Advances in Child Development and Behavior, Volume 36

ADVANCES IN CHILD DEVELOPMENT AND BEHAVIOR VOLUME 36 Contributors to This Volume Zhe Chen David Klahr E. Mark Cumm...

Author: Robert V. Kail

54 downloads 2242 Views 3MB 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

ADVANCES IN CHILD DEVELOPMENT AND BEHAVIOR

VOLUME 36

Contributors to This Volume Zhe Chen

David Klahr

E. Mark Cummings

Melinda S. Leidy

Lisa M. Diamond

Kelly L. Madole

Christopher P. Fagundes

Lisa M. Oakes

Teresa Farroni

Jodie M. Plumert

Roberta Michnick Golinkoff

Alice C. Schermerhorn

Tobias Grossmann

David L. Share

Nancy G. Guerra

Keith E. Stanovich

Kathy Hirsh-Pasek

Maggie E. Toplak

Mark H. Johnson

Richard F. West

ADVANCES IN CHILD DEVELOPMENT AND BEHAVIOR

edited by

Robert V. Kail Department of Psychological Sciences Purdue University West Lafayette, IN, 47907, USA

Volume 36

AMSTERDAM BOSTON HEIDELBERG LONDON NEW YORK OXFORD PARIS SAN DIEGO SAN FRANCISCO SINGAPORE SYDNEY TOKYO Academic Press is an imprint of Elsevier

Academic Press is an imprint of Elsevier 525 B Street, Suite 1900, San Diego, CA 92101-4495, USA 30 Corporate Drive, Suite 400, Burlington, MA 01803, USA 84 Theobald’s Road, London WC1X 8RR, UK Radarweg 29, PO Box 211, 1000 AE Amsterdam, The Netherlands First edition 2008 Copyright r 2008 Elsevier B.V. All rights reserved No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of the publisher Permissions may be sought directly from Elsevier’s Science & Technology Rights Department in Oxford, UK: phone: (+44) 1865 843830, fax: (+44) 1865 853333, E-mail: [email protected]. You may also complete your request online via the Elsevier homepage (http://www.elsevier.com), by selecting ‘‘Support & Contact’’ then ‘‘Copyright and Permission’’ and then ‘‘Obtaining Permissions.’’ Recognizing the importance of preserving what has been written, Elsevier prints its books on acid-free paper whenever possible.

Library of Congress Cataloging-in-Publication Data A catalog record for this book is available from the Library of Congress

British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library ISBN: 978-0-1237-4317-6 ISSN: 0065-2407 (Series)

For information on all Academic Press Publications visit our Web site at www.books.elsevier.com

Printed in the United States of America 08 09 10 11 12

10 9 8 7 6 5 4 3 2 1

Contents Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

ix

Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

xi

King Solomon’s Take on Word Learning: An Integrative Account from the Radical Middle KATHY HIRSH-PASEK AND ROBERTA MICHNICK GOLINKOFF I. The Word Learning Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . II. The Emergentist Coalition Model (ECM) . . . . . . . . . . . . . . . . . . . . . . . III. Validating the Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV. Implications of the ECM for Language Disorder: The Case of Autism . . . V. Implications of a ‘‘Radical Middle’’ Approach: Three Take-Home Messages References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2 3 7 21 23 25

Orthographic Learning, Phonological Recoding, and Self-Teaching DAVID L. SHARE I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . II. The Self-Teaching Theory of Orthographic Learning III. Empirical Findings . . . . . . . . . . . . . . . . . . . . . . . IV. Summary, Conclusions, and the Way Ahead . . . . . Acknowledgement . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

.. . .. .. .. ..

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

31 35 43 73 75 75

. . . . .

85 91 102 113 119

. . . .

135 139 145 154

Developmental Perspectives on Links between Attachment and Affect Regulation Over the Lifespan LISA M. DIAMOND AND CHRISTOPHER P. FAGUNDES I. Review of Attachment Theory . . . . . . . . . . . . . . . . . . II. Affect Regulation . . . . . . . . . . . . . . . . . . . . . . . . . . . III. Attachment and Affect Regulation During Adolescence IV. Implications and Future Directions . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

Function Revisited: How Infants Construe Functional Features in their Representation of Objects LISA M. OAKES AND KELLY L. MADOLE I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . II. The Construct of Object Function . . . . . . . . . . . . . . . . . . . . . . . . . . III. A New Conception of Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV. Our Research on Infants’ Attention to and Representation of Function

. . . .

Contents

vi V.

Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

178 180 180

Transactional Family Dynamics: A New Framework for Conceptualizing Family Influence Processes ALICE C. SCHERMERHORN AND E. MARK CUMMINGS I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . II. Setting the Stage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . III. Transactional Family Dynamics: An Emerging Theme . . . . . . . . IV. Mapping Empirical Work onto a Transactional Family Dynamics Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . V. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

...... ...... ......

187 191 201

. . . .

. . . .

228 240 242 242

. . . . .

251 252 254 259 263

.

267

. .

275 277

. . . . . . .

287 293 298 307 318 321 322

MARK H. JOHNSON, TOBIAS GROSSMANN, AND TERESA FARRONI I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . II. Face Processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

331 334

. . . .

. . . .

. . . .

. . . .

The Development of Rational Thought: A Taxonomy of Heuristics and Biases KEITH E. STANOVICH, MAGGIE E. TOPLAK, AND RICHARD F. WEST I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . II. Experimentally Tractable Deﬁnitions of Rational Thought . . . . . . . . . . III. Dual-Process Models of Cognition . . . . . . . . . . . . . . . . . . . . . . . . . . . IV. A Preliminary Taxonomy of Rational Thinking Errors . . . . . . . . . . . . . V. Classifying Heuristics and Biases . . . . . . . . . . . . . . . . . . . . . . . . . . . . VI. Exemplar Developmental Studies in the Different Categories of the Taxonomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VII. Conclusion: Speciﬁcity and Generality in the Development of Rational Thought . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Lessons Learned: Recent Advances in Understanding and Preventing Childhood Aggression NANCY G. GUERRA AND MELINDA S. LEIDY I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . II. The Adaptive Functions of Aggression . . . . . . . . . . . . . . . III. Aggression and the Ecology of Development . . . . . . . . . . . IV. Risk, Causality, and Prevention . . . . . . . . . . . . . . . . . . . . V. Translating Research to Practice: Building an Evidence Base VI. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

The Social Cognitive Neuroscience of Infancy: Illuminating the Early Development of Social Brain Functions

Contents III. IV. V. VI.

Eye Gaze Processing . . . . . . . . . . Perception of Emotions . . . . . . . . Interactions Between Face Identity, Conclusions . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . .

vii

.................. .................. Eye Gaze, and Emotion .................. ..................

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

343 354 357 362 364

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

373 374 378 401 412 413 413

ZHE CHEN AND DAVID KLAHR I. Introduction: Theoretical, Empirical, and Practical Importance of Research in Remote Transfer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . II. Paradox, Taxonomy, and Paradigms . . . . . . . . . . . . . . . . . . . . . . . . . . III. Evidence of Remote Transfer I: Experimental Approach . . . . . . . . . . . . . IV. Evidence of Remote Transfer II: Naturalistic, Cross-Cultural Approach . . V. Processes Involved in Remote Transfer . . . . . . . . . . . . . . . . . . . . . . . . . VI. Developmental Differences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VII. Educational Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VIII. Conclusions and Future Directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

420 422 426 448 455 458 460 463 465

Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

471

Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

479

Contents of Previous Volumes

493

Children’s Thinking is Not Just about What is in the Head: Understanding the Organism and Environment as a Unified System JODIE I. II. III. IV. V.

M. PLUMERT Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The General Theoretical Framework . . . . . . . . . . . . . . The Development of Spatial Categorization . . . . . . . . . . Explaining the Emergence of Spatial Categorization Skills Limits and Conclusions . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

.... .... .... ... .... .... ....

. . . . . . .

. . . . . . .

Remote Transfer of Scientific-Reasoning and Problem-Solving Strategies in Children

....................................

This page intentionally left blank

Contributors

ZHE CHEN Human Development and Family Studies, University of California, Davis, California 95616, USA E. MARK CUMMINGS Department of Psychology, University of Notre Dame, Notre Dame, Indiana 46556, USA LISA M. DIAMOND Department of Psychology, University of Utah, Salt Lake City, Utah 84112, USA CHRISTOPHER P. FAGUNDES Department of Psychology, University of Utah, Salt Lake City, Utah 84112, USA TERESA FARRONI Centre for Brain and Cognitive Development, School of Psychology, Birkbeck, London WC1E 7HX United Kingdom and University of Padua, Italy ROBERTA MICHNICK GOLINKOFF School of Education, University of Delaware, Newark, Delaware 19716, USA TOBIAS GROSSMANN Centre for Brain and Cognitive Development, School of Psychology, Birkbeck, London, WC1E 7HX United Kingdom NANCY G. GUERRA Department of Psychology, University of California, Riverside, California 92521, USA KATHY HIRSH-PASEK Department of Psychology, Temple University, Philadelphia, Pennsylvania 19122, USA MARK H. JOHNSON Centre for Brain and Cognitive Development, School of Psychology, Birkbeck, London, WC1E 7HX United Kingdom DAVID KLAHR Department of Psychology, Carnegie-Mellon University, Pittsburgh, Pennsylvania 15213, USA MELINDA S. LEIDY Department of Psychology, University of California, Riverside, California 92521, USA

x

Contributors

KELLY L. MADOLE Department of Psychology, Western Kentucky University, Bowling Green, Kentucky 42101-1030, USA LISA M. OAKES Center for Mind and Brain, University of California, Davis, California 95618, USA JODIE M. PLUMERT Department of Psychology, University of Iowa, Iowa City, Iowa 52242, USA ALICE C. SCHERMERHORN Department of Psychology, University of Notre Dame, Notre Dame, Indiana 46556, USA DAVID L. SHARE Department of Learning Disabilities, Faculty of Education, University of Haifa, Mount Carmel, 31905, Haifa, Israel KEITH E. STANOVICH Department of Human Development and Applied Psychology, University of Toronto, Toronto, Ontario M5S 1V6, Canada MAGGIE E. TOPLAK Department of Psychology, York University, Toronto, Ontario M3J 1P3, Canada RICHARD F. WEST Department of Graduate Psychology, James Madison University, Harrisonburg, Virginia 22807, USA

Preface Advances in Child Development and Behavior is designed to provide scholarly technical articles and to provide a place for publication of scholarly speculation. In these critical reviews, recent advances in the ﬁeld are summarized and integrated, complexities are exposed, and fresh viewpoints are offered. These reviews should be useful not only to the expert in the area but also to the general reader. No attempt is made to organize each volume around a particular theme or topic. Manuscripts are solicited from investigators conducting programmatic work on problems of current and signiﬁcant interest. The editor often encourages the preparation of critical syntheses dealing intensively with topics of relatively narrow scope but of considerable potential interest to the scientiﬁc community. Contributors are encouraged to criticize, integrate, and stimulate, but always within a framework of high scholarship. I acknowledge with gratitude the aid of my home institution, Purdue University, which generously provided time and facilities for the preparation of this volume. Robert V. Kail

This page intentionally left blank

KING SOLOMON’S TAKE ON WORD LEARNING: AN INTEGRATIVE ACCOUNT FROM THE RADICAL MIDDLE$

Kathy Hirsh-Paseka and Roberta Michnick Golinkoff b a

DEPARTMENT OF PSYCHOLOGY, TEMPLE UNIVERSITY, PHILADELPHIA, PA 19122, USA b UNIVERSITY OF DELAWARE, NEWARK, DE 19716, USA

I. THE WORD LEARNING PROBLEM II. THE EMERGENTIST COALITION MODEL (ECM) A. A SHORT HISTORY B. CHARTING THE LANDSCAPE IN WORD LEARNING: A CARICATURE C. ASSUMPTIONS OF THE ECM MODEL III. VALIDATING THE MODEL A. DATA FROM THE LEARNING OF OBJECT LABELS B. DATA FROM THE LEARNING OF ACTION LABELS C. STEPPING BACK: WHY VERBS ARE HARD D. INFANTS CAN FIND ACTIONS AND ACTION COMPONENTS IN EVENTS E. INFANTS CAN ALSO FORM CATEGORIES OF ACTIONS F. EVEN TODDLERS HAVE TROUBLE MAPPING WORDS ONTO ACTIONS: THE POWER OF PERCEPTUAL SALIENCE G. BEYOND PERCEPTUAL CUES FOR VERB MAPPING H. THE ECM AND VERB LEARNING IV. IMPLICATIONS OF THE ECM FOR LANGUAGE DISORDER: THE CASE OF AUTISM V. IMPLICATIONS OF A ‘‘RADICAL MIDDLE’’ APPROACH: THREE TAKE-HOME MESSAGES REFERENCES

$

This work was supported by an NICHD grant 5R01HD050199 and by NSF grant BCS-0642529. This chapter is adapted from the Boston Language Conference keynote address delivered in November 2007 and presented in the 2007 Conference Proceedings.

1 Advances in Child Development and Behavior R.V. Kail : Editor

Copyright r 2008 Elsevier B.V. All rights reserved.

2

Kathy Hirsh-Pasek and Roberta Michnick Golinkoff

I. The Word Learning Problem Consider the words ‘‘dog’’ and ‘‘shoe,’’ among the earliest acquired by young children. How do children learn which words go with which objects in their world? Objects are palpable, stable across time, and perceptually consistent. Even so, a child must ﬁgure out that the word ‘‘dog’’ goes with the whole dog and not dog ears or the dog’s bark. Mapping words onto actions and events gets even more complicated. Where do events begin and end? Why do we use the word ‘‘running’’ for a dog and for the neighbor who jogs down the street? This seemingly simple task of linking words to world has eluded philosophical and psychological discussions for centuries. Words are the conceptual building blocks of language. Despite over a millennium of discussion from Plato’s Cratylus to Wittgenstein’s Tractatus Logico-Philosophicus to Brown’s Names for Things, we have yet to understand how we map words to world. In this chapter we examine the question anew by focusing on how children learn their ﬁrst words and come to imbue them with meaning. Do children link the words they hear to the most perceptually salient features of the environment? Do they ‘‘read’’ social cues from master word users who guide them towards the correct referent? Or might an understanding of word meaning be scaffolded by attention to grammatical information, serving to reﬁne the way words represent objects, actions, and events? An abundant body of literature can be marshaled to support each of these possibilities. Perhaps, as King Solomon might have argued, there is no need to choose one over the other. Since the 1990s, there is emerging consensus that the process of word learning will not be best described through deference to one approach or the other. Rather, it will require a comprehensive theory that unites perceptual, social, and linguistic information in the service of word learning. Bloom (1993) wrote, ‘‘y cognitive developments bring the infant to the threshold of language only in conjunction with other developments in expression and social connectedness’’ (p. 52). Woodward and Markman (1998) echoed, ‘‘y word learning depends on an ability to recruit and integrate information from a range of sources’’ (p. 371). Finally, Baldwin and Tomasello (1998) suggested that word learning ‘‘y requires an explanation encompassing both its social and cognitive roots’’ (p. 19). In this chapter we take these charges seriously and propose an integrated model of word learning that is comprehensive, developmental, and empirically testable across time: the Emergentist Coalition Model (ECM, Hollich, Hirsh-Pasek, & Golinkoff, 2000). Our argument unfolds in three parts: First, we present the tenets of the ECM model (e.g., Golinkoff, Hirsh-Pasek, & Hollich, 1999; Hirsh-Pasek, Golinkoff, & Hollich, 2000; Hollich et al., 2000a, 2000b), a hybrid

King Solomon’s Take on Word Learning

3

developmental model describing the factors involved in lexical acquisition and how the importance of those factors might change over time. Second, we present a selection of research conducted in our laboratories that both validate the ECM as well as expand it. We describe studies that explore the learning of object names (nouns) and action names (verbs) as well as studies demonstrating how the hybrid model can be applied to language development in children with autism (Parish-Morris et al., 2007). Finally, we discuss the implications of a radical middle approach for both theory and application.

II. The Emergentist Coalition Model (ECM) A. A SHORT HISTORY The ECM was inspired by three strands of research in the ﬁeld of language development. First, the competition model proposed by Bates and MacWhinney (1982; MacWhinney, Pleh, & Bates, 1985) suggested that children develop language syntax by attending to both grammatical and semantic cues in the input. This groundbreaking work not only enabled researchers to predict how multiple cues might work together to explain language at any given point in time, but also sparked cross-linguistic studies and the examination of individual differences in both typical and atypical children over time (Bates, Bretherton, & Snyder, 1988). In many ways, Bates and MacWhinney were pioneers of systems-based models for language. Second, the ECM was inspired by new methods that allowed us to examine language from the perspective of comprehension rather than production. As Hirsh-Pasek and Golinkoff (1996) argued, comprehension is a window into children’s knowledge of grammar that permits the separate manipulation of children’s differential reliance on prosodic, linguistic, and social features of the input. Finally, the ECM was inﬂuenced by work from Bloom (1970, 1993) who reinforced the need to look at a ‘‘whole active child’’ who draws together information from multiple sources as she constructs her grammar. By 1996, these inﬂuences had coalesced in a theory—forecasting the direction we would take in the study of word learning. We suggested that distinct aspects of language (phonological, semantic, syntactic) served as systems of developing knowledge that were mutually informing and always available, but that were weighted differently across developmental time. Such a vision provided us with a non-linear framework for development (Hirsh-Pasek, Tucker, & Golinkoff, 1996b, p. 464). That is, while a mathematical model of word acquisition might make the process look like a

4

Kathy Hirsh-Pasek and Roberta Michnick Golinkoff

linear accretion of vocabulary (McMurray, 2007), such a model would oversimplify the many factors that jointly contribute to children’s word growth. These factors might prove multiplicative and interactive rather than additive.

B. CHARTING THE LANDSCAPE IN WORD LEARNING: A CARICATURE The ECM (Hollich et al., 2000a, 2000b) offers a model with testable predictions that combines contributions from three competing and leading theoretical approaches: the perceptual, pragmatic, and constraints theories. The ECM is not an either/or theory but rather a when and how account of vocabulary growth, positing that children use different strategies for word learning at different points in time. It addresses two key questions that continue to dominate the study of word acquisition: (1) How do children break the language barrier with their ﬁrst words, and (2) How do they become master word learners within a year’s time? Several theoretical answers to these questions emerged in the 1980s and 1990s. The constraints/principles theories emphasized the importance of cognitive heuristics in word acquisition (e.g., Golinkoff, Mervis, & HirshPasek, 1994; Markman, 1989; Merriman & Bowman, 1989; Waxman & Kosowski, 1990). Children begin the task of word learning with a set of biases that assist them in linking words to objects, actions, and events. For example, as early as 12 months, children take a novel name as referring to the whole object rather than to its parts or properties (Hollich, Golinkoff, & Hirsh-Pasek, 2007). Using a dichotomy created by Hirsh-Pasek and Golinkoff (1996) to characterize theories of language acquisition, word learning is thus guided from the ‘‘inside-out’’ (what the child brings to the task) with domain-speciﬁc principles for word learning rather than from the ‘‘outside-in’’ (from environmental guidance) with domain-general principles of learning. By way of example, under a constraints theory, children might come to the task of word learning with a principle of ‘‘reference’’ (Golinkoff et al., 1994) or with the inherent assumption that words symbolize, or stand for, objects, actions, or events. The social-pragmatic view, in stark contrast, highlights the role of adult– child interaction to prime word learning (e.g., Nelson, 1996). Under this scenario, word learning can be viewed from the ‘‘outside-in’’ as children mine the social environment, noting social cues like eye gaze and pointing (Carpenter, Nagell, & Tomasello, 1998). By just 14 months, toddlers begin to interpret social intent (Gergely & Csibra, 2003; Woodward, Sommerville, & Guajardo, 2001) and by just 18 months, evidence suggests that they can

King Solomon’s Take on Word Learning

5

use social intentional cues to assist in word learning (Hollich et al., 2000a, 2000b). Children have a driving need to share the contents of their mind and to understand others in ways that push them towards a socially informed strategy for word learning (Akhtar & Tomasello, 2000; Bloom, 1993, 2000; Golinkoff, 1993). Under a social-pragmatic theory, children are apprentices to master word users. Finally, a third response to the word-learning challenge is a domaingeneral associationist account that highlights the role of ‘‘dumb attentional mechanisms’’ and memory processes in word development (e.g., Plunkett, 1997; Smith, 2000; Yu & Smith, 2007). Children attach labels to the items that grab their attention and stand out from the context. Calculating the frequency of co-occurrence between words and referents, children form the associations that constitute vocabulary acquisition. Each of these potential answers to the word-learning challenge seems plausible, even in caricature as presented here. Furthermore, each of these theories note that children use multiple mechanisms for word learning; they simply favor one mechanism over the others. The fact that the character of word learning changes over the second year of life (e.g., Clark, 1983; Golinkoff et al., 1994), however, raises interesting questions with respect to the dominance of one theoretical perspective over the other. That is, the novice learner of 12–18 months of age acquires words slowly and often one at a time. The more mature learner of around 24 months, in contrast, is a master word learner acquiring up to 9 words a day (Carey, 1978). Perhaps the mechanisms used by expert word learners are not characteristic of those used by novice learners. Perhaps one learning mechanism is not paramount for all children, such that there are individual differences or multiple pathways for atypical language learners—like those with autism. The ECM is a developmental account hypothesizing that the shift in word learning seen across the ﬁrst two years can be explained as a change in the weighting of multiple factors; attentional, social, and linguistic. The ECM recasts the issue of word learning by asking which components of which theories govern word learning across development, and then combines them in a novel way to allow for testable predictions. Rather than providing snapshots of word learning, it tracks shifting strategies of word learning over time.

C. ASSUMPTIONS OF THE ECM MODEL The ECM makes three fundamental assumptions. First, children are sensitive to multiple cues in word learning: Perceptual, social, and linguistic

6

Kathy Hirsh-Pasek and Roberta Michnick Golinkoff

Fig. 1. The coalition of cues available for establishing word reference is utilized differently across developmental time. Children shift from a reliance on attentional cues (how compelling an object is—perceptual salience—and the coincident appearance of an object and a label-temporal contiguity) to a greater dependency on social and linguistic cues, like eye gaze and grammar, respectively.

(Figure 1). Second, cues for word learning change weights over time. Although a range of cues is always available, not all cues are equally utilized in the service of word learning at any given period of development. Children beginning to learn words rely on a perceptual subset of cues in the coalition, as evidenced by a tendency to label that which is bright or interesting—rather than attending to what the speaker intends to label. This child views reference as a ‘‘goes with’’ relationship. Only later in the second year do children recruit social cues like eye gaze and handling to learn words and to move towards reference as a ‘‘stands for’’ or more symbolic relationship. Third, the principles of word learning are emergent, changing over time. As in Bates and MacWhinney’s competition model (1982; MacWhinney, Pleh, & Bates, 1985), the ECM posits that development progresses as children come to learn which cues (perceptual or social, for example) are more or less reliable for mapping word to world. Thus, infants may start with an immature principle of reference relying on perceptual cues such that a word will be mapped to the most salient object from the infant’s point of view. Later, children become sensitive to speaker intent and map a word onto an object from the speaker’s point of view by noting social cues for word mapping (such as eye gaze and handling). Progress has been made using this hybrid account (e.g., Hollich et al., 2000a, 2000b). By examining infants’ shifting use of associative and social strategies across time, we offer a glimpse of evidence for one piece of the ECM.

King Solomon’s Take on Word Learning

7

III. Validating the Model The ECM makes testable predictions about how words are learned. To evaluate these hypotheses, however, we needed to go beyond existing methods to develop a method that: (a) could be used with young children; (b) made minimal response demands; and crucially, (c) allowed for the manipulation of multiple cues from the attentional, social, and linguistic realms within one paradigm. The ﬁrst method that met these criteria was the Intermodal Preferential Looking Paradigm (IPLP; Golinkoff et al., 1987; Hirsh-Pasek & Golinkoff, 1996). The infant is seated on a blindfolded parent’s lap midway and in front of two laterally spaced video monitors or a single large screen TV. Two images can be shown simultaneously or sequentially. After attracting the child’s attention towards the center of the screen, an audio speaker plays a linguistic stimulus that matches only one of the displays. For example, the display (Figure 2) might portray a boat (screen left) and a shoe (screen right) while the linguistic message is ‘‘Do you see the boat? Find the boat!’’ The total amount of time that the infant spends watching the matching versus the non-matching screen is the dependent variable. The logic of this procedure is that children look longer at the screen displaying the targeted object (the boat) than at the screen displaying the non-targeted object (the shoe). Infants attend more to the video event that matches the linguistic message than to a video event that does not match. This method offers the opportunity to display dynamic stimuli and to examine perceptual and linguistic cues placed in competition with one another. One thing the standard IPLP does not allow, however, is the manipulation of social Video Camera Computer & VCR

Video Monitor Parent & Child

Fig. 2. The intermodal preferential looking paradigm (Golinkoff et al., 1987; Hirsh-Pasek & Golinkoff, 1996).

8

Kathy Hirsh-Pasek and Roberta Michnick Golinkoff

Fig. 3. The interactive intermodal preferential looking paradigm.

cues—even television images of speakers are not the same as live speakers offering social cues about what they are naming. The Interactive Intermodal Preferential Looking Paradigm (Interactive IPLP) overcame this shortcoming. A live, 3-D version of the IPLP can be used with children from 10 to 24 months of age (Hollich et al., 2000a, 2000b). As Figure 3 shows, infants are seated on their blindfolded mother’s lap facing the experimenter and positioned midway in front of the testing apparatus. The experimenter presents real objects for the child to play with and look at—a pair of familiar toys on some trials and a pair of novel toys on others. The toys are then velcroed onto one side of a two-sided black board that can be rotated so that the toys can go in and out of view for a speciﬁed period of time. Hiding behind the board, the experimenter asks the child to look at one or the other of the toys, asking, for example, ‘‘Where’s the boat?’’ Looking time to the matching or non-matching object serves as the dependent variable. Importantly, controls are in place for side of match, salience of the test objects, and side of toys.

A. DATA FROM THE LEARNING OF OBJECT LABELS Equipped with theory and method, we evaluated the principle of reference: that words refer to objects, actions, and events. Two questions framed our investigation. First, how do infants break the word-learning barrier? The ECM predicts that children begin the word-learning process as associative learners guided by perceptual salience when making word mappings. Second, how does the word-learning process change over the

King Solomon’s Take on Word Learning

9

second year of life? The ECM suggests that children should gradually begin to switch to a more reliable, socially informed strategy for word learning that uses information about a speaker’s social intent to determine word-toworld mapping. To investigate this shift in strategy, children were taught new words in the Interactive IPLP paradigm. In one of the two conditions, the coincident condition, we labeled the novel, interesting toy that coincided with children’s preferences (e.g., a sparkling wand). In the other, conflict condition, we labeled the novel, boring toy that did not coincide with the children’s preferences (e.g., a beige plastic bottle opener). Children conﬁrmed our intuitions about whether the object was ‘‘interesting’’ or ‘‘boring’’ in a salience trial. We reasoned that learning the word in the coincident case should be easy for children because all of the ‘‘cues,’’ attentional, social, and linguistic, were aligned. In contrast, learning a novel word in the conﬂict condition should be more difﬁcult because the coalition of cues was not acting in concert (Golinkoff & Hirsh-Pasek, 2006; Hirsh-Pasek et al., 2000) and because perceptual and social cues were placed in conﬂict (Hollich et al., 2000a, 2000b). Children operating at the associative level and who fail to use social cues should simply attach the label to the object they found most interesting regardless of where the speaker was looking. Alternatively, children sensitive to social cues should learn the name for the object that the speaker labeled, even if it was boring. We began our investigation with 10-month-olds, an age at which Fenson et al. (1994) suggest that vocabulary starts to amass. Across ages, the results conﬁrmed our predictions. Ten-month-olds were indeed pure associationists, mapping a novel word onto the object that they found the most interesting, regardless of which object the speaker labeled. Indeed, 10-month-olds acted as if social cues to reference did not exist (Pruden et al., 2006). Just two months later, 12-month-olds showed an entirely different pattern. Social information was necessary, but not sufﬁcient, to ensure word learning. These children only learned the novel word when social and perceptual cues were aligned—when the speaker labeled the interesting object. They failed to learn a word when the speaker labeled the boring object. Had 12-month-olds been pure associationists, they should have mismapped like the 10-month-olds, thinking that every word referenced the interesting object! The fact that they did not do this suggests that they detected the speaker’s social cues. Nineteen-month-olds were attracted to perceptual cues but could use social information to learn the label for the boring object. Finally, by 24 months of age, children convincingly used social information, learning the name for both the interesting and boring objects (Hollich et al., 2000a, 2000b).

10

Kathy Hirsh-Pasek and Roberta Michnick Golinkoff

These results conﬁrm the predictions made by the ECM. Beginning as associationists, children gradually begin to attend to social cues, and then to recruit a speaker’s social cues to afﬁx a label to an object. Children use multiple inputs, differentially weighted, as they become word learners. These ﬁndings suggest that theoretical accounts that appear incompatible (such as the associationist and social-pragmatic accounts) actually capture the process of word learning at different points in developmental time. That is, children assume that words refer, but the process of linking word to referent changes over time.

B. DATA FROM THE LEARNING OF ACTION LABELS Our ﬁrst test of the ECM examined object word learning, arguably a unique word class that allows children to break into language (Waxman & Lidz, 2006). A complete theory of word learning, however, must be able to account for the learning of words from other classes, such as verbs. Is there a shift from reliance on perceptual to social cues when learning labels for actions? Abundant research suggests that verb learning is complicated, often lagging behind noun learning (Bornstein, Hahn, & Haynes, 2004; Gentner, 1982; Hirsh-Pasek & Golinkoff, 2006; but see Tardif, 1996). As Gentner (1982) suggests, verbs do not label actions in the same way that nouns label objects. Nouns commonly refer to objects that are perceived as distinct units. For example, a cup is perceived as distinct from the surface it rests on, can be handled, and persists over time. Verbs, however, are inherently relational. One cannot talk about an action without someone or something performing that action. Furthermore, the actions that verbs label can be conceptualized in terms of a multitude of different components, including, but not limited to, path (the trajectory of an action with respect to some reference point, e.g., approach, enter), manner (how an action is carried out, e.g., walk), or result (e.g., open, ﬁll; Talmy, 1985). Which relation or relations in an event is the verb referent? To learn verbs a child must disentangle a variety of simultaneously occurring components and must choose between a plethora of possible meanings. Are perceptual, social, and syntactic cues weighted differently for verb learning? Are these cues recruited later for verb learning than for noun learning? To test developmental predictions of the ECM for verb learning, we created a situation that was analogous to the noun-learning work of Hollich et al. (2000a, 2000b) and Pruden et al. (2006). By pitting cues against one another, we asked whether young children are biased to learn verbs based on perceptual salience or on speaker information, when speaker information includes both social intent and grammar.

King Solomon’s Take on Word Learning

11

Brandone et al. (2007) presented children with a pair of actions performed on a bright metal box. They were taught the name for one of these actions using a combination of perceptual and speaker cues. Perceptual salience was deﬁned by whether or not the action produced an interesting result (e.g., a switch made a light go on). An action that produced a result was more salient than an action that had no result (e.g., such as pressing a Morse code key that did not produce a sound). Each child was exposed to a pair of actions. To create a conﬂict between perceptual and speaker information, half of the children received a label for the action with a result (the coincident condition), whereas the other half received a label for the action without a result (the conﬂict condition). On the one hand, if perceptual salience is central for verb mapping, it should be easy for children to learn the name of the action with a result (because perceptual cues coincide with speaker information) and more difﬁcult for them to learn the name of the action without a result (because perceptual cues and speaker information conﬂict). On the other hand, if children rely on social intent and language cues for verb learning, they should be able to learn the name of the action regardless of whether it produces a result. Across three studies, our predictions were borne out. Twenty-twomonth-olds attached a verb to one of two actions only when perceptual cues (presence of a result) coincided with speaker cues. That is, they learned the name of the action that produced the result. When these cues conﬂicted (Experiment 1), as when pressing the Morse code key produced no result, children could not learn the name for the action. Nor could children learn the name of the action when both possible referent actions produced a result (Experiment 2). In other words, verb learning only occurred when there was a single salient action that the speaker labeled. When would children be able to override perceptual information to learn the name of an action that was not perceptually salient (Experiment 3)? By 34 months they could do just that. These results demonstrate an early reliance on perceptual information and the emergent use of speaker information for verb mapping. Although most work within the ECM has examined how children attach novel labels to object referents (but see Poulin-Dubois & Forbes, 2006), these ﬁndings extend the story to the domain of verbs. Interestingly, children who are old enough to demonstrate mastery of noun learning nonetheless experience difﬁculty with verb learning. Although the pattern of results in our verb-learning experiments neatly parallel those of the noun-learning experiments (Hollich et al., 2000a, 2000b; Pruden et al., 2006), equivalent verb-learning abilities are not observed until over one year later. These results lend further support to the ECM by extending the model to include the learning of verbs. Yet, the results also force us to ponder why

12

Kathy Hirsh-Pasek and Roberta Michnick Golinkoff

it takes so much longer to attach word to world in the context of verb learning—a question that we now address.

C. STEPPING BACK: WHY VERBS ARE HARD If, as we argued earlier, verbs are more difﬁcult to learn than nouns (Bornstein et al., 2004; Gentner, 1982), then why are some verbs learned quite early (Naigles & Hoff, 2006; Tardif, 1996)? We have called this phenomenon the ‘‘verb learning paradox’’ (Maguire et al., 2006). Why is this so? Unpacking this apparent inconsistency led us to a series of studies that lent more support to the ECM model. To preview the ﬁndings, we learned that in the ﬁrst year of life, children have at least some of the conceptual prerequisites for verb learning. Yet, as has been suggested before (Gentner, 1982; Gillette et al., 1999), inherent ambiguity in word-toworld mapping for verbs makes the task of reference and hence verb learning more difﬁcult on average than noun learning. To solve the reference problem, children ﬁrst appeal to perceptual salience and only later utilize syntactic cues and cues to speaker intent. Many have now suggested that mastery of the verb system requires children to conquer several preliminary tasks (Golinkoff & Hirsh-Pasek, 2006; Golinkoff et al., 2002). First, they must attend to and discriminate between actions in their environment. Second, infants must be able to discriminate and form categories of actions without language. The action of jumping, for example, refers to a decontextualized category of jumping motions that include different kinds of jumps made by the same actor (e.g., Elmo jumping off tables and chairs), and the same action performed by different actors (e.g., Elmo or Lala jumping off the chair). Third, children must be able to map words to actions and action categories in language-speciﬁc ways. We explored each of these skills across a number of studies.

D. INFANTS CAN FIND ACTIONS AND ACTION COMPONENTS IN EVENTS Verbs label a subset of the many, often simultaneously occurring semantic components that exist within motion events. As noted earlier, these components include motion (the general fact that motion is taking place), figure (the prominent entity in the event), manner (the way in which the action/motion is carried out), path (the trajectory of the ﬁgure with respect to some reference point), ground (the reference point for the event’s

King Solomon’s Take on Word Learning

13

path), and cause (the cause of the ﬁgure’s motion), among others (Talmy, 1985). For example, when a man passes under the banner at the end of the New York Marathon, the figure is the man; the manner by which he passed under is running; the path is under the banner; and the ground is the banner itself (Pulverman et al., 2006). All these elements of motion events simultaneously meet the eye. Yet, only some of them will be relevant to learning any particular verb. The story gets even more complicated because what is relevant to verb mapping or ‘‘packaging’’ (Tomasello, 1992) differs across languages. For example, in English, one can say, ‘‘The man limped down the stairs.’’ In this case, the verb encodes the manner of limping and the path appears in a prepositional phrase, down the stairs. In Spanish, the sentence would be, ‘‘El hombre bajo´ las escaleras cojeando,’’ and would be translated as, ‘‘The man went down/descended the stairs limping.’’ In Spanish, the verb encodes the man’s path, and the manner (limping) is an optional addition. Although we know that infants are keenly aware of movement and use movement to individuate objects (e.g., Mandler, 1991) and actions (Sharon & Wynn, 1998; Wynn, 1996), surprisingly little research examines whether infants have the conceptual prerequisites to learn a verb (but see Casasola, Bhagwat, & Ferguson, 2006; Casasola & Cohen, 2002; Hespos & Spelke, 2004; Mandler, 2004). We started our investigations by focusing on two dynamic components of events—path and manner—that appear in most languages and that are packaged differently across languages. Can infants attend to path and manner? Pulverman et al. (2007) asked this question with 14–17-month-olds in a habituation task. Infants viewed silent, computer-animated motion events involving a moving starﬁsh character (the ﬁgure) and a stationary ball (the ground). The starﬁsh performed an action with both a manner ( jumping jacks, spinning, or bending at the ‘‘waist’’) and a path (over the ball, under the ball, or vertically past the ball; Figure 4). After infants habituated to a

Over

Under

Past

Figure 1. Examples of paths and manners to be used in stimuli. Though illustrated as a series of static postures, the manners will be performed as continuous motions.

