149,150, Generalized trust, 466, 468, 469-471, 474-475 Generation effects, 432 Generativity, 56, 291 Genetic basis of neonate temperament, 230 Genetic basis of of temperament and personality, 368-384

Goal properties, properties, 502-505 502-505 Goal life, 485-506

taxonomy, 492 492 taxonomy, for treatment, treatment, 952, 952, 955, 955, 962 962 for Goal-striving behavior, behavior, 496 Goal-striving "Gold standard" standard" criterion, 149-150 149-150

Good and evil, 585-586 Good information, information, 633, 633, 637-638 637-638 Good Good judge, judge, 633 633 Good

Goodness of fit model, 213-214, 217, 256 633, 634-635 Good target, 633, 633, 635-637 Good trait, 633, Gregariousness, 87

Group acceptance, 465, 467, 475; 475; see seealso also Social acceptance Acceptance; Social Group cohesion, 798-799, 802 802 Group Group decision making, 650, 608 Group differences, 44, 44, 45, 54 Group living, 852-854 Group means, 119 Group processes, 798-799 Groupthink, 605 Guilt, 603, 604, 606, 609 Guilt, 58-59, 593, 593, 595, 595, 603,

Genetic components, 368

Habit Habit strength, 592

Genetic investment, 322, 327 Genetic mechanisms, mechanisms, 587-589, 587-589, 598 Genetic Genetic mediation in environment, 245 Genetic variation variation in in environmental environmental measures, Genetic

Hard/soft continuum, 59-63 59-63 Hard synthesis, synthesis, 61-62

Genetic factors, factors, 895-896, 895-896, 950 950 Genetic

381-383 Genetic variation in temperament and personality, 368-371, 378

84, 111 Genotype, 74, 76, 79, 84,

Hardiness, 891, 891, 894, 894, 901-903

Health, physical, 487-488, 497, 504, 504, 891-908

and disease, 487-488, 497, 504 Health behavior, 894-895, 900 900,, 902, 902, 90S, 905, 907

Health Health behavior model, model, 894-895, 900

Health Health outcomes, 902-903, 904-906, 908 Health Health psychology, 23, 23, 891-908

977

INDEX INDEX risks, 894, 898 Health risks, Hedonic capacity, capacity, 788 788 Hedonic 57 Hemispheric dominance, 57 Hereditary factors, factors, 586 586 Hereditary Heritability, 369-371, 374-384, 587-588, 809, Heritability, 831, 840, 895-896, 899, 929 broad-sense, 370, 380 narrow-sense, 370, 377, 380 Heritability coefficients, 369, 537 Heritably induced strategic strategic individual differences, 332-333 Heterostereotypes, Heterostereotypes, 455-456 Hexagonal model, 752 Higher Higher order order factors, 768-776, 768-776, 863, 930 High-risk groups, 898, 898, 900 High-risk Hitler, Hitler, Adolf, Adolf, 58-59, 58-59, 41-64 41-64 passim passim Holism, 87-88 Holism, 4-6, 4-6, 12, 12, 23-25, 23-25, 28-29, 28-29, 87-88 Homicide, 604 Homicide, 588, 588, 603, 603, 604 Honesty, 894, 897, Honesty, 45, 45, 850, 850, 857, 857, 894, 897, 907 907 Hostility, 807, 852-854, 852-854, 858, 858, 861 Hostility, 807, 861 Humanistic psychology, psychology, 43-44, 43-44, 55, 55, 585, 585, 746, 746, 836, Humanistic 836, 949, 952 Human needs; needs; see also also Needs Needs Human Human needs, needs, taxonomy taxonomy of, of, 747 Hypertension, 487, 487, 504, 504, 905 Hyperthymic personality, personality, 921 Hypnotizability, 826-827, 832, 840 Hypochondria, 896 Hypothalamus, 589, 598-600 Hypothesis testing, 54, 130-131, 133, 135, 54,130-131,133,135, 639-640, 907

Id, 585, 585, 595 595

Ideal self, 687

Identification, 593-594, 795 with others, 795

with parents, 593-594

Identity, 56, 291, 294, 358-359, 420, 489, 56, 58, 291, 489, 681-703, 853-854, 865-866

achieved, 695 conflict, 696-697, 702 construct, 698

deficit, 696-697, 702

diffusion, 696

foreclosures, 695-696 moratoriums, moratoriums, 695

negotiation, 467, 467, 619, 619, 664 statuses, statuses, 695-696

Idiocentrism, 444 444,, 448 448 Idiocentrism,

Idiographics, 8, 8,18, 50, 86-87, 86-87,123-125,124n, Idiographics, 18, 50, 123-125, 124n, 130, 130, 485-486, 492, 492, 494, 494, 497, 497, 499, 499, 501 501

Idiothetics, 117-1 18, 130-138 117-118,130-138

Illness, 504; see also Disease; Health; Physical health and disease behavior model, 896, 900-901, 900-901, 904 psychosomatic, 504 504 Imagination, 825, 827, 830, 842 489 Imagoes, 24-25, 489 Imitation, 350 350 Immunocompetence, 487, 504, 892, 905, 907 Implicit personality theory (IPX), ( IPT), 717 Impression formation, 713, 715, 717 Impression management, 18, 26, 85, 86, 86, 662-664, 666, 669, 745, 751, 931, 939 Improvement, 947, 949, 952-953 indices, 947, 952-953 rate, 949 Impulse control, 849, 856, 863-864 Impulsive Stage, 203 Impulsivity, 27, 87, 203, 251-252, 353-353, 360, 390-391, 773, 390-391, 404, 404, 589, 589, 599, 599, 635, 635, 770-771, 770-771, 773, 776, 781, 854, 855, 858; see also also 776, 781, 853, 853, 854, 855, 858; Inhibition Inhibition Inability to perform, perform, 583 Inability Incentives, 494 Incentives, Incest taboo, 588-589 Incest Inconsistency of Inconsistency of values, 601 Indigenous psychology, 442, 457 Indigenous Indirect measurement, 650-651 Indirect Individual differences, 4-5, 7,16, Individual 7, 16, 26, 28, 41-64, 97,117-138, 97, 1 17-138, 209, 222, 259, 275-276, 330-339, 353-354, 374, 388, 465-468, 330-339, 478-480, 619, 634-635, 657, 658, 660, 478-480, 661-662, 664, 666, 669-670, 711, 711, 739, 739, 745, 745, 661-662, 768, 773, 773, 796, 796, 798, 798, 807, 807, 808, 808, 809, 809, 825-842, 768, 825-842, 849-866, 903, 903, 907, 907, 921, 921, 929, 929, 935 849-866, 935 of contexts, contexts, 259 of 259 in primates, primates, 353-354 353-354 in Individual difference difference variables, variables, 1119, Individual 19, 426, 548 Individualism, 685 Individualism, Individualism/collectivism, 444 444,, 451 Individualism/collectivism, Individuality, 683, 683, 685 Individuality, Individual level level of analysis, analysis, 420-422, 420-422, 427-428, 427-428, 430 Individual Individual Individual lives, lives, 41-64 41-64 Industrial/organizational Industrial/organizational psychology, 840-841, 840-841, 857 857 Infant irritability/activity, irritability/activity, 248 Infant memory, memory, 219-220 219-220 Infant Infant temperament temperament and and attachment, 210, Infant 213-216 213-216 Infant temperament temperament and and self-concept, self-concept, 216-220 216-220 Infant Information processes, processes, elementary, elementary, 716 716 Information Information processing processing approach, approach, 490-491, 490-491, 545, Information 548, 556, 556, 562-563, 562-563, 656-661, 656-661, 663, 663, 666 666 548, criteria for for accuracy (cue use), use), 650, 650, 652, criteria accuracy (cue 652, 654-655 654-655 symbolic, symbolic, 714-728