Fig. 4. The paths and manners used by Pulverman and Golinkoff (2004), Pulverman et al. (2006), and Pruden et al. (2006).

14

Kathy Hirsh-Pasek and Roberta Michnick Golinkoff

single event (e.g., jumping jacks over the ball), they were tested on four different types of events: (1) a control event identical to the habituation event (e.g., jumping jacks over the ball); (2) an event with the same manner as the habituation event, but a different path (e.g., jumping jacks under the ball); (3) an event with the same path as the habituation event, but a different manner (e.g., spinning over the ball); and (4) an event in which both the manner and path differed from those in all other events (e.g., waist bends past the ball vertically). The results suggested that 14–17-month-old infants discriminated between paths and between manners. They readily dishabituated to the path change, the manner change, and the both change events. These initial results motivated several additional studies, all of which afﬁrm the main ﬁndings. For example, Pulverman and Golinkoff (2003) found that even 7-month-olds successfully dishabituated to changes of manner, changes of path, and changes of both manner and path. Even before word learning begins, infants notice differences between events that could potentially distinguish one verb from another. For older children, the data also suggest that manner and path are treated as independent elements within a motion event (Pulverman et al., 2006). One additional ﬁnding of interest emerged when Pulverman and Golinkoff (2004) examined how children’s attention to these independent elements in events related to maternal report of their vocabulary. Englishreared children with greater receptive vocabularies were more sensitive to manner changes than their small vocabulary peers. Data from a replication with infants learning Spanish demonstrated that Spanishreared infants who prefer manner changes have relatively lower reported vocabularies than their peers (Pulverman et al., 2005, 2007). Thus, attention to manner has different implications depending on the language to be learned. In English, a child who notices manner is likely to have some advantage in learning verbs because English verbs are more likely to label manners. In Spanish, that same child would be at a disadvantage because Spanish verbs are more likely to encode paths. Note, both languages code manners and paths, but what gets ‘‘packaged’’ within the verb itself differs. Taken together, the data suggest that both English-reared and Spanishreared infants can discriminate the components of events that will be codiﬁed in their verbs. This is among the ﬁrst cross-linguistic demonstrations that infants can ﬁnd components of events that are relevant to language learning. Furthermore, differential attention to these dynamic components of events appears to inﬂuence infants’ lexical acquisition. This is but a ﬁrst step in understanding how infants process dynamic events for language or—as Slobin (2001) called it—master thinking for speaking.

King Solomon’s Take on Word Learning

15

E. INFANTS CAN ALSO FORM CATEGORIES OF ACTIONS Infants are not only capable of noticing changes in both path and manner (Pulverman et al., 2007) and treating those changes as independent elements of events by 14–17 months, but are also capable of categorizing these components prior to acquiring extensive language. Using the same animated and tightly controlled stimuli as Pulverman et al. (2007; see Figure 4), Pruden et al. (2004) asked whether infants could detect an invariant manner across varying paths (e.g., spinning around, spinning past, etc.) and an invariant path among varying manners (e.g., spinning past a ball, twisting past a ball). Three age groups, 7–9-, 10–12-, and 13–15-month-olds were tested using the IPLP (Hirsh-Pasek & Golinkoff, 1996) without language. That is, events were presented individually or simultaneously without accompanying speech. During familiarization, infants viewed an animated starﬁsh performing the same path across four distinct manners. At test, infants were shown two events simultaneously: One depicting a novel exemplar of the familiar category and the other depicting a novel exemplar of a novel category. For example, infants who were familiarized with the path ‘‘over the ball’’ saw the event clips ‘‘touching toe under the ball’’ (i.e., novel manner and novel path) and ‘‘touching toe over the ball’’ (i.e., novel manner and same path). Seven- to nine-month-olds did not have a signiﬁcant preference for either test event. In contrast, older infants showed a signiﬁcant preference for the familiar event during the test phase. Thus, infants as young as 10 months can create a simple category of path. The next study investigated whether infants at the same ages could abstract an invariant manner across multiple exemplars of path. Here, only the 13–15month-olds showed a signiﬁcant preference for the novel test event. Testing the ability to abstract an invariant path or manner only begins to address the question of categorization, as categorization is much richer than simply abstracting invariants from scenes. However, these preliminary studies provide an important advance in our understanding of the categorization of dynamic action and are key to understanding infants’ prerequisite skills for verb learning. They also allow us to examine the parameters that enhance or hinder action category formation (Pruden, Hirsh-Pasek, & Golinkoff, 2008; Pulverman, 2006). For example, results from Pruden and Hirsh-Pasek (2006) showed that adding a label (the nonsense word ‘‘javing’’) to the same task assisted 7–9-month-olds who could not form a category to do so. In another study, when the stimuli during familiarization were shown side-by-side, enabling comparison, 7–9-month-olds could form a category of path. Could children form categories of events in more complicated ‘‘real life’’ scenes? Song et al. (2006) examined infants’ ability to categorize manner using four distinct human actors performing the same action in four distinct

16

Kathy Hirsh-Pasek and Roberta Michnick Golinkoff

ways in outdoor settings. Ten- to twelve-month-olds indicated that they could distinguish between old and new actions, thereby evincing category formation. Results from studies on spatial constructs such as containment and support as well as on manner and path parallel those found in our lab (Casasola & Cohen, 2002; Casasola et al., 2006; Choi, 2006). In sum, it appears that children can (1) isolate event properties that will be encoded in verbs such as changes in manner and path (e.g., Pulverman et al., 2007); (2) form invariant categories of manner and path (e.g., Pruden & Hirsh-Pasek, 2006); and (3) form a category of manner across much variation with human actors (e.g., Song et al., 2006). Although research is only starting to unveil infants’ processing of event structure, evidence suggests that trouble with conceptual underpinnings does not explain children’s difﬁculty in learning verbs.

F. EVEN TODDLERS HAVE TROUBLE MAPPING WORDS ONTO ACTIONS: THE POWER OF PERCEPTUAL SALIENCE If children have the requisite conceptual structure to learn verbs, then the root of the verb-learning problem must come from mapping words to world. Several lines of evidence (including our own) converge to suggest that verb mapping is particularly difﬁcult—even for adults. One compelling study comes from Gleitman and her colleagues’ Human Simulation paradigm (Gillette et al., 1999; Snedeker & Gleitman, 2004). In their studies, adults viewed a series of silent video clips depicting a mother and child playing. The participants’ task was to guess what word the speaker might have used in place of a beep. Adults correctly guessed the missing nouns in 45% of the cases; correct guesses for verbs were a paltry 15%. In fact, if responses for mental verbs were considered alone, the proportion of correct verb ‘‘guesses’’ dropped to zero! Adults undoubtedly have all the requisite concepts on which verbs rest. Nonetheless, they have difﬁculty inferring which verbs are being spoken in a natural interaction between mother and child. Mapping from action or mental state to word is considerably more challenging than mapping from object to word. Verb mapping even proves difﬁcult in languages that are reputed to be ‘‘verb friendly,’’ where verbs may appear in isolation or at the end of sentences as in Japanese or Korean. For example, Imai et al. (2006) examined fast mapping and extension of nouns and verbs in English, Chinese, and Japanese in a laboratory setting. Whereas all children performed well with nouns at the age of 3, readily learning and extending the name of a novel object, they were at chance performance for verb learning and extension. Children did not readily map and extend a new verb

King Solomon’s Take on Word Learning

17

until the age of 5. Even then, they mapped verbs to correct actions only when they were provided with the appropriate grammatical supports afforded by their language. English-speaking children, for example, solved the task when the verb was introduced with full syntax (as in ‘‘She blorked it!’’) and not otherwise. Japanese children solved the mapping problem when the word was offered alone (consistent with their language’s frequent dropping of subject and objects). When they were offered a full sentence, they failed. Finally, Chinese children learn a language in which the verb structure is ambiguous with the noun structure. The only way to map a word onto a verb is through extra-linguistic cues. Thus, the Chinese children did not successfully map actions to words until they were 8 years old! What might help children link words and meanings? Until they are savvy enough to utilize social intent and grammatical cues in the service of word meaning, the ECM suggests that perceptual salience offers a natural starting point for word reference. Evidence that children rely heavily on perception in early verb learning comes from a number of sources. Recall the study by Brandone et al. (2007) suggesting that social cues do not ‘‘win out’’ over perceptual cues until children are 34 months of age. That the appearance of an action is crucially important for early verb meaning is further supported by Behrend (1990) and Forbes and Farrar (1995). First verbs seem to map onto perceptually salient actions and are only narrowly extended (e.g., Tomasello, 1995). Research by Snedeker, Li, and Yuan (2003) provides further evidence that the appearance of an action is partially determinative of ease with which the name for that action is learned. Snedeker et al. showed Mandarin and English-speaking adults videos of Mandarin mothers and children interacting in the Human Simulation paradigm. Mandarin verbs tend to be used in a way that is more contextually narrow (e.g., not simply carry but carry on the back or carry with arms at the side) and hence more perceptually distinct. In this study, both adult groups were better at guessing the Mandarin than the English verbs! If the difference between nouns and verbs diminished when Snedeker et al.’s participants guessed words from Mandarin scenes, we can entertain the idea that form class per se (e.g., noun and verb) is not determinative of when a word is learned. Rather, perceptual salience might rule. Based on ﬁndings supporting a strong role for perception in early word learning, we hypothesized that words labeling more perceptually salient objects and actions would be learned faster than those with lower salience. ‘‘Imageability,’’ or the ease with which a word evokes a mental image, seems to be a reasonable measure of perceptual salience. Words that are relatively more imageable (e.g., cat and run versus idea and think) tend to label referents that are easily detectable and individuable (Paivio, Yuille, & Madigan, 1968).

18

Kathy Hirsh-Pasek and Roberta Michnick Golinkoff

Indeed, in Gleitman and colleague’s work, concreteness, a correlate of imageability, served as a kind of proxy for perceptual salience. Ratings of concreteness were highly correlated with the number of participants who guessed a word correctly (Gillette et al., 1999; Snedeker & Gleitman, 2004). McDonough et al. (in preparation) and Ma, Golinkoff, Hirsh-Pasek, McDonough, and Tardif (under review) expanded on prior work by offering a direct test of the imageability hypothesis that words with higher imageability ratings would be learned earlier than words without imageability ratings. Imageability ratings in English were extracted from published rating sets and correlated with the age of acquisition of nouns and verbs on the MacArthur Communicative Development Inventory (CDI; Fenson et al., 1994). McDonough et al. found that early nouns were signiﬁcantly more imageable than early verbs in English. Thus, the noun advantage may not be a function of form class per se but of the differences in imageability in words from these classes. Furthermore, the earliest acquired nouns and earliest learned verbs were indeed those that were most imageable. Ma et al. (under review) investigated the scope of this imageability hypothesis by testing it in a language very different from English and with a different pattern of early acquisition than English, viz., Chinese. Their research posed two questions: (1) Can imageability ratings predict when a word will be learned in Chinese? and (2) Why do early Chinese vocabularies contain a higher proportion of verbs than early English vocabularies (Tardif, 2006)? Ma et al. obtained imageability ratings in China from native speakers. Importantly, there was no difference in overall imageability ratings between English and Chinese participants on a subset of 31 verbs with close meanings across languages. Ma et al. found that imageability was negatively and strongly correlated with age of acquisition on the Chinese version of the CDI (Tardif et al., in preparation). This study was the ﬁrst to show that imageability is a reliable predictor of age of acquisition across languages. Chinese and English children’s nouns do not differ in imageability ratings. Chinese children’s early verbs, however, were more imageable than English children’s ﬁrst verbs. The ﬁnding that early Chinese verbs have higher imageability ratings than early English verbs is in accord with the observation that Chinese-speaking children learn more verbs and learn them earlier than their English-speaking counterparts (Tardif, 2006). Although nouns are easier to learn than verbs because they are, on average, more imageable than verbs, some verbs are highly imageable and hence are learned early. Why are the early Chinese verbs more imageable and learned earlier than English verbs? One possibility is that imageability is correlated with the individuability of the action labeled by the verb. This is consistent with the ﬁnding that the ﬁrst verbs children produce or understand usually describe actions or events that encode physical motion rather than the invisible

King Solomon’s Take on Word Learning

19

mental status of an agent (Bloom, Lightbown, & Hood, 1975; Snedeker & Gleitman, 2004). High imageability might also make it easier for children to extract the invariant action across multiple instances of an action or the ‘‘verbal essence’’ (Golinkoff et al., 2002). For example, to learn the highly imageable verb drink, one has to abstract the common relation of bringing a vessel to one’s lips from a range of drinking actions that include different agents, different drinks, different locations, etc. But to learn the much less imageable verb think, extracting the invariance would be more difﬁcult. That is, high imageability may be correlated with context speciﬁcity. For children faced with the task of abstracting a common relation, a verb that encodes a more speciﬁc set of actions should be easier to learn than a verb that names a broader set of actions (see also Tardif, 2006). Chinese, for instance, has 26 verbs for carry, each encoding a different— more context speciﬁc—way of carrying (e.g., on the back versus with the hand extended downward). These verbs received imageability ratings from Chinese-speaking adults that ranged from 5.60 to 6.27 on a 7-point scale and were acquired at a mean age of 17.25 months. In English, there is only one verb for carry, regardless of the particular way in which the carrying is done. It received an adult imageability rating of 3.81 (Masterson & Druks, 1998; McDonough et al., in preparation) and is acquired at 23 months. Imageability is, in part, a measure of the perceptual availability of a concept. It is not the only factor that explains the verb learning paradox (see results on word frequency combined with imageability by Ma, Golinkoff, & Hirsh-Pasek, 2007). It does, however, offer a toehold on a problem that has plagued researchers for the last 25 years. It also feeds into a more comprehensive conceptualization of early word learning. According to the ECM, the words children initially learn will be perceptually available and contextually bound. This will be the case irrespective of syntactic word class because the ECM is blind to word class and operates as a general framework for explaining all vocabulary acquisition. Note that we are not claiming that linguistic form class does not exist for the young child. For the development of early vocabulary, however, we are suggesting that it may not be syntactic differences between nouns and verbs that cause nouns to be learned earlier (Maguire et al., 2006). Rather, children’s ﬁrst words (nouns or verbs) might be constructed from a perceptual base.

G. BEYOND PERCEPTUAL CUES FOR VERB MAPPING Up until now we have emphasized the ﬁrst phase of the ECM—children’s reliance on perceptual information for word learning. But eventually,

20

Kathy Hirsh-Pasek and Roberta Michnick Golinkoff

children learn words in circumstances in which perceptual cues are not available. Words like the noun ‘‘idea’’ and the verb ‘‘think’’ have weak perceptual links and are also weak in the imagery they generate. Thus, to learn any word—noun or verb—children must coordinate perceptual, social, and linguistic inputs to uncover more precise word meanings. Gleitman (1990) long argued that after children know some nouns and can use the arrangement of the argument structure around the verb, they can avail themselves of a process she calls ‘‘syntactic bootstrapping.’’ Syntactic bootstrapping rests on the assumption that reliable rules link argument structures to meanings. For example, causatives in English typically (although not invariably) occur with an argument on either side of the verb. Thus, encountering ‘‘John blorked Mary’’ one can assume that this means John might have done something to Mary. By now a number of studies have shown that at around 2 years of age children become able to exploit these syntax-to-semantics regularities to learn more about verb meaning (e.g., Fisher & Song, 2006; Hirsh-Pasek, Golinkoff, & Naigles, 1996a; Imai et al., 2006; Naigles, 1990). Additionally, many verbs turn on intention—either of the speaker or of the actor. Subtle social cues discriminate between, for example, the verbs ‘‘pour’’ and ‘‘spill’’ or ‘‘kill’’ and ‘‘die.’’ These distinctions are difﬁcult for children to learn (Bowerman, 1974) because they are only distinguished by whether the action was done on purpose or accidentally. Poulin-Dubois and Forbes (2002, 2006) demonstrated that before age 21, toddlers do not distinguish between perceptually similar but pragmatically distinct verbs. As Poulin-Dubois and Forbes (2006) wrote, ‘‘It would appear that between 21 and 28 months of age, children’s verb learning strategy transitions from a reliance on the overall appearance of verb referent events to a reliance on behavioral and linguistic cues about others’ behavioral and verbal semantic intentions’’ (p. 277). Clearly, attention to the grammatical and social/ pragmatic cues of input will help children reﬁne word meaning and learn more about the regularities in word-to-world mapping. Furthermore, attention to these kinds of input cues might prove particularly important for discerning verb meaning.

H. THE ECM AND VERB LEARNING The process of mapping words onto actions and events contains a number of stumbling blocks for young language learners. A burgeoning literature (as well as our own studies) conﬁrms Gentner’s (1982) suggestion that it is ‘‘not perceiving relations, but packaging and lexicalizing them that is difﬁcult.’’ For words that are more relational

King Solomon’s Take on Word Learning

21

(Gentner & Boroditsky, 2001) and more referentially ambiguous (e.g., think versus car), the ability to link words to world requires stronger support from grammar and from cues to social intent. That is, verbs in general, but not invariably, will be more difﬁcult to learn than nouns and hence will require more scaffolding from grammatical and social cues. The ECM suggests that when children are learning their ﬁrst words, be they nouns or verbs, they rely on perceptual salience as a cue to reference. Reference at this point is more of a ‘‘goes with’’ than a ‘‘stands for’’ relationship. The research we reviewed strongly endorses this claim, revealing that across word categories, those words that are more perceptually salient and contextually bound are learned ﬁrst. This is true when learning nouns (object words) (Hollich et al., 2000a, 2000b), appears in a parallel task with verbs (Brandone et al., 2007), and is evident in the imageability studies of both Chinese and English nouns and verbs (Ma et al., 2007; McDonough et al., in preparation). The ECM not only predicts that mapping will begin with deference to perceptual cues, but also that perceptual information will give way to children’s reliance on social intent and grammar. This must be true for two reasons. First, one cannot get through life learning words for only attractive and salient objects and events. Second, both nouns and verbs encode abstractions and relations (e.g., idea and promise). Thus, for word learning, the ECM predicts that more concrete words—generally nouns—will be learned ﬁrst and hence nouns will often appear before verbs. The ECM not only offers an explanation of the verb-learning paradox, but also provides some account for why the pace of word learning seems to increase during the second year. When children use social and grammatical input in the service of word learning they become not only more reliable word learners, but also faster word learners.

IV. Implications of the ECM for Language Disorder: The Case of Autism The ECM provides a systems-based model for word learning that can also serve as a framework for examining the nature of language disorders. Because the ECM is a hybrid model of word learning, it implicitly states that words can be acquired through a number of pathways. If perceptual cues, social cues, and linguistic cues are all available for word learning, then it may be possible to (a) determine which set of cues is not being utilized and (b) supplement the input for word learning by strengthening other cues that autism children can recruit (Hennon, 2000). Using this logic, we

22

Kathy Hirsh-Pasek and Roberta Michnick Golinkoff

reasoned that children with autism, who tend to have problems noting social intent but may have average perceptual abilities, should be able to learn words that have perceptually salient referents. Thus, during the ﬁrst year of life, when perceptual learning dominates, children with autism and typically developing children should follow the same word acquisition trajectories. After that time, however, when typical children move to a greater reliance on social cues, their word-learning trajectories should begin to diverge from those of autistic children. Furthermore, we hypothesized that within a population of children with autism, those who are better at mining social intentional cues should be superior word learners than those who are worse at using social cues to infer another individual’s intent. Guided by these testable predictions from the ECM, Parish-Morris et al. (2007) explored word learning in children diagnosed with autism (AD). Children with autism were matched to two groups of typically developing children on (1) language outcome (PPVT) and (2) non-verbal intelligence. Four experiments were conducted using tasks that tapped children’s ability to (1) attend to social cues outside the word-learning situation (2 separate tasks), (2) discern the intention of another in a non-verbal enactment task, and (3) interpret intention in a word-learning task. Results demonstrated considerable variability in the performance of all three groups, but particularly in the AD group. The ﬁndings suggest that all children pay attention to social cues to some degree. Social attentional cues are manifest when a child follows another person’s line of regard and uses social information (e.g., eye gaze, pointing, handling) as a ‘‘perceptual ﬂashlight’’ to focus on an object or event. AD children could use social attentional information to not only follow a speaker’s line of regard towards an object, but also to learn the word for a perceptually interesting object indicated by the speaker. Consistent with the literature, these same AD children had more trouble learning words for objects that were not interesting to them even when the speaker indicated that object through eye gaze, pointing, or handling. That is, social intentional cues were more opaque for the AD children. Only typically developing children consistently utilized a speaker’s social intentions to label a ‘‘boring’’ object. Typically developing children were also much more likely to complete the experimenter’s intended but unfulﬁlled action (e.g., ﬁnding a named but never seen object in a search task). For the AD group, performance on these social intentional tasks was the most powerful predictor of vocabulary, accounting for a very large percentage of the variance (68%). Taken together, these ﬁndings suggest that children can learn words relying on perception cues alone, and that children with autism can learn new words when a referent is highly salient. Social cues are not necessary for word learning—although using them to infer a speaker’s intent clearly

King Solomon’s Take on Word Learning

23

facilitates word learning. AD children who demonstrated access to a speaker’s social intent also had bigger vocabularies.

V. Implications of a ‘‘Radical Middle’’ Approach: Three Take-Home Messages The research from noun and verb learning as well as from word acquisition in children with autism suggests that the ECM is a valid model generating clear predictions about how children learn their ﬁrst words and why the character of word learning changes over time (Golinkoff & HirshPasek, 2006). As such, the model is quintessentially developmental. The ECM embraces dominant theories of word learning and simultaneously integrates these theories into a broader framework. In so doing, the ECM stands at the radical middle of language theories not by default, but rather, by design. Although the ECM is a start, it is not without its limits. For one, it can be argued that it is but a description of how word learning progresses rather than an explanation of mechanisms. What makes something perceptually salient? What accounts for the shift from perceptual governance to social mediated word learning? We are currently working on computer models to address some of these questions. At present, however, we can only speculate about how a child comes to rely more on certain cues over others during development. To date all of our work has been cross-sectional rather than longitudinal. Such an approach falls short of considering change over time and short of providing an explanation for the source of individual differences in word learning. Here the model has enormous—though unrealized—potential. The alternative pathways for word learning that can be used by different children across different contexts and ages will only be revealed after we understand children’s individual proﬁles. The model is also limited in that we have only tested some of the very preliminary predictions made by the ECM about when children move from perceptual to socially informed strategies of word learning. Much more needs to be done to ﬂesh out the model by documenting the change to a reliance on grammatical information. Furthermore, factors like frequency of input, phonological form, etc., are also important for word learning and have yet to be built into the model. Even with the current limitations, however, the ECM and the research accumulated to date leave us with several take-home messages. First, complex problems like language learning are best examined by looking at multiple factors. In the interest of parsimony, researchers often examine one

24

Kathy Hirsh-Pasek and Roberta Michnick Golinkoff

contributor to development to the exclusion of others, paying only lip service to other equally important and interacting motors for growth (Hollich et al., 2000a, 2000b). The science that is reﬂected in Bates and MacWhinney’s (1982) competition model and Bronfenbrenner’s (1993) systems theory is taking hold across much of developmental psychology. Language development—and word learning—is likely inﬂuenced not only by the perceptual biases children bring to the task, but also by the social and linguistic information that children construct and construe across various contexts. Second, when a developmental approach is taken seriously, it offers a compelling way to unite theories that may appear incompatible. The associative word-learning theories of Smith et al. (2000), the more socialpragmatic theories of Akhtar and Tomasello (2000), and Bloom and Tinker (2001), each have theoretical merit and empirical support. Viewed across time, the mechanisms proposed by these approaches can be seen as laying on a developmental continuum wherein strategies become more reﬁned and more reliable as the child moves from novice to expert. In contrast, a myopic approach on a single factor does not promote examination of the possible interactions among cues. Had the experiments by Hollich et al. (2000a, 2000b) and Brandone et al. (2007) been motivated by mutually exclusive hypotheses, there would have been no opportunity to observe the trajectory of word learning as children begin to coordinate the many cues at their disposal. Thus, to understand the word-learning process, the key is to study the interactions among multiple cues. Third, a developmental model emphasizing multiple factors holds the promise of revealing individual differences and pathways of development. Our tradition (though see Bates et al., 1988) is to look at small sample sizes whose data is treated in the aggregate. Such an approach might mask different trajectories of development. Furthermore, a move towards an individual difference approach might help us re-envision how we think about typical versus atypical development. Perhaps atypical development manifests itself as an extreme version of individual differences in typical development. As such, after we determine how multiple factors interact in word learning we can better understand what components are heightened or depressed in certain populations of children and what interventions might prove most useful for optimizing language growth. These take-home messages conﬁrm King Solomon’s approach. Rather than sacriﬁcing the baby for one theoretical mother, the true story of word learning requires a powerful theoretical stance that integrates elements of many theories. Word learning is accomplished by a whole baby who uses perceptual, social, and linguistic information with differential weighting over developmental time. Herein lies the beneﬁt of a radical middle theory like the ECM. The question of how we map words to world and how these

King Solomon’s Take on Word Learning

25

words become imbued with meaning is a complex one that deserves a complex answer. The hybrid model that we call the ECM offers one small step towards embracing that complexity.

REFERENCES Akhtar, N., & Tomasello, M. (2000). The social nature of words and word learning. In R. M. Golinkoff, K. Hirsh-Pasek, L. Bloom, L. Smith, A. Woodward, N. Akhtar, M. Tomasello, & G. Hollich (Eds.), Becoming a word learner: A debate on lexical acquisition (pp. 3–17). New York, NY: Oxford University Press. Baldwin, D. A., & Tomasello, M. (1998). Word learning: A window on early pragmatic understanding. In E. V. Clark (Ed.), Proceedings of the Stanford Child Language Research Forum 3–23. Stanford, CA: Center for the Study of Language and Information. Bates, E., & MacWhinney, B. (1982). Functionalist approaches to grammar. In L. Gleitman & E. Wanner (Eds.), Language acquisition: The state of the art (pp. N173–N218). New York: Cambridge University Press. Bates, E., Bretherton, L., & Snyder, L. (1988). From first words to grammar: Individual differences and dissociable mechanisms. New York: Cambridge University Press. Behrend, D. A. (1990). The development of verb concepts: Children’s use of verbs to label familiar and novel events. Child Development, 61, 681–696. Bloom, L. (1970). Language development: Form and function in emerging grammars. Cambridge, MA: The MIT Press. Bloom, L. (1993). The transition from infancy to language: Acquiring the power of expression. New York: Cambridge University Press. Bloom, P. (2000). How children learn the meaning of words. Cambridge, MA: MIT Press. Bloom, L., Lightbown, P., & Hood, L. (1975). Structure and variation in child language. Monographs of the Society for Research in Child Development, 40, (2, Serial No. 160). Bloom, L., & Tinker, E. (2001). The intentionality model and language acquisition: Engagement, effort, and the essential tension in development. Monographs of the Society for Research in Child Development, 66 (4, Serial No. 267). Bornstein, M. H., Hahn, C. S., & Haynes, M. O. (2004). Speciﬁc and general language performance across early childhood: Stability and gender considerations. First Language, 24, 267–304. Bowerman, M. F. (1974). Discussion summary: Development of concepts underlying language. In R. L. Schiefelbusch & L. L. Lloyd (Eds.), Language perspectives: Acquisition, retardation, and intervention (pp. 191–209). Baltimore: University Park Press. Brandone, A., Pence, K. L., Golinkoff, R. M., & Hirsh-Pasek, K. (2007). Action speaks louder than words: Young children differentially weigh perceptual social and linguistic cues to learn verbs. Child Development, 78, 1322–1342. Bronfenbrenner, U. (1993). The ecology of cognitive development: Research models and fugitive ﬁndings. In R. H. Wozniak & K. W. Fischer (Eds.), Development in context (pp. 3–44). Hillsdale, NJ: Lawrence Erlbaum. Carey, S. (1978). The child as a word learner. In M. Halle, J. Bresnan, & G. A. Miller (Eds.), Linguistic theory and psychological reality (pp. 264–295). Cambridge, MA: MIT Press. Carpenter, M., Nagell, K., & Tomasello, M. (1998). Social cognition, joint attention, and communicative competence from 9 to 15 months of age. Monographs of the Society for Research in Child Development, 63.

26

Kathy Hirsh-Pasek and Roberta Michnick Golinkoff

Casasola, M., & Cohen, L. (2002). Infant categorization of containment, support and tight-ﬁt spatial relations. Developmental Science, 5, 247–264. Casasola, M., Bhagwat, J., & Ferguson, K. (2006). Precursors to verb learning: Infants’ understanding of motion events. In K. Hirsh-Pasek & R. M. Golinkoff (Eds.), Action meets words: How children learn verbs (pp. 160–191). New York: Oxford University Press. Choi, S. (2006). Preverbal spatial cognition and language-speciﬁc input: Categories of containment and support. In K. Hirsh-Pasek & R. M. Golinkoff (Eds.), Action meets words: How children learn verbs (pp. 544–565). New York: Oxford University Press. Clark, E. V. (1983). Meanings and concepts. In J. H. Flavell & E. M. Markman (Eds.), Handbook of child psychology, Vol. III: Cognitive development (pp. 787–840). New York: Wiley. Fenson, L., Dale, P. S., Reznick, J. S., & Bates, E. (1994). Variability in early communicative development. Monographs of the Society for Research in Child Development, 59, (Serial 242). Fisher, C., & Song, H. (2006). Who’s the subject? Sentence structure and verb meaning. In K. Hirsh-Pasek & R. M. Golinkoff (Eds.), Action meets words: How children learn verbs (pp. 394–429). New York: Oxford University Press. Forbes, J. N., & Farrar, J. M. (1995). Learning to represent word meaning: What initial training events reveal about children’s developing action verb concepts. Cognitive Development, 10, 1–20. Gentner, D. (1982). Why are nouns learned before verbs: Linguistic relativity versus natural partitioning. In S. A. Kuczaj II (Ed.), Language development, Vol. 2: Language, thought and culture (pp. 301–334). Hillsdale, NJ: Lawrence Erlbaum Associates. Gentner, D., & Boroditsky, L. (2001). Individuation, relational relativity and early word learning. In M. Bowerman & S. Levinson (Eds.), Language acquisition and conceptual development (pp. 215–256). Cambridge, UK: Cambridge University Press. Gergely, G., & Csibra, G. (2003). Teleological reasoning in infancy: The naı¨ ve theory of rational action. Trends in Cognitive Science, 7, 287–291. Gillette, J., Gleitman, H., Gleitman, L., & Lederer, A. (1999). Human simulations of vocabulary learning. Cognition, 73, 135–176. Gleitman, L. (1990). The structural sources of verb meanings. Language Acquisition, 1, 3–55. Golinkoff, R. M. (1993). When is communication a ‘‘meeting of minds’’? Journal of Child Language, 20, 199–207. Golinkoff, R. M., & Hirsh-Pasek, K. (2006). Baby wordsmith: From associationist to social sophisticate. Current Directions in Psychological Science, 15, 30–33. Golinkoff, R. M., Hirsh-Pasek, K., Cauley, K. M., & Gordon, L. (1987). The eyes have it: Lexical and syntactic comprehension in a new paradigm. Journal of Child Language, 14, 23–45. Golinkoff, R. M., Hirsh-Pasek, K., & Hollich, G. (1999). Emerging cues for word learning. In B. MacWhinney (Ed.), The emergence of language (pp. 305–330). Hillsdale, NJ: Lawrence Erlbaum Associates. Golinkoff, R. M., Mervis, C. V., & Hirsh-Pasek, K. (1994). Early object labels: The case for a developmental lexical principles framework. Journal of Child Language, 21, 125–155. Golinkoff, R. M., Chung, H. L., Hirsh-Pasek, K., Liu, J., Bertenthal, B. I., Brand, R., Maguire, M. J., & Hennon, E. A. (2002). Young children can extend motion verb labels to point-light displays. Developmental Psychology, 38, 604–614. Hennon, E. (2000). Integrating the field of language development: How middle ground theories are embracing both normal and atypical development. Unpublished manuscript. Hespos, S., & Spelke, E. (2004). Conceptual precursors to language. Nature, 430, 453–456. Hirsh-Pasek, K., & Golinkoff, R. M. (1996). The preferential looking paradigm reveals emerging language comprehension. In D. McDaniel, C. McKee, & H. Cairns (Eds.), Methods for assessing children’s syntax (pp. 105–124). Cambridge, MA: MIT Press.

King Solomon’s Take on Word Learning

27

Hirsh-Pasek, K., & Golinkoff, R. (Eds.). (2006). Action meets word: How children learn verbs. New York: Oxford University Press. Hirsh-Pasek, K., Golinkoff, R. M., & Naigles, L. (1996a). Young children’s use of syntactic frames to derive meaning. In K. Hirsh-Pasek & R. Golinkoff (Eds.), The origins of grammar (pp. 123–158). Cambridge, MA: MIT Press. Hirsh-Pasek, K., Tucker, M., & Golinkoff, R. M. (1996b). Dynamical systems: Reinterpreting prosodic bootstrapping. In J. Morgan & K. Demuth (Eds.), Signal to syntax: Bootstrapping from speech to grammar in early acquisition (pp. 449–466). Hillsdale, NJ: Lawrence Erlbaum Associates. Hirsh-Pasek, K., Golinkoff, R. M., & Hollich, G. (2000). An Emergentist Coalition Model for word learning: Mapping words to objects is a product of the interaction of multiple cues. In R. M. Golinkoff, K. Hirsh-Pasek, L. Bloom, L. Smith, A. Woodward, N. Akhtar, M. Tomasello, & G. Hollich (Eds.), Becoming a word learner: A debate on lexical acquisition (pp. 136–164). New York: Oxford University Press. Hollich, G. J., Hirsh-Pasek, K., & Golinkoff, R. M. (with Hennon, E., Chung, H. L., Rocroi, C., Brand, R. J., & Brown, E.). (2000a). Breaking the language barrier: An emergentist coalition model for the origins of word learning. Monographs of the Society for Research in Child Development, 65 (3, Serial No. 262). Hollich, G., Hirsh-Pasek, K., Tucker, M. L., & Golinkoff, R. M. (2000b). A change is afoot: Emergentist thinking in language acquisition. In P. Anderson, C. Emmeche, N. O. Finnemann, & P. V. Christiansen (Eds.), Downward causation (pp. 143–178). Oxford, England: Aarhus University Press. Hollich, G., Golinkoff, R. M., & Hirsh-Pasek, K. (2007). Young children prefer to attach labels to whole objects over their parts. Unpublished manuscript. Imai, M., Li, L., Haryu, E., Hirsh-Pasek, K., Golinkoff, R. M., & Shigematsu, J. (2006). Novel noun and verb learning in Chinese, English, and Japanese children: Universality and languagespecificity in novel noun and verb learning. Unpublished manuscript. Ma, W., Golinkoff, R. M., & Hirsh-Pasek, K. (2007). Why do Chinese children learn more verbs? Presented at the biennial meeting of the Society for Research in Child Development, March. Boston, MA. Ma, W., Golinkoff, R., Hirsh-Pasek, K., McDonough, C., & Tardif, T. (under review). Imagine that! Imageability predicts verb learning in Chinese children. MacWhinney, B., Ple´h, C., & Bates, E. (1985). The development of sentence interpretation in Hungarian. Cognitive Psychology, 17, 178–209. Maguire, M., Hirsh-Pasek, K., & Golinkoff, R. (2006). A uniﬁed theory of word learning: Putting verb acquisition in context. In K. Hirsh-Pasek & R. Golinkoff (Eds.), Action meets word: How children learn verbs (pp. 364–392). Oxford University Press. Mandler, J. M. (1991). Prelinguistic primitives. In L. A. Sutton & C. Johnson (Eds.), Proceedings of the seventeenth annual meeting of the Berkeley Linguistics society (pp. 414–425). Berkeley, CA: Berkeley Linguistics Society. Mandler, J. M. (2004). The foundations of mind: Origins of conceptual thought. New York: Oxford University Press. Markman, E. (1989). Categorization and naming in children: Problems of induction. Cambridge, MA: MIT Press. Masterson, J., & Druks, J. (1998). Description of a set of 164 nouns and 102 verbs matched for printed word frequency and familiarity and age of acquisition. Journal of Neurolinguistics, 11, 331–354. Merriman, W. E., & Bowman, L. (1989). The mutual exclusivity bias in children’s early word learning. Monographs of the Society for Research in Child Development, 54, (Serial No. 220).

28

Kathy Hirsh-Pasek and Roberta Michnick Golinkoff

McDonough, C., Hirsh-Pasek, K., Golinkoff, R. M., Lannon, R., & Song, L. (in preparation). An image is worth a thousand words: Why nouns tend to dominate verbs in early word learning. McMurray, B. (2007). Defusing the childhood vocabulary explosion. Science, 317, 631. Naigles, L. (1990). Children use syntax to learn verb meanings. Journal of Child Language, 17, 357–374. Naigles, L., & Hoff, E. (2006). Verbs at the very beginning. In K. Hirsh-Pasek & R. M. Golinkoff (Eds.), Action meets word: How children learn verbs (pp. 336–364). New York: Oxford Press. Nelson, K. (1996). Language in cognitive development. New York: Cambridge University Press. Paivio, A., Yuille, J. C., & Madigan, S. A. (1968). Concreteness, imagery and meaningfulness values for 925 nouns. Journal of Experimental Psychology, 76, 1–25. Parish-Morris, J., Hennon, E. A., Hirsh-Pasek, K., Golinkoff, R. M., & Tager-Flusberg, H. (2007). Children with autism illuminate the role of social intention in word learning. Child Development, 78, 1265–1287. Poulin-Dubois, D., & Forbes, J. N. (2002). Toddlers’ attention to intentions-in-action in learning novel action words. Developmental Psychology, 38, 104–114. Poulin-Dubois, D., & Forbes, J. N. (2006). Word, intention, and action: A two-tiered model of action word learning. In K. Hirsh-Pasek & R. M. Golinkoff (Eds.), Action meets word: How children learn verbs (pp. 262–286). Oxford, England: Oxford Press. Plunkett, K. (1997). Theories of language acquisition. Trends in Cognitive Sciences, 1, 146–151. Pruden, S. M., & Hirsh-Pasek, K. (2006). Foundations of verb learning: Labels promote action category formation. In D. Bamman, T. Magnitskaia, & C. Zaller (Eds.), Proceedings of the 30th Annual Boston University Conference on Language Development (pp. 476–488). Somerville, MA: Cascadilla Press. Pruden, S. M., Hirsh-Pasek, K., & Golinkoff, R. M. (2008). Current events: How infants parse the world events for language. In T. Shipley & J. Zacks (Eds.), Understanding events: How humans see, represent, and act on events. New York, NY: Oxford University Press. Pruden, S. M., Hirsh-Pasek, K., Golinkoff, R. M., & Hennon, E. A. (2006). The birth of words: Ten-month-olds learn words through perceptual salience. Child Development, 77, 266–280. Pruden, K., Hirsh-Pasek, K., Maguire, M., & Meyer, M. (2004). Foundations of verb learning: Infants categorize path and manner in motion events. Proceedings of the 28th Annual Boston University Conference on Language Development, Somerville, MA: Cascadilla Press. Pulverman, R., & Golinkoff, R. M. (2003). Starting out on the right path: Seven-month-olds’ attention to potential verb referents in nonlinguistic events. Proceedings of the 27th Annual Boston University Conference on Language Development, Somerville, MA: Cascadilla Press. Pulverman, R., & Golinkoff, R. M. (2004). Seven-month-olds’ attention to potential verb referents in nonlinguistic events. In A. Brugos, L. Micciulla, & C. E. Smith (Eds.), Proceedings of the 28th Annual Boston University Conference on Language Development (pp. 473–481). Somerville, MA: Cascadilla Press. Pulverman, R., Golinkoff, R. M., Hirsh-Pasek, K., & Maldonado, D. (2005). Linguistic relativity in one-year-olds? English and Spanish learning infants’ attention to manner and path in silent events. Presented at the biennial meeting of the Society for Research in Child Development, Atlanta, GA. Pulverman, R., Hirsh-Pasek, K., Golinkoff, R. M., Pruden, S., & Salkind, S. (2006). Conceptual foundations for verb learning: Celebrating the event. In K. Hirsh-Pasek & R. M. Golinkoff (Eds.), Action meets word: How children learn verbs (pp. 134–159). New York: Oxford University Press. Pulverman, R., Sootsman, J., Golinkoff, R., & Hirsh-Pasek, K. (2007). Manners matter: Infants’ attention to manner and path in non-linguistic dynamic events. Unpublished manuscript.

King Solomon’s Take on Word Learning

29

Sharon, T., & Wynn, K. (1998). Individuation of actions from continuous motion. Psychological Science, 9, 357–362. Slobin, D. (2001). The child learns to think for speaking: Puzzles of crosslinguistic diversity in form-meaning mappings. Paper presented at the biennial meeting of the Society for Research in Child Development, April. Minneapolis, MN. Smith, L. (2000). Learning how to learn words: An associative crane. In R. M. Golinkoff, K. Hirsh-Pasek, L. Bloom, L. B. Smith, A. L. Woodward, N. Akhtar, M. Tomasello, & G. Hollich (Eds.), Becoming a word learner: A debate on lexical acquisition (pp. 51–80). New York: Oxford University Press. Snedeker, J., & Gleitman, L. (2004). Why it is hard to label our concepts. In G. Hall & S. Waxman (Eds.), Weaving a lexicon (pp. 257–295). Cambridge, MA: MIT Press. Snedeker, J., Li, P., & Yuan, S. (2003). Cross-cultural differences in the input to early word learning. Proceedings of the Twenty-Fifth Annual Conference of the Cognitive Science Society (pp. 214–219). Mahwah, NJ: Lawrence Erlbaum. Song, L., Golinkoff, R. M., Seston, R., Ma, W., Shallcross, W., & Hirsh-Pasek, K. (2006, November). Action stations: Verb learning rests on constructing categories of action. Poster session presented at the 31st Annual Boston University Conference on Language Development, Boston, MA. Talmy, L. (1985). Lexicalization patterns: Semantic structure in lexical forms. Language typology and syntactic description (pp. 57–149). Cambridge, MA: Cambridge University Press. Tardif, T. (1996). Nouns are not always learned before verbs: Evidence from Mandarin speakers’ early vocabularies. Developmental Psychology, 32, 492–504. Tardif, T. (2006). But are they really verbs? Chinese words for action. In K. Hirsh-Pasek & R. M. Golinkoff (Eds.), Action meets words: How children learn verbs (pp. 477–499). New York: Oxford University Press. Tardif, T., Fletcher, P., Liang, W. L., Zhang, Z. X., Leung, C. S., & Kaciroti, N. (in preparation). Early vocabulary development in Cantonese and Putonghua: Using the MacArthur-Bates Communicative Development Inventories to compare development across languages. Tomasello, M. (1992). First verbs: A case study in early grammatical development. Cambridge, England: Cambridge University Press. Tomasello, M. (1995). Pragmatic contexts for early verb learning. In M. Tomasello & W. E. Merriman (Eds.), Beyond names for things: Young children’s acquisition of verbs (pp. 115–146). Hillsdale, NJ: Lawrence Erlbaum Associates. Waxman, S. R., & Kosowski, T. D. (1990). Nouns mark category relations: Toddlers’ and preschoolers’ word-learning biases. Child Development, 61, 1461–1473. Waxman, S. R., & Lidz, J. (2006). Early word learning. In D. Kuhn & R. Siegler (Eds.), Handbook of Child Psychology (6th ed., Vol. 2, pp. 299–335). Hoboken, NJ: Wiley. Woodward, A. L., & Markman, E. M. (1998). Early word learning. In W. Damon, D. Kuhn, & R. Siegler (Eds.), Handbook of child psychology, Vol. 2: Cognition, perception and language (pp. 371–420). New York: Wiley. Woodward, A. L., Sommerville, J. A., & Guajardo, J. J. (2001). How infants make sense of intentional action. In B. F. Malle, L. J. Moses, & D. A. Baldwin (Eds.), Intentions and intentionality (pp. 149–169). Cambridge, MA: MIT Press. Wynn, K. (1996). Infants’ individuation and enumeration of actions. Psychological Science, 7, 164–169. Yu, C., & Smith, L. (2007). Rapid word learning under uncertainty via cross-situational statistics. Psychological Science, 18, 414–420.

This page intentionally left blank

ORTHOGRAPHIC LEARNING, PHONOLOGICAL RECODING, AND SELF-TEACHING

David L. Share DEPARTMENT OF LEARNING DISABILITIES, FACULTY OF EDUCATION, UNIVERSITY OF HAIFA, MOUNT CARMEL, 31905, HAIFA, ISRAEL

I. INTRODUCTION A . PROLOGUE B . THE CRUCIAL ROLE OF ORTHOGRAPHIC LEARNING IN THE ACQUISITION OF SKILLED READING II. THE SELF-TEACHING THEORY OF ORTHOGRAPHIC LEARNING A . SOME HISTORICAL BACKGROUND B . KEY FEATURES OF SELF-TEACHING III. EMPIRICAL FINDINGS A . THE EARLY PIONEERS B . STUDIES OF ORTHOGRAPHIC LEARNING WITHIN THE SELFTEACHING FRAMEWORK C . INDIVIDUAL DIFFERENCES IN SELF-TEACHING, DYSLEXICS, AND OTHER POOR READERS D . EARLY ONSET E . OTHER SELF-TEACHING MECHANISMS? IV. SUMMARY, CONCLUSIONS, AND THE WAY AHEAD ACKNOWLEDGEMENT REFERENCES

I. Introduction A. PROLOGUE First, I would like to ask the reader to read the following short passage. Please read at your own natural reading pace—much as you would read a light novel. In the middle of Australia is the hottest town in the world. This town is called Sloak and it’s right in the middle of the desert. In Sloak, the temperature can reach 60 degrees. It’s so hot that even the ﬂies drop dead and the rubber tires on the cars start to melt. You can even fry an egg on the roof of your car.

31 Advances in Child Development and Behavior R.V. Kail : Editor

Copyright r 2008 Elsevier B.V. All rights reserved.

David L. Share

32

The houses in Sloak are under the ground, far away from the heat of the sun. The people also dig for gold deep under the ground. In Sloak, they drink lots of beer to stay cool. They drink beer in the morning, in the afternoon, and in the evening. The beer in Sloak is very strong. If you’re not used to drinking beer you’d better watch out! Would you like to live in Sloak?

It will immediately strike the reader that this text is aimed not at beerlovers but at young children—actually second graders. And those familiar with the Australian ‘‘outback’’ will appreciate that this imaginary town is not entirely ﬁctitious. Now that I’ve managed to ﬁlibuster for a sentence or two, we can get back to the story of orthographic learning, but ﬁrst, one other small request. Without going back to the text, look carefully at the following two spellings—SLOAK/SLOKE. Which do you think was the name of the town in the story? Because both are what we call ‘‘pseudowords’’—invented or possible words that do not actually exist, no prior knowledge is available to help out. And because the two spellings sound exactly the same, in weighing up the two alternatives you have only your memory to rely upon for the speciﬁc spelling seen a few moments ago. So, how did you do? It turns out that skilled adult readers and young inexperienced readers alike demonstrate a surprisingly strong preference for the original spelling. It might be objected, however, that this spelling somehow looks ‘‘right’’ or more ‘‘wordlike,’’ and would have been favored even if it had not appeared in the passage. But even when the alternative spelling Sloke replaces Sloak throughout the passage, the advantage for this spelling is just as strong (see, e.g., Share, 1999). In any case, these preexisting ‘‘orthographic preferences’’ can be checked independently to ensure that both spellings look equally plausible or ‘‘wordlike’’ for a particular group of readers. The evidence reviewed in this chapter shows that not only are able readers remarkably adept at distinguishing the original spelling from its socalled ‘‘homophonic foil,’’ but this newly acquired knowledge is manifest in a variety of tasks including regular pen-and-paper spelling, and also naming times—the original spelling is consistently pronounced more quickly and with fewer errors than the alternate unseen spelling. This learning also seems to occur surprisingly rapidly; there were only six occurrences of the ‘‘target’’ word (Sloak) in the passage presented previously, but even a single exposure may be sufﬁcient to get the learning process underway (Share, 2004). No less remarkable is the durability of this learning. The data collected by a number of different researchers working in a variety of languages (not just English) clearly show that if the spelling choice task you just performed

Orthographic Learning, Phonological Recoding, and Self-Teaching

33

had been delayed until after you had ﬁnished reading this chapter—the choices would be no less impressive. In fact, in another week (or even another month!) you would probably still be able to recognize the original spelling—not necessarily with unwavering certainty, but at levels well beyond the 50% chance level. Summing up, a growing number of studies now indicate that this process of picking up these visual (or more correctly ‘‘visual-orthographic’’) forms of new words—‘‘orthographic learning’’—is a surprisingly rapid and durable process. In this chapter, I ask how this learning comes about, focusing on a particular theory of orthographic learning—the self-teaching hypothesis. In the ﬁrst section, I provide an exposition of the theory. This is followed by a review of the empirical studies that others and I have undertaken in attempting to test this idea. But why all the fuss about spellings in an age where ‘‘correct’’ spelling seems out of vogue. Today’s writers—young and old—have automatic spelling-correction routines at their ﬁngertips, spell-checkers if need be, and even if it is an old-fashioned pencil-and-paper production, we are urged to tolerate creative (‘‘invented’’) spellings particularly among novice readers/ writers. The ﬁnal straw, of course, is the SMS text-messaging jargon which seems to throw all known spelling conventions to the wind. So what difference does it make if we remember Sloak rather than Sloke? The answer is that it is not really about spelling but about reading, indeed, the very foundations of skilled reading, text understanding, and ultimately literacy.

B. THE CRUCIAL ROLE OF ORTHOGRAPHIC LEARNING IN THE ACQUISITION OF SKILLED READING Perhaps the single most distinctive characteristic of skilled reading is the sheer speed and effortlessness of the word identiﬁcation process. While there is no disputing that comprehending the meaning of complete units of text (from brief phrases such as Staff Only through to a 2-volume Harry Potter novel) is the ultimate goal of reading acquisition, the component word recognition processes apparently constitute a unique and crucial ingredient in this process. Consider the existence of severely dyslexic yet highly intelligent and literate individuals (see, e.g., Eileen Simpson’s inspiring autobiography) as well as underprivileged groups such as women in traditional societies who, despite normal intelligence and linguistic capabilities, were (and in some countries still are) denied the opportunity to learn to read and write (see, e.g., Lukatela et al., 1995). What is lacking in both these groups are not the capabilities needed to comprehend written language but the speciﬁc ability to deal with print, namely, the ability to

34

David L. Share

recognize the printed word. Of course, although word recognition proﬁciency is a pre-requisite for literacy learning, it is no guarantee of good text understanding. Highly skilled word decoders—so-called hyperlexics (the mirror image of dyslexia)—are distinguished by their exceptional word reading ability, yet their low intelligence and particularly poor language comprehension renders them unable to comprehend the material they decode so well (Nation, 1999). The ability to instantly and effortlessly recognize the printed word is, in many ways, the quintessence of reading skill. Such ‘‘technical’’ expertise or ‘‘automaticity’’ seems to be a deﬁning feature of most human skills—not just reading. Consider the envy of the novice driver marveling at how the veteran motorist manages to carry on a near-seamless conversation while negotiating city trafﬁc, or the novice pianist who observes (often in despair) at how easily a technically challenging passage currently beyond her current level of prowess is executed by the teacher with ﬁnesse and apparent effortlessness. The art of skilled word recognition, when practiced by a highly literate individual is so efﬁcient or ‘‘automatized’’ it seems almost ‘‘unstoppable’’—occurring whether we will it or not. It has even been shown that printed words can be recognized without conscious awareness (see, e.g., the subliminal phenomenon called ‘‘masked priming,’’ Forster & Davis, 1984). Indeed, roadside billboard advertising seems predicated on precisely this automaticity—if the print is clearly visible and the graphics sufﬁciently eye-catching, the driver just cannot avoid taking in the advertiser’s message (see Logan, 1988).1 Such accomplished word recognition depends on the accumulation of a large mental store of printed word forms (‘‘representations’’), each encapsulating the knowledge that a particular conﬁguration of letters such as Sloak (as opposed to other similar-looking ones such as Slook, Sleek, Slack, Cloak, Soak, Look, etc.) is the written form of a word identiﬁed in speech as ‘‘sloak’’ and referring to the name of a mythical outback Australian town. The skilled reader has at his or her disposal thousands of these internal so-called ‘‘orthographic representations,’’ not to mention a good general knowledge of orthographic (spelling) conventions. In the classroom these representations are called a reader’s ‘‘sight’’ vocabulary, in the reading researcher’s laboratory—the ‘‘orthographic lexicon’’ or ‘‘orthographic processor.’’ Whether this knowledge is best captured in individually stored word-level units—so-called localist models (see, e.g., 1 The reader familiar with Gough and Hoover’s so-called ‘‘Simple Model’’ of Reading will no doubt recognize the well-known notion that reading comprehension consists of two components—decoding and linguistic comprehension, but even this oversimpliﬁed model still seems a useful way of framing the issue.

Orthographic Learning, Phonological Recoding, and Self-Teaching

35

Coltheart et al., 1993, 2001)—or as a distributed (i.e., non-localist) set of connections in a multi-layered network of simple neuron-like units (so-called ‘‘triangle’’) models (Harm & Seidenberg, 2004; Plaut et al., 1996), there is a broad consensus that orthographic representations specifying the identity and order of a word’s letters (Adams, 1990; Ehri, 1980, 1992; Perfetti, 1992) and tightly linked to word identity—phonology, meaning (and more)—are the key ‘‘skill’’ in skilled reading and that compiling this orthographic knowledge is the main challenge facing the novice reader. To sum up, orthographic learning is the very nucleus of print processing. But how do readers manage to compile such impressive repositories of orthographic knowledge? This brings us to the self-teaching hypothesis. The self-teaching hypothesis is simply a theory about how readers build up this knowledge—a process that takes place not over months but many years and probably never stops.

II. The Self-Teaching Theory of Orthographic Learning The self-teaching hypothesis (Firth, 1972; Jorm, 1979; Jorm & Share, 1983; Share, 1995) proposes that the ability to translate unfamiliar printed words into their spoken equivalents (‘‘phonological recoding’’ or simply ‘‘decoding’’) is the central means by which orthographic representations are acquired.2 Each successful decoding of a new word is assumed to provide an opportunity to acquire the word-speciﬁc orthographic information that is the foundation of skilled visual word recognition. Exhaustive phonological recoding is assumed to be critical for the formation of well-speciﬁed orthographic representations because it draws the reader’s attention to the graphemic detail—the order and identity of the letters and how they map onto the phonological representation—the spoken form (see Ehri, 1992, 2005; Perfetti, 1992). In this way, phonological recoding functions as a self-teaching mechanism or built-in teacher enabling a child to 2 As originally pointed out (Share, 1995, footnote 1, p. 152) the term phonological recoding does not imply any particular procedure but is used as an umbrella term for the process of print to sound conversion by whatever means this is accomplished. This covers several possibilities including (but not necessarily limited to) explicit letter-by-letter application of grapheme– phoneme correspondence rules, an analogical activation-synthesis mechanism, an implicit statistical learning mechanism, or automatic activation of a distributed (connectionist) network of simple neuron-like units. Although often misinterpreted in this way, the self-teaching hypothesis does not imply that orthographic learning is solely the product of the ﬁrst of these procedures. Furthermore, the nature of this process will undoubtedly vary developmentally and across orthographies.

36

David L. Share

independently develop the word-speciﬁc orthographic knowledge necessary for skilled reading. The self-teaching idea is a little unconventional in that the process of orthographic learning is assumed to take place unintentionally, as a byproduct of the process of decoding—readers do not usually aim to analyze and remember spellings, it just happens and probably without our being aware of the process. It is important to note too that the self-teaching idea rejects the intuitively appealing notion that identities of most new printed words can be directly taught (by teachers, parents, or peers) or can be guessed on the basis of contextual information (see Share, 1995 for detailed discussion). Only decoding seems to offer a sufﬁciently reliable means for identifying novel letter strings (owing to the fundamentally alphabetic nature of the written code) thereby providing the opportunities for (incidental) learning of the visual form (spellings) of these items.

A. SOME HISTORICAL BACKGROUND At the time that the self-teaching notion was ﬁrst conceived, the role of phonological recoding was seen solely as a back-up mechanism for word recognition when ‘‘direct’’ visual recognition failed (e.g., Coltheart, 1978; McCusker, Hillinger, & Bias, 1981). The dominant conceptual framework at the time was the classic ‘‘either-or’’ dual-route notion of two independent routes (now eschewed by Coltheart in favor of a more synergistic ‘‘twohoses-ﬁlling-a-bucket’’ notion, see Coltheart, 2005). Because the focus of traditional dual-route theorizing was the skilled reader who already possesses extensive orthographic knowledge, this back-up route clearly had only subsidiary status in skilled word recognition, although there was broad agreement that for the beginning reader phonology was somehow more important (e.g., Barron, 1986; Doctor & Coltheart, 1980), because among less experienced readers many more words are unfamiliar. The self-teaching notion was originally conceived by a doctoral student named Ian Firth—one of a long line of researchers (both preceding and succeeding him) to discover, or rather re-discover, the remarkable power of pseudoword naming to discriminate good and poor readers.3 Firth proposed that the ability to convert letters into sounds as a means to pronouncing unfamiliar words was the way a child built up the range of words recognized 3 Even today, I stand in awe of Firth’s colossal dissertation which included one of the ﬁrst (if not the very ﬁrst) computational models of reading—a computer simulation remarkably similar to Ans, Carbonnel, and Valdois’ (1998) contemporary multi-trace model. Unfortunately, this thesis was never published.

Orthographic Learning, Phonological Recoding, and Self-Teaching

37

by ‘‘sight.’’ Thus, decoding was the key to reading acquisition and the development of a child’s sight or reading vocabulary. At that point in the history of developmental reading research, the phonological deﬁcit hypothesis was unknown and visual/perceptual theories (eventually laid to rest some years later by Vellutino, 1979) were pre-eminent. Firth’s pseudoword naming data, therefore, was an unexpected ﬁnding that needed explaining in an era of visual/perceptual theories of reading difﬁculties. Developing the self-teaching notion, Jorm and Share (1983) pointed to a growing body of ﬁndings indicating a major role for phonological recoding and, more generally, information processing (perception, learning, and memory) in the speech domain (see also Jorm, 1983). As already noted, the dominant thinking at the time regarded printed word recognition as a mainly visual affair. Like Firth, Jorm and Share (1983) were puzzled by the apparent contradiction, or paradox, between predominantly visual/nonphonological skilled word recognition favored at the time (Coltheart, 1978; Glushko, 1979; Kay & Marcel, 1981; McCusker et al., 1981) and a growing body of developmental evidence pointing to a central role for phonology in reading. The self-teaching idea was a way to resolve this (apparent) paradox. It was conceded that visual word recognition, at least among skilled readers, and at least for familiar words, was largely if not exclusively a visual process4 but argued that the acquisition of this knowledge base was largely the result of decoding encounters with new words—phonological recoding. Jorm and Share stressed the fact that children are continually encountering unfamiliar words, and consequently require a means for independently identifying these orthographic newcomers, which are simply too abundant to be taught on a one-by-one rote basis (the direct instruction option) nor could be guessed accurately enough on the basis of contextual information (for reasons elaborated in Share, 1995). This left only the recoding mechanism to fulﬁll the ‘‘self-teaching’’ function of enabling children to independently decipher novel letter strings and, in the spirit of paired-associate learning, permit bonding of the visual form of the word to its spoken and semantic form. At ﬁrst, the proposed self-teaching process was conceptualized in rather mechanistic/behavioristic terms as a ‘‘learning trial’’ in which successful decoding of a printed word permitted associative pairing of the new visual form with its spoken form and meaning (see Jorm & Share, 1983). This paired-associate learning process—like the dominant dual-route model—was rather vague about what actually was acquired.

4 Today, the pendulum seems to have swung back to a much stronger phonological position (see, e.g., Frost, 1998; Perfetti, 2003; Van-Orden, Pennington, & Stone, 1990).

38

David L. Share

The emphasis was on the how of orthographic learning, rather than what was actually learned. Jorm and Share (1983) were at pains to point out that natural text is replete with low-frequency words not only for beginning readers but also for more experienced readers (see also Foorman et al., 2004). Share (1995) however, noted that, at some point, every printed word, even a child’s own name, is unfamiliar visually.5 Thus, self-teaching, it was argued, was relevant to the process of learning every printed word.6 A subsequent conception of the self-teaching idea recast the self-teaching function of phonological recoding within the broader more universalistic framework of a transition from initial identiﬁcation (decoding) of unfamiliar words to their rapid recognition as familiar units. Share (2007) argues that this ‘‘unfamiliar-to-familiar’’ transition (seen from the perspective of individual items) or (from the reader’s perspective) ‘‘novice-to-expert’’ transition represents a fundamental and overarching duality in word reading that applies to all words in all possible orthographies. On the one hand, because all words are novel at some point in reading development, the reader must possess some algorithm, albeit imperfect, yet nonetheless functional for independently identifying words encountered for the ﬁrst time in everyday reading. Secondly, and again a literacy universal, the reader must eventually be able to achieve a high degree of automatization in word recognition—rapid and effortless recognition of familiar words and morphemes (LaBerge & Samuels, 1974; Logan, 1997, 2002; Perfetti, 1985; Rayner, 1998; van der Leij & van Daal, 1999) perceived as whole units via a direct-retrieval mechanism (see Ans et al., 1998; Weekes, 1997). Here is where the well-speciﬁed ‘‘autonomous’’ orthographic representations discussed previously are crucial (Perfetti, 1992). This ability to automatize or ‘‘modularize’’ word identiﬁcation (Adams, 1990; Stanovich, 1990, 2000) is probably the quintessence of reading skill (Perfetti, 1985, 1994). As with the decoding algorithm, this high-speed direct-retrieval mode applies to all words in all orthographies. This universalistic dualism has several advantages over the traditional Coltheart/Baron dual-route approach that focuses primarily on the peculiarly English-language distinction between regular and irregular words. First, it merges the study of reading with the study of human skill learning across a range of domains (see, e.g., Anderson, 1981; Goldstone, Until recently my own name, printed in the original Czech, complete with hacek—Sˇer, was visually unfamiliar. 6 This is almost but not quite correct. The most frequent 100 or so words in printed English (was, is, are, to, of ) are extraordinarily irregular but pop up in almost every phrase—hence may well be learned or taught whole-word rote style. 5

Orthographic Learning, Phonological Recoding, and Self-Teaching

39

1998; Karni, 1996; LaBerge & Brown, 1989; Logan, 1988; Newell & Rosenbloom, 1981; Shiffrin & Schneider, 1977; Siegler, 1988; Venezky, 2006). The dualism common to all skill learning is a contrast between, or transition from slow, deliberating, step-by-step unskilled performance to rapid automatized one-step or ‘‘unitized’’ skilled performance without which the ‘‘skill’’ of reading would probably never have made so profound an impact on modern knowledge-based cultures (Olson, 1994). Second, this broader ‘‘novice/expert’’ or ‘‘unfamiliar/familiar’’ dualism also converges with the dualistic nature of an efﬁcient orthography. Speciﬁcally, an efﬁcient script can be conceptualized as a compromise between the often competing needs of the novice and the expert reader (Rogers, 1995; Venezky, 2007). This orthographic dualism might be termed the ‘‘decipherability/automatizability’’ criterion. An effective orthography must ﬁrst provide the reader with a means for deciphering new words independently. This applies to both the young child new to the world of print, and to the skilled reader encountering a new or unfamiliar word. Furthermore—and this is crucial to skill learning in all domains—this algorithmic process must lay the foundations for the rapid direct-retrieval mechanism. This ‘‘do-it-yourself ’’ or ‘‘self-teaching’’ function of decoding is probably the chief virtue of alphabetic scripts— supplying not only an economical means for identifying new words (via print-to-sound translation), but, critically, establishing the detailed orthographic representations on which rapid fully unitized skilled word recognition is founded. Secondly, a successful script must also answer to the needs of the expert by providing visually distinctive word-speciﬁc (or morpheme-speciﬁc) visual-orthographic conﬁgurations required for the unitization and automatization of skilled word recognition. Ideally, each morpheme should have one and only one representation (morpheme ‘‘constancy’’) without showing morphophonemic variation (e.g., electric/ electricity/electrician), with different morphemes represented differently (morpheme ‘‘distinctiveness’’) (Rogers, 1995).7,8 A script catering primarily to the needs of skilled readers, such as the pre-communist Chinese characters (and in many respects English 7

English is faithful to the distinctiveness principle in its heterographic homophones (blue/blew) but not in the large numbers of polysemous (homographic) homophones (well/well/well ) or the relatively rare heterophonic homographs (wind/wind ). 8 There are clearly advantages for a script that also maintains morpheme ‘‘constancy’’—the same morpheme always written the same way—but it may be morpheme distinctiveness that is crucial for the automatization of word recognition. In other words, it’s not that the w in two is important for revealing morphemic relations (e.g., twelve, twice, twilight)—a highly doubtful assumption for the young reader—but that this etymological quirk provides distinct spellings for potentially confusable homophones (too/two/to).

40

David L. Share

orthography—see Chomsky & Halle, 1968) will pose enormous challenges for the novice. Conversely, a script providing maximum decipherability for the novice—Korean hangul, Japanese kana, or highly regular pedagographies such as i.t.a—will often fail (as a stand-alone script) to meet the needs of the skilled reader, primarily owing to homophony. (Consider the problems incurred if two, too, and to all shared the same spelling—tu.) To sum up, the alphabetic code furnishes the necessary algorithm—a selfteaching mechanism permitting independent identiﬁcation by means of decoding that lays the foundations for the skilled reader’s word recognition expertise via the establishment of autonomous orthographic representations. Four central features of the self-teaching theory are reviewed next.

B. KEY FEATURES OF SELF-TEACHING 1. Self-teaching is Item-based, Not Stage-based In their earlier review of the (inconclusive) evidence for and against stage models of word recognition development, Jorm and Share (1983) proposed that many of the conﬂicting ﬁndings might be resolved by considering item familiarity—high-frequency words can be rapidly recognized visually but unfamiliar items depend more on phonology. They suggested that the pertinent question was not how children identify words, but how they identify which words. Because word-speciﬁc orthographic knowledge is acquired so quickly (Hogaboam & Perfetti, 1978; Reitsma, 1983a; Share, 1999, 2004), even among inexperienced readers, words seem to be rapidly assimilated to a child’s reading or so-called sight vocabulary. This implies that at any one point, a child will be reading some words (the most common) rapidly via primarily direct visual/orthographic recognition, whereas other less familiar words are processed primarily phonologically. Furthermore, this item-based learning appears to begin very early (consistent with the early onset hypothesis, described subsequently) and may well be a never-ending process since unfamiliar printed words are continually being encountered even by skilled readers. The unfamiliar-tofamiliar transition discussed previously implies experiential item-by-item learning. This aligns well with instance-based theories of learning such as Logan’s (1988) and the multi-trace computational model of skilled reading of Ans et al. (1998), which see the process of learning as highly dependent on ‘‘episodic’’ encounters with speciﬁc stimuli. In the Ans et al. model, for example, (as with Firth’s simulation) the ‘‘default’’ mode of word recognition for familiar words is global (whole word or lexical), but analytic

Orthographic Learning, Phonological Recoding, and Self-Teaching

41

(either at the syllable or letter level) in the case of unfamiliar words. (It remains to be seen to what extent this global/analytic dichotomy aligns with the phonological/orthographic dichotomy.) Once again, the key element in this learning process, as with all skill learning is the unfamiliarto-familiar, unskilled-to-skilled transition. Of course, recognition strategies at the level of individual words could, if one insists, be conceived of as a ‘‘stage-like’’ progression from one strategy or mode to another, but not at the level of reading in general as in traditional stage theories (see, e.g., Frith, 1985). The emphasis on item-level changes, however, does not preclude developmental changes in the process of deciphering new words— these are captured by the lexicalization idea. 2. Lexicalization Many discussions of the reading process seem to imply that phonological recoding is a single unvarying routine or set of routines. In contrast, the notion of lexicalization regards phonological recoding as a developmental process, particularly among English-language speakers/readers for whom this learning process probably stretches on indeﬁnitely. The evidence reviewed by Share (1995) indicates that most English-language readers seem to start out with a relatively simple set of one-to-one letter–sound correspondences that are relatively insensitive to orthographic and morphemic context. These initial correspondences are often invariant and therefore, strictly speaking, incorrect. Yet, they offer the novice a manageable set of correspondences capable of generating an approximation to an identiﬁable pronunciation, one (and later, perhaps more) that can be checked for contextual ‘‘goodness of ﬁt.’’ With increased print exposure, these ‘‘beginner’’ letter–sound correspondences become ‘‘lexicalized,’’ that is, modiﬁed in the light of lexical constraints imposed by a growing body of orthographic knowledge. The growing print lexicon alerts the child to regularities beyond the level of simple one-to-one correspondences, such as context-sensitive (soft and hard g and c), positional (ﬁnal versus initial y), and morphemic constraints (missed rather than misst). Thus, contrary to prevailing opinion, this view posits no single decoding procedure or routine, but an ever-changing and self-reﬁning process that at ﬁrst appears be very ‘‘bottom-up,’’ with little sensitivity to higherorder regularities but over the course of print experience becomes increasingly attuned (‘‘lexicalized’’) to the given orthography in a two-way interplay between decoding abilities and orthographic knowledge. 3. Early Onset Another key feature of self-teaching is early onset; beginning reading is assumed to be beginning self-teaching. Several studies suggest that some

42

David L. Share

decoding skills may exist at the very earliest stages of learning to read, even before a child possesses any decoding skill in the conventional sense of being able to sound out and blend even simple pseudowords (Ehri & Wilce, 1985, 1987; Morris, 1992; Stuart & Coltheart, 1988). This view runs counter to many contemporary accounts of printed word learning that propose an initial logographic or visual stage prior to later-developing phonological recoding (see, e.g., Frith, 1985; Gough & Juel, 1991). The early onset hypothesis proposes that even some rudimentary decoding ability may be sufﬁcient for the establishment of primitive or partial orthographic representations of the kind discussed by Ehri (1992) and Perfetti (1992). This early self-teaching depends on three factors; letter–sound knowledge, some minimal phonological sensitivity in the form of awareness of initial sounds (or initial and ﬁnal sounds which are typically the more regular consonants), and the ability to utilize contextual information to determine exact word pronunciations on the basis of an incomplete or inaccurate decoding. On the thorny question of the role of context, there is a critical distinction to be drawn between the identification of an unfamiliar letter string (the early self-teaching discussed here) as opposed to the recognition of a familiar string. The early onset idea proposes an important supplementary or facilitative role for context in the initial learning of newly encountered words (especially in English)—this is quite unlike its role in the largely autonomous recognition of familiar (well-unitized) words that no longer require ‘‘outside’’ (supra-lexical) assistance. That is, the reliance on context in the case of unfamiliar words, especially irregular ones, can be helpful in orthographic learning by resolving decoding ambiguity, but, later, as word recognition becomes more modularized (Shatil & Share, 2003) can be discarded much like the young bicycle rider’s training wheels; disabled readers, however, seem unable to discard the contextual ‘‘crutch’’ owing to poorly automatized word recognition.

4. Two Components to Self-teaching—Phonological and Orthographic and the Phonology-Primary/Orthography-Secondary Hypothesis The process of self-teaching is not just about decoding but seems to involve at least two component processes—phonological and orthographic. Both components are assumed to make independent contributions to printed word learning although the phonological component is considered primary, accounting for the largest portion of the variance in individual differences in reading ability. The orthographic component represents an additional, independent but secondary component. The phonological component is simply the ability to use knowledge of spelling–sound

Orthographic Learning, Phonological Recoding, and Self-Teaching

43

relationships to identify unfamiliar words. Phonological recoding is, of course, a reading sub-skill, and as such is assumed to reﬂect both instructional/environmental factors such as teaching method and print exposure as well as basic underlying cognitive/linguistic processes such as phonological memory and phonological awareness. The ability to translate print to speech is a necessary but not sufﬁcient condition for orthographic learning. This point is worth reiterating: decoding skill creates opportunities for self-teaching but does not guarantee that orthographic learning will take place. Over and above the ability to decode unfamiliar words, there exist individual differences in the speed and accuracy with which word-speciﬁc (and general orthographic) knowledge is assimilated (Cunningham, Perry, & Stanovich, 2001). The common metric of orthographic ability is typically spelling knowledge (routinely assessed with tasks such as orthographic choice or homophone choice). These measures of what might be termed ‘‘crystallized’’ orthographic ability reﬂect not only the basic visual analysis or visual attention and memory abilities that presumably determine how quickly and accurately orthographic representations are established9 but also instructional/environmental and print exposure variables. The contribution of visual/orthographic ability to the development of word-speciﬁc orthographic representations, however, will depend heavily on the successful operation of the phonological component. Thus, visual/orthographic ability is seen not merely as a second source of variance, but as a secondary source of individual differences in printed word learning, hence the ‘‘phonology-primary/orthography-secondary’’ hypothesis.

III. Empirical Findings A modest number of studies have now been undertaken on the general topic of orthographic learning; some of the later work focused speciﬁcally on the self-teaching hypothesis addressing issues raised in the previous section. Some of the ﬁndings have proven surprisingly robust, others remain intriguing puzzles, while many more are merely research questions awaiting investigation. Although no direct test of the self-teaching hypothesis was carried out until the end of the 20th century (see Share, 9

This question remains the ‘‘black box’’ or perhaps even the bette noire of orthographic learning—the present literature is still in resplendent disarray, (see, e.g., Bosse, Tanturier, & Valdois, 2007; Burt, 2006; Castles & Coltheart, 1996; Castles & Nation, 2006; Cestnick & Coltheart, 1999; Goulandris & Snowling, 1991; Hawelka & Wimmer, 2005; Williams, et al., 2003).