978 578

INDEX

Information storage, storage, 723-724 723-724 Information Inhibited Inhibited power power motive motive syndrome syndrome (IPMS), (IPMS), 487 487 Inhibition, 487, 581-609, 635, 698-701, S05, 805, 816 internal, 487, 581-609 psychological sources of, 590-590-597, 600-606 Inhibition/impulsion, 635 Inhibitory mechanisms, 589 Inner experience, 835 Inner self, 683 Instigation, 524-527, 533, 533, 592 Instinct, 26, 520-521, 921-922 Institutional level of analysis, 420-422, 427-428 Instrumental acts, 524, 530, 532 Instrumental behavior, 925 Instrumentality, 499-500 Instrumental polarity, 926 Instrumental tasks, tasks, 294 Integrated running text, 488 Integrated Stage, 204-205 Integration, 504-505, 950-951, 964 Integrity, 291, 291, 849-866 Intellect, 826-828, 831-832 Intellectance, InteUectance, 751-754, 827 Intellectualizing, 837 Intelligence measurement, 56,199,299, 56, 199, 299, 828, 831, 831, 832-835 Intensity, 827, 835, 838 Intentional context, 495 Intentionality, 514 Intentions, 853 Interactional model, 243 Interactional processes (linkage), 425, 430, 432 167, 169-172, 185-187, Interactionism, 22, 43, 43,167,169-172,185-187, 603, S08; 808; see also also Social interactionism; Symbolic interactionism Interactionist strategy, 167-187 dynanric, 169-172, 185-187 dynamic, 167, 167,169-172,185-187 mechanistic, 167 Interactive measurement, 133 Internal conflicts, 603 Internal consistency, 144-147, 501, 650, 652, 655, 661 criteria for accuracy, 650, 652, 655, 661 Internal controls, 581-609 Internal dispositional variables, 22, 627-628; see see also Traits also Traits Internal dynamics, 952 952 Internal dynanrics, Internal inhibition, inhibition, 487, 487, 581-609 581-609 Internal instigation, instigation, 592 Internal Internal 581-609 Internal restraints, restraints, 581-609 Internal systems systems of of control, control, 593-596 593-596 Internal

Internal working working models, models, 215-216 215-216 Internal Interpersonal aspect/dimension, 492, 682 Interpersonal behavior, behavior, 748, 748, 849, 936-938 936-938 Interpersonal Interpersonal circumplex, circumplex, 27, 27,112, 361, 747, 747, 748, Interpersonal 112, 361, 748, 751-754, 796, 796, 800-803, 800-803, 927-928 927-928 751-754, Interpersonal competence, competence, 750-754 750-754 Interpersonal Interpersonal disruption, disruption, 928, 928, 932, 932, 936 936 Interpersonal Interpersonal openness, openness, 828 828 Interpersonal Interpersonal skill skill deficit, deficit, 933 933 Interpersonal Interpersonal style, style, 928 Interpersonal 928 Interpersonal theory, 927-928 927-928 Interpersonal theory, Intervention, see Treatment Treatment Intervention, Intimacy, 56, 56, 291, 291, 295, 295, 450-451, 450-451, SOl-802 801-802 Intimacy, Intrasexual competition, competition, 321-322, 321-322, 325-327, 329 Intrasexual 325-327, 329 Introjection of of values, values, 593, 593, 595 595 Introjection Introspectiveness/impulsivity, 770 770 Introspectiveness/impulsivity, Introversion/extraversion, see see Extraversion Extraversion IntroversiOn/extraversion, Intrusions, 562-563 562-563 Intrusions, Ipsative changes, changes, 281 Ipsative 281 Ipsative measures, measures, 962 962 Ipsative Ipsative reasoning, reasoning, 132, 132,132n Ipsative 132n Irrational beliefs, beliefs, 952, 952, 959 959 Irrational Item content, content, 16, 16,151-153, Item 151-153, 866 866 Item-level analysis, analysis, 631-637 631-637 Item�level

James, 48, 219, James, VVilliam, William, 5-6, 5-6,48, 219, 279, 279, 526, 526, 529, 529, 712, 712, 769 769 Jealousy, 347 347 Jealousy, Job performance performance behavior behavior prediction, prediction, 850, 850, 857, Job 857, 860-861, 860-861, 863-864 863-864 Job performance performance criteria, criteria, 850, 850, 854, 854, 860-861, 860-861, 866 866 Job Job proficiency, proficiency, 858 858 Job Johari window, window, SO-83, 80-83^ 85 85 Johari Judgability, 634-635 634-635 Judgability, Judgment heuristics, heuristics, 714 Judgment 714 Jung, Carl, Carl, 4, 4, 9, 9, 45, 45, 48, 48, 49, 49, 54, 54, 83, 83, 291, 291, 532, 532, 747, 747, Jung, 769-770, 921 769-770, 921

Knowledge structure, structure, 715, 715, 723 Knowledge 723 Kohlberg's stages stages of of moral moral development, development, Kohlberg's 201-202 201-202

Labeling, 19, 19, 27, 27, 603, 603, 828-835 828-835 Labeling, Language and and emotion, emotion, 513-517, 513-517, 530-532, 530-532, 534 Language Language Language use use and and self-concept, self-concept, 221, 221, 227 227 Latent classes, 144 Latitudes Latitudes of of acceptance, acceptance, 590 590 Latitudes of of rejection, rejection, 590 Latitudes 590 Learned helplessness, helplessness, 694, 694, 904 Learned