44

David L. Share

1999), a small number of pioneering studies of orthographic learning reported experimental data consistent with the self-teaching hypothesis.

A. THE EARLY PIONEERS The ﬁrst study of orthographic learning among developing readers was undertaken by Hogaboam and Perfetti (1978, Experiment 2). Fourth graders were taught a set of printed pseudowords presented either aurally or in print, then practiced these items at least 15 times over a period of three days. No child was asked to read the printed pseudowords independently— all pronunciations were given and the child asked to repeat them. A day later, both groups named the items that had been seen and pronounced in printed form faster than the items that were only heard and spoken, and this advantage was maintained at 10-week retest. There was no effect of meaning. The orthographic learning effect was further explored in a followup study with third graders who received 0, 3, 6, 9, 12, or 18 exposures to pseudowords presented either aurally or visually (meaning was dropped). Evidence of orthographic learning was obtained at each of the ﬁve exposure conditions although the differences in naming latencies did not reach signiﬁcance possibly owing to the small sample sizes (n ¼ 5). In another trail-blazing study, Ehri and Roberts (1979) taught ﬁrst graders to pronounce eight pairs of printed homonyms (e.g., rows/rose), each practiced 16 times over the course of several sessions either in context or in isolation. The pre- to post-test gains for word reading accuracy and spelling choice suggested that orthographic learning had occurred. A later study (Ehri & Wilce, 1980) replicated and extended these ﬁndings to a set of context-dependent function words (e.g., might, while, must, from, enough). The most comprehensive series of pioneering studies into orthographic learning was carried out in Dutch by Pieter Reitsma. In his ﬁrst experiment, Reitsma (1983a) taught third graders pseudoword names for ﬁctitious animals and fruits. Half of these items were presented auditorily and half both auditorily and visually. Test items were then presented visually in a semantic (animal/fruit) categorization task in which all items appeared six times. Classiﬁcation times for the items not seen in printed form were signiﬁcantly slower only for the ﬁrst three presentations. That is, by the fourth trial, response latencies had effectively converged with the items learned in both spoken and visual form, suggesting rapid orthographic learning in the context of (silent) reading for meaning. In a second study (Reitsma, 1983a, Experiment 2), second graders were ﬁrst familiarized with the spoken forms of a set of pseudowords before being taught to read the printed word. Each item was then practiced (in isolation)

Orthographic Learning, Phonological Recoding, and Self-Teaching

45

either four or eight times. Three days later, target spellings were named signiﬁcantly faster than homophonic spellings but only for the group who practiced reading the targets eight times. A third study (Experiment 3) compared word learning in skilled and unskilled ﬁrst grade readers and an older reading age-matched group of disabled readers. Twenty words judged to be familiar in spoken form but unfamiliar in print were presented in meaningful sentences which were read and reread two, four, or six times over two successive days. Three days later, both the original spellings and homophonic spellings were presented for naming. Both groups of ﬁrst graders read the target spellings more quickly and more accurately for words practiced at least four times. There was no evidence of orthographic learning among disabled readers either in naming speed or accuracy. In two further follow-up studies summarized in Reitsma (1989), normal ﬁrst graders and older reading age-matched disabled readers again practiced reading unfamiliar real words 0, 2, 4, or 6 times. As before, the younger beginning readers, but not the older disabled readers, showed the familiar divergence in target/homophone naming times with increasing practice. In contrast to the earlier study, a naming time difference was already apparent after only two exposures. Response time differences for words and homophones as a function of practice were also correlated signiﬁcantly with scores on a word reading ﬂuency test for both groups of readers. A signiﬁcant correlation between naming time differences and performance on a test of oral pseudoword repetition led Reitsma to conclude that acquisition of word-speciﬁc knowledge depends partly on efﬁcient phonological processing. In a second experiment, normal ﬁrst graders and older disabled readers practiced reading unfamiliar real words 0, 3, 9, or 18 times. In this study, the naming time effect was again evident among the normal readers but only after nine (but not three) exposures. Once again, there was no evidence of orthographic learning among the older disabled readers even after 18 exposures. The effects of phonemic (cross-modal) priming were found to decline with increasing practice for the normal beginners suggesting that their acquisition of word-speciﬁc orthographic information was accompanied by a diminishing reliance on phonology. For disabled readers, the beneﬁts of a related phonemic prime were consistent across all exposures. The ﬁnal ‘‘pre-self-teaching’’ study to be summarized here was carried out by Manis (1985). In the course of four sessions, normal and disabled Grade 5 and 6 children were ﬁrst taught the meanings and pronunciations of low-frequency (real) words varying in regularity and length then presented with their printed forms for pronunciation. All errors were corrected and reread. In three further sessions, children

46

David L. Share

were retrained brieﬂy on the test items before being given two naming tasks (immediate and delayed). By the third test session (after at least eight visual exposures), naming accuracy and speed among the normal readers (but not the disabled readers) had effectively converged on the naming times for a set of high-frequency control words. Declining regularity and length effects also suggested that word-speciﬁc orthographic information had become rapidly assimilated by the normal readers. Collectively, these studies demonstrate impressive convergence—a rare phenomenon in pioneering work yet one indicative of highly robust effects: surprisingly few exposures appear to be sufﬁcient for the acquisition of word-speciﬁc orthographic information among normal readers, but not disabled readers. These training outcomes, furthermore, are consistent with several earlier studies examining the issue of phonological versus visual ‘‘routes’’ among beginning readers which found that even beginning readers are apparently able to read highly familiar words via direct visual recognition with minimal involvement of phonological assembly processes (Barron & Baron, 1977; Condry, McMahon-Rideout, & Levy, 1979; Rader, 1975). Not only did this early research attain consensus regarding the rapidity and durability of early printed word learning, it also suggested that as children acquire word-speciﬁc orthographic representations the role of phonology diminishes, consistent with the item-based transition from word identiﬁcation that is heavily dependent on phonology to more visual-orthographic recognition (Manis’s declining regularity and length effects, and Reitsma’s, 1989, sound priming data, see also Harm & Seidenberg, 2004). The fact that training/learning effects revealed in these investigations were uniformly itemspeciﬁc and did not extend to untrained control items represents strong support for the item-based view of orthographic learning espoused previously. Summing up, all the experimental investigations reviewed here are certainly consistent with the hypothesis that word-speciﬁc orthographic representations are acquired by virtue of the self-teaching opportunities afforded by successful decoding. These data are nonetheless inconclusive on the self-teaching issue for reasons that will become apparent in the presentation of the next set of studies, all of which examined orthographic learning through the lens of the self-teaching hypothesis.

B. STUDIES OF ORTHOGRAPHIC LEARNING WITHIN THE SELF-TEACHING FRAMEWORK The ﬁrst test of the self-teaching hypothesis was a non-experimental longitudinal study carried out by Jorm et al. (1984). In the context of a

Orthographic Learning, Phonological Recoding, and Self-Teaching

47

large-scale longitudinal study of over 500 Australian children followed from the end of Kindergarten—the ﬁrst year of formal reading instruction in the state of Victoria—28 children who had a score of 0 on a test of pseudoword reading but a score of 5 or more on a composite measure of sight word reading were individually matched to children scoring 4 or more on pseudoword reading. To guard against regression to the mean, children were also compared on a second sight word reading measure (on which they had not been matched) but were still found to be well matched. A year later, the ‘‘early decoders’’ had drawn ahead on the second sight word list and were signiﬁcantly superior on the Neale test of word reading accuracy. The two groups were found to diverge even more a year later at the end of Grade 2—the gap in reading age doubling between Grade 1 and Grade 2. (It was unfortunately not possible to include pairs of groups matched on decoding but differing in sight word reading.) Although this study clearly demonstrated that decoding is valuable in reading acquisition, as the authors acknowledged, it provided no direct evidence for the self-teaching interpretation of these data. Direct evidence was only supplied some years later in a series of studies following in the footsteps of the earlier experimental work reviewed above. The ﬁrst experimental study within the self-teaching framework was carried out in Hebrew by Share (1999) employing the same basic experimental paradigm used by Reitsma (1983a) but with some signiﬁcant variations. First, in contrast to prior studies in which target words were presented either in isolation or in isolated sentences (read and often reread), Share aimed to present targets in as natural a setting as possible by constructing short ‘‘stories’’ like the one at the beginning of this chapter. Children were speciﬁcally instructed to read for meaning and told they would be questioned about the content of the story following reading. They were also asked to decide which story they liked the most. A second innovation was the switch to unassisted reading—previous work had either explicitly taught participants the pronunciation of target strings or had corrected errors, often asking the child to repeat the corrected pronunciation. Among young readers of Dutch—a moderately regular orthography— this is not a devastating problem (see Seymour, Aro, & Erskine, 2003) because the vast majority of test items would be expected to be decoded correctly without assistance (probably between 80% and 90%). In the English-language investigations, by contrast, many—perhaps most items (see Seymour et al., 2003)—would probably not have been decoded without help. Thus the generalizability of the English-language ﬁndings to naturally occurring unassisted text reading is in doubt. It is also important to note that in all this work, the experimental procedure obliged the child to decode all target strings. Self-teaching is

48

David L. Share

assumed to operate when a child is independently reading connected text for meaning. In everyday reading therefore, children may (a) choose to ignore unfamiliar words which can often be skipped without penalizing overall comprehension, (b) may guess (correctly or incorrectly) on the basis of prior context and/or prior knowledge, or (c) make uncorrected misreadings. Share (1999) observed that a massive body of evidence shows that normal young readers are able to phonologically recode novel letter strings such as pseudowords when obliged to, but not a single study directly demonstrates that this knowledge is actually applied in independent reading. It is essential therefore to demonstrate that phonological recoding is actually employed even when the child is free to resort to these alternative options. Accordingly, in this ﬁrst direct test of the self-teaching hypothesis, no mention was made of the target words either before or after text reading—even the comprehension questions took care not to mention the target words but focused on the semantic content of the text. Needless to say, all targets appeared in the same typeface as the rest of the text. Children who sought help from the experimenter were encouraged to read the word as best they could by themselves. The Share (1999) study also addressed a second shortcoming of the pioneering word-learning studies reviewed above. Orthographic learning may be attributable to mere visual attention to the target strings rather than to the decoding process per se. Although Reitsma’s (1989) ﬁnding of signiﬁcant correlations between orthographic learning on the one hand and both oral reading ﬂuency and oral pseudoword repetition on the other is certainly suggestive, this alternative hypothesis is difﬁcult to rule out in any of the early studies. The visual inspection hypothesis was directly investigated by Share (1999, Experiments 2, 3, and 4) and, subsequently, by Kyte and Johnson (2006). The targets in Share (1999) were all novel letter strings (pseudowords) embedded in passages such as the one at the beginning of this chapter. These pseudowords appeared either 4 or 6 times per text in two homophonic spellings; half the sample saw one version in the relevant text (e.g., Sloak) while the other half saw the alternate spelling (Sloke). As a further precaution, pre-existing spelling preferences were examined with a spelling preference or ‘‘wordlikeness’’ task administered to a comparable group of children who did not participate in the main study. Only pairs of spellings that demonstrated a balanced preference—close to 50:50—were used. According to the self-teaching hypothesis, even when unassisted, children will apply their knowledge of letter–sound correspondences in order to derive the pronunciation of these unfamiliar words and, if successful, will begin to acquire knowledge of their orthographic forms such that the correct form will be recognized and recalled beyond chance on future occasions.

Orthographic Learning, Phonological Recoding, and Self-Teaching

49

Three days later, orthographic learning was assessed in three ways; orthographic choice (from among four alternatives; the original spelling, a homophonic foil, and two non-homophonic spellings), naming accuracy and speed for the target and homophonic foil, and spelling production. The second graders participating in this experiment correctly decoded the targets (at the consonantal level) on 84% of occasions. This level of consonantal accuracy will seem rather high to readers familiar with the reading accuracy of young readers of English, but this ﬁgure is standard for Hebrew’s highly regular pointed script (see Share & Levin, 1999) or, indeed, for most other regular orthographies (see Seymour et al., 2003). In the 4-alternative orthographic choice task, the original target spellings were correctly selected well beyond the 25% chance level (74%). Naming times for the original spellings were also signiﬁcantly faster for the learned spellings than for the homophonic spellings, with no differences in error rates. Finally, on the spelling measure, the correct spelling was reproduced in its entirety on 41% of occasions compared to only 10% for the homophonic spelling. When spelling was scored on a per-letter basis as opposed to a whole-word basis, correct target letters were reproduced on 67% of occasions compared to only 29% for homophonic letters. The overall pattern of results held for both the 4- and 6-exposure conditions with only small non-signiﬁcant advantages for the 6-exposure condition. These results replicate the pioneering work on orthographic learning and indicate that very few exposures (four or even less) to a target spelling are sufﬁcient for orthographic learning to occur. This study also extends this ﬁnding to unassisted (oral) reading of connected text. Experiment 2 examined the alternative visual exposure hypothesis by presenting target strings under conditions designed to allow visual inspection but minimize phonological processing. Children performed a lexical decision task (deciding whether a letter string is a real word or not— a task known to induce relatively shallow, primarily orthographic processing)—with irrelevant concurrent vocalization (saying the pseudoword dubba over and over) with target exposure limited to 300 ms. Second graders showed themselves quite capable of performing this task (lexical decision accuracy averaged 90%) but it was clear that phonological recoding was only reduced, not eliminated. Brief visual inspection was sufﬁcient to produce reliable orthographic learning in orthographic choice and spelling, but there was no signiﬁcant advantage for original spellings in naming accuracy or times. The levels of orthographic learning were signiﬁcantly inferior to the levels observed in Experiment 1 with unlimited exposure to orthographic targets. There still remained the possibility that the attenuated but signiﬁcant orthographic learning in this second experiment might be attributable to

50

David L. Share

visual inspection alone but diminished owing either to brief visual exposure and/or presentation without supporting context. Experiment 3 evaluated this possibility by asking children to name (i.e., phonologically recode) the same brieﬂy presented items. Comparison of the outcomes of Experiments 2 and 3—orthographically relevant pronunciation (Experiment 3) versus irrelevant articulation under the same conditions (brief decontextualized exposure)—provided a relatively clean test of the contribution of phonology to orthographic learning. The results showed that second graders, once again, proved themselves up to the challenge—correctly pronouncing the brieﬂy exposed targets on 72% of trials compared to 86% for the corresponding (6-exposure) condition in the non-time-limited story-reading condition in Experiment 1. Phonological recoding of the same targets under similar conditions to Experiment 2 produced signiﬁcantly greater orthographic learning than orthographically irrelevant vocalization in Experiment 2, pointing to a unique contribution of phonological recoding and indicating that the results of Experiment 1 cannot be attributed solely to visual exposure, although some role for pure visual attention per se cannot be ruled out by this pair of studies. As pointed out by Kyte and Johnson (2006), however, the comparison between Experiments 2 and 3 is confounded by the different task requirements of reading aloud and lexical decision. In a fully within-participant design, Kyte and Johnson (2006) had fourth and ﬁfth graders perform a lexical decision for real words and pseudoword targets under two conditions: a read-aloud condition designed to maximize phonological recoding of target strings and a concurrent articulation condition designed to minimize phonological recoding yet allowing sufﬁcient visual-orthographic processing to occur for lexical decision. Pseudoword targets appeared six times in two alternate spellings: half the sample saw one spelling and half the other. Brief (and masked) exposures were also used (400 ms). Lexical decisions were very accurate and very similar in the two conditions: pronunciation accuracy in the naming condition was 95%. A day later, orthographic choice revealed a signiﬁcant advantage for the read-aloud condition with condition accounting for around 20% of the variance. Spelling produced stronger effect sizes, ranging from 37% to 45%. The post-test naming latency data revealed a signiﬁcant advantage of condition for item means but not participant means. Kyte and Johnson’s (2006) more rigorous design provides strong support for the role of phonology in orthographic learning. Returning to the question of the contribution of ‘‘pure’’ visual exposure to orthographic learning, Share (1999) conducted a fourth experiment using non-alphabetic symbol strings (e.g., ?+$) that offer no possibility of

Orthographic Learning, Phonological Recoding, and Self-Teaching

51

recoding. These strings were presented in a combined letter search and string length judgment task in an effort to approximate the item (letter) level and string (word) level processing assumed to occur in phonological recoding. Three days later, orthographic choice indicated that target strings were recognized correctly on 33% of occasions. This ﬁgure was not signiﬁcantly superior to the foils designed to be parallel to the homophone foils in Experiment 1 (28%), but it was signiﬁcantly beyond chance (25%). Clearly, however, the magnitude of this effect, if indeed reliable, signiﬁes only a small proportion of the overall learning effect observed following phonological recoding. This result accords with other studies underscoring the extraordinary difﬁculty involved in memorizing strings of nonalphabetic symbols containing common visual elements (e.g., Ehri & Wilce, 1987; Jorm, 1981). The basic ﬁndings from Share’s ﬁrst experiment—with targets appearing in short texts—were replicated and extended to the English language by Cunningham et al. (2002). Using the same procedure and materials adapted from Hebrew to English with targets appearing 6 times in each story, overall decoding accuracy for these second graders was 74%. Replicating Share (1999, Experiment 1), the original target spellings were correctly recognized more often, reproduced with greater accuracy and named more rapidly than homophonic foils. Conﬁrming a key prediction of the self-teaching hypothesis, Cunningham et al. reported a signiﬁcant correlation (r ¼ .52) between orthographic learning and the number of target words correctly decoded during story reading. Kyte and Johnson (2006) also reported a signiﬁcant positive correlation (.40) between target decoding in their read-aloud condition and their orthographic learning composite (combining spelling, orthographic choice, and naming) for their ﬁfth and sixth graders. It should be kept in mind, however, that target decoding accuracy was at near-ceiling levels (averaging 95%) in these older more proﬁcient readers. In a later study by Cunningham (2006) among ﬁrst graders, the correlation was .66. To complete the picture, Share (2007) found a correlation of .43 between Hebrew target word decoding (which averaged 89% in his Grade 3 sample) and a composite measure of orthographic learning (orthographic choice and spelling), which remained signiﬁcant (.31) even after controlling for age, general intelligence, and two measures of phonological recoding ability (pseudoword naming and Olson’s phonological choice). This latter ﬁnding is especially important as it shows that these decoding-orthographic learning correlations do not merely reﬂect a general association between decoding and orthographic skills but an item-speciﬁc relationship as speciﬁed by the self-teaching hypothesis.

52

David L. Share

1. Orthographic Learning is Rapid and Durable The evidence reviewed thus far suggests that orthographic learning is both rapid and highly robust. Only a few exposures seem sufﬁcient to produce detectable orthographic learning; indeed, four exposures have been shown to produce reliable outcomes in a number of studies (Ehri & Saltmarsh, 1995; Reitsma, 1983a, 1989; Share, 1999), three exposures too (Hogaboam & Perfetti, 1978), but results for two exposures are mixed (Reitsma, 1983a, 1983b, 1989). These data suggest a ‘‘threshold’’ model of orthographic learning with signiﬁcant learning occurring only after some threshold level of experience. In contrast, connectionist models (e.g., Harm & Seidenberg, 1999, 2004; Plaut et al., 1996; Woollams et al., 2007) as well as more general instance-based learning theories (e.g., Logan, 1988) predict signiﬁcant learning from the very ﬁrst trial; indeed the standard connectionist learning algorithm (the delta rule) predicts that the most ‘‘powerful’’ learning trial is necessarily the ﬁrst trial owing to the fact that changes in connection weights are directly proportional to the magnitude of the discrepancy between the current (initial) values and target values. The tantalizing possibility that a single learning trial is sufﬁcient to produce signiﬁcant orthographic learning was investigated by Share (2004). In this study targets were presented (again in meaningful text) either once, twice, or four times. This study also pursued the question of durability by comparing orthographic learning after 3, 7, and 30 days. According to Hogaboam and Perfetti (1978), orthographic learning is maintained for at least 10 weeks! Hebrew-speaking third graders participated in a fully crossed 3 (exposures) by 3 (post-test intervals) design. Even a single exposure produced signiﬁcant post-test learning, with 61% success spelling the critical (homophonic) letters where chance is 50%, z ¼ 3.30. And remarkably, this newly acquired orthographic knowledge was retained for up to 30 days after exposure (collapsing across all 3 exposures)—59%, z ¼ 2.66. The orthographic choice data reproduced this pattern of outcomes. There was a small but non-signiﬁcant 16 ms advantage in naming latencies for targets compared to homophones. These data clearly support the item-speciﬁc logistic learning functions posited by the connectionist models. Furthermore, they subsume the printed word-learning data within the broader context of skill learning as discussed previously (Logan, 1988, 2002; Newell & Rosenbloom, 1981). Nation, Angells, and Castles (2007) also addressed the question of rapidity and durability of orthographic learning in their (Year 3 and Year 4) English-speaking sample. Of special interest here was their attempt to replicate the single-exposure ﬁnding of Share (2004, Experiment 1). Their study compared 1, 2, and 4 exposures and evaluated retention after 1 or 7 days. Pseudoword targets were presented for oral reading either in story

Orthographic Learning, Phonological Recoding, and Self-Teaching

53

context or without context: overall decoding accuracy averaged 78%. Orthographic choice was the sole measure of orthographic learning. Outcomes indicated superior recognition at the 1-day delay compared to the 7-day delay, and 4 exposures was superior to a single exposure (with intermediate outcomes at 2 exposures). Of primary interest here are the ﬁndings for single-exposure learning. Although the overall result was (just) signiﬁcantly beyond the 25% chance level, it is apparent from Table I (Nation et al., 2007, p. 76) that only the 1-day result (41%) but not the 7-day outcome (30%) was beyond chance (25%) performance. This result raises doubts about the reliability of single-trial orthographic learning. Nation et al., like others, found signiﬁcant albeit modest correlations (.38, 1-day; .28, 7-day) between target decoding accuracy and orthographic learning among these Year 3 and Year 4 children. Examining the relation at the item level, Nation et al. used logistic regression to show that orthographic learning was greater for items decoded correctly, at both the 1- and 7-day interval. In contrast, a within-item regression showed no effect of decoding accuracy on orthographic choice at 1 day and a marginal effect at 7 days. It is important to note, however, that the dependent measure was not a continuous variable but a forced choice between four alternative spellings with chance performance expected to average 25%. Whereas performance on orthographic choice for correctly decoded items was clearly well above chance (51%; 1-day, and 43%; 7-day)—indicating reliable orthographic learning—performance on targets decoded incorrectly was just above chance at the 1-day interval—37% ( p ¼ .035, 1-tailed) and unequivocally at chance (26%) at 7 days. Considering the repeated ﬁnding that orthographic learning is not a transitory 1-day phenomenon but a long-lasting effect (over weeks, Bowey & Muller, 2005; Landi et al., 2006, and even months, Hogaboam & Perfetti, 1978; Share, 2004), it seems fair to conclude that the Nation et al. data demonstrate reliable orthographic learning only when targets are correctly decoded. And this is surely strong item-level support for the self-teaching hypothesis. A further point needs to be made here too. As already emphasized previously, the self-teaching hypothesis argues that successful decoding does not guarantee orthographic learning—but only provides opportunities for orthographic learning. Other factors (see the earlier phonologyprimary/orthography-secondary section) will also inﬂuence the assimilation of new word-speciﬁc orthographic information. In remarking that ‘‘there were many instances where items were decoded correctly but not recognized y’’ Nation et al. are misrepresenting the self-teaching hypothesis to imply a perfect one-to-one (and causal) correlation between target decoding and orthographic learning. As emphasized in this chapter and in previous discussions (Jorm & Share, 1983; Share, 1995), decoding is a

54

David L. Share

necessary but not sufﬁcient condition for orthographic learning. It is therefore unreasonable to expect that all targets decoded successfully will result in correct orthographic choices. However, the self-teaching hypothesis clearly predicts signiﬁcant (i.e., beyond chance) and superior orthographic learning for correctly decoded targets. In a later section on individual differences in orthographic learning, I return to the intriguing question of what else accounts for success in orthographic learning beyond decoding.

2. Pseudowords and Real Words Although pseudowords guarantee that a printed word is visually unfamiliar, there are several reasons why it may by unwise to generalize the results from pseudoword learning to naturalistic reading of real words in connected text. First, the availability of phonological knowledge in the form of well-speciﬁed phonological representations may permit ‘‘early closure’’ of the decoding process (consider the word teleph) especially if this word is encountered in meaningful (although not necessarily predictive) contexts. This would be expected to reduce decoding exhaustiveness thereby diverting attention from orthographic detail. Offsetting these ‘‘orthographic’’ disadvantages, the availability of spoken forms should help resolve any decoding ambiguities particularly for irregular words (see Tunmer & Chapman, 1998, 2006) leading to fewer word identiﬁcation failures and improved orthographic learning. These considerations are maximally applicable to beginning readers, for whom most visually novel words are likely to be familiar in spoken form. Later, of course, an increasing number of new words will also be unfamiliar in spoken form—hence phonological learning will accompany orthographic learning as is the case for pseudowords. In most of the studies discussed previously target words were pseudowords but for the few studies using real words (Ehri & Roberts, 1979; Ehri & Wilce, 1980; Reitsma, 1983a, Experiment 3, 1989) the outcomes for orthographic learning seem indistinguishable from the pseudoword data. Subsequently, Cunningham (2006) examined orthographic learning of real words in unassisted (oral) reading of connected text in normal ﬁrst grade English-speakers/readers. Both the original spelling ( piece) and a pseudo-homophonic spelling ( peece) were presented. Prior knowledge of the reading accuracy of these items was evaluated on a comparable sample of ﬁrst graders—only words that 95% of the pilot sample could not identify correctly in print were included in the experiment. (It should be noted that, for ﬁrst grade readers, these items are clearly challenging as regards spelling–sound relationships—bored, chews, course, groan, pause,

Orthographic Learning, Phonological Recoding, and Self-Teaching

55

piece, prince, and thirsty.) Targets appeared six times either in context (short stories) or in a decontextualized condition (i.e., the same short stories scrambled). There was clear evidence of above-chance orthographic learning on the orthographic choice post-test with similar results for both context and scrambled conditions. The spelling data were less clear although Cunningham noted that this task proved very difﬁcult for these beginner readers/writers. To summarize, the rapidity and durability of orthographic learning seems to apply to both real and pseudowords, although the underlying processes may not be identical.

3. Silent and Oral Reading Bowey and Muller (2005) raised an important concern regarding prior work using oral reading. Even when a child engages in unassisted oral reading before an adult there is an implicit obligation to read all the text. Truly independent reading is a private activity not performed either in a dyadic or public setting and furthermore silent rather than oral—at least among young readers who are no longer ‘‘beginners.’’ Truly silent reading among beginners is questionable (Wright, Sherman, & Jones, 2004) but among older children, from around third grade at least, silent reading is the norm and therefore represents the most ‘‘authentic’’ mode of independent reading. Accordingly, Bowey and Muller (2005) examined self-teaching during silent reading in a large sample of third graders. Targets (and homophonic foils) were presented either 4 or 8 times with post-test orthographic learning evaluated either immediately or at a 6-day delay. Bowey and Muller elegantly matched the targets and homophones for grapheme–phoneme correspondences by switching component vowel graphemes—( ferd/surn and furd/sern). Orthographic learning was evaluated using list-based naming times (target lists versus homophone lists) as well as 3-choice orthographic choice ( furd/ferd/fard ). Target lists were read faster than non-target lists (with no difference in accuracy), and orthographic choices for the original spelling far outnumbered homophonic foils. Bowey and Muller argued that the presence of rapid orthographic learning not only in the orthographic choice task but also in the list-reading task indicated that rapid orthographic learning must have involved phonological recoding during silent reading of stories. ‘‘Phonological recoding can be inferred most safely when children name target nonwords faster than homophone alternatives. Children can do so only if they have already phonological recoded them within story reading and established functional orthographic representations that amalgamate graphemic and phonological information y Our ﬁnding that target nonwords were read

56

David L. Share

faster than homophones thus provides a strong case for self-teaching through phonological recoding’’ (p. 218). In a follow-up study of silent reading with third graders, using immediate and 2-day post-test delays, Bowey and Miller (2007) again found faster listnaming times for targets compared to foils (with no difference in accuracy), but while orthographic choice at immediate delay was highly signiﬁcant, the 2-day delay effect was only marginal. De Jong and Share (2007) also found evidence of signiﬁcant orthographic learning in silent reading in a Dutch study designed to directly compare the two reading modes—oral and silent—in a within-participants design. They predicted superior orthographic learning in oral compared to silent reading owing to the greater involvement of phonological processes in oral reading. Third graders read short texts with target pseudowords appearing either twice or six times: orthographic choice, spelling and naming (vocalization onset latencies for individual items) were evaluated three days later. Replicating Bowey’s work, there were signiﬁcant and similar levels of orthographic learning as assessed by orthographic choice and spelling in both oral and silent reading. An advantage for naming latencies was obtained for targets in oral but not silent reading. De Jong and Share concluded that although there was some evidence for stronger orthographic learning in oral reading, the differences were not profound. The reader may recall that Reitsma’s (1983a) very ﬁrst study in orthographic learning was a semantic categorization task (in a third grade sample)—a silent reading task in which items appeared six times and that revealed reliable orthographic learning after four trials. Putting all these three studies together makes a good case for self-teaching in silent reading.

C. INDIVIDUAL DIFFERENCES IN SELF-TEACHING, DYSLEXICS, AND OTHER POOR READERS As with the enormous variability in acquired skills such as reading, orthographic learning also encompasses profound individual differences. Although some children just keep decoding new words laboriously over and over as if seen anew for the ﬁrst time, for others orthographic learning seems virtually instantaneous—the switch from effortful phonological recoding to seemingly direct orthographically-based retrieval appears to occur in a single encounter. Because cultural tools such as reading lack a dedicated brain basis anchored in evolution, the topic of individual differences is probably as important as the ‘‘normative’’ architecture of the system. Perhaps for this

Orthographic Learning, Phonological Recoding, and Self-Teaching

57

reason, in almost all research in this ﬁeld—from the earliest pioneering efforts through to contemporary studies—the issue of individual variability has been high on the agenda. This question has been pursued in a variety of research designs including group-wise comparisons of high- and low-ability readers (or diagnosed dyslexics) as well as correlational approaches examining the full spectrum of variability. Turning ﬁrst to comparisons between designated groups of good and poor (or dyslexic) readers, Hogaboam and Perfetti’s (1978) second experiment found signiﬁcant orthographic learning but no group-by-condition interactions when comparing above-average Grade 4 readers to below-average readers. The same result emerged in Reitsma’s (1983a, Experiment 3) study of ﬁrst graders split into more- and less-skilled sub-groups which found signiﬁcant overall orthographic learning but no group interactions. In contrast, the ﬁndings for disabled readers tell a very different story. Reitsma also included an older group of third grade disabled readers (diagnosed dyslexics) who were matched to the ﬁrst graders. Among these disabled readers, he found no evidence of orthographic learning (even after 6 exposures) either in naming speed or accuracy, although sensitivity to word-speciﬁc information was evident in a set of very high-frequency words that were read signiﬁcantly faster than their corresponding homophonic (mis)spellings. In the later studies summarized by Reitsma (1989) using 6 (Experiment 1) and 18 exposures (Experiment 2), disabled readers (but not normal readers) once again showed no evidence of orthographic learning. A similar result for disabled readers was also obtained by Manis (1985). By the third training session (after six visual exposures), naming times (and errors) for the normal readers (but not the disabled readers) had converged on the naming times for a set of high-frequency control words. Error and latency data clearly indicated that learning had occurred among the disabled-reader group, but not at the same pace as the normal readers. As in Reitsma’s work, which found evidence of orthographic knowledge for familiar (high-frequency) words, it seems fair to conclude that orthographic learning among disabled readers is not entirely absent but much slower or less efﬁcient. To date, no study has yet pursued the issue of how many exposures (or the type and context of exposure) are required for disabled readers to demonstrate reliable orthographic learning of the type that occurs so rapidly among most readers. Further evidence for deﬁcient orthographic learning by disabled readers was reported by Ehri and Saltmarsh (1995) who taught skilled and lessskilled ﬁrst graders and older disabled readers to read simpliﬁed phonetic spellings for a set of real words (e.g., mesngr, stupd ). Following Reitsma (1983a), original and altered spellings were presented in a naming task 3

58

David L. Share

days later. Altered spellings included both phonetic equivalents (e.g., cradl/ kradl ) and phonetically close but non-equivalent spellings (e.g., bambu/ pambu, stupd/stup). All target spellings were directly taught by the experimenter who also explained the meaning of each word. The test list was then practiced between 10 and 12 times over consecutive days, with all mispronunciations corrected and reread. Skilled ﬁrst grade readers required only four practice trials to achieve errorless performance on the entire list, but the other two groups each required over twice this number of trials. Three days later, both non-disabled groups, but not the older disabled readers, read the original spellings signiﬁcantly faster than the fully homophonic spellings. For phonetically divergent spellings, all three groups were signiﬁcantly faster on the original spellings. All these data are consistent with the self-teaching hypothesis, which holds that poor decoding skill should impair orthographic learning. Hence disabled readers who, as a group, have well-known decoding deﬁcits, would be expected to demonstrate inferior orthographic learning. In contrast, it is often claimed that disabled readers have ‘‘compensatory’’ abilities in nonphonological aspects of reading such as visual-orthographic skills. Supporting the notion of compensatory processing is the common ﬁnding that orthographic knowledge is often less impaired (relative to reading age-matched controls) than phonological skills—primarily pseudoword naming (see Share, 1995 for a review). If this is more than just a reﬂection of greater print exposure among the older more experienced readers and constitutes an acquired processing strategy, then orthographic learning tested in an experimental paradigm in which print exposure is a constant should be better than expected on the basis of decoding ability alone. This compensatory hypothesis (and several related issues) were examined in a study by Share and Shalev (2004) comparing the self-teaching of four groups of children: dyslexics in Grades 4, 5, and 6, non-dyslexic gardenvariety poor readers, age-matched normal readers, and a younger group of normal readers matched to the garden-variety poor readers on both reading and mental age. Decoding deﬁcits were expected to impair identiﬁcation of target words leading to inferior orthographic learning, but, if the compensatory hypothesis is correct, disabled readers’ orthographic learning was predicted be somewhat less impaired than non-disabled readers. Consistent with their pre-existing levels of reading ability, target decoding levels were highest for the non-disabled (normal) readers, lowest for the dyslexics with the garden-variety sandwiched in between (see Figure 1). Contrary to the English and Dutch data, however, the Hebrew data revealed lower but reliable levels of orthographic choice and spelling among the disabled readers that were closely tied to levels of target decoding (see the three near-parallel lines in Figure 1). A non-signiﬁcant repeated

Orthographic Learning, Phonological Recoding, and Self-Teaching

59

100 Control GV Dyslexic

Percentage %

90

GV control

80

70

60

50 Decoding

Choice

Fig. 1. Target decoding and orthographic choice (in percentages) in four groups of readers [chronological level (CA)] controls, garden-variety poor readers, dyslexics, and garden-variety [reading level-mental age (RL-MA)] controls (reprinted with permission from Share & Shalev, 2004).

measures group-by-measures interaction conﬁrmed the parallelism portrayed in Figure 1. Even the main effect of group on both target decoding and orthographic learning vanished after overall reading level was controlled, thereby conﬁrming the picture of developmental delay rather than deviance. (The highly divergent result for the younger reading-agematched group is discussed subsequently in the section on early onset.) Outcomes also indicated that IQ-discrepant and non-discrepant poor readers differed quantitatively but not qualitatively. This remarkably close match between overall (group-wise) levels of target decoding and orthographic learning reinforces the experimentally induced differences reported in Share (1999) and also Kyte and Johnson (2006). As regards the ‘‘compensatory processing’’ hypothesis, there was no hint of evidence that either group of poor readers was able to compensate for poor decoding via ‘‘superior’’ orthographic skills. Experiential factors such as print exposure rather than inherent or cultivated processing skills may be the source of the orthographic ‘‘advantage’’ enjoyed by poor readers in reading-age comparisons. The next group of studies examined the individual-difference issue from a somewhat different angle: each was speciﬁcally motivated by the