979 979

INDEX Learning theory, 10, 43; see also Biosocial learning; Cognitive learning; learning; Social learning Leisure situations and dispositions, 181 Lens model, 654, 657 Lexical analysis, 27, 754-758, 796-798 Lexicons of personality, 739-744, 754-757 Lexicons of traits, 849, 865 Liberal thinking, 827, 839 Life course, 257 Life cycle effects, 432 Life goals, goals, 485-506 485-506 Life Life history, history, 553, 553, 934 934 Life Life space, space, 10, 10,12 Life 12 Life-span development, development, 23, 269 Life-span 23, 269 Life story, 23, 24-25, 49, 291, 291, 293, 293, 489 Life Life tasks, tasks, 485, 485, 492, 492, 495-497, 495-497, 499-500, 499-500, 908 Life Life transitions, transitions, 496 496 Life Likability, 751-754, 751-754, 796 Likability, 796 Limbic system, system, 401-403, 406 Limbic 401-403, 406 Linearity, 200 200 Linearity, Linguistic analysis, analysis, 755 Linguistic 755 Linkage processes, processes, 432 432 Linkage Living systems systems framework framework (LSF), (LSF), 489-491 489-491 Living Locus of of control, control, internal-external, internal-external, 152, 362,423, Locus 152, 362, 423, 552, 813, 813, 860, 895 895 426, 552, Longitudinal stability in adult personality, 269-285 in psychodiagnosis, 919, 932, 932, 933-934 of traits, 745 Longitudinal studies, studies, 247-252, 255-256, 269-285, 292-294, 296, 299-302, 304-306, 309-310, 432-433, 553-554 Love, 748, 774 Magnitude of effects, 254 Main-effect models, 241-243 Mammalian behavior, 345-349, 354-355 Manie sans delire, 920 Manipulation, Manipulation, 349-350, 356 Marital 251, 253-254 Marital relationship, 251, Market pricing (MP), 448 Mastery, 773, 801; see also also Competence 773, 786, 801; Maternal stability, 248 Mate selection and retention, 321-323, 327, 329-330, 335-336 Maturation, 274, 275, 281 Mean level stability, 275 Measurement error, error, 146 Measurement Measurement validity, validity, 651-652 651-652 Measurement Mediated effects, effects, 426-433 426-433 Mediated Mediated form, 418 418 Mediated form, Medical model, model, 96 96 Medical

285; see also also Medicine, psychosomatic, 285; Psychosomatics Memory, 276-278, 562, 562, 712, 712, 714-715, 723 in childhood, 224-225 infant, 219-220 person, 722-726 repressed and child sexual abuse, 546 repression effects on, 546, 550, 550, 556 social, 723, 723, 725 working, 715-716 Mental processes, 712-728 Mental representations, 712-728 analysis, 420-421, Meso-situational level of analysis, 427-428 150, 850, 856-863, 856-863, 866, 898-899, Meta-analysis, 150, 956 Methodological techniques, 43 experimental, 43 quantitative-correlational, 43 Microgenesis, 558 558 Microgenesis, Mistrust, 601 601 Mistrust, Model-fitting approaches, 373-374 Modeling, 593-594, 593-594, 602, 602, 609, 609, 950 950 Modeling, Model testing, testing, 906-907 906-907 Model Moderator variables, variables, 21, 21, 631, 631, 633-639, 815, 816, Moderator 633-639, 815, 816, 840,860 840, 860 Monoamine oxidase, oxidase, 401-402, 401-402, 405 405 Monoamine Monotony avoidance, avoidance, 404 404 Monotony Mood, 772, 772, 783-785, 783-785, 832 832 Mood, Mood disorders, disorders, 924, 924, 959 959 Mood Moral development, development, 201-202, 201-202, 480-481, 480-481, 553, 553, 585, 585, Moral 589-597, 607-608, 839; 839; see also Gilligan model; Kohlberg model Moral insanity, insanity, 920 Morality, 585, 585, 588, 592 815, 826, 861 Moral reasoning, 57, 427, 813-814, 815, Moral sensitivity, 589, 599 Moratoriums, 695 Morphogenies (idiographics), 124n, 130 Morphogenics Morphs, 293 899, 900 900,, 905 Mortality, 899, Motivated unawareness, 547-548 4-6,12, Motivation, 4-6, 12, 25, 28, 421-422, 485-501, 525, 564, 583, 583, 584, 584, 586, 586, 608, 608, 656, 656,657, 657,658, 658, 659-660, 661, 661, 663, 663, 665-666, 687, 690-691, 865 701, 773, 826, 832, 839, 852, 865 Motivational concepts, concepts, 485-489 485-489 Motivational Motivational conflict, conflict, 503-504 503-504 Motivational Motivational dynamic, dynamic, 905-906 905-906 Motivational Motivational state, state, 584, 584, 586, 586, 608 608 Motivational Motive dispositions, dispositions, 485-489, 485-489, 494, 494, 496-497, 496-497, 505 Motive 505 Motive measurements, measurements, 487-488 487-488 Motive

980 980

INDEX INDEX

Motive patterns, patterns, 487 487 Motive Motives, 73, 73, 485-506, 485-506, 853 853 Motives, explicit, explicit, 505 implicit, implicit, 505

Motive systems, systems, 486 486 Motive

Multiaxial approach, approach, 924 Multiaxial

Multiple criteria, criteria, 655 655 Multiple Multiple selves, selves, 24 24 Multiple Multitrait-multimethod matrix matrix (MTMM), (MTMM), Multitrait-multimethod 155-159

Normative Normative perspective, 296 Normative reasoning, reasoning, 131-135 Novelty, Novelty, 839

Nuclear Nuclear episodes, 24-25 24-25 Null 133, 135, Null hypothesis hypothesis testing, testing, 130-131, 130-131,133,135, 639-640 639-640

Nurturance, 347, 347, 352-353, 352-353, 360, 360, 748-750, 813, Nurturance, 815; 815; see see also also Agreeableness Nurturance/communionllove, Nurturance/communion/love, 748

models, 738, 738, 740, 740, 745, 745, 747, 747, 756 Multivariate models, 666, 669-673 669-673 Narcissism, 666,

Obedience, 57, 856 Obedience to authority, 605, 608

tone, 24-25 24-25 Narrative tone,

Objective, Objective, 524-527

National character differences, 418-419, 421,

Object Object relations theory, 55, 295, 295, 544, 544, 555

Natural language, language, 755-757, 755-757, 796-804, 796-804, 808, Natural

640 640 Observational perspective, 626

370, 377, 377, 380 380 Narrow-sense heritability, 370,

Objectivity-projectivity, 635

439-440

traits, 624-625, 627, 627, 632, 632, 636, Observability of traits,

815-816, 826, 831-832; 831-832; see also Ordinary language

Observer, 750-754, 853-854, 865

cognition perspective, 803-804 personality descriptions, descriptions, 849, 849, 863 863 personality

Natural selection, 317-340

Need for variety, variety, 832, 832, 839-840 839-840 Need for see also also Human Human needs needs Needs, 9, 9,12, 45, 485; see 12, 45, Negative affect, affect, 772, 772, 778, 778, 780, 780, 783 Negative Negative change, 953 Negative constructs, constructs, 582, 582, 608 Negative Negative emotionality, emotionality, 770 770 Negative Negative outcome anticipation, anticipation, 583, 583, 591, 591, Negative 596-597 Neural substrate, 716 Neuroendocrine system, 892 Neurology, 805 Neuroses, 921 Neuroticism, 12, 27, 361, 405, 406, 737, 746, 825, 827, 836, 841, 841, 858, 861, 861, 862-863, 866, 894, 896, 899-901, 902, 903, 904, 906, 929-930 Niche picking, 384 145, 154, 803, 856, 907 Nomological network, 145,154, 117-118, 119, Nomothetics, 8, 51, 51, 86-87, 86-87,117-118,119, 122-138, 124n, 485, 492, 497, 499, 501, 949 122-138,124n, Nonpersonality variables, 783 Nonshared environment, 244-245, 258, 371, 371, 379-381 Nonsocial cognition, 714 Nonsocial perception, 717 Nonverbal Nonverbal conceptions, conceptions, 231 231 Nonzero Nonzero variance, variance, 119 119 Normative Normative culture, culture, 434 434 Normative Normative models models criteria criteria for for accuracy, accuracy, 650, 650, 652-654, 668n