60

David L. Share

self-teaching hypothesis and each used a correlational or regression-based approach surveying the full range of ability rather than selected groups or sub-groups differing in overall reading ability. The ﬁrst study exploring variability in orthographic learning was Cunningham et al. (2002). In addition to the signiﬁcant correlation (r ¼ .52) already mentioned between second graders’ target decoding and composite orthographic learning, Woodcock Word Attack skills also correlated .58 with orthographic learning. Rapid serial naming (RAN) correlated weakly (.35) but not general cognitive ability (Raven, Peabody and digit-span). Hierarchical regression also established that after target decoding accuracy was partialled out, neither cognitive ability nor RAN contributed additional unique variance to orthographic learning. Orthographic knowledge (Olson’s spelling choice test), in contrast, contributed a signiﬁcant and substantial 20% of variance. Cunningham (2006) also examined sources of variability in orthographic learning in her study of ﬁrst graders: predictors included rapid naming, general cognitive ability, and orthographic knowledge (a composite of spelling choice—take-taik, homophone choice—Which is a flower; rows/ rose, and wordlikeness—Which looks most like a word; fage/fayj). Simple correlations with orthographic learning were again found for Woodcock Word Attack but not RAN or general intelligence. Prior orthographic knowledge again accounted for additional unique variance in orthographic learning (11%) after the contribution of target decoding accuracy, but not rapid naming or general cognitive ability. In their study of self-teaching in silent reading, Bowey and Miller (2007) sought to evaluate Cunningham et al.’s (2002) individual-difference ﬁndings. Using a similar set of predictors, this study examined variability in orthographic learning (assessed by orthographic choice alone) immediately following exposure and again after a 2-day interval. As noted earlier, the relevant orthographic choice outcomes were highly signiﬁcant at the immediate post-test, but the data for the 2-day delay were inconclusive—a ﬁnding at variance with almost all the studies reviewed here and one that places a question mark over the dependent variable. Turning to the results, Bowey and Muller found a signiﬁcant correlation (.42) between general phonological recoding efﬁciency ( pseudoword naming accuracy and speed combined) and orthographic choice, but no evidence for a relation between rapid naming and orthographic learning—a purely accuracy measure in this study. A signiﬁcant but weak (.30) association between orthographic knowledge and orthographic learning was also obtained, but after partialling out general phonological recoding efﬁciency, orthographic knowledge only explained a marginal 6% of the variance. A similar re-analysis of Cunningham et al.’s data (controlling for Word

Orthographic Learning, Phonological Recoding, and Self-Teaching

61

Attack) also revealed a weak and marginal association between orthographic knowledge and orthographic learning. The contribution of RAN to orthographic learning was also evaluated in a study of the role of decoding ﬂuency in orthographic learning by Lurie and Share (2007) presented in fuller detail subsequently. This study examined the relations between several phonological processing tasks (RAN included), target decoding accuracy and speed, and orthographic learning (a composite of spelling production and orthographic choice) in a sample of Hebrew-speaking third graders. Instead of the conventional RAN, which usually requires RAN of single digits and letters (and sometimes pictured objects and colors), this study presented sub-syllabic CV units for rapid naming—combinations of consonantal letter and appended vowel diacritic that are the basic building blocks of printed Hebrew words (see Share & Blum, 2005). Whereas non-RAN tasks such as pseudoword repetition (accuracy) correlated more with target decoding accuracy than target decoding speed, Lurie and Share’s RAN measure correlated more with decoding speed than accuracy. Turning to the decoding-orthographic learning relation, the RAN task correlated signiﬁcantly with orthographic learning and, furthermore, this relation was sustained even after partialling out target decoding accuracy. However, when target decoding speed was partialled out, the RAN measure was no longer signiﬁcant, indicating that the RAN task contributes to orthographic learning via its contribution to decoding speed. This ﬁnding supports the view that RAN contributes unique variance to orthographic learning (Bowers & Wolf, 1993), but suggests that this contribution derives not from non-phonological sources of variance but from the speed dimension of phonological processing not tapped in traditional un-timed measures of phonological processing. The Lurie and Share data afﬁrm that for self-teaching to operate effectively, the constituent elements of novel letter strings must not only be decoded accurately but also sufﬁciently fast for the establishment of word-speciﬁc orthographic representations. Share (2007) followed up the Cunningham et al. (2002) study with a wideranging battery of predictors aimed at comparing sources of individual differences in orthographic learning in two scripts—shallow pointed Hebrew and the relatively deeper unpointed script. One of the aims of this study was to test the phonology-primary/orthography-secondary hypothesis outlined previously. In addition to multiple measures of phonology and orthography, potential predictors of orthographic learning included general intelligence, working memory, cognitive style (Kagan’s test of reﬂection-impulsivity—the Matching Familiar Figures Test), morphological knowledge, semantics, and syntax. Eighty third graders proﬁcient in

62

David L. Share

reading both pointed and unpointed text read a series of short passages (half pointed, half unpointed) each containing a target pseudoword appearing four times. Post-test measures (orthographic choice and spelling production) were administered one week later. Levels of consonantal decoding accuracy for the two types of targets (pointed and unpointed) as well as post-test orthographic choice and spelling were very similar. To test the phonology-primary/orthographysecondary hypothesis, hierarchical regression was employed by ﬁrst entering age and general intelligence followed by a block of phonological variables (pseudoword naming accuracy, time-limited phonological choice, target decoding accuracy, phonological awareness, and oral pseudoword repetition) and a block of orthographic variables (Olson’s orthographic choice, Kleiman’s time-limited word boundaries measure, WISC-III Symbol Search, and Stanford-Binet Bead Memory) each entered, in turn, at Steps 2 and 3 (see Figure 2). As can be seen in Figure 2, two mirror images emerged; the phonological block was the strongest predictor of orthographic learning in shallow pointed script (see the middle segment of the leftmost column labeled with the variance ﬁgure of 19%), but the visualorthographic block the foremost predictor of unpointed orthographic learning—the middle segment (20%) of the rightmost column. Thus, the phonology-primary/orthography-secondary hypothesis was upheld in the case of Hebrew’s highly regular fully voweled (pointed) orthography, but rejected for the deeper unpointed text. Replicating the ﬁndings of Cunningham (Cunningham, 2006; Cunningham et al., 2002) and Bowey and Miller (2007), conventional RAN digits and letters did not contribute reliably to orthographic learning, nor did meaning or syntax although it should be kept in mind that all targets were pseudowords. On the positive side, there was evidence for a role of working memory and cognitive style (reﬂectiveness–impulsiveness). An interesting sidelight in this individual-difference study was the examination of predictors of target decoding accuracy. Consistent with the orthographic depth hypothesis (Frost, 2005; Katz & Frost, 1992), the pronunciation of pointed pseudoword targets in text was associated only with sub-lexical phonology and working memory, whereas the decoding of the deeper unpointed targets was related to a wide range of sub-lexical, lexical, and supra-lexical factors (phonology, working memory, visualorthographic processing, morphology, semantics, and syntax). Summing up, evidence from multiple independent studies converges in showing that differences between readers in orthographic learning, as posited by the self-teaching hypothesis, are partly dependent on the success of phonological recoding. Beyond this, some intriguing ﬁndings have been uncovered but the search for sources of variability in orthographic learning

Orthographic Learning, Phonological Recoding, and Self-Teaching

63

35

30 4% 25

2% 12% 13%

20 19%

20%

15 11% 9%

10

5 7%

7%

6%

6%

0 shallow pointed

deep unpointed

Visual-orthographic block

Phonological block

Age & IQ

Fig. 2. Variance (R2 change) in shallow pointed and deep unpointed orthographic learning explained by phonological and visual-orthographic variables (after controlling (Step 1) for age and general intelligence). In each tri-partite column, the first segment represents the variance accounted for by age and IQ (Step 1), the second (i.e., middle segment) is the variance attributable to the block of variables (either phonological or visual-orthographic) entered at Step 2, and the topmost segment the third and remaining block’s Step 3 variance.

has clearly only just begun and, like the whole question of orthographic ‘‘processing’’ or ‘‘ability,’’ is still largely a black box. Ultimately, there will surely be developmental and script-dependent differences.

D. EARLY ONSET How early does self-teaching begin? Indirect support for the early onset hypothesis can be seen in the fact that even beginning readers display a certain level of word-speciﬁc knowledge. For example, Reitsma (1983b, Experiment 3) found that spellings of familiar words were read faster than homophonic spellings. And there is even evidence for knowledge of general orthographic conventions (Cassar & Treiman, 1997;

64

David L. Share

Pacton et al., 2001) in that even beginners can recognize illegal spellings such as ckat, mmom. As regards the early acquisition of orthographic knowledge, Reitsma (1983a, Experiment 3) found signiﬁcant orthographic learning (manifest in naming times) among ﬁrst grade Dutch readers after only four exposures. In the ﬁrst study reported in his 1989 chapter, Reitsma again found signiﬁcant learning among ﬁrst graders (even at 2 exposures); the second study found learning among ﬁrst graders but only after 9 exposures. Three studies by Ehri and coworkers (Ehri & Roberts, 1979; Ehri & Saltmarsh, 1995; Ehri & Wilce, 1980) also obtained reliable orthographic learning among ﬁrst grade English readers. As discussed previously, however, all these studies either taught children the target words or corrected any misreadings—often asking children to repeat the correct pronunciation— something foreign to independent reading. Share (2004, Experiments 2 and 3) examined the issue of self-teaching in beginning readers’ unassisted (oral) reading of Hebrew’s highly regular pointed Hebrew. At the outset it needs to be stressed that pointed Hebrew, in contrast to the moderately regular Dutch spelling, is perfectly regular as far as grapheme–phoneme correspondence is concerned. Hebrew also has a relatively simple syllable structure with mostly open CV syllables and few consonant clusters. Share’s Grade 1 study employed a fully repeated-measures design with two and four exposures and 3-day and 7-day post-test intervals. Consonantal decoding of the pseudoword targets was quite accurate (93%). Nevertheless, there was a surprisingly uniform lack of evidence for reliable orthographic learning—all three measures, spelling, naming, and orthographic choice were close to chance. It must be acknowledged that these ﬁndings were derived entirely from pseudoword stimuli, whereas the English and Dutch studies employed real words or pseudohomophones of real words judged to be familiar to children in spoken but not written form. A second study therefore used both types of items, increasing the number of target exposures to 4 and 8 (Share, 2004, Experiment 3). All post-testing was conducted after an interval of seven days. The real words selected to be familiar in spoken but not printed form included words ranging in length from two to four syllables (3–6 consonant letters). The spelling data were consistently at chance and the orthographic choice data (a composite of spelling and orthographic choice) revealed only ‘‘glimmerings’’ of orthographic learning with some results marginally signiﬁcant, others non-signiﬁcant. Additional independent ﬁndings from the Share and Shalev (2004) study of self-teaching among older disabled readers corroborated this unexpected result. These data were collected from a group of 20 children tested early

Orthographic Learning, Phonological Recoding, and Self-Teaching

65

in their second school year (Grade 2) who were selected as reading-level controls for older disabled readers. Normal Grade 2 readers were matched to the older readers on speed and accuracy of reading both real words and pseudowords. These young Grade 2 readers decoded the targets at the same level of accuracy as did the Share (2004) Grade 1 children (93%), yet orthographic choices failed to exceed chance levels (57%). No effects were found on either naming accuracy or naming times, and target letter spelling was also very close to chance (52%) (witness the egregiously non-parallel line in Figure 1). Note, however, that all other groups in the Share and Shalev (2004) study demonstrated reliable and signiﬁcant orthographic learning despite the fact that decoding accuracy was somewhat lower (80% for dyslexics and 88% for garden-variety poor readers). The presence of statistically signiﬁcant crossover interactions between decoding success, on the one hand, and orthographic choice and spelling on the other, conﬁrmed that these normal Grade 2 readers were unable to recognize or recall orthographic detail despite high levels of decoding success. Taken together, these three data sets suggest that the early onset hypothesis must be rejected in Hebrew’s highly regular pointed script. The evidence of rapid early orthographic learning reported in the deeper orthographies of Dutch and English does not appear to generalize to the shallow pointed Hebrew script. Share and Shalev (2004) advanced two alternative hypotheses to account for the negligible orthographic learning among Hebrew ﬁrst graders. First, it was suggested that the near-perfect one-to-one letter–sound relations in Hebrew may induce a highly ‘‘bottomup’’ letter-by-letter (‘‘surface’’) decoding strategy which is relatively insensitive to higher-order word-level orthographic information. Consistent with this possibility, readers of highly regular orthographies such as Italian and German are often reported to display more exhaustive letter-by-letter decoding (Landerl, 2000; Thaler et al., 2004; Thorstad, 1991)—an observation supported by eye-movement studies (e.g., De Luca et al., 2002; Hutzler & Wimmer, 2004) and brain imaging work (Paulesu et al., 2000). Further supporting evidence comes from a variety of ﬁndings reviewed by Ziegler and Goswami (2005) indicating that relative to readers of more transparent orthographies, beginning readers of English are more sensitive to higher-order (larger ‘‘grain size’’) information—at the level of letter patterns and whole words. The orthographic depth hypothesis (Frost, 2005; Katz & Frost, 1992) also asserts that English’s deep orthography obliges the reader to look beyond low-level phonology and consider higherorder regularities that are often word-speciﬁc. A second related explanation for the unexpected divergence of the Hebrew ﬁndings relates to the often-neglected dimension of decoding speed and efficiency. The characteristically laborious, letter-by-letter decoding

David L. Share

66

reported among beginning readers of shallow orthographies may simply be too slow to support orthographic learning. The role of decoding ﬂuency as a contributor to orthographic learning over and above decoding accuracy was directly addressed in a study by Lurie and Share (2007). In a sample of 42 third grade Hebrew readers, accuracy of (pseudoword) target decoding correlated only weakly with orthographic learning (r ¼ .24), but decoding times as measured from target presentation onset through to pronunciation offset were much more strongly (and signiﬁcantly) related to orthographic learning (r ¼ .52). This ﬁnding coheres with a large number of studies indicating that reading speed rather than accuracy (which reaches ceiling levels very early in regular orthographies) is a more potent discriminator of developmental and individual differences in regular orthographies (de Jong & van der Leij, 2003; Wimmer, 1993). This study also examined the relation between decoding speed and orthographic learning in light of the evidence showing that the phonological store (i.e., ‘‘phonological loop’’) in working memory is time-limited, speciﬁcally, that memory span is limited to the number of items that subjects can articulate in a 2-second time window (Baddeley, Thomson, & Buchanan, 1975; Naveh-Benjamin & Ayres, 1986; Standing et al., 1980). Consistent with the 2-second notion, children who took longer than approximately 2 s (from target presentation until completion—offset—of pronunciation) displayed signiﬁcantly poorer orthographic learning, as seen in Table I. By splitting the sample according to the overall naming times Table I Orthographic Learning for Children with Overall Naming Times (Onset Time+Naming Duration) above and below Speciﬁed Cutoff Points Naming time (duration) of target words (in seconds) 1.8 1.9 2.0 2.1 2.12 2.13 2.14 2.15 2.16 2.17 2.18 2.19 2.2

n at or above cutting point

n below cutting point

p value (spelling production)

p value (orthographic choice)

6 11 11 13 13 13 14 14 14 14 14 14 15

34 29 29 27 27 27 26 26 26 26 26 26 25

.9688 .3020 .3020 .1524 .1524 .1524 .0741 .0741 .0741 .0741 .0741 .0741 .0218

.6991 .1420 .1420 .0573 .0573 .0573 .0322 .0322 .0322 .0322 .0322 .0322 .0095

Orthographic Learning, Phonological Recoding, and Self-Teaching

67

(leftmost column) it can be seen that only at a cutoff of 2.14 s was there a signiﬁcant difference in orthographic choice between groups whose average naming times fell either above or below this value. For spelling, this ‘‘critical’’ point was located at a similar value of 2.2 s. These data support the hypothesis that slow laborious letter-by-letter decoding characteristic of novice readers of shallow orthographies may impair orthographic learning in the same way that slow word recognition can impair text integration processes (Perfetti, 1985). If correct, this decoding ﬂuency hypothesis would explain the lack of orthographic learning among novices.

E. OTHER SELF-TEACHING MECHANISMS? 1. Context Ever since Goodman’s, controversial pioneering study in 1967, the role of context in word identiﬁcation and word learning has remained disputations. Contemporary investigations of the role of context in orthographic learning continue this tradition. The early study by Ehri and Roberts (1979) found that learning new printed words in context reduced memory for word-speciﬁc orthographic detail compared to an isolation condition. The context-trained group was better at identifying the target words’ meanings but acquired less wordspeciﬁc orthographic detail; the isolation-trained group, in contrast, displayed superior word-speciﬁc orthographic knowledge revealed in faster naming times, superior pre- to post-test gains in spelling choice, and letter-level spelling production. A follow-up study by Ehri and Wilce (1980) extended these ﬁndings to context-dependent function words (e.g., might, while, must, from, enough). The context effect was found to be stronger among the poorer readers (see also Landi et al., 2006). Archer and Bryant (2001) selected a set of words that ﬁrst grade children had difﬁculty decoding (in a list of isolated words). Half of these items were then presented in context (a short meaningful sentence) and half in isolation (the experimenter supplying the pronunciation for items not successfully read alone). Next, the items were re-presented in the original word list, and again a day later. Although children were more successful reading the words in context than in isolation, the experience of words in meaningful context and in isolation led to equivalent improvements in later reading. Unlike the previous studies, Cunningham (2006) examined the role of context in unassisted reading of target words selected to be familiar in spoken but not printed form. Once again, context helped (English) ﬁrst

68

David L. Share

graders to decode these items (84% in context versus 67% in contextscrambled condition) but no advantage emerged in post-test orthographic learning. It is important to note however, that the targets were also selected not only on the basis of novelty but also decoding difﬁculty—all had irregular or complex spellings (e.g., piece, thirsty). Hence for ﬁrst graders, context might be expected to assume a greater role relative to less regular words. (Lacking is a study comparing the role of context for regular and irregular words.) Landi et al. (2006) also reasoned that learning new words in a meaningful context draws attention away from orthographic detail—detracting from long-term retention. Their ﬁrst experiment compared more- and less-skilled ﬁrst and second graders’ orthographic learning as a function of context and isolation. Following Archer and Bryant (2001), Landi et al. selected words children were unable to read correctly in isolation. Half were assigned to a context condition and half to an isolation condition. Context consisted of two-sentence paragraphs read aloud by the experimenter. Reading of the target word (presented once only) was unassisted and no feedback provided. A week later, children were asked to read the words aloud (in isolation). Words were read more successfully when presented in context but there was no overall difference in overall post-test reading accuracy. This implies (as was conﬁrmed) that of the words correctly identiﬁed at initial exposure, a lower proportion of the words identiﬁed correctly in context (69%) than in isolation (47%) were retained a week later. Similar effects were obtained for both more- and less-skilled readers. A second experiment replicated and extended these results to a larger sample of children using three presentations of each target word appearing either in isolation, in a single passage or in multiple passages. Although presenting targets in a single context (three times) or in (three) multiple contexts did not alter the outcomes, the context effect was signiﬁcantly greater for the less-skilled readers as found by Ehri and Wilce (1980). Martin-Chang, Levy, and O’Neill (2007) threw a spanner in the works by arguing that in all these studies the measurement of practice and retention are confounded. Words practiced and post-tested in isolation had been recalled under the same conditions, context-trained words were test under different conditions (i.e., transfer from context to isolation). In their study, difﬁcult-to-decode words were trained either in context or in isolation, followed by a retention task administered again 8 days later in which trained items were presented under the same conditions—isolation or context respectively—as well as a transfer task (trained words presented again in a novel story). Second graders taught via whole-language methods practiced reading test words (with assistance) in isolation or story context

Orthographic Learning, Phonological Recoding, and Self-Teaching

69

over the course of 3 days.10 Words were read more accurately in context than isolation and this advantage was maintained at the 8-day post-test (keep in mind that context-trained items were tested in context, isolationtrained items tested in isolation). Reading the test items in a novel passage in the transfer task, context evinced a much larger advantage compared to isolation. Turning the tables, and conﬁrming the authors’ ‘‘congruency’’ hypothesis, a second study found a clear advantage at transfer for words practiced in isolation when post-test transfer was tested in a novel isolated word reading task. Martin-Chang et al. concluded that transfer is greatest when congruency of training and transfer is maximized. Several general comments are pertinent to all these studies. With the exception of Cunningham (2006), the fact that all these studies involved explicit teaching or corrective feedback and practice limits the conclusions that can be drawn regarding unassisted independent reading/self-teaching. As already discussed, this is a genuine problem in English. The use of words selected not on the basis of ascertained unfamiliarity but of demonstrated decoding difﬁculties is problematic from the point of view of the selfteaching hypothesis—these items may be creating problems for reasons such as spelling complexity or sheer length, thereby necessitating greater than usual reliance on context (see Share, 1995; Tunmer & Chapman, 1998, 2006). It is important to consider various levels of word complexity but in order to reach generalizable conclusions regarding orthographic learning it is paramount that researchers ensure that test items are unfamiliar. Experimenter-supplied context or shared reading techniques may be valuable pedagogical tools but are divorced from unassisted everyday reading. Stanovich (1986) drew an important distinction between ‘‘effective’’ context (the context actually gained from unassisted reading) and ‘‘nominal’’ context (the maximum possible contextual support). Also crucial is the level of predictability of the sentences or passages which, in several of these studies, is often far beyond the norm for connected text (e.g., Archer & Bryant, 2001; Landi et al., 2006). Natural text has very low predictability (Share, 1995). Yet another complication is the instructional regime that children have experienced. The Martin-Chang et al. sample was recruited from whole-language schools where primary emphasis is placed on contextual prediction rather than decoding skills.

10

It should be remarked that the training set in this particular study was rather different from the items used in the studies described previously. A total of 85 words (selection criteria unspeciﬁed) were embedded in a single story—each appearing twice. Many of this pool would appear to be highly familiar even at this age and also largely regular (e.g., small, tree, seen, help, caught, looked, screeching, firecrackers).

70

David L. Share

Traditional phonics-emphasis instruction might not necessarily yield the same result. These reservations notwithstanding, results from all these investigations agree that contextual support for identifying hard-to-decode words is helped by context but at the expense of attention to orthographic detail. Hence, when only the print is available on a later encounter, reading suffers. This conclusion also accords with earlier work by Samuels and colleagues (Samuels, 1967; Singer, Samuels, & Spiroff, 1974) on the role of illustrations in printed word learning. ‘‘Picture and context cues deter acquisition of reading responses because they enable the child to identify the word in practice without focusing on its graphic features’’ (Singer et al., 1974, p. 555). The Martin-Chang et al. (2007) study raises an important concern—to which context do we wish to generalize the ﬁndings of printed word learning or, for that matter, word training studies? Meaningful text is indisputably the most common context in which new words are ﬁrst encountered and subsequently read, and, of course, the most important context for expanding knowledge of word meaning and syntactic function (Ehri & Roberts, 1979; Ehri & Wilce, 1980), especially as children encounter more and more words beyond their spoken vocabularies. Ultimately, however, we want children to be able to read and understand words in all possible circumstances—this includes both context and isolation. But we want readers who are writers too—so spelling facility is also a crucial element of word learning. Here context is of no consequence. (It is noteworthy that many word-learning studies rely solely on oral reading accuracy—or reading speed—ignoring orthographic knowledge per se.) It seems that reading hard-to-decode words with contextual support is likely to perpetuate a level of contextual reliance suitable for text reading that may be detrimental to decontextualized reading and to spelling and writing (consider the case of the good reader—poor speller; Frith, 1980). The ﬂip side of the argument that context detracts from attention to orthographic detail is, of course, the case of spelling which obliges close and thorough attention to orthographic detail and sub-lexical print-to-sound correspondence—possibly even more than does decoding.

2. Spelling Although a number of studies have demonstrated the beneﬁcial inﬂuence of spelling on reading and printed word learning, Shahar-Yames and Share (2008) noted that this evidence could be interpreted simply in terms of the general beneﬁts of superior working knowledge of the alphabetic code

Orthographic Learning, Phonological Recoding, and Self-Teaching

71

among children still learning the rudiments of print-to-sound decoding rather than word-speciﬁc orthographic knowledge of the kind on which automatized skilled word recognition depends. In a study focusing speciﬁcally on the role of spelling in orthographic learning, Shahar-Yames and Share presented third grade Hebrew speakers with target pseudowords in three conditions; reading, spelling (actually reading-plus-spelling), and an unseen control condition. In the spelling condition, participants ﬁrst read meaningful sentences containing the target word and were then asked to write down the target from memory with the sentences removed from sight. In this condition each participant read the target twice and wrote the target twice. In the reading-only condition, the participant merely read (decoded) the target (4 times). Post-test orthographic learning was assessed one week later with orthographic choice and spelling production—these two measures were also combined into a composite measure. Spelling was predicted to produce superior orthographic learning compared to reading owing to the additional processing demands invoked, and, furthermore, this advantage was expected to be greatest for lateroccurring orthographic detail in view of the fact that spelling obliges the writer to process each and every letter in a word on every occasion whereas decoding encounters, although likely to be quite exhaustive initially, are probably less exhaustive on subsequent occurrences— particularly in connected text. During the learning phase, similar and high levels of decoding and spelling success were observed in both conditions ensuring that comparison of orthographic learning across the two conditions was not confounded by differences in initial learning. Orthographic learning outcomes are presented in Table II. Reading led to signiﬁcant orthographic learning (relative to unseen controls), but spelling led to more powerful and more consistent learning outcomes. The difference between reading and spelling in the case of noninitial letters was almost twice the difference for the initial letter (9.5% versus 5.0%) although formally this interaction failed to reach signiﬁcance possibly because many of the non-initial letters were not word-ﬁnal. Like most ‘‘laboratory’’ training studies, this investigation shows that spelling can but not necessarily does perform a self-teaching function. The question of applicability to orthographic learning in vivo will depend very much on the literacy curriculum and the role of writing as well as factors such as attention to errors, toleration of misspellings, the availability of automatic spelling-correction routines and more. As we saw in the previous section on the contribution of context, the role of spelling in word-speciﬁc orthographic learning brings us back to the perennial instructional trade-off between meaning and mechanics.

72

Table II Post-Test Orthographic Learning (Means and Standard Deviations in Percentages) across Three Conditions (Control, Reading, Reading/Spelling) with t Values and Effect Sizes (Z2) (n=45) (reprinted from Shahar-Yames and Share, 2008) Post-test measure

Condition Control

Letter 2 Orthographic choice (4-choice) Composite orthographic learning(z) po.05; po.01.

Spelling

Reading/control 2

Reading/spelling

Spelling/control

t

Z

t

Z2

.19

2.36

.11

3.39

.21

2.05

.09

2.04

.09

4.00

.27

1.80

.07

.99

.02

2.51

.13

1.08

.03

1.68

.06

3.48

.22

2.82

.15

.86

.02

3.46

.21

2.67

.14

2.12

.09

4.81

.35

M (SD)

M (SD)

M (SD)

t

Z

22.2 (19.38) 47.8 (14.91) 46.7 (23.60) 48.9 (20.61) 23.9 (21.29) 0.79 (1.39)

26.1 (19.91) 54.7 (14.67) 55.0 (18.92) 54.4 (25.16) 37.4 (20.61) 0.07 (1.68)

37.8 (25.35) 61.9 (19.02) 60.0 (29.39) 63.9 (22.96) 41.9 (26.02) 0.67 (1.36)

.93

2

David L. Share

Spelling production Whole word (%) Target letters (n=2) Letter 1

Reading

Pairwise comparisons

Orthographic Learning, Phonological Recoding, and Self-Teaching

73

IV. Summary, Conclusions, and the Way Ahead A good deal has been learned about the basic parameters of orthographic learning, but as many researchers in this ﬁeld have observed (see, e.g., Castles & Nation, 2006; Landi et al., 2006), less is known about how this learning comes about. It is encouraging to see that much of this work has offered support to the self-teaching hypothesis. This can be seen in several sets of ﬁndings. First, levels of orthographic learning appear to be closely tied to levels of decoding success whether induced experimentally (Kyte & Johnson, 2006; Share, 1999) or occurring naturally (Ehri & Saltmarsh, 1995; Manis, 1985; Reitsma, 1983a, 1989; Share & Shalev, 2004). Second, at the individual level, the data have been quite consistent in showing a signiﬁcant positive association between target decoding success and orthographic learning—a relation furthermore that does not appear to be simply the offshoot of the general relation between pre-existing decoding ability and orthographic learning (Cunningham, 2006; Cunningham et al., 2002; Kyte & Johnson, 2006; Lurie & Share, 2007; Share, 2007; but see Nation et al., 2007). And third, the role of context in reducing attention to orthographic detail and spelling in enhancing it also conﬁrms the basic self-teaching premise that phonological recoding contributes to orthographic learning by drawing attention to letter detail and word-speciﬁc spelling– sound relations. Much work in this ﬁeld lies ahead. The orthographic learning function needs to be systematically mapped for both normal and disabled readers. Most studies to date have understandably used a small number of selected points on the learning curve, and although more exposures tend to elicit stronger outcomes, only a comprehensive mapping effort can determine whether orthographic learning adheres to the standard power function common to a wide variety of skill learning. Share’s (2004) ﬁnding of singletrial learning sorely needs replicating in view of the inconsistencies in Nation et al.’s (2007) single-exposure ﬁndings. If conﬁrmed, this would establish yet another tie between printed word learning and spoken word learning given the well-known phenomenon of fast mapping in spoken language development (see, e.g., Baldwin & Tomasello, 1998; Carey & Bartlett, 1978; Dollaghan, 1987; Woodward, Markman, & Fitzsimmons, 1994). Another potentially fruitful avenue for future research emerges from the earlier discussion of the unfamiliar–familiar/novice–expert dualism in which I argued that an efﬁcient orthography must supply the reader not only with a means for identifying novel words—a self-teaching device—but also with visually distinct conﬁgurations for morphemes necessary for the unitization and automatization of word recognition. It might be

74

David L. Share

hypothesized that the greater the distinctiveness of individual morphemes (operationalized, e.g., via orthographic neighborhood density or by distinctive graphic features such as capital letters, descenders, or ascenders), the faster and more efﬁcient the unitization process. Yet another important but untouched topic in the study of orthographic learning is the acquisition of knowledge of general orthographic conventions. To date, all the work referred to in this chapter has dealt exclusively with word-speciﬁc knowledge. This is a crucial yet very speciﬁc (i.e., minimally productive) type of knowledge relating to individual lexical items (or individual lexical families)—typically root morphemes. More general insights into (highly productive) orthographic conventions concerning grammatical and derivational (typically bound) morphemes is another critical dimension of orthographic learning (see, e.g., Chliounaki & Bryant, 2007) that remains to be investigated from the self-teaching point of view. Research into the determinants of individual differences in orthographic learning has only just begun. Some intriguing and puzzling ﬁndings (such as the evidence against the phonology-primary/orthography-secondary hypothesis in deep unpointed Hebrew) will remain enigmas until further research offers clariﬁcation. The early onset ﬁndings provide cause for optimism. Conﬂicting outcomes in English and Hebrew have helped point the way to several fascinating new research directions on the question of decoding ﬂuency, word length, and morphemic distinctiveness. Studies of context effects showing high levels of word identiﬁcation success that do not translate into superior orthographic learning as well as the initial Hebrew ﬁrst grade data indicating negligible orthographic learning despite near-ceiling levels of decoding accuracy all provide a warning to researchers regarding the value of using ‘‘gross’’ decoding accuracy that overlook the nature of decoding. The ‘‘how’’ of decoding promises to be a painstaking but crucial avenue for future research. We have mountains of research on decoding yet no ﬁne-grained studies revealing how this takes place for different readers at different levels of development and across diverse orthographies. We need to scrutinize the nature and location of the errors and not rest content with crude measures of overall accuracy. No study has yet gone beyond overall target decoding success and selectively examined errors on the critical homophonic letters. The entire process of decoding remains unchartered waters—is this performed letter-by-letter followed by one-step whole-string blending, maximally ‘‘chunked’’ (i.e., blended at each point along the entire string), or perhaps chunked at some intermediate point such as the sub-syllabic (e.g., onset-rime or body-coda sub-division), or whole syllable level? Does

Orthographic Learning, Phonological Recoding, and Self-Teaching

75

this process vary by script possibly aligning with the grain sizes postulated by the psycholinguistic grain size theory (Ziegler & Goswami, 2005)? Is the process foreshortened in the case of a phonologically familiar word? Where are the differences between regular and irregular words (and pseudowords) in context and in isolation—might this shed light on how context mediates orthographic learning? Above all, how does the process of decoding impinge on the assimilation of orthographic information? It should be apparent that simply recording raw decoding accuracy is merely scratching the surface of one of the most complex (and poorly understood) processes children accomplish. We know a great deal about the role of phonology in reading acquisition but the nature and origins of (‘‘crystallized’’) orthographic knowledge and the factors underlying its development (let us call it ‘‘ﬂuid’’ orthographic ability or aptitude) remains a black box (Burt, 2006; Castles & Nation, 2006; Share & Stanovich, 1995; Venezky, 2006). Symptomatic of this lack of progress are the heated debates over the nature and deﬁnition of surface-subtypes of acquired and developmental dyslexia (see, e.g., Coltheart, 2005; Woollams et al., 2007) and the ongoing controversy over the role of RAN in reading (see, e.g., Katzir et al., 2008; Vukovic & Siegel, 2006). The subject of orthographic learning offers researchers a ﬁeld of study lying at the very heart of one of the most important and challenging skills that children are expected to master. It is a promising young ﬁeld replete not only with fresh discoveries, enigmatic and contradictory ﬁndings but brimming with research questions awaiting investigation.

Acknowledgement A number of the Hebrew studies reviewed here were supported by grants from the Israel Science Foundation and the Israeli Ministry of Education—Ofﬁce of the Chief Scientist. Special thanks to Liat Butbul for assistance in preparing this manuscript.

REFERENCES Adams, M. J. (1990). Beginning to read. Cambridge, MA: Bradford. Anderson, J. R. (1981). Cognitive skills and their acquisition. Hillsdale, NJ: Erlbaum. Ans, B., Carbonnel, S., & Valdois, S. (1998). A connectionist multiple-trace memory model for polysyllabic word reading. Psychological Review, 105, 678–723.

76

David L. Share

Archer, N., & Bryant, P. (2001). Investigating the role of context in learning to read: A direct test of Goodman’s model. British Journal of Psychology, 92, 579–591. Baddeley, A. D., Thomson, N., & Buchanan, M. (1975). Word length and the structure of short-term memory. Journal of Verbal Learning and Verbal Behavior, 14, 575–589. Baldwin, D., & Tomasello, M. (1998). Word learning: A window on early pragmatic understanding. In E. Clark (Ed.), The Proceedings of the Twenty-Ninth Annual Child Language Research Forum (Vol. 7, pp. 3–23). Chicago: Center for the Study of Language and Information. Barron, R. W. (1986). Word recognition in early reading: A review of the direct and indirect access hypotheses. Cognition, 24, 93–119. Barron, R. W., & Baron, J. (1977). How children get meaning from printed words. Child Development, 48, 587–594. Bosse, M. L., Tainturier, M. J., & Valdois, S. (2007). Developmental dyslexia: The visual attention span deﬁcit hypothesis. Cognition, 104, 198–230. Bowers, P. G., & Wolf, M. (1993). Theoretical links among naming speed, precise timing mechanisms, and orthographic skill in dyslexia. Reading and Writing, 5, 69–85. Bowey, J. A., & Miller, R. (2007). Correlates of orthographic learning in third-grade children’s silent reading. Journal of Research in Reading, 30, 115–128. Bowey, J. A., & Muller, D. (2005). Phonological recoding and rapid orthographic learning in third-graders’ silent reading: A critical test of the self-teaching hypothesis. Journal of Experimental Child Psychology, 92, 203–219. Burt, J. S. (2006). What is orthographic processing skill and how does it relate to word identiﬁcation in reading? Journal of Research in Reading, 29, 400–417. Carey, S., & Bartlett, E. (1978). Acquiring a single new word. Papers and Reports on Child Language Development, 15, 17–29. Cassar, M., & Treiman, R. (1997). The beginnings of orthographic knowledge: Children’s knowledge of double letters in words. Journal of Educational Psychology, 89, 631–644. Castles, A., & Coltheart, M. (1996). Cognitive correlates of developmental surface dyslexia: A single case study. Cognitive Neuropsychology, 13, 25–50. Castles, A., & Nation, K. (2006). How does orthographic learning happen? In S. Andrews (Ed.), From inkmarks to ideas: Challenges and controversies about word recognition and reading (pp. 158–179). London: Psychology Press. Cestnick, L., & Coltheart, M. (1999). The relationship between language-processing and visualprocessing deﬁcits in developmental dyslexia. Cognition, 71, 231–255. Chliounaki, K., & Bryant, P. (2007). How children learn about morphological spelling rules. Child Development, 78, 1360–1373. Chomsky, N., & Halle, M. (1968). The sound pattern of English. New York: Harper & Row. Coltheart, M. (1978). Lexical access in simple reading tasks. In G. Underwood (Ed.), Strategies of information processing (pp. 151–216). London: Academic Press. Coltheart, M. (2005). Modeling reading: The dual-route approach. In M. Snowling & C. Hulme (Eds.), The science of reading: A handbook (pp. 6–23). Oxford: Blackwell. Coltheart, M., Curtis, B., Atkins, P., & Haller, M. (1993). Models of reading aloud: Dual-route and parallel-distributed-processing approaches. Psychological Review, 100, 589–608. Coltheart, M., Rastle, K., Perry, C., Langdon, R., & Ziegler, J. (2001). DRC: A dual-route cascaded model of visual word recognition and reading aloud. Psychological Review, 108, 204–256. Condry, S. M., McMahon-Rideout, M., & Levy, A. A. (1979). A developmental investigation of selective attention to graphic, phonetic and semantic information in words. Perception & Psychophysics, 25, 88–94. Cunningham, A. E. (2006). Accounting for children’s orthographic learning while reading text: Do children self-teach? Journal of Experimental Child Psychology, 95, 56–77.

Orthographic Learning, Phonological Recoding, and Self-Teaching

77

Cunningham, A. E., Perry, K. E., & Stanovich, K. E. (2001). Converging evidence for the concept of orthographic processing. Reading and Writing, 14, 549–568. Cunningham, A. E., Perry, K. E., Stanovich, K. E., & Share, D. L. (2002). Orthographic learning during reading: Examining the role of self-teaching. Journal of Experimental Child Psychology, 82, 185–199. de Jong, P. F., & Share, D. L. (2007). Orthographic learning during oral and silent reading. Scientific Studies of Reading, 11, 55–71. de Jong, P. F., & van der Leij, A. (2003). Developmental changes in the manifestation of phonological deﬁcit in dyslexic children learning to read a regular orthography. Journal of Educational Psychology, 95, 22–40. De Luca, M., Borrelli, M., Judica, A., Spinelli, D., & Zoccolotti, P. (2002). Reading words and pseudowords: An eye movement study of developmental dyslexia. Brain and Language, 80, 617–626. Doctor, E. A., & Coltheart, M. (1980). Children’s use of phonological encoding when reading for meaning. Memory & Cognition, 8, 195–209. Dollaghan, C. (1987). Fast mapping in normal and language-impaired children. Journal of Speech and Hearing Disorders, 52, 218–222. Ehri, L. C. (1980). The development of orthographic images. In U. Frith (Ed.), Cognitive processes in spelling (pp. 311–338). San Diego, CA: Academic Press. Ehri, L. C. (1992). Reconceptualizing the development of sight word reading and its relationship to recoding. In P. B. Gough, L. C. Ehri, & R. Treiman (Eds.), Reading acquisition (pp. 107–144). Hillsdale, NJ: Erlbaum. Ehri, L. C. (2005). Development of sight word reading: Phases and ﬁndings. In M. S. Snowling & C. Hulme (Eds.), The science of reading: A handbook (pp. 135–154). Oxford, UK: Blackwell. Ehri, L. C., & Roberts, K. T. (1979). Do beginners learn printed words better in contexts or in isolation? Child Development, 50, 675–685. Ehri, L. C., & Saltmarsh, J. (1995). Beginning readers outperform older disabled readers in learning to read words by sight. Reading and Writing, 7, 295–326. Ehri, L. C., & Wilce, L. S. (1980). The inﬂuence of orthography on readers’ conceptualization of the phonemic structure of words. Applied Psycholinguistics, 1, 371–385. Ehri, L. C., & Wilce, L. S. (1985). Movement into reading: Is the ﬁrst stage of printed word learning visual or phonetic? Reading Research Quarterly, 20, 163–179. Ehri, L. C., & Wilce, L. S. (1987). Cipher versus cue reading: An experiment in decoding acquisition. Journal of Educational Psychology, 79, 3–13. Firth, I. (1972). Components of reading disability. Unpublished doctoral dissertation. University of New South Wales. Foorman, B. R., Francis, D. J., Davidson, K. C., Harm, M. W., & Grifﬁn, J. (2004). Variability in text features in six grade 1 basal reading programs. Scientific Studies of Reading, 8, 167–197. Forster, K. I., & Davis, C. (1984). Repetition priming and frequency attention in lexical access. Journal of Experimental Psychology: Learning, Memory, and Cognition, 10, 680–698. Frith, U. (Ed.). (1980). Cognitive processes in spelling. New York: Academic Press. Frith, U. (1985). Beneath the surface of developmental dyslexia. In K. E. Patterson, J. C. Marshall, & M. Coltheart (Eds.), Surface dyslexia (pp. 301–322). London: Lawrence Erlbaum Associates. Frost, R. (1998). Toward a strong phonological model of reading: True issues and false trails. Psychological Bulletin, 123, 71–99. Frost, R. (2005). Orthographic systems and skilled word recognition processes in reading. In M. S. Snowling & C. Hulme (Eds.), The science of reading: A handbook (pp. 272–295). Oxford: Basil Blackwell.

78

David L. Share

Glushko, R. J. (1979). The organization and activation of orthographic knowledge in reading aloud. Journal of Experimental Psychology: Human Perception and Performance, 5, 674–691. Goldstone, R. L. (1998). Perceptual learning. Annual Review of Psychology, 49, 585–612. Goodman, K. S. (1967). Reading: A psycholinguistic guessing game. Journal of the Reading Specialist, 6, 126–135. Gough, P. B., & Juel, C. (1991). The ﬁrst stages of word recognition. In L. Rieben & C. A. Perfetti (Eds.), Learning to read: Basic research and its implication (pp. 47–56). Hillsdale, NJ: Lawrence Erlbaum Associates. Goulandris, N. K., & Snowling, M. (1991). Visual memory deﬁcits: A plausible case of developmental dyslexia? Evidence from a single case study. Cognitive Neuropsychology, 8, 127–154. Harm, M. W., & Seidenberg, M. S. (1999). Phonology, reading acquisition, and dyslexia: Insights from connectionist models. Psychological Review, 106(3), 491–528. Harm, M. W., & Seidenberg, M. S. (2004). Computing the meaning of words in reading: Co-operative division of labor between visual and phonological processes. Psychological Review, 111, 662–720. Hawelka, S., & Wimmer, H. (2005). Impaired visual processing of multi-element arrays is associated with increased number of eye movements in dyslexic reading. Vision Research, 45, 855–863. Hogaboam, T. W., & Perfetti, C. A. (1978). Reading skill and the role of verbal experience in decoding. Journal of Educational Psychology, 70, 717–729. Hutzler, F., & Wimmer, H. (2004). Eye movements of dyslexic children when reading in a regular orthography. Brain and Language, 89, 235–242. Jorm, A. F. (1979). The cognitive and neurological basis of developmental dyslexia: A theoretical framework and review. Cognition, 7, 19–33. Jorm, A. F. (1981). Children with reading and spelling retardation: Functioning of wholeword and correspondence-rule mechanisms. Journal of Child Psychology and Psychiatry, 22, 171–178. Jorm, A. F. (1983). Speciﬁc reading retardation and working memory: A review. British Journal of Psychology, 74, 311–342. Jorm, A. F., & Share, D. L. (1983). Phonological recoding and reading acquisition. Applied Psycholinguistics, 4, 103–147. Jorm, A. F., Share, D. L., Maclean, R., & Matthews, R. G. (1984). Phonological recoding skills and learning to read: A longitudinal study. Applied Psycholinguistics, 5, 201–207. Karni, A. (1996). The acquisition of perceptual and motor skills: A memory system in the adult human cortex. Cognitive Brain Research, 5, 39–48. Katz, L., & Frost, R. (1992). Reading in different orthographies: The orthographic depth hypothesis. In R. Frost & L. Katz (Eds.), Orthography, phonology, morphology, and meaning (pp. 67–84). Amsterdam: North Holland. Katzir, T., Kim, Y., Wolf, M., Morris, R., & Lovett, M. (2008). The varieties of pathways to dysﬂuent reading: Comparing subtypes of children with dyslexia at letter, word and connected-text reading. Journal of Learning Disabilities, 41, 47–66. Kay, J., & Marcel, T. (1981). One process, not two, in reading aloud: Lexical analogies do the work of nonlexical rules. Quarterly Journal of Experimental Psychology, 33A, 397–413. Kyte, C. S., & Johnson, C. J. (2006). The role of phonological recoding in orthographic learning. Journal of Experimental Child Psychology, 93, 166–185. LaBerge, D., & Brown, V. (1989). Theory of attentional operations in shape identiﬁcation. Psychological Review, 96, 101–124.

Orthographic Learning, Phonological Recoding, and Self-Teaching

79

LaBerge, D., & Samuels, J. (1974). Toward a theory of automatic information processing in reading. Cognitive Psychology, 6, 293–323. Landerl, K. (2000). Inﬂuences of orthographic consistency and reading instruction on the development of nonword reading skills. European Journal of Psychology of Education, 15, 239–257. Landi, N., Perfetti, C. A., Bolger, D. J., Dunlap, S., & Foorman, B. R. (2006). The role of discourse context in developing word representations: A paradoxical relation between reading and learning. Journal of Experimental Child Psychology, 94, 114–133. Logan, G. D. (1988). Toward an instance theory of automatization. Psychological Review, 95, 492–527. Logan, G. D. (1997). Automaticity and reading: Perspectives from the instance theory of automatization. Reading and Writing Quarterly, 13, 123–146. Logan, G. D. (2002). An instance theory of attention and memory. Psychological Review, 109, 376–400. Lukatela, K., Carello, C., Shankweiler, D., & Liberman, I. Y. (1995). Phonological awareness in illiterates: Observations from Serbo-Croation. Applied Psycholinguistics, 16, 463–487. Lurie, S., & Share, D. L. (2007). Decoding fluency and orthographic learning. Manuscript in preparation. Manis, F. R. (1985). Acquisition of word identiﬁcation skills in normal and disabled readers. Journal of Educational Psychology, 77, 78–90. Martin-Chang, S. L., Levy, B. A., & O’Neil, S. (2007). Word acquisition, retention, and transfer: Findings from contextual and isolated word training. Journal of Experimental Child Psychology, 96, 37–56. McCusker, L. X., Hillinger, M. L., & Bias, R. G. (1981). Phonological recoding and reading. Psychological Bulletin, 89, 217–245. Morris, D. (1992). Concept of word: A pivotal understanding in the learning-to-read process. In S. Templeton & D. R. Bear (Eds.), Development of orthographic knowledge and the foundation of literacy: A memory festschrift for Edmund H. Henderson (pp. 53–57). Hillsdale, NJ: Erlbaum. Nation, K. (1999). Reading skills in hyperlexia: A developmental perspective. Psychological Bulletin, 125, 338–355. Nation, K., Angells, P., & Castles, A. (2007). Orthographic learning via self-teaching in children learning to read English: Effects of exposure, durability, and context. Journal of Experimental Child Psychology, 96, 71–78. Naveh-Benjamin, M., & Ayres, T. J. (1986). Digit span, reading rate, and linguistic relativity. The Quarterly Journal of Experimental Psychology A: Human Experimental Psychology, 38, 739–751. Newell, A., & Rosenbloom, P. S. (1981). Mechanisms of skill acquisition and the law of practice. In J. R. Anderson (Ed.), Cognitive skills and their acquisition (pp. 1–55). Hillsdale, NJ: Erlbaum. Olson, D. R. (1994). The world on paper. Cambridge: Cambridge University Press. Pacton, S., Perruchet, P., Fayol, M., & Cleeremans, A. (2001). Implicit learning out of the lab: The case of orthographic regularities. Journal of Experimental Psychology: General, 130, 401–426. Paulesu, E., McCrory, E., Fazio, F., Menoncello, L., Brunswick, N., Cappa, S. F., Cotelli, M., Cossu, G., Corte, F., Lourusso, M., Pesenti, S., Gallagher, A., Perani, S., Price, C., Frith, C. D., & Frith, U. (2000). A cultural effect on brain function. Nature Neuroscience, 3, 91–96. Perfetti, C. A. (1985). Reading ability. New York: Oxford University Press.

80

David L. Share

Perfetti, C. A. (1992). The representation problem in reading acquisition. In P. B. Gough, L. C. Ehri, & R. Treiman (Eds.), Reading acquisition (pp. 145–174). Hillsdale, NJ: Erlbaum. Perfetti, C. A. (1994). Psycolinguistics and reading ability. In M. A. Gernsbacher (Ed.), Handbook of psycholinguistics (pp. 849–894). San Diego, CA: Academic. Perfetti, C. A. (2003). The universal grammar of reading. Scientific Studies of Reading, 7, 3–24. Plaut, D. C., McClelland, J. L., Seidenberg, M. S., & Patterson, K. (1996). Understanding normal and impaired word reading: Computational principles in quasi-regular domains. Psychological Review, 103, 56–115. Rader, N. (1975). From written words to meaning: A developmental study. Unpublished doctoral dissertation, Cornell University. Rayner, K. (1998). Eye movements in reading and information processing: 20 Years of research. Psychological Bulletin, 124, 372–422. Reitsma, P. (1983a). Printed word learning in beginning readers. Journal of Experimental Child Psychology, 36, 321–339. Reitsma, P. (1983b). Word-speciﬁc knowledge in beginning reading. Journal of Research in Reading, 6, 41–56. Reitsma, P. (1989). Orthographic memory and learning to read. In P. G. Aaron & R. M. Joshi (Eds.), Reading and writing disorders in different orthographic systems (pp. 51–73). New York: Kluwer Academic/Plenum Publishers. Rogers, H. (1995). Optimal orthographies. In I. Taylor & D. R. Olson (Eds.), Scripts and literacy: Reading and learning to read alphabets, syllabaries and characters (pp. 31–44). New York: Kluwer. Samuels, S. J. (1967). Attentional progress in reading: The effect of pictures on the acquisition of reading responses. Journal of Educational Psychology, 58, 337–342. Seymour, P. H. K., Aro, M., & Erskine, J. M. (2003). Foundation literacy acquisition in European orthographies. British Journal of Psychology, 94, 143–174. Shahar-Yames, D., & Share, D. L. (2008). Spelling as a self-teaching mechanism in orthographic learning. Journal of Research in Reading, 31, 22–39. Share, D. L. (1995). Phonological recoding and self-teaching: Sine qua non of reading acquisition. Cognition, 55, 151–218. Share, D. L. (1999). Phonological recoding and orthographic learning: A direct test of the selfteaching hypothesis. Journal of Experimental Child Psychology, 72, 95–129. Share, D. L. (2004). Orthographic learning at a glance: On the time course and developmental onset of self-teaching. Journal of Experimental Child Psychology, 87, 267–298. Share, D. L. (2007). Sources of individual differences in orthographic learning: A comparison of deep and shallow scripts. Manuscript submitted for publication. Share, D. L., & Blum, P. (2005). Syllable splitting in literate and preliterate Hebrew speakers: Onsets and rimes or bodies and codas? Journal of Experimental Child Psychology, 92, 182–202. Share, D. L., & Levin, I. (1999). Learning to read and write in Hebrew. In M. Harris & G. Hatano (Eds.), Learning to read and write (pp. 89–111). Cambridge: Cambridge University Press. Share, D. L., & Shalev, C. (2004). Self-teaching in normal and disabled readers. Reading and Writing, 17, 769–800. Share, D. L., & Stanovich, K. E. (1995). Cognitive processes in early reading development: Accommodating individual differences into a model of acquisition. Issues in Education, 1, 1–57. Shatil, E., & Share, D. L. (2003). Cognitive antecedents of early reading ability: A test of the modularity hypothesis. Journal of Experimental Child Psychology, 86, 1–31.

Orthographic Learning, Phonological Recoding, and Self-Teaching

81

Shiffrin, R. M., & Schneider, W. (1977). Controlled and automatic human information processing: II. Perceptual learning, automatic attending, and a general theory. Psychological Review, 84, 127–190. Siegler, R. S. (1988). Strategy choice and the development of multiplication skill. Journal of Experimental Psychology: General, 117, 258–275. Singer, H., Samuels, S. J., & Spiroff, J. (1974). The effect of pictures and contextual conditions on learning responses to printed words. Reading Research Quarterly, 9, 555–567. Standing, L., Bond, B., Smith, P., & Isely, C. (1980). Is the immediate memory span determined by the subvocalization rate? British Journal of Psychology, 71, 525–539. Stanovich, K. E. (1986). Matthew effects in reading: Some consequences of individual differences in the acquisition of literacy. Reading Research Quarterly, 21, 360–406. Stanovich, K. E. (1990). Concepts in developmental theories of reading skill: Cognitive resources, automaticity, and modularity. Developmental Review, 10, 72–100. Stanovich, K. E. (2000). Progress in understanding reading: Scientific foundations and new frontiers. New York: Guilford. Stuart, M., & Coltheart, M. (1988). Does reading develop in a sequence of stages? Cognition, 30, 139–181. Thaler, V., Ebner, E. M., Wimmer, H., & Landerl, K. (2004). Training reading ﬂuency in dysﬂuent readers with high reading accuracy: Word speciﬁc effects but low transfer to untrained words. Annals of Dyslexia, 54, 89–111. Thorstad, G. (1991). The effect of orthography on the acquisition of literacy skills. British Journal of Psychology, 82, 527–537. Tunmer, W., & Chapman, J. W. (1998). Language prediction skills, phonological recoding ability and beginning reading. In R. M. Joshi & C. Hulme (Eds.), Reading and spelling: Development and disorders (pp. 33–67). Hillsdale, NJ: Erlbaum. Tunmer, W., & Chapman, J. W. (2006). Metalinguistic abilities, phonological recoding skill and the use of context in beginning reading development: A longitudinal study. In R. M. Joshi & P. G. Aaron (Eds.), Handbook of orthography and literacy (pp. 617–636). Mahwah, NJ: Erlbaum. van der Leij, A., & van Daal, V. H. P. (1999). Automatization aspects of dyslexia: Speed limitations in word identiﬁcation, sensitivity to increasing task demands, and orthographic compensation. Journal of Learning Disabilities, 32, 417–428. Van-Orden, G. C., Pennington, B. F., & Stone, G. O. (1990). Word identiﬁcation in reading and the promise of subsymbolic psycholinguistics. Psychological Review, 97, 488–522. Vellutino, F. R. (1979). Dyslexia: Theory and research. Cambridge, MA: MIT Press. Venezky, R. (2006). Foundations for studying basic processes in reading. In R. M. Joshi & P. G. Aaron (Eds.), Handbook of orthography and literacy (pp. 735–758). Mahwah, NJ: Erlbaum. Venezky, R. (2007). In search of the perfect orthography. Written Language and Literacy, 7, 139–163. Vukovic, R. K., & Siegel, L. S. (2006). The double deﬁcit hypothesis: A comprehensive analysis of the evidence. Journal of Learning Disabilities, 39, 25–47. Weekes, B. S. (1997). Differential effects of number of letters on word and nonword naming latency. The Quarterly Journal of Experimental Psychology A: Human Experimental Psychology, 50A, 439–456. Williams, M. J., Stuart, G. W., Castles, A., & McAnally, K. I. (2003). Contrast sensitivity in subgroups of developmental dyslexia. Vision Research, 43, 467–477. Wimmer, H. (1993). Characteristics of developmental dyslexia in a regular writing system. Applied Psycholinguistics, 14, 1–33.

82

David L. Share

Woodward, A., Markman, E., & Fitzsimmons, C. (1994). Rapid word learning in 13–18 month olds. Developmental Psychology, 30, 553–566. Woollams, A. M., Ralph, M. A. L., Plaut, D. C., & Patterson, K. (2007). SD-squared: On the association between semantic dementia and surface dyslexia. Psychological Review, 114, 316–339. Wright, G., Sherman, R., & Jones, T. B. (2004). Are silent reading behaviors of ﬁrst graders really silent? International Reading Association, 57, 546–553. Ziegler, J. C., & Goswami, U. (2005). Reading acquisition, developmental dyslexia and skilled reading across languages: A psycholinguistic grain size theory. Psychological Bulletin, 131, 3–29.

DEVELOPMENTAL PERSPECTIVES ON LINKS BETWEEN ATTACHMENT AND AFFECT REGULATION OVER THE LIFESPAN

Lisa M. Diamond and Christopher P. Fagundes DEPARTMENT OF PSYCHOLOGY, UNIVERSITY OF UTAH, SALT LAKE CITY, UT 84112, USA

I. REVIEW OF ATTACHMENT THEORY A . DEVELOPMENTAL TRANSITIONS B . INDIVIDUAL DIFFERENCES C . BRIDGING THE INFANT-CHILD AND ADULT TRADITIONS II. AFFECT REGULATION A . CAREGIVERS AND THE DEVELOPMENT OF AFFECT REGULATION B . ATTACHMENT ANXIETY AND AVOIDANCE C . PHYSIOLOGICAL PROCESSES LINKING ATTACHMENT AND AFFECT REGULATION D . TOWARD A PROCESS-ORIENTED, BIOBEHAVIORAL APPROACH III. ATTACHMENT AND AFFECT REGULATION DURING ADOLESCENCE A . BACKGROUND AND METHODS B . PATTERNS OF ATTACHMENT TRANSFER C . BRIDGES TO AUTONOMY: EXPLORATION AND DEPENDENCY D . ATTACHMENT STYLE AND AFFECT REGULATION E . PHYSIOLOGICAL CORRELATES OF AFFECT REGULATION IV. IMPLICATIONS AND FUTURE DIRECTIONS A . INTEGRATION OF ATTACHMENT AND AFFECT REGULATION WITH OTHER REGULATORY PROCESSES B . A DYADIC APPROACH TO ADOLESCENCE C . THE SPECIFIC IMPORTANCE OF POSITIVE AFFECT D . CONCLUSION REFERENCES

In the late 1950’s since John Bowlby’s initial reﬂections on infant– caregiver attachment ﬁrst appeared in print, attachment theory has arguably revolutionized research on affectional bonding and its role in psychological health and development. One of the most compelling aspects of attachment theory is its lifespan perspective. Although Bowlby focused primarily on infant–caregiver attachment, he argued that attachment

83 Advances in Child Development and Behavior R.V. Kail : Editor

Copyright r 2008 Elsevier B.V. All rights reserved.

84

Lisa M. Diamond and Christopher P. Fagundes

processes remain central to mental and physical well-being ‘‘from the cradle to the grave’’ (Bowlby, 1988, p. 62). Accordingly, the extension of attachment theory to adult love relationships (Hazan & Shaver, 1987; Shaver, Hazan, & Bradshaw, 1988) has created opportunities for building comprehensive developmental models that use the same core principles to explain the nature, dynamics, and effects of intimate human relationships at all stages of life. Yet the promise of such sweeping lifespan models has largely gone unfulﬁlled. Rather, contemporary attachment research remains bifurcated between developmental investigations of infant–caregiver bonds and socialpsychological investigations of adult romantic bonds. Researchers from each ‘‘camp’’ emphasize different aspects of attachment and use different methods to capture and evaluate attachment phenomena. With some exceptions (for example Grossmann, Grossmann, & Waters, 2005), few researchers have attempted to integrate ﬁndings—and empirical investigations—across these domains into a broader analysis of the attachment system and its implications for social, psychological, and physical wellbeing over the entire life course. We think that a greater emphasis on the affect- and emotion-regulation functions of attachment can ameliorate this problem and provide a powerful unifying framework for integrative, biobehavioral, process-oriented models of the attachment system from birth through adulthood. In this chapter, we make this case by reviewing prior theory and research linking attachment to affect-regulation processes, and also reviewing some of our own empirical data on linkages between these domains during the adolescent years. As a bridge between childhood and adulthood, adolescence presents special challenges and opportunities for investigating the functioning of the attachment system and its implications for mental and physical well-being. We hope to demonstrate how investigations of multiple processes of affect regulation, and the multiple origins of individual differences in these processes, can contribute to research on attachment not only during the adolescent years, but across the life course. We begin with a brief review of attachment theory and some of the current challenges facing attachment research, speciﬁcally with regard to integrating the child and adult literatures, and integrating research on normative processes vs individual differences. We then review research linking affect and emotion regulation to attachment processes, highlighting both psychological and physiological aspects of affect regulation, and showing how these processes potentially explain the mental and physical health implications of attachment relationships. We then spend the remainder of the chapter reviewing some of our own research on these topics, and highlighting promising directions for future study.

Developmental Perspectives

85

I. Review of Attachment Theory Bowlby (1958, 1982) conceptualized attachment as an evolved behavioral system designed to regulate infants’ proximity to caregivers and thereby maximize chances for survival. When an infant experiences distress, he or she immediately attempts to seek contact with the attachment ﬁgure. In normative cases, this proximity reassures and soothes the infant, who subsequently comes to associate the presence of the attachment ﬁgure with emotional security and distress-alleviation. Even when the attachment ﬁgure is not consistently successful at alleviating distress, infants typically develop a unique, exclusive, emotionally primary relationship with the attachment ﬁgure, such that this person becomes the preferred target for security-seeking. Normative attachments are characterized by the presence of four distinct forms of behavior: seeking and maintaining physical closeness to the attachment ﬁgure (‘‘proximity seeking’’), turning to the attachment ﬁgure for comfort and reassurance (‘‘safe haven behavior’’), experiencing distress as a result of separations from the attachment ﬁgure (‘‘separation distress’’), and using the attachment ﬁgure as a reliable, dependable base of support from which to explore the world (‘‘secure base behavior’’) (Ainsworth et al., 1978; Bowlby, 1982). According to attachment theory, infants develop nonconscious mental representations of their bond with the caregiver—termed internal working models—which encode expectations of caregiver behavior (Sensitive? Trustworthy? Dependable? Consistent?) and corresponding views of one’s self as worthy or unworthy of love and attention (Bowlby, 1973, 1980, 1982; Kobak & Sceery, 1988; Main, Kaplan, & Cassidy, 1985). As the child matures, these working models provide organizing frameworks for relationship skills and expectations (Bartholomew & Horowitz, 1991; Cassidy et al., 1996; Sroufe, 2005; Sroufe & Fleeson, 1986; Sroufe et al., 2005) and provide the child with an ‘‘inner resource’’ of security that allows him or her to seek increasing independence from the caregiver and to explore his/her environment (Ainsworth et al., 1978).

A. DEVELOPMENTAL TRANSITIONS The attachment system remains active over the life course, but undergoes important developmental changes. As argued by Hazan and Shaver (1987), adults do not typically continue to utilize parents as primary bases of emotional security, but instead turn to romantic partners for that function. Hence, adult romantic relationships are thought to be functionally analogous to infant–caregiver attachments, and based in the same

86

Lisa M. Diamond and Christopher P. Fagundes

neurobiologically mediated social-behavioral system. The fundamental correspondence between infant–caregiver attachment and adult romantic ties is supported by extensive research documenting that these phenomena share the same core emotional and behavioral dynamics: heightened proximity maintenance, resistance to separation, and utilization of the partner as a preferred target for comfort- and security-seeking (reviewed in Hazan & Zeifman, 1999). Even more powerful evidence is provided by the voluminous animal research documenting that two types of affectional bonding are mediated by the same opioid- and oxytocin-based neural circuitry (Carter, 1998). This view suggests that the basic purpose and processes of the attachment system remain largely continuous over the lifespan, although the ‘‘target’’ of the system changes quite dramatically, from parents to romantic partners. The foundation for this transition is laid during adolescence, as part of the normative, well-documented developmental process through which youths seek progressively more independence and differentiation from parents, and correspondingly seek more intimacy, support, and companionship from friends and dating partners (Laursen, 1996; SavinWilliams & Berndt, 1990). In the general adolescent developmental literature, these processes are described in terms of autonomy and differentiation (Allen & Hauser, 1996; Steinberg & Silverberg, 1986); in the attachment literature, they are discussed in terms of ‘‘transfer’’ of attachment from parents to peers, and the ‘‘reshufﬂing’’ of the attachment hierarchy such that parents no longer occupy the preeminent positions they once did (Cooper et al., 2004; Hazan & Zeifman, 1994; Markiewicz et al., 2006; Trinke & Bartholomew, 1997). Speciﬁcally, by the time youths are 17–20, they are no longer expected to maintain a primary sense of security through proximity and contact with parents, but instead through contact with a romantic partner. Parents remain important, but function more in the background, as ‘‘attachment ﬁgures in reserve’’ (Weiss, 1982). The ﬁrst systematic study of this process was conducted by Hazan and Zeifman (1994), who sought to document age-related changes in attachment behavior from late childhood to young adulthood. They asked children and adolescents, ranging in age from 6 to 17, to ﬁll out a self-report questionnaire that assessed who they primarily utilized for the attachment functions of proximity maintenance (sample item: ‘‘who do you most enjoy hanging out with?’’), safe haven (‘‘who provides you with support when you’re under stress?’’), and secure base (‘‘who do you know will always be there for you, no matter what?’’). Consistent with attachment theory, younger children typically listed their mother or father as the primary target for each of these components. Yet with advancing age, youths increasingly listed peers rather than parents as primary targets. Importantly, Hazan and

Developmental Perspectives

87

Zeifman (1994) described a speciﬁc developmental sequence to this process of ‘‘transferring’’ attachment functions to peers: proximity seeking, followed by safe haven, and ﬁnally secure base (which was often not completely transferred to peers until the adolescent had developed a serious romantic relationship). Analogous ﬁndings using similar methods were reported by Fraley and Davis (1997) and Trinke and Bartholomew (1997). Yet Markiewicz and colleagues (2006) found a somewhat different pattern. They compared youths between 12 and 15, between 16 and 19, and between 20 and 28. Similar to the aforementioned studies, proximity seeking and safe haven behaviors were directed to mothers less often with age. Yet contrary to the notion that peers also supplanted parents as primary sources of security by young adulthood, mothers continued to serve as targets for the secure-base component of attachment youths across the entire age range, even among those with serious romantic partners. Also, their ﬁndings regarding safe haven behavior were not entirely consistent with the classic ‘‘transfer’’ perspective: Younger rather than older adolescents preferentially directed safe haven behaviors to best friends, suggesting that youths might vacillate back and forth over time in the targeting of their attachment behavior, perhaps redirecting comfort- and security-seeking back to parents as they confront the progressively more challenging and complex developmental transitions of later adolescence. Hence, the notion of progressively ‘‘transferring’’ attachment from parents to peers over the course of adolescence might be oversimpliﬁed, and might fail to capture the ways in which adolescents become increasingly peer-directed in their attachment behavior without necessarily relinquishing the primary role of parents. Because previous studies of attachment transfer have used a forced-choice method, in which only one person could be nominated as the primary target for each attachment behavior, they do not reﬂect the extent to which youths may utilize parents and peers simultaneously and equally for attachment-related functions during certain stages of development. Such simultaneity would be more consistent with the published research on adolescent autonomy, in which the healthiest development trajectories combine increasing behavioral and psychological independence with continued warmth, emotional connectedness, and emotional security (reviewed by Allen et al., 1994; Allen & Land, 1999). Many unanswered questions remain regarding the normative development of attachment in adolescence, which is perhaps not surprising in light of the overall underinvestigation of normative features of the attachment system (Berlin & Cassidy, 1999; Hazan, Gur-Yaish, & Campa, 2004; Marvin & Britner, 1999; Simpson & Rholes, 1998), in contrast to the extensive body of research on individual differences in attachment style

88

Lisa M. Diamond and Christopher P. Fagundes

(reviewed subsequently). Future research would particularly proﬁt from greater attention to questions about the speciﬁc interpersonal and intrapsychic processes through which youths begin to perceive peers as attachment ﬁgures, parents’ perceptions of—and behavioral reactions to— this shift, and the multiple situational, temperamental, and relational factors which moderate this overall process.

B. INDIVIDUAL DIFFERENCES The aforementioned discussion concerns the normative component of attachment theory, which focuses on species-typical processes of affectional bonding, and their normative development. Yet attachment theory also concerns individual differences in experiences and expectations of attachment relationships, which inﬂuence psychological and interpersonal functioning over the life course. These individual differences are called attachment styles, and are conceived as trait-like expectations concerning the responsiveness of attachment ﬁgures, established through early infant– caregiver interactions (Ainsworth et al., 1978). These expectations not only inﬂuence relationship experiences, but also come to organize the encoding, storage, retrieval, and manipulation of information related to affective states and—in particular—experiences of stress vs security (see reviews in Bartholomew & Horowitz, 1991; Mikulincer, 1998a; Mikulincer, Shaver, & Pereg, 2003). Speciﬁcally, ‘‘secure’’ infants are those with sensitive and responsive caregivers, who consistently experienced proximity to these caregivers as distress-alleviating. As a result, they come to view themselves as competent and worthy of love and to view others as willing and able to provide comfort and support. Individuals with an anxious attachment style experienced inconsistent caregiving and consequently seek repeated reassurance of the availability of their attachment ﬁgures. Individuals with an avoidant attachment style did not receive adequate, sensitive care from their attachment ﬁgures and therefore learned not to seek contact with them when distressed. Although these styles were originally hypothesized to describe children’s orientations toward their attachment ﬁgures, researchers have found that they also describe adults’ orientations toward romantic partners, consistent with the notion that romantic partners function as adult attachment ﬁgures (Hazan & Shaver, 1987; Shaver et al., 1988). Hundreds of studies have detected associations between adult attachment style and individuals’ feelings and behaviors toward romantic attachment ﬁgures, including disclosure and communication (Feeney, Noller, & Callan, 1994), supportseeking and support-provision (Collins & Feeney, 2000; Simpson, Rholes, &

Developmental Perspectives

89

Nelligan, 1992), conﬂict (Feeney et al., 1994; Senchak & Leonard, 1992), and overall relationship satisfaction and stability (Kirkpatrick & Davis, 1994; Senchak & Leonard, 1992; Simpson, 1990). Yet there continues to be extensive debate about whether adolescent and adult attachment styles are really infant-child styles ‘‘grown up,’’ or whether attachment anxiety and avoidance represent different phenomena—with different antecedents—in infancy, childhood, adolescence, and adulthood. Longitudinal research has detected signiﬁcant evidence for continuity and discontinuity in attachment security from infancy and childhood to adolescence and adulthood (Allen & Land, 1999; Hamilton, 2000; Roisman et al., 2005; Waters et al., 2000; Weinﬁeld, Sroufe, & Egeland, 2000), and there is also evidence suggesting the continued capacity for change in adulthood as a function of participation in different types of romantic relationships (Davila, Karney, & Bradbury, 1999; Kirkpatrick & Hazan, 1994). Importantly, interpretation of these research ﬁndings is complicated by the fact that attachment security at different stages of life is typically assessed with different methods (reviewed by Allen & Land, 1999; Crowell, Fraley, & Shaver, 1999; Jacobvitz, Curran, & Moller, 2002; Solomon & George, 1999). Yet the overall picture suggests that although early individual differences in attachment security have lasting effects on psychological and interpersonal functioning, individuals’ cumulative experiences in attachment relationships over time can enhance or disrupt stability in anxiety and avoidance. For example, Allen and colleagues (2003) found that as much as 40% of variation in adolescents’ attachment security was reﬂected in the current quality of youths’ interpersonal interactions with parents. Weinﬁeld, Sroufe, and Egeland (2000) found that stressful life events could signiﬁcantly disrupt family functioning and precipitate longitudinal transitions from security to insecurity. Accordingly, researchers have increasingly emphasized the importance of assessing both generalized attachment styles and speciﬁc experiences of need-fulﬁllment within current attachment relationships in order to accurately model individual differences in attachment style at different stages of life (Cook, 2000; La Guardia et al., 2000).

C. BRIDGING THE INFANT-CHILD AND ADULT TRADITIONS Attachment theory has provided a powerful and comprehensive model of the inﬂuence of intimate relationships on social and psychological functioning over the life course, and it is currently the preeminent theory underlying research on child–caregiver relationships and adult romantic

90

Lisa M. Diamond and Christopher P. Fagundes

relationships. Yet research on adult attachment has developed and evolved quite separately from research on infant–child attachment, despite their common heritage in Bowlby’s seminal work. To some extent, this can be attributed to straightforward disciplinary boundaries: Research on infant– child attachment is typically conducted by developmental psychologists, for whom the identiﬁcation of adult manifestations of the parent–child processes they study may not be a primary topic of interest. Research on adult attachment is typically conducted by social and personality psychologists, who may possess a basic familiarity with the purported developmental origins of attachment styles, but who are typically far more interested in probing their implications for adult functioning. Each of these ‘‘camps’’ has produced tremendously valuable investigations into the functioning of the attachment system in infancy-childhood and adulthood, but the developmental bifurcation of attachment research has hampered our understanding of how the system itself develops and changes over time. Even the aforementioned longitudinal studies, which have followed individuals from infancy to adulthood, have focused on basic questions of continuity in attachment security from childhood to adulthood, and do not permit close investigation of developmental changes in attachment-related processes. Perhaps the most vivid manifestation of this blind spot in attachment research is the continued underinvestigation of attachment processes during adolescence rather than infancy, childhood, and adulthood. As reviewed by Allen and Land (1999), adolescence is a critically important period of life from the lens of attachment theory. Adolescents’ increasing capacities for complex reasoning, abstraction, and executive functioning (Blakemore & Choudhury, 2006; Keating, 1990) promote the progressive consolidation of internal working models of attachment, and their integration with concrete, current interpersonal experiences. Adolescents must also balance the normative developmental press for differentiation from parents with continued needs for parental support and assistance, especially in light of the increasingly complex social, emotional, and psychological challenges that accompany this stage of life. Finally, adolescents’ increasing interest and participation in romantic and sexual relationships lays the groundwork for the signature developmental transformation in the attachment system: the shift from unilaterally seeking security from parents to reciprocally seeking and providing security to romantic partners. We propose that the best way to integrate the growing body of research on adolescent attachment processes with the existing infant-child and adult traditions is to focus more systematically on the affect- and emotionregulation functions of attachment. Affect and emotion regulation are critically implicated in both the normative and individual difference

Developmental Perspectives

91

components of attachment theory, and have been found to mediate and moderate attachment processes in both adulthood and infancy-childhood. Furthermore, the growing body of research on the multiple physiological mechanisms underlying affect and emotion regulation helps to elucidate the basic psychobiology of the attachment system, and the speciﬁc physiological processes through which attachment experiences and relationships shape both mental and physical health over the life course (Diamond, 2001; Diamond & Hicks, 2004). In the next section, we provide an overview of affect and emotion regulation and their associations with attachment processes. We show that at all stages of life, affect and emotion regulation remain primary functions of the attachment system, although the speciﬁc processes through which they are effected change over time. We then turn to our own research on linkages among attachment, affect regulation, and well-being during early adolescence.