Observer assessment, assessment, 74, 74, 76, 76, 79-83 79-83 Observer reliability, 278-279 Occupational groups, 858-860 Occupational performance, performance, 751, Occupational 751, 753 Occupational position, 430-432 751, 753 Occupational roles, 751, Occupational situations and dispositions,

181-182 Openness to experience, 27, 79, 737, 746, 863, 930 825-842, 858, 863, Operant conditioning, 594, 950 Operational accuracy, 620-621 Operational criteria for accuracy, 650, 652-653, 670-671, 673 Optimal arousal hypothesis, 399

Optimal personality, 827 Optimism/pessimism, 903-905 Orderliness, 855 Ordinary language, 79, 81, 81, 98-108, 110-111, 113, 98-108,110-111,113, 516, 529, 739, 744, 751, 754-757; see also also

Natural language Organismic integrity, 327 Organizational delinquency, 850, 858 Organizational Organizational/industrial Organizational/industrial psychology, 840-841, 857 Organizational performance, 751, 753 Orientations, 769 action, action, 491 491 psychosocial, psychosocial, 486 486 state, state, 491 491 toward others, 810, 813 toward work, 856

981 981

INDEX

782, Orthogonal models, 738, 742, 748-750, 752, 782, 796, 801-803, 927-928 Other-orientation, 810, 813 Outcome assessment practices, 947-964 Outcome quality, quality, 949 Outcome Outcome variables, variables, 786 Oversampling, 151 Overtness, 450 450 Overtness, Overt tests, tests, 857-858, 857-858, 861-862 861-862 Overt Panic disorder, 936-937 Parallel forms, 146 Paranoid, 926-927, 931 Parental investment, 322, 327 Parental personality and attachment, 210-211 and infant temperament, 210-211 Parent effect on infants, 249 Parent inftuence influence on child, 210-211 Parent perception of child, 210-211, 217 Passion, 825 Passion, Passivity, 514, 514, 521, 521, 959 Passivity, Pathogenic 892-894 Pathogenic processes, processes, 892-894 Pathological syndromes, 926 Pathological personality personality syndromes, 926 Pathophysiological processes, processes, 893, 895-897, 900, Pathophysiological 893, 895-897, 900, 905,908 905, 908 Patient, 947-964 ego strength, 947 personality, 947-948 947-948 personality, population, 956 population, reports, 949 949 reports, Pearson product-moment product-moment correlation correlation Pearson coefficients, 119 Percept-genetic approach (PG), 558-559, 562-563 Perception, 712 direct, 718 nonsocial, 717 person, 717 social, 717 Perceptual defense, 556 Perceptual synaesthesia, 838 Peripheral physiological change, 524, 526-528, 532 530, 532 Perpetrators, 476-477 Personal construct construct theory, theory, 657, 657, 949-950 949-950 Personal Personal documents, documents, 487-488 487-488 Personal Personal goals, 485-506 Personal goals, 485-506 dimensions, 499 Personal illness (clinical syndrome), 938 Personality assessment, assessment, 43, 43, 95-96, 95-96, 619-621, 619-621, 628, 628, Personality 737-758, 849-850, 849-850, 852, 852, 860, 860, 906-907, 906-907, 737-758, 930-932, 952, 952, 953, 953, 956-957, 956-957, 964 930-932, 964

Personality-based tests, tests, 857-858, 861-862 Personality change, 43, 282, 283, 291, 291, 337-339, 949, 958, 962, 963 963 gender-related, 304-307 Personality constructs, 16-17, 143-160, 657, 16-17,143-160, 897-908, 949-950 Personality description, 619-620, 637 Personality development, 43, 78, 184-185, 78,184-185, 199-207 early, 209-234 model, 228-233 Personality deviance, 938 Personality dimensions, 27, 767, 795-816, 825, 894-908 across cultures, 440, 441-450, 454-459 Personality disorder, 750, 767, 841-842, 919-941, 924, 929, 929, 959; 959; see seealso alsoAxis Axis II II 924, assessment, 930-932 930-932 assessment, definition, 932 932 definition, diagnosis, 919-941, 930 diagnosis, 919-941, 930 history, 919-922 history, 919-922 taxonomy, 925 925 taxonomy, Personality disposition, disposition, see Trait Trait Personality Personality dynamics, dynamics, 43, 43, 467, 467, 485, 485, 836, 836, 960 960 Personality Personality integration, integration, 951-951, 951-951, 964 964 Personality Personality inventories, inventories, 15, 15, 83, 83, 84, 84,109,144, Personality 109, 144, 270, 270, 272, 278, 278, 279, 279, 283, 283, 552, 552, 854-857, 854-857, 859, 859, 865, 865, 272, 957, 962 962 957, Personality measurements, measurements, 15, 27, 45, 45,143-160, Personality 15, 27, 143-160, 199, 849, 948 948 Personality processes, 892-894 Personality profiles, 628-629, 631-632 Personality psychology, 3-29, 466, 485, 825, 830, 840, 849, 849, 852, 852, 919-941, 919-941, 934-935 934-935 840, 4-5 definition, 4-5 history, 3-29, 3-29, 43-47 43-47 history, theories, 3-29, 44, 54, 56, 96-113, 118, 285 96-113,118, Personality stability, 337-339 Personality stages, 199-207 Personality structure, structure, 43, 43, 595, 595, 737-758, 826-827, Personality 737-758, 826-827, 851, 930, 948, 950, 960 835-839, 851, Personality taxonomies, taxonomies, 906 906 Personality variables, variables, 422-423, 422-423, 427 427 Personality Personal projects, projects, 25, 25, 485, 485, 492, 492, 494-496, 494-496, 497, Personal 497, 499, 500 500 499, Personal strivings, strivings, 25, 25, 484, 484, 492, 492, 495, 495, 496-497, 496-497, Personal 499, 500 500 Person categories, 722 Person concepts, 720 720 Person Person memory, memory, 722-726 722-726 Person Personnel research, research, 857 Personnel 857 Personnel selection, selection, 85, 85, 89, 89, 753, 753, 840 840 Personnel

982 981

INDEX INDEX

Personology, 99 Personology, Person perception, 717 Person Person schemas, schemas, 555 Person-situation-structure interaction, 424 Person variables, variables, 813-814 Pharmacological therapy, therapy, 958-960 958-960