II. Affect Regulation The terms ‘‘emotion regulation’’ and ‘‘affect regulation’’ are often used interchangeably, but there are slight differences between them that bear attention: ‘‘Emotion regulation’’ is usually used to refer to internal and transactional processes through which individuals consciously or unconsciously modulate the experience or expression of emotions elicited by environmental events (Eisenberg et al., 2000; Gross, 1999; Thompson, 1994). Affect regulation refers to similar processes of modulation, but the regulated ‘‘output’’ includes broader, ongoing affective states and moods, and not just discrete, situationally triggered emotions (Larsen, 2000). Because both affect and emotion regulation are thought to be shaped by the attachment system, for the purposes of this chapter we use the term ‘‘affect regulation’’ in a broad sense to refer to both. The progressive mastery of a diverse range of strategies for affect regulation is considered a core developmental task for both children and adolescents (Cooper, Shaver, & Collins, 1998; Denham, 2006; Eisenberg & Fabes, 1992; Eisenberg, Spinrad, & Morris, 2002; Fox, 1994a; Halberstadt, Denham, & Dunsmore, 2001; Masten, 2001; Repetti, Taylor, & Seeman, 2002; Saarni, 1992). Because powerful emotions have the potential to disorganize and/or disrupt multiple psychological processes, modulation of their experience and expression (through both intrapsychic and interpersonal processes) has been considered essential for basic state regulation, behavioral exploration, cognitive processing, and social competence (reviewed in Fox, 1994b). Accordingly, inability to effectively regulate

92

Lisa M. Diamond and Christopher P. Fagundes

one’s own emotions, as well as one’s cognitions and behaviors in emotionally arousing situations, has been linked to a range of psychological and behavioral problems in both childhood and adolescence (Cicchetti, Ackerman, & Izard, 1995; Cooper et al., 1995; Cooper et al., 1998; Eisenberg et al., 2001; Frick & Morris, 2004; Kobak & Ferenz Gillies, 1995; Silk, Steinberg, & Morris, 2003). Furthermore, studies of adolescents and adults have found that high and unregulated levels of negative affect are associated with general maladjustment (Gross & Munoz, 1995), anxiety and depressive disorders (Fabes & Eisenberg, 1997; Nolen-Hoeksema, Parker, & Larson, 1994), substance use (Colder & Chassin, 1997; Cooper et al., 1995; Pandina, Johnson, & Labouvie, 1992; Wills, Windle, & Cleary, 1998), and even impaired neuroendocrine, autonomic, and immune functioning (Repetti et al., 2002; Ryff & Singer, 2001; Taylor, Dickerson, & Klein, 2002; Taylor, Repetti, & Seeman, 1997). Hence, investigating both the normative development of affect regulation and also individual differences in affectregulation capacities and strategies is important for understanding how healthy trajectories of socioemotional development can be established and maintained through childhood, adolescence, and adulthood.

A. CAREGIVERS AND THE DEVELOPMENT OF AFFECT REGULATION Attachment ﬁgures have been theorized to play a fundamental role in the initial development and ongoing maintenance of infant and children’s affect regulation because of the centrality of distress-alleviation and securityprovision in the attachment system (see Berlin & Cassidy, 1999; Simpson & Rholes, 1994). During emotionally laden interactions, the caregiver continuously modulates the infant’s affective and attentional state and aligns it with his/her own through changes in facial expression, behavioral activation, and direct engagement with different features of the immediate environment (Kopp, 1989). This process of caregiver-managed engagement and disengagement of attentional and stress-regulatory systems in the orbitofrontal cortex is thought to provide the foundation for effective selfregulation more generally, and affect regulation in particular (Schore, 1996a; Siegel, 2001). This view is supported by a growing body of biologically oriented research showing that in humans and other mammals, early experiences with nurturant vs neglectful caregiving ‘‘tune’’ stressregulatory processes in the autonomic and neuroendocrine systems (Glaser, 2000; Gunnar & Donzella, 2002; Repetti et al., 2002; Schore, 1996a; 2000).

Developmental Perspectives

93

Caregiver-managed affect regulation continues to play an important role in early childhood. Both observational and experimental studies of children have found that caregiver sensitivity/responsiveness is associated with less expression of negative affect (Eisenberg et al., 1991; Fabes et al., 1994; Morris et al., 2007) and, at later ages, greater ability to employ a range of different self-regulatory strategies across different situations (Cohn & Tronick, 1983; Gable & Isabella, 1992; Hardy, Power, & Jaedicke, 1993; Kliewer, Fearnow, & Miller, 1996). In direct contrast, maternal behavior that is neglectful or hostile has been associated with a range of deﬁcits in both affect and behavioral regulation, which are thought to be a primary mechanism through which troubled (or ‘‘risky’’) family environments impair children’s long-term social and emotional functioning (comprehensively reviewed in Repetti et al., 2002). The primary developmental transition from infancy to adolescence involves internalization of affect regulation. Whereas infants and children must rely on direct contact with the attachment ﬁgure to regulate distress, older children gradually learn to modify their own affective states independently of such contact (Calkins et al., 1998; Thompson, 1994) through strategies such as self-soothing, attention shifting, reappraisal, active coping, or simply avoiding certain stimuli (Kobak et al., 1993; Rothbart, 1991). Despite this progressive internalization, children and adolescents continue to seek assistance with managing affective states from a variety of different social partners (Gross & Munoz, 1995; Thompson, 1994). Attachment theory predicts that attachment ﬁgures remain the most preferred and most effective providers of this function at all stages of life, particularly when regulatory demands are high. This is supported by research demonstrating that both infants and adults prefer to seek contact with attachment ﬁgures over other social partners in times of extreme distress (Cassidy, 1994; Hazan & Zeifman, 1994; Thompson, 1994; Trinke & Bartholomew, 1997). This renders adolescence a particularly notable period of life for investigating the interpersonal context of affect regulation: Although parents might remain the most effective providers of distress-alleviation, youths seek progressively more companionship, support, and comfort from peers rather than parents during this period as part of their normative developmental transition to greater autonomy and differentiation (Laursen, 1996; Savin-Williams & Berndt, 1990; Silk et al., 2003; Steinberg & Morris, 2001). Hence, although adolescents’ needs for assistance with affect regulation remain high, they may often be seeking such assistance from less effective providers. At the same time, their own friends (and eventually romantic partners) increasingly call upon them to provide empathy, comfort, and

94

Lisa M. Diamond and Christopher P. Fagundes

distress-alleviation, putting them in new caregiving roles that require effective management of their own emotions (Bandura et al., 2003). Adolescents’ increasingly complex and intimate relationships with friends and—eventually—romantic partners may expose them to newly intense emotions (Larson, Clore, & Wood, 1999): Even a routine romantic breakup may prove to be an adolescent’s ﬁrst substantive exposure to the helplessness and depression associated with interpersonal loss and loneliness (Steinberg & Silk, 2002). Combined with the heightened tension and negativity that often accompanies conﬂicts with parents over issues of independence and autonomy, adolescents often experience heightened and more variable positive and negative emotions, thereby increasing their day-to-day self-regulatory demands (Allen & Land, 1999; Conger & Ge, 1999; Cooper et al., 1998; Larson et al., 1999; Steinberg & Morris, 2001). Finally, contrary to the notion that youths’ basic physiological capacities for affect regulation are ‘‘ﬁnished’’ developing, maturation of neural regions in the prefrontal cortex involved in affect regulation continues to undergo development and maturation well into late adolescence (Spear, 2000). Given these normative developmental challenges, it is not surprising that youths with deﬁciencies in affect regulation show a range of psychological and social problems, many of which persist into young adulthood. Here, we review the extensive evidence that a primary basis for such deﬁciencies is attachment insecurity.

B. ATTACHMENT ANXIETY AND AVOIDANCE Much initial research on individual differences in attachment security focused on the cognitive-representational aspects of infants’ internal working models: As noted earlier, anxious and avoidant working models encapsulate speciﬁc mental expectations of caregiver behavior and corresponding views of one’s self as worthy or unworthy of care. These representations function as mental prototypes for future relationship experiences. Yet consistent with Bowlby’s (1973) original writings, researchers studying attachment (in both children and adults) increasingly view internal working models as also providing an organizing framework for affective experience, expression, and regulation (Brennan & Shaver, 1995; Kobak & Sceery, 1988; Mikulincer & Shefﬁ, 2000; Simpson et al., 1992). Hence, both infant and adult attachment styles are thought to encode not only expectations of caregiver behavior, but consistent capacities and strategies for affect regulation derived from early stress-regulating

Developmental Perspectives

95

interactions with caregivers (Cooper et al., 1998; Rholes et al., 1999; Simpson et al., 1992). Speciﬁcally, children who did not receive adequate ‘‘external’’ affect regulation from their caregivers are thought to sustain developmental deﬁcits in their own self-regulatory capacities (see Glaser, 2000), and consequently come to rely on anxiety and avoidance as secondary—and suboptimal—affect-regulation strategies. Individuals with high attachment anxiety have been found to maximize the experience and expression of negative affect, to be hypervigilant to threat cues, and to show patterns of spreading emotional reactivity such that one negative thought or memory triggers many others (Shaver & Mikulincer, 2002). Furthermore, anxious adults have been found to respond cognitively to positive affect inductions with the same reduced cognitive ﬂexibility and creativity traditionally associated with negative affect (Mikulincer & Shefﬁ, 2000), suggesting that their regulatory deﬁcits extend beyond experiences of distress. Individuals with high attachment avoidance, to the contrary, tend to minimize experiences of negative affect and to direct attention away from threat cues (Mikulincer et al., 2003). These ‘‘deactivating’’ strategies involve the denial or suppression of affective experience, the inhibition of affective expression, and distortion of encoding of affective experiences (BeckerStoll, Delius, & Scheitenberger, 2001; Kobak et al., 1993; Mikulincer et al., 2003). Importantly, both types of attachment insecurity are associated with the inability to derive affect-regulating beneﬁts from contact with attachment ﬁgures (Feeney, 1999). The affect-regulation conceptualization of attachment style is consistent with research on adults demonstrating that attachment anxiety and avoidance are associated with distinct patterns of affect-related appraisals and experiences over the life course (reviewed in Mikulincer et al., 2003). For example, securely attached adults report more positive and benign interpretations of others’ facial expressions (Magai et al., 2000), endorse more positive and less negative interpretations of both hypothetical and actual relationship events (Collins, 1996; Simpson, Ickes, & Grich, 1999; Simpson, Rholes, & Phillips, 1996), and make less hostile attributions of others’ motives (Mikulincer, 1998b) and more positive interpretations of others’ supportive behavior (Lakey et al., 1996). Correspondingly, securely attached individuals tend to report more frequent and more intense positive affect, whereas insecurely attached individuals report more negative affect (Feeney, 1999; Feeney & Ryan, 1994; Mikulincer & Orbach, 1995). These patterns have long-term implications for mental health. Numerous studies have detected strong associations between attachment anxiety and avoidance and numerous affective disorders, including depression, mania, dysthymia, panic disorder, agoraphobia, social phobia, generalized anxiety,

96

Lisa M. Diamond and Christopher P. Fagundes

and substance use (Mickelson, Kessler, & Shaver, 1997). Similar associations between attachment style and emotional adjustment and functioning have been found among adolescents (Cooper et al., 1998, 2004; Doyle & Markiewicz, 2005; Moris, Meesters, & van den Berg, 2003; Torquati & Vazsonyi, 1999). Speciﬁcally, negative affective predispositions and affectregulation problems in childhood and adolescence have been associated with maladaptive peer behavior, low social functioning, and conduct problems (Allen et al., 1998; Cassidy et al., 1996; Cooper et al., 1998; Repetti et al., 2002). Interestingly, attachment anxiety has been found to be reliably associated with internalizing problems such as anxiety and depression (Allen et al., 1998; Cooper et al., 1998; Kobak & Sceery, 1988; Kobak, Sudler, & Gamble, 1991; Rosenstein & Horowitz, 1996), whereas avoidance is consistently associated with externalizing behaviors such as aggression, rule-breaking, and peer hostility (Fagot & Kavanagh, 1990; Goldberg, 1997; Renken et al., 1989). Attachment insecurity also has important implications for the development of adolescent autonomy. As noted previously, a critical normative developmental transition during the adolescent years involves increased autonomy, independence, and differentiation from parents. However, the achievement of adolescent independency and autonomy does not preclude continued emotional support and connection, but in fact requires it (reviewed in Allen & Land, 1999), consistent with the notion that the provision of a secure base by parents facilitates normative processes of exploration (Feeney, 2007). Relatedly, one developing strain of research on adolescent autonomy suggests the particular importance of emotion- and self-regulatory processes. Speciﬁcally, Deci and Ryan’s (2000) self-determination theory postulates that autonomy is the degree to which behaviors are enacted with a sense of volition and choicefulness. Accordingly, adolescent autonomy is not achieved simply by separating and individuating from parents, but requires that youths develop the capacity to competently and knowledgeably select and endorse their actions, based on self-awareness of their motives, goals, and abilities. According to this perspective, the opposite of autonomy is not dependence but rather heteronomy (i.e., the feeling of being controlled in one’s actions by external forces or by internal compulsions). Clearly, youths’ abilities to reﬂect on, understand, and regulate affective experience should contribute directly to the sense of autonomous volition described by Deci and Ryan (2000). Hence, by providing the foundation for successful affect regulation, attachment security directly promotes healthy, age-appropriate trajectories of autonomy and self-determination. Finally, as noted subsequently, individual differences in attachment anxiety and avoidance also have long-term implications for adolescent and

Developmental Perspectives

97

adult physical health, via patterns of maladaptive physiological functioning associated with affect- and stress-regulation deﬁcits (Diamond & Hicks, 2004; Diamond, Hicks, & Otter-Henderson, 2006; Feeney, 2000; Powers et al., 2006; Repetti et al., 2002). We review such pathways in the next section.

C. PHYSIOLOGICAL PROCESSES LINKING ATTACHMENT AND AFFECT REGULATION Despite the decidedly non-biological slant of most research on infant and adult attachment (reviewed in Diamond, 2001; Spangler & Zimmermann, 1999), Bowlby conceptualized the attachment system as a fundamentally psychobiological system, especially with regard to its affect-regulating functions. Speciﬁcally, he posited two different ‘‘rings’’ of homeostasis that assist the individual in responding to major and minor stressors so that emotional security could be maintained and environmental exploration resumed (Bowlby, 1973). The inner ring comprises life-maintaining biological systems that govern ongoing physiological adaptation to external demands. The outer ring comprises behavioral (and particularly, interpersonal) strategies for coping and adaptation. From Bowlby’s perspective, the integrated functioning of these two levels is critical for optimal self-regulation. Numerous studies of both animals and humans have conﬁrmed Bowlby’s view. As noted earlier, early experiences of nurturant care appear to play a critical role in ‘‘tuning’’ multiple stress-regulatory systems in the orbitofrontal cortex that provide a foundation for adaptive affect regulation (see Glaser, 2000; Repetti et al., 2002; Schore, 1996a, 2000; Taylor et al., 2002). Hence, deﬁcits in infant–caregiver attachment not only disrupt children’s social and behavioral development, but also their biological capacities for maintaining homeostasis in the face of threat. These early regulatory problems create potential cascades of related dysregulation in immunological, endocrinological, and autonomic functioning (Cacioppo & Berntson, 2007; Gunnar, 2003; Kiecolt-Glaser et al., 2002b; Repetti et al., 2002; Ryff et al., 2001) with direct implications for long-term risks for a variety of pathophysiological processes and outcomes, including cardiovascular disease, diabetes, hypertension, and cancer (Croiset et al., 1990; Grossman, Brinkman, & de Vries, 1992; Hessler & Fainsilber Katz, 2007; Irwin, Hauger, & Brown, 1992; McEwen & Stellar, 1993; Munck & Guyre, 1991). To identify the developmental, intra-familial origins of such risk trajectories, researchers are increasingly adopting ‘‘biosocial’’ approaches

98

Lisa M. Diamond and Christopher P. Fagundes

to family life and child development (Booth, Carver, & Granger, 2000; Repetti et al., 2002) in which individuals and families are understood as the product of reciprocal inﬂuences among environmental, interpersonal, behavioral, psychological, and biological processes, unfolding over time (Cairns, Gariepy, & Hood, 1990; Gottlieb, 1991). In this view, biological predispositions set the stage for certain types of behavioral and psychological adaptation to environmental challenge. These adaptational patterns, which become increasingly regularized over time, have both immediate and long-term effects on physiological functioning. Although such dynamics involve numerous biological processes, we focus here on two systems that have particular relevance for affect regulation: the parasympathetic branch of the autonomic system and the hypothalamicpituitary-adrenocortical (HPA) axis of the endocrine system.

1. Parasympathetic Regulation of Heart Rate The functioning of the parasympathetic nervous system (PNS) in maintaining chronotropic control of the heart (sometimes called vagal regulation) has become one of the most widely researched physiological indices of affect regulation. The speciﬁc relevance of this physiological system for attachment-related phenomena is discussed at length elsewhere (Diamond, 2001), but key elements are reviewed here. Brieﬂy, both the PNS and the sympathetic nervous system (SNS) are involved in the moment-by-moment physiological changes triggered by environmental demands—changes in heart rate, blood pressure, sweating, and the like. Yet the SNS and the PNS have antagonistic effects on autonomic functioning, and thus stress responses such as heart rate acceleration can be brought about by activation of the SNS, withdrawal of the PNS, or some combination of the two. This has important implications for affect regulation because autonomic changes that are driven by adjustments in the PNS appear to be more rapid, more ﬂexible, and easier to disengage than SNS-dominated changes (Berger, Saul, & Cohen, 1989; Saul, 1990; Spear et al., 1979). Hence, individuals with greater PNS regulation of heart rate are conceptualized as having nervous systems that ﬂexibly react to and recover from environmental stressors, facilitating more effective affect regulation (Calkins, 1997; DeGangi et al., 1991; Porges, 1992; Porges, Doussard-Roosevelt, & Maiti, 1994). This is borne out by studies relating tonic levels of PNS chronotropic control (indexed by resting levels of respiration-related variability in heart rate, also known as respiratory sinus arrhythmia or RSA) to regulatory outcomes. For example, infants with greater PNS regulation of heart rate (i.e., greater RSA) are more facially expressive, more reactive to novel

Developmental Perspectives

99

events, and better able to sustain attention and avoid distraction (Porges, 1992; Stifter & Fox, 1990). In contrast, infants and children with lower RSA show a compromised capacity for self-soothing after psychological stress and are less easily and effectively soothed by others (reviewed in Porges, 1991). They also show poorer emotional control and higher behavioral inhibition (Fox, 1989; Snidman, 1989). In adults, higher RSA is associated with more effective emotional and behavioral responses to stress (Fabes & Eisenberg, 1997), whereas lower levels are associated with depression, anger, mental stress, generalized anxiety, and panic anxiety (reviewed in Brosschot & Thayer, 1998; Friedman & Thayer, 1998; Horsten et al., 1999). Historically, far fewer studies of PNS functioning and affect regulation have been conducted among adolescents than among adults, infants, or children, but several studies of adolescents have conﬁrmed associations between problems with affective, attentional, and behavioral regulation and low PNS regulation of heart rate (Beauchaine, Kopp, & Mead, 2007; Kibler, Prosser, & Ma, 2004; Tobin & Graziano, 2006), and individual differences in PNS regulation in childhood appeared to be preserved into adolescence (El-Sheikh, 2005). Future coordinated assessment of PNS regulation, affect regulation, and attachment dimensions can make important contributions to research on the biopsychological context of adolescent psychosocial development.

2. HPA Axis Activity As reviewed by Seeman (2001), interpretations of environmental demands and one’s resources (both social and nonsocial) for meeting these demands are processed ﬁrst by the neocortex and then fed to the amygdala and hippocampus, leading to systemwide neuroendocrine activation (Bovard, 1985; Gray, 1995; LeDoux, 1995; McEwen, 1995; Schneiderman, 1983; Williams Jr., 1985). Speciﬁcally, the hypothalamus signals the anterior pituitary to release a cascade of neurochemicals that operate in concert to increase blood glucose levels and modulate immune activity in response to the perceived demand. Thus, information-processing biases that consistently favor negative and threat-related interpretations of environmental events and stimuli can produce patterns of ‘‘warped emotion processing’’ (Repetti et al., 2002, p. 351) that trigger maladaptive proﬁles of physiological activation. Accordingly, multiple studies of animals and humans have documented associations between HPA hyperreactivity in response to stress and patterns of cognitive and behavioral affect regulation. For example, individuals with exaggerated HPA reactivity show deﬁcient coping strategies and exaggerated experiences of negative affect (reviewed in Scarpa & Raine, 1997; Stansbury & Gunnar, 1994).

100

Lisa M. Diamond and Christopher P. Fagundes

There are also individual differences in tonic levels of HPA activity, which have been linked to chronic stress. Yet importantly, these studies have found that chronic stress can result in either disproportionately high or low cortisol levels (Miller, Chen, & Zhou, 2007), depending on a variety of factors. Consistently high levels of tonic HPA activity among individuals exposed to chronic stress and strain suggest that failures to down-regulate stress and negative affect are associated with dysregulation of the normal feedback processes through which HPA activation is typically ‘‘shut down’’ once sufﬁcient levels of cortisol are present in the bloodstream to meet environmental demands. Yet studies have also found associations with stress and chronically low or ‘‘blunted’’ HPA activity, which have been interpreted as a potentially adaptive mechanism for protecting the brain from the detrimental effects of sustained exposure to cortisol, which include deﬁcits in immune functioning (Coe et al., 1988) as well as memory and attentional process (Kirschbaum et al., 1996; Lupien et al., 1994; McEwen et al., 1992). Yet this ‘‘blunting’’ response may entail long-term regulatory ‘‘costs’’ in the forms of psychobiological dysfunctions in stress-regulation and immune function (Buske-Kirschbaum et al., 1997; Hart, Gunnar, & Cicchetti, 1995). If secure attachment fosters effective affect regulation, this may be reﬂected in adaptive patterns of HPA axis functioning. Sure enough, high levels of physical affection and warmth between caregivers and their infants during stressful circumstances have been tied to normal HPA activation proﬁles in response to environmental demands (Chorpita & Barlow, 1998; Hertsgaard et al., 1995). Additionally, Flinn and England (1995) found that HPA activation in response to such normal demands varied as a function of family environment—but most notably, high vs low levels of maternal care—in children aged 2–18. Hence, both tonic- and stress-related patterns of HPA activity provides a potential window into attachment-related disruptions in affect regulation.

D. TOWARD A PROCESS-ORIENTED, BIOBEHAVIORAL APPROACH To summarize, a growing body of research suggests that the welldocumented associations between adolescent attachment insecurity and deﬁcits in psychosocial and psychological functioning (Allen et al., 1998; Cooper et al., 1998, 2004; Hauser, Gerber, & Allen, 1998; Kobak et al., 1991; Rosenstein & Horowitz, 1996; Sroufe, 2005; van Ijzendoorn, Schuengel, & Bakermans-Kranenburg, 1999) may be mediated by attachment-related deﬁcits in affect regulation. Given the abrupt status transitions and the

Developmental Perspectives

101

increased emotional challenges of the adolescent years (Buchanan, Eccles, & Becker, 1992; Larson & Richards, 1994; Spear, 2000), biobehavioral investigations of attachment and affect regulation during this period would seem to provide powerful new insights into pathways of physical and psychological risk vs resilience from childhood to adulthood. Yet this requires that we identify the specific cognitive, physiological, and behavioral mechanisms through which affect regulation and attachment shape adolescent adjustment. This is no easy feat: As many researchers have noted, the concept of affect and emotion regulation has been deﬁned and operationalized in many different ways by theorists and researchers focusing on different stages of life (Bridges, Denham, & Ganiban, 2004; Campos, Frankel, & Camras, 2004; Cole, Martin, & Dennis, 2004; Gross, 1999). Even differentiating between regulated and unregulated affect is problematic. As Gross (1999) noted, the distinction between these concepts implies that an emotional experience or expression ‘‘after’’ regulation is fundamentally different from its ‘‘unregulated’’ state, yet some have argued that affect and emotion are always regulated to some degree (Fridja, 1986; Tomkins, 1984). Gross (1999) adopts a middle ground between these two extremes and argues for an emphasis on relative regulation of different aspects of emotional phenomena under different circumstances. However, he cautions that an ongoing and critical challenge for research on such processes involves specifying whether affect regulation has even occurred, what components of emotion have been regulated, and how regulation has altered such components. Then there is the question of which regulatory processes: Different researchers have emphasized a range of different conscious and unconscious regulatory capacities and strategies, both adaptive and maladaptive: for example reactivity, recovery, suppression, disengagement, rumination, reappraisal, attention shifting, distraction, problem-focused coping, selfsoothing, support-seeking, and the like (Belsky, Friedman, & Hsieh, 2001; Broderick, 1998; Glynn, Christenfeld, & Gerin, 2002; Gross, 1998). Furthermore, different combinations of these processes, in concert with other individual differences in affective experience, appear to produce a range of different affective phenomena. For example, Gohm (2003) demonstrated important differences between the regulatory strategies of emotionally reactive individuals as a function of whether such individuals also had high emotional clarity, representing their capacity to understand and interpret the basis for their emotional responses. Speciﬁcally, reactive individuals with low emotional clarity were more easily ‘‘overwhelmed’’ by their emotions and hence made greater efforts to avoid and attenuate strong emotional experiences, whereas those with high emotional clarity appeared better able to tolerate these experiences and their immediate effects. These

102

Lisa M. Diamond and Christopher P. Fagundes

ﬁndings support emerging views of ‘‘emotional intelligence’’ (Salovey & Mayer, 1990) that emphasize how psychological well-being and interpersonal functioning are facilitated by the accurate perception, appraisal, and expression of emotion, effective utilization of emotion in the service of cognitive processing, effective comprehension and communication of emotion-relevant concepts, and the capacity to regulate one’s own emotions and those of others. Clearly, if greater emphasis on the affect-regulation functions of attachment, particularly during the adolescent years, holds promise for bridging the infant-child and adult attachment literatures, this emphasis must include more systematic delineation of speciﬁc affect-regulation processes, including motivated processes such as suppression and reappraisal, basic capacities such as emotional clarity and efﬁcient recovery, and physiological indices of regulatory capacities, such as parasympathetic functioning and HPA activity. Up until now, studies adopting such an integrative, differentiated approach have tended to focus on either adults or infants (with notable exceptions, such as Granger, Weisz, & Kauneckis, 1994; Moss et al., 1999; Walker, Walder, & Reynolds, 2001), and little of this research has speciﬁcally investigated links to attachment-related processes and their normative changes over the adolescent years. Thus, comprehensive biobehavioral models of normative affect-regulation processes across the adolescent years, individual differences in these processes, and their links to attachment phenomena remain to be developed. We have sought to contribute to this long-term goal in our own longitudinal, biobehavioral research on adolescent attachment to parents and peers, affect regulation, and physical and mental health. In the next section, we review ﬁndings from this ongoing program of research suggesting that normative developmental processes and individual differences in attachment-related patterns of affect regulation help to elucidate adolescent trajectories of well-being. We hope that this type of biosocial research can provide a model of how a greater emphasis on affectregulation processes can contribute to lifespan models of attachment and affect regulation that integrate research ﬁndings from both the infant-child and adult literatures.

III. Attachment and Affect Regulation During Adolescence A. BACKGROUND AND METHODS The ﬁndings presented here come from an ongoing longitudinal investigation of 103 14-year-old youths (51 boys and 52 girls) that examines

Developmental Perspectives

103

linkages among the quality of youths’ relationships with their parents and peers, normative developmental transitions in attachment hierarchies, individual differences in attachment style and affect regulation, and individual differences in physiological indices of affect regulation. Participants were recruited from public and private high schools around Salt Lake City, Utah. Each youth visited our laboratory with both parents (37 youths were not currently living with their father, and completed the laboratory visit with their mother only), where they completed questionnaires assessing the following constructs: 1.

2.

Attachment anxiety and avoidance: To measure adolescents’ attachment anxiety and avoidance with respect to their parents, we administered a revised version of Miller and Hoicowitz’s (2004) Adolescent Attachment Scale, which is based on the widely used Experiences in Close Relationships Inventory, designed to assess romantic attachment anxiety and avoidance (Fraley, Waller, & Brennan, 2000). Sample items assessing attachment anxiety include ‘‘I sometimes wonder of my mother really loves me’’ and ‘‘I worry that my mother doesn’t care about me as much as I care about her.’’ Sample items assessing avoidance include ‘‘it’s hard for me to let myself count on my mother,’’ and ‘‘I don’t feel comfortable opening up to my mother.’’ Youths completed the measure separately in relation to their mother and father: For the analyses reported here, responses were averaged across parents (ancillary analyses found that this did not change the major ﬁndings). Parenting characteristics: We administered Uchino and colleagues’ Social Relations Index (Uchino et al., 2001) to assess the degree of unpredictability and helpfulness-enthusiasm that youths perceived in their parents. This measure asks individuals to rate social partners’ unpredictability and helpfulness-enthusiasm speciﬁcally in response to times when (1) you achieve something good and (2) you need help or assistance. This item was administered separately for mothers and fathers, and the responses were averaged. To assess parenting style, we administered Schaefer’s (1965) widely used assessment of parental warmth, discipline, and psychological control (the latter indexing the degree to which parents use strategies like love-withdrawal, expression of their own negative affect, and manipulation in order to control their child’s behavior). To measure emotional climate in the family, we administered Halberstadt’s Self-Expressiveness in the Family Scale (Halberstadt et al., 1995), which measures the degree of positive and negative emotional expressiveness in youths’ families.

104

3.

4.

5.

6.

7.

Lisa M. Diamond and Christopher P. Fagundes

Adjustment: We used the following subscales from the Youth Self Report (Achenbach, 2001), a widely used measure of adolescent adjustment: attentional problems, internalizing problems, and externalizing problems. We also administered the ‘‘self’’ subscale of Patrick, Edwards, and Topolski’s (2002) Adolescent quality of life scale to measure youths’ overall satisfaction with their sense of self. To measure feelings of loneliness in the family context, we administered the family subscale of the Measure of Social and Emotional Loneliness (DiTommaso & Spinner, 1993). To measure depression, we used the CESD-R (Radloff, 1977). Affect regulation: Our measures of affect regulation included measures of recovery (Rothbart & Derryberry, 1981), indexing how quickly individuals’ calm down and recovery after intense emotional activation, suppression (Gross, 1998) indexing the strategy of dealing with negative emotions intentionally trying to dampen them, repair, a composite measure indexing the use of positive reframing and reappraisal to moderate one’s emotions, comprised of items from the ‘‘repair’’ subscale of the trait meta-mood scale (Salovey et al., 1995), Gross’ (1998) measure of reappraisal, and Moos’ (1988) measure of positive reframing. To assess individuals’ ability to detect and identify their emotions, we administered Salovey et al.’s (1995) measures of emotional clarity, assessing the perceived clarity with which respondents experience their emotions, and emotional attention, measuring the extent to which respondents attend to and value their emotional experiences. Exploration and dependency: To measure exploration, we used selected items from Block and Kremen’s (1996) Ego-Resiliency scale. Sample items included ‘‘I like to do new and different things,’’ and ‘‘I enjoy dealing with new and unusual situations.’’ To measure dependency, we included selected items from the aforementioned Youth Self-Report, including ‘‘I depend too much on adults’’ and ‘‘I act young for my age.’’ Physical affection with parents: We measured the degree to which adolescents hugged their parents, cuddled with them, and engaged in routine forms of physical contact such as adjusting hair or behavior and patting arms and legs (Diamond, 2000). This scale was based on previous reviews of the forms of physical contact considered most speciﬁc to attachment relationships (Hazan & Zeifman, 1994). Attachment components: To measure the dimensions of proximity seeking, safe haven, and secure base, we used items from the companionship, intimacy, and reliable alliance subscales of the Network of Relationships Inventory (Furman & Buhrmester, 1992). These NRI subscales are highly similar to the aforementioned WHOTO (Hazan & Zeifman, 1994), but the NRI has been well-validated and is more

Developmental Perspectives

8.

105

widely used than the WHOTO, permitting more direct comparisons with previous research. Also, the NRI permits ratings of the specific degree to which certain social partners are utilized for speciﬁc attachment-related functions, and hence allows two different individuals to have the same rating, whereas the WHOTO uses a forcedchoice format. A sample companionship (i.e., proximity seeking) item is ‘‘how much do you play around with and have fun with this person?’’ A sample intimacy (i.e., safe haven) item is ‘‘how much do you share your secrets and private feelings with this person?’’ A sample reliable alliance (i.e., secure base) item is ‘‘how sure are you that this relationship will last no matter what?’’ Participants completed this measure with regard to both parents, their best friend, their romantic partner (if applicable), a close relative, a close but unrelated adult, and a sibling. Physiological stress regulation: Two indices of physiological stress regulation were measured: RSA, a measure of parasympathetic control over heart rate, and HPA axis activity, measured via salivary cortisol. RSA was measured during a quiet baseline assessment during which youths’ breathed in time with a recorded tape (following recommendations of Grossman, Stemmler, & Meinhardt, 1990), because variations in respiratory rate can interfere with measurement of individual differences in RSA (Berntson et al., 1997; Grossman, Karemaker, & Wieling, 1991). Details of the physiological recording equipment can be found in previous published reports (Diamond & Hicks, 2005; Diamond et al., 2006). RSA was assessed on the basis of ECG and respiration data. Interbeat intervals (IBIs) were calculated as the time in milliseconds between successive R waves in the electrocardiogram, and the ‘‘peak-to-valley’’ method (Grossman & Svebak, 1987) was used to derive RSA on the basis of these IBIs. This method computes the difference between the heart period between inspiration onset and expiration onset. Following standard practice, RSA values were logged before analysis in order to normalize their distribution. Additionally, several days after the laboratory visit, participants provided a saliva sample at home, taken at the same time as their laboratory session had begun. Because of diurnal variation in cortisol, all laboratory sessions were scheduled to begin between 4 and 6 pm.