Press, 9, 9, 489, 489, 801 801 Press, Primary Primary affects, affects, 25 25 Primary Primary process process thinking, thinking, 837 Primate Primate behavior, behavior, 345-355 345-355 development, development, 350-352 350-352 personality personality traits, traits, 352-354 352-354

Phenotype, 74, 74, 76, 76, 79, 79, 84, 84, 111, 369, 369, 497, 497, 742, 742, 755 755 Phobias, Phobias, 45

Priming Priming effect, effect, 716

Physical conditions, conditions, 924; 924; see also also Axis III

Private self-statements, 449-450

Physical energy, energy, 717 Physical health and disease, 487-488, 497, 497, 504,

891-908 Physical traits, traits, 369

Physiological arousal, arousal, 892-893 Physiological dimension, 958-959, 958-959, 961-962

Physiological mechanisms, 586-590, 598-600,

607-608

Physiological reactivity, 897, 897, 899

Piaget, Jean, 199-201, 565, Jean, 23, 23,199-201, 565, 590, 590, 607 Picture-story exercise, 486-488 Planfulness, Planfulness, 299-309 Pleasure-pain dimension, 926 Point-light technique, 719 Point of of view, 626, 628 Political culture, 429 Political psychology, 488 Politician, 650, 650, 656, 656, 658, 658, 662-665, 668-669 Polythetic approach, 923 Popularity, 465 Positive affective affective relationship, 486 Positive change, change, 953 Positive illusions, 650, 665, 667, 671-672 Possible selves, 686-687, 702 Potentiality, 682 Power, 466, 469, 662 agentic, 488 distribution, 295-296 motives, 56, 486-488, 801-802 inhibited, 905-906 stressors, 487 81 Pragmatic rules, 81 Pragmatic utility accuracy, 651 Preconscious, 551 Predictability, 473, 750 750 preferenCb for, preference for, 856 of behavior, 108, 108, 640-641, 652, 935 Prediction of performance, 850, 857, 860-861 in job performance, Predictive limitations, 326-327 Predictive utility, 109-110, 906 Predictive validity, 862-863 Preference, Preference, 584, 586 Prescribed Prescribed behaviors, 591

Private self-consciousness, 362, 362, 826, 826, 840 Probabilistic approaches, 720-722

Probabilistic statements, 120-121 statements, 104, 104,120-121 Problem-focused Problem-focused strategies, 561, 566 Problem-solving strategies, 557 Professed Professed probity, probity, 856 856

Profile Profile partial partial correlations, correlations, 631 631

Progress, 51-53, 54 54 Progress, 3, 3, 27-29, 27-29, 51-53, Projection, Projection, 836-837 836-837

Projective 951, 953 953 Projective tests, tests, 83, 83, 86, 86, 951, Proper-set, Proper-set, 721 721

Proprioceptive Proprioceptive feedback, feedback, 525, 525, 529 529 Proscribed Proscribed behaviors, behaviors, 591 591

Prosocial Prosocial behavior, behavior, 57, 57, 804, 804, 808-816 808-816 Prosocial Prosocial personality, personality, 795, 795, 808-816 808-816 Prototype, Prototype, 23, 23, 923-924 923-924 Prudence, Prudence, 750n, 750n, 751 751

Psychiatric Psychiatric disturbance, disturbance, 938 938

Psychoanalytic Psychoanalytic approach, approach, 26, 26, 43, 43, 44, 44, 45, 45, 55, 55, 58, 58,

295, 590, 295, 418, 418, 466, 466, 543-546, 543-546, 551, 551, 555, 555, 557, 557, 590, 594-596, 601, 745, 745, 795-796, 795-796, 827, 827, 836, 836, 851, 851, 594-596, 601, 921, 921, 948-949, 948-949, 951, 951, 953, 953, 963 963 Psychobiography, 506 Psychobiography, 41-64, 41-64, 426, 426, 488, 488, 506 Psychodiagnosis, Psychodiagnosis, 45, 45, 47, 47, 56, 56, 58, 58, 554-555, 554-555, 721, 721, 748-750, 923-941, 952, 958 952, 958 748-750, 923-941, Psychodynamics, 41, 57, 57, 553, 553, 555, 555, 927-928 927-928 Psychodynamics, 41, Psychohistory, 433 Psychohistory, 48, 48, 426, 426, 433 Psycho lexical analysis, Psycholexical analysis, 796-798 796-798 Psycholinguistics, 740, 754 Psychological automatisms, 545 Psychological centrality processes, 432 Psychological magnification, magnification, 25 Psychological measurements of of defense, 563 Psychological mechanisms, 323-325 Psychological sources of of inhibitions, 590-597, 600-606 Psychological terminology, 73 389, Psychometrics, 41, 43-44, 96-113, 200, 206, 389, 486, 488, 500, 651, 903, 929, 951 Psychopathic personality, 920-921 Psychopathology, 606, 750, 825, 836, 933-934, 936-937, 959 Psychopathy, 585, 600, 852, 920 920

983 983

INDEX

Psychophysiology, 387-408 indicator measurement, 393-399, 407 reactivity, 907 907 Psychosis, 920 Psychosocial orientations, 486 also Axis IV Psychosocial stressors, 924; see also Psychosocial theory, 54-55, 58 Psychosomatic hypothesis, 891 Psychosomatic illness, 504 Psychosomatic medicine, 285 Psychosomatics, 891-908 Psychotherapy, 748-750, 826, 840, 947-964 outcome assessment, 947-964 history, 948-953 929 Psychoticism, 12, 12, 929 Psychotropic medications, medications, 589-590, 589-590, 599 Psychotropic Public area, SO-83 80-83 Public area, Public self-awareness, 358 Public self-awareness, 358 Public self-statements, self-statements, 449-450 449-450 Public Quale, or feeling tone, 526 Range of convenience, 744, 754 Range of interests, 832 Rational/emotive therapy, 950-951 Rationalization, 604 604 Reaction range, 246-247 Reactivity, 391, 391, 929 929 Rebelliousness, 852 Reciprocal influences, 167, 247 Reference criteria, 497-498 Reference value, 497-498 Reflected appraisal process, 432 Reflexes, 514 Reflexivity, 522, 524, 532 Regression in service of the ego, 827, 837 Reinforcement, 11, 167, 591, 11,167, 591, 592, 594, 925-927 Reinforcement matrix, 926 Rejection, 470-471, 475-476, 590 Relational theory, 295, 304, 306 Relational trust, 466, 468-475 Relationships, 347-348, 486, 681 487 Relaxed affiliative affiliative syndrome (RAS), 487 Reliability, 145-148, 149-150 145-148,149-150 Religion, 596 Repressed memories, 545-546; 545-546; see also also Child sexual abuse; Memory �emory Repression, 545-546, 548-551, 556, 556. 562, 562, 584, 584, 635, 836-837; see also �emory Memory effects effects on memory, 546, 550, 556 ' Repressive style, 548-551

Repressor construct, 551 Reproductive success, 322, 322, 467, 852-853 Reputation, 467-469, 751-752, 853-854, 865; 467-469,751-752, 865; see see also also Social reputation Reputational consensus, 798 Resentment, 852 Response class, class, 102 Response competition, 583 Response sets, 277, 441-442 Response styles, styles, 16, 629 Responsibility, 860-861 Restraints, internal, 581-609 Retrieval, 712, 715, 715, 725 Rhathymia, 770 851, 933 933 Rigidity, 851, 480-481 Risks, 469-470, 472, 474-475, 4SO-481 Risk-taking, 468-469, 770-771 Ritualized social social interaction, interaction, 26, 26, 750; 750; see also also Ritualized Social interaction Role relationship models, 555 Roles, see see Social Social roles roles Roles, Role-taking, 851-852 851-852 Role-taking, Role theory, theory, 26, 26, 662-663, 662-663, 713-714 713-714 Role Rule-compliance, 852, 852, 856 856 Rule-compliance, Sampling bias, 272 Scene, 25 25 23, 555, 661, 681, 685, 800-801, Schema, 23, 661, 681, 803 803 926, 933 Schizoid, 926, Schizophrenia, 924, 924, 933, 933, 937 937 Schizophrenia, Schizotypal, 927, 927, 933 Schizotypal, 933 Scientist, 650, 650, 656, 656, 657-660, 657-660, 667-668 667-668 Scientist, Script theory, theory, 25, 25, 55, 55, 443, 443, 715, 715, 720 720 Script Secondary process thinking, 837 Secret area, area, 80-83 Selective breeding of animals, 588 SeU, Self, 23, 449-450, 617-641, 649-673, 681-703 Self-actualization, 54, 291, 291, 585 Self-affectivity, Self-affectivity, 219 Self-affirmation, Self-affirmation, 665 Self-agency, 219 Self-aggrandizement, 669 Self-assertion, 492 Self-assessment, 74, 76, 79-83 Self-awareness, 358, 694, 697-700, 702 public, 358 SeU-Aware Self-Aware Stage, 204 Self-care, 295 Self-centeredness, 685 Self-coherence, 219 219 Self-coherence,

984 984

INDEX

Self-concept, Self-concept,

303, 420, 603, 603, 619, 619, 209, 215-234, 303, 655, 660-663, 666, 681-703, 681-703, 726, 960-961 960-961 age for for development, 218-220 in children, 209, 215-234 clarity, 662 development in preschool and early school schoolage children, 223

Self-confidence, Self-confidence,

685 357-359, 362 private, 362, 362, 826, 840 Self-control, 609 Self-deception, 547, 565, 565, 649, 683, 683, 686, 853 Self-deceptive Self-deceptive positivity, 666 Self-defeating behavior, 693-695, 701-702, 928 Self-defeating Self-definition, 295, 302 Self-definition, 295, 302 631 Self-description, 224, 224, 619-621, 626, 631 children's, 224 Self-destructive behavior, 59, 59, 694-695, 699, Self-destructive 699, 702 828 Self-disclosure, 634, 745, 745, 828 Self-efficacy, 593, 593, 895, 895, 961 Self-efficacy, Self-enhancement, 622-626, 650, 654, 664-673, 664-673, Self-enhancement, 693 358, 362, 432, 469, 543, 543, 547, 547, Self-esteem, 303, 303, 358, 557, 560, 564, 566, 665-666, 671, 681, 687-693, 702, 897, 940, 961 Self-expression, 685 Self-history, 219 Self-image, 278, 618-634, 665, 665, 669, 751 751 Self-importance, 666, 670 653, 656, 667, Self-insight, 649-650, 653, 667, 671, 673 Self-interest, 694 Self-interpretation, Self-interpretation, 751 Self-judgments, 617-641 617-641 Self-knowledge, 635, 635, 649, 659, 683, 683, 685-689, 693, 702 Selflessness, 795 Self-monitoring, 22, 152, 664-651 22,152, 664-651 Self-narrative, 532 Self-other Self-other congruence, 617-641 Self-other Self-other dimension, 926 Self-perception, 622, 649-673 bias, 657, 667, 668, 672 672 processes, 432, 656-667 Self-presentation, Self-presentation, 85, 85, 86, 362, 467, 632, 656, 663-665, 668-669, 689, 751-752, 865-866, 931 931 Self-preservation, Self-preservation, 694 Self-processes, 657, 657, 666 Self-protection, 693 Self-Protective Self-Protective Stage, 203 Self-psychology, 55 Self-ratings, 621, 623-625, 627-629, 635, 637, 637, 651, 653, 740

Self-consciousness,

Self-regulation,

805, 816, 805, 816, 903 270-271, 653, 653, 660, 663-666, 742, 745, 745, 931-932, 936, 936, 937-938 children's, 222, 276, 278 of of emotion, 530 of -901, 903 of illness, 487, 487, 896, 896, 900 900-901, of of physiological activity, 527 Self-restraint, 487, 905 661 Self-schema, 661 Self-serving bias, bias, 622 Self-serving 902, 959 Self-statements, 449-450, 902, collective, 449-450 negative, 959 positive, 902, 902, 959 private, 449-450 public, 449-450 Self-talk, 952 Self-verification, 619, 661 Self-verification, 619, 661 Self-view, 649, 649, 655-656, 660-661, 660-661. 663, 663, 665, 665, 751 751 Self-view, 656, 665-666, 897 Self-worth, 656, differential model, 718 Semantic differential Semantic relations, 755 504, 828 Sensation, 504, vs. intuition, 828 399-404. 771, 776, 781, 781, Sensation-seeking, 390, 399-404, 839, 842; 842; see also also Adventure Adventure825, 827, 839, seeking; Excitement-seeking; Excitement-seeking; Sensitivity; Sensitivity; seeking; Thrill and and adventure adventure seeking seeking Thrill Sensory surfaces, surfaces, 717 Sensory Sentiments, 73 Sentiments, Sex, 543, 551 differences, biological, 292-294, 296, 322-323. differences, 322-323, 335-336 infant, 248 of infant, 45, 57, 57, 59 Sexual behavior, 45, reproduction, 321-323, 327-330, 335-336 Sexual reproduction, envu*onment, 371, 376, 379-381 Shared environment, Sibling differences, 244, 257-258 Sibling Sibling influences, influences. 244 Simulation models, models, 714, 725 Simulation 714, 725 Single factors, 152, 243, 862 152, 243, 862 Situational attributes, 626-627 Situational attributes, 626-627 Situational choices, choices, 165-187 165-187 Situational and self-conceptions, 178-179 and social attitudes, 179-180 and social relationships, 180-181 180-181 609, 905-907 Situational factors, 609, Situational press, 801 801 Situational trust, 469 Situational variables, variables, 430 430 Situational Situationism, 18, 74, 421-423, 421-423, 425-428, 425-428, 713 713 Situationism, 18, 74, Situation taxonomies, taxonomies, 167 Situation 167 Self-reports and reliability,

98S 985

INDEX Six-factor model, 751-754 Sociability, 251-252, 347, 352-353, 375, 451, 451, 750-752, 770-771, 814, 929 Social acceptance, 852-854, 865 Social anxiety, 362 Social attitudes, 179-180, 826, 836 Social behavior, 850, 937 Social categories, 720-722 Social class, 430-432, 830 Social clock, 296-297 Social closeness, 773, 773, 776-777 Social cognition, 422, 656, 661, 712-728 Social-cognitive approach, 496 Social comparison theory/process, 432, 659, 665, 716 Social competence, 813-814 Social consensus, consensus, 650-653, 650-653, 668n, 668n, 670, 670, 671, 671, 673, 673, Social 798-799 criteria for for accuracy, accuracy, 650-653, 650-653, 668n, 668n, 670, 670, 671, 671, criteria 673 673 Social construction of of self, self, 618-619, 618-619, 636 Social construction 636 Social context, context, 20, 20, 74 Social 74 Social desirability, desirability, 16, 16, 270, 548, 666, 666, 745, 752, Social 270, 548, 745, 752, 939 Social evaluations, 739 Social exchange, 748 Social exclusion, 691 Social history of women, 292 Social insensitivity, 852, 854, 854, 858 Social insensitivity, 852, 858 Social interaction, interaction, 715, 750, 851-854, 865 Social 715, 750, 851-854, 865 Social/interpersonal component, component, 772, 772, 783 Social/interpersonal 783 Socialization, 13, 359-361, 359-361, 521, 521, 590-593, 590-593, 595, Socialization, 13, 595, 601, 607, 609, 713, 805, 810-811, 596, 601, 851-852, 854, 864 and gender, 293, 293, 294 and Socialization models, 242 Social judgment theory, 714 585, 590, 808, 810, 961 Social learning, 585, 723, 725 Social memory, 723, Social milieu, 928, 937 Social motives, 487-489, 505 Social network, 479 Social perception, 652, 717 717 Social perception, 652, Social performance, performance, 853-854 853-854 Social Social planning, planning, 283-284 283-284 Social Social potency, potency, 773, 773, 777 Social 777 Social potentials, 521, 523 523 Social potentials, 521, Social psychology, psychology, 13, 13, 23, 23, 26, 26, 275, 275, 417-434, 417-434, Social 547-548, 563-565, 618-619, 629, 650, 711-713, 798-799, 840 psychological, 420, 422-423 sociological, 420, 420, 422-423 422-423

Social relationships, 180-181 Social relations model, model, 630-631 Social responsibility, 604, 609 Social roles, 275, 523, 523, 535-537, 681-684, 713, 751, 962 Social stereotypes, 722 Social structural variables, 501 Social structure, 417 Social systems of behavior, 521 Social Social validation, 687 Societal maintenance systems, 418 Societal motive profiles, 488 Socioanalytic theory, 26, 451, 466, 752-754, 851, 26,451,466, 852-854, 865-866 Sociobiology, 26, 43, 43, 44, 808-809 Sociocognitive development, 810-811, 813-814 Sociocultural context, 495 Sociological approach, 618, 620, 636 Sociological role theory, see Role theory Sociometries, Sociometrics, 937 Sociosexuality, 637 Sociotypes, 456 Soft psychology, 130-131 Soft synthesis, 42, 59-63 Somatopsychic processes, 907 Source polarity, 926 Source traits, 113, 521-522, 929 113, 521-522, Spatial images, 723 Spatial models, 713 Special processes, 533-534 Specialties, 439, 442 Stability/instability (neurosis), 827, 950; 950; see also also Neurosis of behavior, 166 of personality, 337-339 Stage theories, 23, 199-207 23,199-207 Standard nomenclature, 922-925, 931 Standard reduced personality sphere, sphere, 797 State and trait variables, 154 State anxiety, 959-960 State manifestations, manifestations, 924 State measures, 958 State orientation, 491 States, 225, 739, 772 and dispositions, 225 Status, 26, 349, 430-432, 466, 467, 475, 486, 662-663, 748, 750, 752-753, 774, 852-854, 865 Status hierarchy, 852, 852, 865 Status-seeking, 752-753 Stimulus, 524 Stimulus array, 716-718

986

INDEX INDEX

Stimulus ensemble, ensemble, 726 726 Stimulus Stimulus-response behaviorism, behaviorism, 74, 74, 97-98, 97-98, 592, 592, Stimulus-response 712, 712, 714 Strategic Strategic individual individual differences, differences, 331-333 331-333 Stress, 57, 57, 487, 487, 551, 551, 556, 556, 560-561, 560-561, 562, 562, 566, 566, Stress, 892-895, 902, 902, 907 907 892-895, appraisal, 902 902 appraisal, hormones, 487 487 hormones, traumatic, 556 556 traumatic, Stress-moderation model, model, 892-895, 892-895, 897, 897, 901, 901, 903, 903, Stress-moderation 905,907 905, 907 Stressors, 487, 487, 892-893, 892-893, 897, 897, 899, 899, 902-903 902-903 Stressors, external, 487 487 external, power, 487 487 power, psychosocial, 924 924 psychosocial, Structural change, change, 951, 951, 961 961 Structural Structured interviews, interviews, 931-932 931-932 Structured Subjective culture, culture, 443, 443, 457 457 Subjective Subjective visibility, visibility, 636 636 Subjective Subjectivity, 50, 50, 514, 514, 515 Subjectivity, 515 Subliminal perception, perception, 563 563 Subliminal Subordination, 450-451, 454 454 Subordination, 450-451, Superego, 582, 582, 595, 595, 601, 601, 851 851 Superego, Superordination, 450-451, 450-451, 454 454 Superordination, Suppressed behaviors, behaviors, 582, 582, 591 Suppressed 591 Suppression, 635, 635, 837 Suppression, 837 Surface traits, traits, 113 Surface 113 Surgency, 452, 452, 454, 454, 748-750, 748-750, 774, 774, 776, 776, 780, 780, 784; 784; Surgency, also Extraversion Extraversion see also Survival, 321, 321, 327, 327, 328, 328, 853 853 Survival, Suspension of of individual individual values, values, 605 Suspension 605 Symbolic Symbolic gestures, gestures, 687 687 Symbolic information processing, 714-728 Symbolic interactionism, 423, 618, 662-663 Sympathetic nervous system, 892, 897 Symptomatic disorders, 921, 936 936 Symptoms, 921-922, 928, 936, 938-941, 949, 954, 958-964 Symptom-specifi Symptom-specificc measures, 956 Syndromal coherence, 940 Systemic context, context, 495 Systemic 495 Systems of of control, control, internal, 593-596 Systems internal, 593-596 Taboos, 582, 584, 588-589, 613 Taboos, 582, 584, 586, 586, 588-589, 613 Target, Target, 524-527 524-527 956, 958 . behaviors, behaviors, 954, 954, 956, 958 symptoms, symptoms, 958 958 variance, 630-631 variance, 630-631 Taxonomic 515-516 Taxonomic tree, tree, 515-516 Taxonomies, 89, 95-96 95-96 Taxonomies, 89, class inclusion, 516n class inclusion, 516n of 566-568 of defense, defense, 566-568

of ego actions, actions, 552 552 of emotion, emotion, 515 of human human needs, needs, 747 of knowledge structures, structures, 715 715 of personality, 906 of personality attributes, 739, 739, 742 742 of personality disorder, disorder, 925 of situations, situations, 167 of social behavior, 937 of trait descriptive terms, 739-744, 754-758 of traits, 27, 28, 89, 768, 906 Teleonomic trends, 497 Temperaments, see Personality descriptions; Traits Temporal strings, strings, 723 723 Temporal Temptation, 246, 246, 584, 584, 603, 603, 609 Temptation, Tension, 10, 10,12,17 Tension, 12, 17 Terror management (1M), (TM), 546-547, 691 Test-retest correlation, correlation, 147, 147, 276, 276, 278, 278, 501 501 Test-retest Thema, 99 Thema, Theme, 497, 497, 726 726 Theme, Therapeutic process, process, 948 948 Therapeutic Therapist personality, personality, 947-948 947-948 Therapist Thought disorder, disorder, 939 Thought 939 Three-factor model, 746, 768 Three-factor model, 746, 768 Three-factor semantic differential model, model, 718 718 Three-factor semantic differential Three-stage model model for for women's women's growth, growth, 295 Three-stage 295 Thrill and and adventure adventure seeking (TAS), Thrill seeking (T AS), 399-400; 399-400; also Adventure-seeking; Adventure-seeking; Excitement Excitementsee also seeking; Sensation-seeking Sensation-seeking seeking; Tolerance, 831 Tolerance, 831 for ambiguity, 444, 444, 832, 840 832, 838, 838, 840 Tonic arousal arousal system, system, 392-393 392-393 Toxic substances, 589-590, 599, 608 Trait anxiety, 959-960 Trait attribution, 655, 748, 751 Trait disturbances, 924 Trait language/descriptive terms, 74, 79, 86, 441, 441, 739-744, 754-758, 796-797, 796-797, 800-801, 800-801. 849, 853, 853, 865 865 Traits, Traits, 8,11,18, 8, 11, 18, 20-22, 20-22, 26-27,28, 26-27, 28, 41, 41, 73-89, 73-89, 95-113,166,167n, 362, 486, 486, 520, 95-113, 166, 167n, 270, 270, 309, 309, 362, 520, 537, 685, 686, 686, 711, 711, 537, 548-551, 548-551, 655, 655, 681, 681, 683, 683, 685, 717-718, 717-718, 723-725, 723-725, 738-758, 738-758, 767-790, 767�790, 825-842, 894-896, 899-900, 825-842, 894-896, 899-900, 903-907, 903-907, 919-941,959-960 919-941, 959-960 of persons, 102-108 as attributes of as categorical summaries, sunmiaries, 103-108 as causal dispositions, 102-103 explanations of of behavior, 110-113 as explanations favorability, 624-625 640 observability, 624-625, 627, 632, 636, 640

987

INDEX INDEX

physical, 309 of behavior, 108 as predictors of taxonoDlY of, 27, 28, 89, 768, 906 taxonomy of, Traitlsituationist debate, 18, 21-22, 26, 74-76, Trait/situationist 95-113 Trait variation, 369-374 Dlodels of development, developDlent, 242-243 Transactional models Transactional views, 893-894 TrauDla, 599, 836 Trauma, TrauDlatic stress, 556 Traumatic TreatDlent course, 928, 939, 947-964 Treatment duration, 956 effects, 956 focus, 955 goals, 952, 955, 962 outcoDles, outcomes, 934 type, 956, 959 True score, 146 Trust, 58, 466, 468-475, 480-481, 488, 3-89, 95-113,166,167n, 95-113, 166, 167n, 270, 309, 362, 486, 520, 537, 548-551, 655, 681, 711, 681, 683, 685, 686, 711, 537, 717-718, 723-725, 738-758, 767-790, 899-900,, 903-907, 825-842, 894-896, 899-900 919-941, 959-960 959-960 affiliative, 488 cynical, 488 generalized, 466, 468-471, 474-475 relational, 466, 468-475 situational, 469 vs. Dlistrust, mistrust, 601 Trustworthiness, 469, 472 Twin studies, studies, 244-245, 249, 257-258, 369, 551, 588, 809, 370-374, 375-378, 379, 382, 551, 831 831 Two-factor Dlodel, model, 740 Type A pattern, 891, 891, 897-899, 906 Type B, B, 897 Type language, 74, 74, 86 Types, 73-89, 144, 200, 921-922 73-89,144, Typological model of repressive style, 549-551 549-551 Typologies of personality, 747, 747, 751-754, 768, 768, 925

Unconscious, 951 951 Unconscious area, area, SO-83 80-83,, 85

Uniqueness/individuality Uniqueness/individuality distinction, 125 Units of of analysis, 73-89 Unity thema, 9 Universals, 439, 450 Utility maximizing, maxiDlizing, 657n Valence, 10, 772 Validity, 15, 500-501, 650n, 65On, 651-652 convergent, 15, 154-155, 619, 6.1 9, 641, 745, 750, 904 898, 902, 904 15,154-155, discriminant, 15, 154-155, 750, 789, 898, 902, 904 Dleasurement, measurement, 651-652 predictive, 862-863 Value development, 601-603, 606-608 606-608 Value developDlent, 601-603, Values, 78,499-500, 584-585, 590, 590, 593, Values, 78, 499-500, 582-583, 582-583, 584-585, 593, 595-596, 600-608, 682, 683, 683, 684-685, 684-685, 686, 595-596, 600-608, 682, 686, 827, 830, 830, 832, 832, 839 827, 839 Variability of of values, values, 582-583 582-583 Variability Variants, 439 Veiled purpose purpose tests, tests, see Personality-based Personality-based tests tests Veiled Venturesomeness, 775-776, 775-776, 786, 786, 825, 825, 929 VenturesoDleness, Verbal behavior, 524, 530-532 Verbal code, 723 Verbal conceptions, 232 Verisimilitude, 131, 131, 133 VictiDls, Victims, 476-477 Virtuousness, 856, 866 Visceral activity, 526, 529 women, 299-309 Vocational goals for WODlen, Vocational preference, preference, 751 751 Vocational Within culture culture variance, variance, 441, 441, 443, 443, 448 448 Within Within family family effects, effects, 243-244 243-244 Within Women, adult adult developDlent development in, in, 291-309; 291-309; see see also also WODlen, Gilligan model model Gilligan social history, history, 292 social 861 Work ethics, 861 Working DleDlory, memory, 715-716 715-716 Working Work orientation, orientation, 856 856 Work Workplace testing, testing, 857 857 Workplace variance, 1119 Zero variance, 19