B. PATTERNS OF ATTACHMENT TRANSFER Previous research on the process of transferring attachment from parents to peers suggests that in early to mid-adolescence, most youths will continue

106

Lisa M. Diamond and Christopher P. Fagundes

to utilize parents as a primary base of security, but will preferentially seek companionship and safe haven from peers. Yet as critiqued previously, the reliance of previous studies on forced-choice methodologies has not allowed for investigation of mixed patterns of parent- and peer-directedness in attachment functions. Accordingly, we were interested in revisiting the question of attachment transfer and speciﬁcally investigating such mixed patterns. In short, does transfer appear to occur ‘‘in order,’’ as suggested by Hazan and Zeifman (1994), beginning with proximity seeking, followed by safe haven and then secure base? Overall, 53% of youths rated one of their parents higher on the reliable alliance subscale of the NRI (which, for the sake of efﬁciency, we will call ‘‘security’’) than they rated a peer (either their best friend or their romantic partner); 39% gave equal ratings to both, and 8% rated a peer higher. As for the intimacy subscale, 65% of youths rated one of their parents higher than they rated a peer, 13% gave equal ratings to both, and 22% rated a peer higher. For companionship, 45% rated a parent higher than a peer, 17% gave equal ratings, and 37% gave peers higher ratings. This pattern of results supports some aspects of the classic picture of attachment transfer, but clearly demonstrates the importance of investigating patterns of mixed peer–parent orientation. Speciﬁcally, consistent with previous conceptualizations of normative attachment, very few 14-year-olds sought security from peers. Yet they were actually more parent-oriented for intimacy than for security. This may reﬂect subtle differences between interpersonal intimacy and ‘‘safe haven,’’ strictly deﬁned. Youths might preferentially nominate peers for disclosure of secrets, and yet still consider parents better sources of emotional support in times of stress. Notably, research by Reis and Franks (1994) has shown that interpersonal intimacy and support are distinct relationship dimensions, and that the mental and physical health beneﬁts of interpersonal intimacy are actually mediated by support-provision. Hence, peerdirectedness in interpersonal intimacy does not fully capture transfer of the safe haven component of attachment. As for companionship, although the classic portrait of attachment transfer suggests that youths should be most peer-oriented in proximity seeking, only about a third of youths reported more companionship with peers than with parents. This suggests that even in mid-adolescence, parents remain important targets for even the most elemental attachment functions. Also, contrary to the notion that romantic involvement is the impetus for youths to transfer security-seeking from parents to peers (Hazan & Zeifman, 1994), youths with current romantic partners (25% of the sample) were more likely—rather than less likely—to rate parents higher than peers

Developmental Perspectives

107

as sources of security (62% vs 50%). Youths without romantic partners were more likely to give peers and parents equal ratings (45% vs 20%). As for companionship and intimacy, there were no signiﬁcant differences between youths with and without romantic partners. Of course, given the age of our participants, their romantic relationships simply may lack the depth and intimacy that characterizes the more ‘‘attachment-like’’ romantic involvements of late adolescence and early adulthood; hence, in later years, romantic involvement will probably be more directly associated with peerdirectedness in attachment functions. Clearly, there is substantial variation in the normative pattern of ‘‘attachment transfer’’: What are the implications of these variations? To answer this question, we compared the three security groups—parentoriented, parent-and-peer-oriented, and peer-oriented—on internalizing and externalizing problems, attention problems, depression, and quality of life, and found a signiﬁcant effect of group membership, controlling for gender. Inspection of the univariate tests revealed that for each outcome except for quality of life, youths who were peer-oriented reported the poorest adjustment, whereas the other two groups did not differ from one another. Hence, seeking security from peers while continuing to seek security from parents does not appear to be detrimental. The ideal pattern, then, may not be to transfer attachment, but to broaden it, beginning to explore new attachments to peers while still actively maintaining functional relationships to parents. Even Weiss’ (1982) notion of ‘‘attachment ﬁgures in reserve’’ might underestimate the importance of parental ties for youths in this age range, given that they continued to view parents as primary targets for companionship as well as security. The group ‘‘to watch,’’ then, is the group that proved to be disproportionately peer-oriented with respect to security. What speciﬁc processes and mechanisms explain their maladjustment? One possibility is that the link between peer-orientation and maladjustment is simply a function of the fact that peer-oriented youths start out with unusually low perceptions of security in their parental relationships, so that as they increasingly—and normatively—seek security from peers over the course of adolescence, peers inevitably end up ‘‘ranking higher’’ than parents. If this is the case, then the association between peer-orientation and maladjustment should no longer be signiﬁcant after controlling for overall levels of parental security. Sure enough, this was the case. When we entered youths’ ratings of parental security into the aforementioned model predicting maladjustment from peer-vs-parent-orientation, the effect of peer-vs-parent-orientation was no longer signiﬁcant. Inspection of the univariate tests revealed that after controlling for overall security levels, peer-vs-parent-orientation remained signiﬁcantly

108

Lisa M. Diamond and Christopher P. Fagundes

associated with externalizing problems, and marginally so with attention problems, but not with internalizing problems or depression. In contrast, overall parental security was signiﬁcantly associated with internalizing problems, depression, and quality of life and not with externalizing or attention problems. These results suggest that one potential reason that ‘‘premature’’ transfer of security-seeking to peers from parents is potentially maladaptive is that youths who take this developmental route have disproportionately low levels of parental security to begin with, which appears to have detrimental associations with adolescent psychosocial functioning regardless of whether youths attempt to ‘‘compensate’’ by seeking greater security from peers. Consistent with the ﬁndings of other research (reviewed in Repetti et al., 2002), these results clearly indicate that security from peers cannot, in fact, compensate for low parental security.

C. BRIDGES TO AUTONOMY: EXPLORATION AND DEPENDENCY As noted earlier, historical perspectives on adolescent autonomy suggest that individuation from parents is a precursor for age-appropriate exploration; if so, then one would expect that youths who are peerorientated in security-seeking, companionship, and intimacy should report greater tendencies toward exploration and also less dependency on parents. We tested this hypothesis, and found that although peer-vs-parent orientation in security was associated with both exploration and dependency, peer-vs-parent orientation in intimacy and companionship were not. Speciﬁcally, youths who were parent-oriented in security had the lowest levels of exploration, whereas those who were peer-oriented or equally parent- and peer-oriented reported comparable—and higher—levels of exploration. For dependency, there was an interaction between peer-vsparent orientation in security and gender: Among boys, peer-orientation was actually associated with higher levels of dependency, whereas the lowest levels of dependency were observed among youths who were equally oriented to parents and peers. In girls, groups did not differ. These ﬁndings provide further support for the notion that adolescent adaptation is best facilitated by a pattern of security-seeking from both parents and peers, rather than an exclusive focus on either parents or peers. In concert with the ﬁndings on overall adjustment, we think that this pattern of results harkens back to Deci and Ryan’s (2000) self-determination theory. As reviewed earlier, this theory suggests that autonomous adolescents are not merely those who successfully separate and individuate from their parents, but those whose conﬁdence in their own judgment, and

Developmental Perspectives

109

awareness of their own values and motives, leads them to conﬁdently endorse and defend their actions. According to this theory, the optimal exercise of adolescent volition, judgment, and self-reliance is fully compatible with—and potentially enhanced by—the continued use of parents as bases for emotional security (Soenens et al., 2007). Adolescents who are disproportionately peer-oriented regarding security may begin to assert independence without developing volition and choicefulness, which may hamper both exploration and overall adjustment, and facilitate continued dependency.

D. ATTACHMENT STYLE AND AFFECT REGULATION We were also interested in whether attachment style moderated these patterns, and so we added attachment anxiety and avoidance (measured separately for each parent and averaged, unless noted otherwise) to the models described previously. The results were striking: Contrary to previous research suggesting that attachment anxiety has stronger associations than avoidance with psychological adjustment in both adolescence and adulthood (Cooper et al., 1998, 2004; Mickelson et al., 1997), we found that attachment avoidance was the only signiﬁcant overall predictor of youths’ psychological adjustment and was strongly related to each of the adjustment outcomes. Perhaps most notably, parental security and parentvs-peer orientation were no longer signiﬁcantly associated with any of the adjustment outcomes after controlling for attachment avoidance, indicating that both of these effects were mediated by avoidance. Why—and through what psychological mechanisms—does avoidance have such detrimental implications for adjustment? Attachment theory suggests two possibilities: First, avoidant adolescents may have affect-regulation deﬁciencies that predispose them to poor adjustment; second, avoidant adolescents have such poor quality family relationships that their resulting feelings of loneliness and isolation predispose them to poor adjustment. To examine each of these possibilities, we added the set of affect-regulation measures (repair, clarity, suppression, recovery, and attention) to the model, and found that they were strongly related to adjustment, whereas the effect of avoidance was no longer signiﬁcant. In contrast, when loneliness within the family was added to the model, it did not make a unique contribution. The preeminence of affect regulation in this model prompted us to return to our initial peer-orientation groupings, and examine whether youths who (1) were disproportionately peer-oriented regarding security, or (2) had low overall ratings of security to their parents, were characterized by affectregulation deﬁcits. In this model, we began by predicting repair, clarity,

110

Lisa M. Diamond and Christopher P. Fagundes

attention, recovery, suppression from parent-vs-peer orientation. There was a strong omnibus effect, and inspection of the univariate tests showed that peer-oriented youths had poorer affect regulation on all of the dimensions except for suppression. After adding parental security to the model, parentvs-peer orientation remained strongly associated with affect regulation, whereas parental security was not. Hence, unlike the adjustment outcomes, youths’ speciﬁc capacities and strategies for affect regulation are uniquely associated with peer-orientation in security, regardless of youths’ overall levels of parental security. Given our previous ﬁndings regarding the signiﬁcance of attachment avoidance for youths’ adjustment, we then added attachment avoidance to the affect-regulation model to see whether it was mediating the effect of parent-vs-peer orientation. This proved to be the case: Avoidance had signiﬁcant unique associations with each of the affect-regulation measures, and neither parent-vs-peer orientation nor parental security was now associated with these outcomes. Although our data do not permit causal inferences, our ﬁndings are consistent with a potential developmental pathway in which avoidantly attached youths’ poor affect-regulation predisposes them to low parental security, high levels of peer-orientation, and maladjustment. Although these youths’ relative security with parents and peers also contribute to their levels of maladjustment, affect-regulation deﬁcits appear to play the most preeminent role. This provides strong evidence for the importance of affect regulation in understanding normative and non-normative trajectories of attachment during the transition from childhood to adulthood, and their implications for youths’ well-being.

E. PHYSIOLOGICAL CORRELATES OF AFFECT REGULATION Our ﬁndings of consistent associations among attachment domains, affect regulation, and overall adjustment ﬁnd further support in our analyses of physiological correlates of affect regulation. Speciﬁcally, boys’ afternoon levels of cortisol, measured at home, were associated with their degree of attachment anxiety to mothers and their perceptions of maternal support-enthusiasm. That is, higher levels of attachment anxiety were associated with lower cortisol levels, whereas perceptions of supportenthusiasm were associated with higher levels. These ﬁndings are consistent with previous research documenting that adolescent boys with affect regulation and conduct problems tend to show dampened cortisol levels (Loney et al., 2006; Ramirez, 2003; Shoal, Giancola, & Kirillova, 2003). Also, another study found that boys whose parents had divorced before the

Developmental Perspectives

111

age of 10 showed dampened cortisol responses in young adulthood after stimulation with corticotropin releasing hormone, in comparison to controls, suggesting the speciﬁc importance of attachment-related stressors for HPA dysregulation (Bloch et al., 2007). Contrary to the ﬁndings with boys, we found that girls who described their parents as unpredictable had higher evening levels of cortisol, at the trend level; yet notably, this effect was mediated by girls’ capacities for emotional clarity (which was signiﬁcantly associated with both parental unpredictability, and girls’ home cortisol levels). When added to the model predicting home cortisol levels, emotional clarity was marginally associated with lower cortisol but the effect of parental unpredictability was no longer signiﬁcant. These ﬁndings suggest intriguing possibilities regarding the familial origins of affect regulation and their associations with HPA functioning. One possibility is that girls whose parents display unpredictable reactions to their own positive and negative emotions develop difﬁculties with affective awareness and understanding that are manifested in heightened HPA activation at home. This is consistent with prior research showing that unpredictability and negativity in the home places children at risk for heightened emotional reactivity and, eventually, chronic emotional insecurity (Cummings & Davies, 1996). Girls’ heightened HPA levels might stem from these chronic appraisals of uncontrollability and insecurity in the family’s emotional dynamic, consistent with prior research on links among interpersonal experiences, emotional appraisals, and HPA activity (Seeman, 2001). Our ﬁnding of such linkages among girls, but not boys, is consistent with previous studies documenting similar gender differences (McCormick & Mathews, 2007; Schiefelbein & Susman, 2006). Alternatively, patterns of heightened HPA activity may ‘‘drive’’ interpersonal problems, predisposing girls to difﬁculties with affective self-awareness which eventually lead them to perceive their parents as unpredictable. If the latter were so, one would expect that when predicting girls’ emotional clarity from their cortisol levels and their perceptions of parental unpredictability, heightened cortisol levels should mediate the association between emotional clarity and perceptions of parental unpredictability. Yet when we tested this possibility, we found that it was not the case. Rather, both parental unpredictability and home cortisol levels made unique contributions to emotional clarity. As for PNS functioning, it also showed gender-speciﬁc patterns. In both girls and boys, higher baseline RSA was associated with greater externalizing problems, less physical affection from parents, less positive emotional expression in the family, and less emotional repair, clarity, and attention. The strong association with emotional clarity is particularly

112

Lisa M. Diamond and Christopher P. Fagundes

notable, given the aforementioned association between girls’ emotional clarity and their HPA activation. Given that emotional clarity was associated with each of the other variables (with the exception of physical affection), we ran a series of analyses to determine whether the link between RSA and affect regulation mediated these associations, and found this to be the case. These ﬁndings pose a contrast to our earlier work with adults, in which we found lower resting RSA among adult men who reported high attachment anxiety and low perceptions of security in their current romantic relationships (Diamond & Hicks, 2005). Notably, these studies did not include the detailed assessments of adjustment and of individual differences in affect regulation that we included in the present study (although we did ﬁnd that high-anxious men’s low RSA mediated their tendencies to show poor self-reported emotional recovery from laboratoryadministered stress). One possible explanation for the different pattern of results for adults vs youths is that as these youths develop and as their individual differences in affect regulation and adjustment canalize them into speciﬁc types of intimate relationships, their differential capacities for PNS regulation will eventually be made manifest in poorer quality romantic relationships and—accordingly—both heightened attachment anxiety and lower perceptions of relationship-speciﬁc security. The youths in our present sample may simply not have had enough romantic experience for such effects to emerge. Yet the pattern of ﬁndings reviewed previously suggests the possibility that we are detecting, at an early developmental stage, precursors to the patterns of association between attachment anxiety and PNS functioning that we detected in our previous studies of adult men. The notion of a developmental component to associations between affect regulation and attachment style is also consistent with the fact that although most studies of adults ﬁnd robust associations between romantic attachment anxiety and maladjustment, we did not ﬁnd this pattern among our adolescent participants. Instead, we found avoidance to be more robustly associated with negative psychological outcomes. Of course, this may be attributable to differences between the mental health implications of maternal vs romantic attachment anxiety. Perhaps these youths’ current affect regulation and adjustment problems will eventually propel them into the types of romantic attachments which serve to reinforce their anxieties and emotional difﬁculties, regardless of whether they are also carrying over residual attachment-related anxieties from their parental attachments. Hence, by the time they are adults, romantic anxiety may be ‘‘codetermined’’ by both affect regulation problems and concrete romantic experiences.

Developmental Perspectives

113

IV. Implications and Future Directions The results reviewed here suggest the promise of integrative, biosocial research on adolescent attachment, affect regulation, and well-being. In particular, developmental research focusing on the affect-regulation functions of attachment provides a fruitful way to build theoretical and empirical bridges between the infant-child and adult literatures, toward the eventual goal of developing integrative lifespan models of the attachment system and its impact on physical and mental functioning from the cradle to the grave. Of course, the data we have presented focus on one wave of assessment, and are therefore unable to answer fundamental questions of causation and to examine patterns of reciprocal linkage over time: For example, to what extent do early-appearing, temperamental deﬁciencies in affect regulation ‘‘drive’’ the development of attachment insecurity (instead of vice versa) via strained, asynchronous parent–child interactions? Are there speciﬁc developmental moments during which problematic trajectories of attachment insecurity, affect regulation, and adolescent internalizing and externalizing problems can be fruitfully redirected? Answers to these questions will come with continued longitudinal investigation. For now, we want to conclude by highlighting several additional avenues for future research which can make signiﬁcant contributions toward the development of generative lifespan models of attachment and affect regulation.

A. INTEGRATION OF ATTACHMENT AND AFFECT REGULATION WITH OTHER REGULATORY PROCESSES Historically, links between affect regulation and broader processes of selfregulation have been more explicitly discussed by researchers focusing on infant-child development than those focusing on regulatory processes in adolescents and adults. For example, Siegel (1999) noted that because affect and emotion reﬂect the mind’s assignment of value to internal and external events, and because they consequently direct the distribution of attentional resources to engage these events, affect regulation in infancy ‘‘can be seen at the center of the self-organization of the mind’’ (p. 245). Dodge (1991) placed speciﬁc emphasis on cognitive processes, contending that ‘‘y all information processing is emotional, in that emotion is the energy that drives, organizes, ampliﬁes, and attenuates cognitive activity and in turn is the experience and expression of this activity’’ (p. 159). In their research on infants and children, Fox and Calkins (2003) have highlighted links between affect regulation and behavioral self-regulation, noting that effective management of affective reactivity is critical for motivating approach

114

Lisa M. Diamond and Christopher P. Fagundes

behaviors and inhibiting withdrawal behaviors that might otherwise interfere with children’s goal pursuit or their compliance with rules and/or expectations. Investigating such interconnections among the organizing and valuing functions of affect and emotion, information processing, social behavior, and goal pursuit at all stages of life holds great promise for elucidating the multiple interconnecting mechanisms through which attachment security promotes adjustment and well-being across the life course. Another advantage of such an approach is that it would provide a valuable corrective to views of attachment and affect regulation which implicitly presume that the role of attachment ﬁgures is to unilaterally down-regulate intense (and presumably disruptive) experiences of affect so that goal pursuit can proceed or resume. This view reﬂects outmoded conceptualizations of emotion and cognition as separate and often opposing processes (i.e. hot vs cold processing, thinking vs feeling, emotion-focused vs problem-focused coping—for reviews and critiques see Isen, 2003; Isen & Hastorf, 1982; Stanton et al., 1994, 2000). We would advocate, instead, investigation of how attachment ﬁgures assist with more subtle and nuanced forms of affective regulation toward the goal of optimizing affective ‘‘input’’ into a range of situation-speciﬁc cognitive and behavioral processes. As we noted earlier, we ﬁnd self-determination theory (Deci & Ryan, 2000) to be a useful framework for developing integrative conceptualizations of the reciprocal linkages among affect regulation, self-regulation, attachment, and autonomy during adolescence. Speciﬁcally, we expect that the backdrop of attachment security will promote the ability of adolescents to develop a sense of volition and choicefulness regarding a range of goals, requiring a range of regulatory skills, and to take the necessary exploratory risks to achieve their goals. Feeney (2007) has most explicitly articulated the argument that in the realm of attachment, ‘‘dependence’’ in the form of emotional security promotes rather than hinders eventual independence and autonomy. Future investigation of adolescents’ development of selfregulation across a broader variety of domains, and how each developmental trajectory is facilitated (perhaps at different maturational stages) by attachment security, can contribute to a comprehensive, process-oriented understanding of the web of interconnections among attachment and multiple regulatory processes from infancy to adulthood.

B. A DYADIC APPROACH TO ADOLESCENCE One of the major developments within relationship research since the late 1990s involves the increased emphasis on dyadic approaches to modeling

Developmental Perspectives

115

and measuring interpersonal phenomena (Gable & Reis, 1999; Lyons & Sayer, 2005). This approach can make important contributions to investigations of attachment and affect regulation during the adolescent years, particularly given the complex negotiation between intimacy and autonomy which characterizes this stage of life. In particular, we think that dyadic approaches pull for greater speciﬁcity in conceptualizing the relative roles of self and other in the context of affect regulation. As reviewed previously, the development of affect regulation from childhood to adolescence can be cast as a gradual transition from reliance on a sensitive, responsive ‘‘other’’ for regulatory assistance to reliance on one’s own regulatory skills and capacities, such as attention shifting, active coping, or selective approach and avoidance (Calkins et al., 1998; Kobak et al., 1993; Rothbart, 1991; Thompson, 1994). This conceptualization presumes meaningful boundaries between regulatory processes that reside in the ‘‘self’’ vs those that reside in the ‘‘other.’’ However, research increasingly suggests that such boundaries might be relatively ﬂuid, and that future developmental research should more closely attend to the multiple bidirectional, co-regulatory processes that unfold in different contexts, with different constraints, at different stages of life. This approach has already been adopted by researchers investigating the development of self-regulation in infancy. Beebe and Lachmann (1998), for example, have argued for greater attention to how ‘‘dyadic process may (re-)organize both inner and relational processes, and reciprocally, how changes in self-regulation in either partner may alter the interactive process’’ (p. 481). Fogel (1992) has similarly emphasized that social behavior, communication, and emotions do not reside ‘‘in’’ the infant, but are continuously constructed in the course of direct interaction with the caregiver. Interestingly, neurobiological research provides converging support for this dyadic approach. The cascade of psychobiological effects of infant–caregiver interactions — from experience-expectant and experiencedependent proliferation and pruning of neural circuits (Schore, 1996a, 1996b) to endocrinological responses to stress and soothing (Chorpita & Barlow, 1998; Gunnar & Donzella, 2002; Hertsgaard et al., 1995) — suggests that especially in early stages of development, the infant – caregiver dyad can be viewed as a mutually regulating psychobiological unit (Schore, 2000). The extent to which this model also characterizes adolescents’ and adults’ most intimate and important relationships is unknown. Pipp and Harmon (1987) speculated that ‘‘homeostatic regulation between members of a dyad is a stable aspect of all intimate relationships throughout the lifespan’’ (p. 651), and Hofer (1984) has similarly argued that the psychological effects of interpersonal loss and bereavement can be interpreted as concomitants of multisystem dysregulation stemming from the removal of

116

Lisa M. Diamond and Christopher P. Fagundes

one member of the dyad. Dyadic approaches are also directly relevant to investigations of the health consequences of close relationships: Cacioppo (1994), for example, argued that an individual’s overall patterns of cardiovascular and neuroendocrine activity could be conceptualized as a function of his/her most important interpersonal relationship. Such approaches have been more consistently applied to studies of infants (in the context of infant–caregiver relationships) and adults (in the context of romantic ties) than to adolescents. Yet we expect that studies of adolescent social and psychological development, particularly regarding linkages between attachment relationships and affect-regulation processes, would beneﬁt greatly if researchers shifted toward treating the dyad—rather than the isolated adolescent—as the unit of analysis. Of course, one obvious complication is that more than one type of dyad is likely to be developmentally signiﬁcant: Although mother–adolescent pairs have received the most extensive attention in prior research (consistent with the fact that mothers are typically adolescents’ primary attachment ﬁgures), research has devoted increasing attention to fathers’ roles as attachment ﬁgures (Kerns & Barth, 1995; Youngblade, Park, & Belsky, 1993) and to the unique roles that fathers play in psychosocial development (Cabrera et al., 2000; Marsiglio et al., 2000; Phares & Compas, 1992). Similarly, different dynamics might characterize peer–peer dyads, depending on the nature of the relationship (best friends? romantic couple?) and the unique contributions of each peer’s temperament and behavior. Finally, all of these dynamics might be developmentally speciﬁc, undergoing notable maturational changes from early adolescence to young adulthood. Clearly, a comprehensive dyadic approach to adolescent development introduces numerous logistical and methodological challenges, yet such an approach may help to elucidate relationship-speciﬁc processes through which adolescents’ intimate relationships shape—and are shaped by—affectregulation capacities and strategies over time.

C. THE SPECIFIC IMPORTANCE OF POSITIVE AFFECT Most research on affect regulation, particularly in the context of attachment, focuses on attenuating negative affect, and particularly on alleviating psychological stress. This is not without cause: Both acute and chronic negative affectivity has been found to impede children’s and adults’ social functioning, empathy, exploratory behavior, cognitive processing, and the quality of their close relationships (Cooper et al., 1998; Eisenberg et al., 2000; Kim et al., 2001; Mikulincer et al., 2003). Studies of adolescence, in particular, have suggested that normative increases in the

Developmental Perspectives

117

frequency and intensity of negative emotions (Buchanan et al., 1992; Larson, Csikszentmihalyi, & Graef, 1980; Larson & Richards, 1994; Richards et al., 1998) often set the stage for adjustment and behavioral problems (Cooper et al., 1998). Yet, researchers have increasingly focused on the multiple psychological and physiological beneﬁts of positive affective experience. Positive and negative affect operate through distinct neural pathways (Lane et al., 1997) and appear to inﬂuence physical and mental functioning through different psychological mechanisms (Isen, 2002; Taylor et al., 2002). In particular, positive affect is associated with approach-oriented behavior (Cacioppo, Gardner, & Berntson, 1999), active engagement with the environment (reviewed in Fredrickson, 2001), more creative and ﬂexible decision-making (reviewed in Isen, 1993, 2000), generating multiple potential solutions to one’s problems (Fredrickson & Joiner, 2002), the effective processing of negative—but useful—problem-relevant information (reviewed in Aspinwall, 1998), anticipation and management of stressors before they occur (known as ‘‘proactive coping,’’ Aspinwall & Taylor, 1997), and positive reframing of one’s problems to emphasize the meaning that can be gleaned from adversity (Afﬂeck & Tennen, 1996; Davis, Nolen-Hoeksema, & Larson, 1998). This rich constellation of beneﬁts not only promotes everyday cognitive and social competence, but also fosters adaptive coping to both major and minor stressors (Folkman & Moskowitz, 2000; Park, Cohen, & Murch, 1996). On the basis of such ﬁndings, Fredrickson (2001) developed the broadenand-build theory of positive emotions, which maintains that positive emotional experiences ‘‘broaden people’s momentary thought-action repertoires and build their enduring personal resources, ranging from physical and intellectual resources to social and psychological resources’’ (p. 219). This theory is supported by empirical research demonstrating that not only can positive emotions offset or ‘‘undo’’ some of the immediate negative psychological and physiological effects of negative emotional arousal (Fredrickson & Levenson, 1998; Fredrickson et al., 2000; Fujita, Diener, & Sandvik, 1991), but that they appear to foster future increases in coping resources and psychological resilience (Fredrickson & Joiner, 2002). At the current time, these intriguing new conceptualizations of positive affect have not been systematically integrated into attachment–theoretical perspectives on affect regulation (with some exceptions, such as Mikulincer et al., 2003). Nor have they received extensive attention in the adolescent literature. Yet they have important implications for understanding the processes through which attachment security promotes adolescent affect regulation and well-being. Notably, positive affect experienced in the context of close interpersonal relationships appears to be particularly

118

Lisa M. Diamond and Christopher P. Fagundes

inﬂuential and beneﬁcial. Reis (2001) has argued that socially derived affect plays a unique role in shaping both day-to-day and global well-being, and Ryff and Singer (2001) have shown that trajectories of interpersonal affective experience, beginning with parent–child ties and continuing through adult marital relationships, show robust associations with both physical and mental well-being over the long term. In adolescence, positive affect experienced during interactions with attachment ﬁgures may prove to be particularly important for preventing escalation of the heightened negativity and conﬂict that often characterizes these relationships (Conger & Ge, 1999; Kim et al., 2001; Laursen, Coy, & Collins, 1998). This is consistent with emerging perspectives on regulatory beneﬁts associated with coactivation of negative and positive affective states (Larsen, McGraw, & Cacioppo, 2001; Larsen et al., 2003). Speciﬁcally, experiencing positive affect in concert with negative affect is thought to bolster individuals’ psychological and physiological resources for processing and coping with negative events, thereby preventing acute episodes of negative affect from becoming solidiﬁed into defensive and maladaptive regulatory patterns. Hence, such coactivation in youths’ attachment relationships may provide them with direct and immediate examples of how mobilization of positive affect can assist with the process of coping with both major and minor environmental demands. Clearly, greater investigation of such possibilities, and of the broader psychological concomitants of positive affect, can make important contributions to our understanding of the multiple, developmentally speciﬁc processes linking attachment to affect regulation from childhood to adulthood.

D. CONCLUSION The importance of attachment relationships in fostering psychological and physical well-being at all stages of the life course makes it all the more important to bridge the long-standing bifurcation between infant-child and adult attachment research. The development of integrative, lifespan, biobehavioral models of the attachment system should be a priority for future research, and greater emphasis on the affect-regulation functions of attachment, particularly during the critical developmental transitions of the adolescent years, can make an important contribution to this goal. From our perspective, affect regulation is not a developmental task to be mastered at a certain age (after which attention turns to the psychological and behavioral implications of one’s relative success or failure at this task), but rather a ‘‘moving target’’ that is continually sensitive to changing goals

Developmental Perspectives

119

and contexts. The optimal developmental outcome, therefore, is not complete regulatory independence from attachment ﬁgures and other social partners, but rather a ﬂexible and enduring capacity to adapt one’s affect-regulation strategies to the context at hand, to engage the assistance of social partners when needed, and to develop a sense of autonomy and self-determination that is based in the psychological resources fostered by attachment security. Our own research on attachment and affect regulation during the adolescent years shows the importance of attending to multiple affectregulation processes, multiple components of attachment relationships, and multiple domains of adjustment in order to capture dynamic linkages among these domains over time. The results demonstrate that the quality of youths’ parental attachments has implications for both subjective and physiological aspects of affect regulation, opening up a host of fascinating questions regarding the basic biopsychology of the attachment system and its potential developmental changes over the lifespan. Addressing these questions can help to integrate the increasingly sophisticated bodies of knowledge on social relationships and mental–physical health that have developed within the social-psychological, developmental, and behavioral medicine traditions. Such an integration is critical for elucidating how and why attachment bonds play such a fundamental role in well-being over the life course.

REFERENCES Achenbach, T. M. (2001). Manual for the ASEBA youth self-report and profiles. Burlington, VT: University of Vermont, Research Center for Children, Youth, & Families. Afﬂeck, G., & Tennen, H. (1996). Construing beneﬁts from adversity: Adaptational signiﬁcance and dispositional underpinnings. Journal of Personality, 64, 899–922. Ainsworth, M. D. S., Blehar, M. C., Waters, E., & Wall, S. (1978). Patterns of attachment: A psychological study of the strange situation. Hillsdale, NJ: Lawrence Erlbaum Associates, Inc. Allen, J. P., & Hauser, S. T. (1996). Autonomy and relatedness in adolescent-family interactions as predictors of young adults’ states of mind regarding attachment. Development and Psychopathology, 8, 793–809. Allen, J. P., & Land, D. (1999). Attachment in adolescence. In J. Cassidy & P. R. Shaver (Eds.), Handbook of attachment: Theory, research, and clinical applications (pp. 319–335). New York: Guilford. Allen, J. P., Hauser, S. T., Bell, K. L., & O’Connor, T. G. (1994). Longitudinal assessment of autonomy and relatedness in adolescent-family interactions as predictors of adolescent ego development and self-esteem. Child Development, 65, 179–194. Allen, J. P., Moore, C., Kuperminc, G., & Bell, K. (1998). Attachment and adolescent psychosocial functioning. Child Development, 69, 1406–1419.

120

Lisa M. Diamond and Christopher P. Fagundes

Allen, J. P., McElhaney, K. B., Land, D. J., Kuperminc, G. P., Moore, C. W., O’Beirne-Kelly, H., & Kilmer, S. L. (2003). A secure base in adolescence: Markers of attachment security in the mother-adolescent relationship. Child Development, 74, 292–307. Aspinwall, L. G. (1998). Rethinking the role of positive affect in self-regulation. Motivation and Emotion, 22, 1–32. Aspinwall, L. G., & Taylor, S. E. (1997). A stitch in time: Self-regulation and proactive coping. Psychological Bulletin, 121, 417–436. Bandura, A., Caprara, G. V., Barbaranelli, C., Gerbino, M., & Pastorelli, C. (2003). Role of affective self-regulatory efﬁcacy in diverse spheres of psychosocial functioning. Child Development, 74, 769–782. Bartholomew, K., & Horowitz, L. M. (1991). Attachment styles among young adults: A test of a four-category model. Journal of Personality and Social Psychology, 61, 226–244. Beauchaine, T. P., Kopp, L., & Mead, H. (2007). Polyvagal theory and developmental psychopsychology: Emotion dysregulation and conduct problems from preschool to adolescence. Biological Psychology, 74, 174–184. Becker-Stoll, F., Delius, A., & Scheitenberger, S. (2001). Adolescents’ nonverbal emotional expressions during negotiation of a disagreement with their mothers: An attachment approach. International Journal of Behavioral Development, 25, 344–353. Beebe, B., & Lachmann, F. M. (1998). Co-constructing inner and relational processes: Selfand mutual regulation in infant research and adult treatment. Psychoanalytic Psychology, 15, 480–516. Belsky, J., Friedman, S. L., & Hsieh, K. H. (2001). Testing a core emotion-regulation prediction: Does early attentional persistence moderate the effect of infant negative emotionality on later development? Child Development, 72, 123–133. Berger, R. D., Saul, J. P., & Cohen, R. J. (1989). Transfer function analysis of autonomic regulation: I. The canine atrial rate response. American Journal of Physiology, 256, H142– H152. Berlin, L. J., & Cassidy, J. (1999). Relations among relationships: Contributions from attachment theory and research. In J. Cassidy & P. R. Shaver (Eds.), Handbook of attachment: Theory, research, and clinical applications (pp. 688–712). New York: Guilford. Berntson, G. G., Bigger, J. T. J., Eckberg, D. L., Grossman, P., Kaufmann, P. G., Malik, M., Nagaraja, H. N., Porges, S. W., Saul, J. P., Stone, P. H., & van der Molen, M. W. (1997). Heart rate variability: Origins, methods, and interpretive caveats. Psychophysiology, 34, 623–648. Blakemore, S.-J., & Choudhury, S. (2006). Development of the adolescent brain: Implications for executive function and social cognition. Journal of Child Psychology and Psychiatry, 47, 296–312. Block, J., & Kremen, A. M. (1996). IQ and ego-resiliency: Conceptual and empirical connections and separateness. Journal of Personality and Social Psychology, 70, 349–361. Bloch, M., Peleg, I., Koren, D., Aner, H., & Klein, E. (2007). Long-term effects of early parental loss due to divorce on the HPA axis. Hormones and Behavior, 51, 516–523. Booth, A., Carver, K., & Granger, D. A. (2000). Biosocial perspectives on the family. Journal of Marriage and the Family, 62, 1018–1034. Bovard, E. W. (1985). Brain mechanisms in effects of social support on viability. In R. B. Williams, Jr. (Ed.), Perspectives on behavioral medicine. Volume 2. Neuroendocrine control and behavior (pp. 103–129). Orlando, FL: Academic Press. Bowlby, J. (1958). The nature of the child’s tie to his mother. International Journal of Psychoanalysis, 39, 350–373. Bowlby, J. (1973). Attachment and loss. Volume 2. Separation: Anxiety and anger. New York: Basic.

Developmental Perspectives

121

Bowlby, J. (1980). Attachment and loss. Volume 3. Loss: Sadness and depression. New York: Basic. Bowlby, J. (1982). Attachment and loss. Volume 1. Attachment. (2nd ed.). New York: Basic. Bowlby, J. (1988). A secure base: Parent–child attachment and healthy human development. New York: Basic. Brennan, K. A., & Shaver, P. R. (1995). Dimensions of adult attachment, affect regulation, and romantic relationship functioning. Personality and Social Psychology Bulletin, 21, 267–283. Bridges, L. J., Denham, S. A., & Ganiban, J. M. (2004). Deﬁnitional issues in emotion regulation research. Child Development, 75, 340–345. Broderick, P. C. (1998). Early adolescent gender differences in the use of ruminative and distracting coping strategies. Journal of Early Adolescence, 18, 173–191. Brosschot, J. F., & Thayer, J. F. (1998). Anger inhibition, cardiovascular recovery, and vagal function: A model of the link between hostility and cardiovascular disease. Annals of Behavioral Medicine, 20, 326–332. Buchanan, C. M., Eccles, J. S., & Becker, J. B. (1992). Are adolescents victims of raging hormones: Evidence for activational effects of hormones on moods and behavior at adolescence. Psychological Bulletin, 111, 62–107. Buske-Kirschbaum, A., Jobst, S., Psych, D., Wustmans, A., Kirschbaum, C., Rauh, W., & Hellhammer, D. (1997). Attenuated free cortisol response to psychosocial stress in children with atopic dermatitis. Psychosomatic Medicine, 59, 419–426. Cabrera, N. J., Tamis-LeMonda, C. S., Bradley, R. H., Hofferth, S., & Lamb, M. E. (2000). Fatherhood in the twenty-ﬁrst century. Child Development, 71, 127–136. Cacioppo, J. T. (1994). Social neuroscience: Autonomic, neuroendocrine, and immune responses to stress. Psychophysiology, 31, 113–128. Cacioppo, J. T., & Berntson, G. G. (2007). The brain, homeostasis, and health: Balancing demands of the internal and external milieu. In H. S. Friedman & R. C. Silver (Eds.), Foundations of health psychology (pp. 73–91). New York: Oxford University Press. Cacioppo, J. T., Gardner, W. L., & Berntson, G. G. (1999). The affect system has parallel and integrative processing components: Form follows function. Journal of Personality and Social Psychology, 76, 839–855. Cairns, R. B., Gariepy, J. L., & Hood, K. E. (1990). Development, microevolution, and social behavior. Psychological Review, 97, 49–65. Calkins, S. D. (1997). Cardiac vagal tone indices of temperamental reactivity and behavioral regulation in young children. Developmental Psychobiology, 31, 125–135. Calkins, S. D., Smith, C. L., Gill, K. L., & Johnson, M. C. (1998). Maternal interactive style across contexts: Relations to emotional, behavioral, and physiological regulation during toddlerhood. Social Development, 7, 350–369. Campos, J. J., Frankel, C. B., & Camras, L. (2004). On the nature of emotion regulation. Child Development, 75, 377–394. Carter, C. S. (1998). Neuroendocrine perspectives on social attachment and love. Psychoneuroendocrinology, 23, 779–818. Cassidy, J. (1994). Emotion regulation: Inﬂuences of attachment relationships. In N. Fox (Ed.), The development of emotion regulation: Biological and behavioral considerations. Monographs of the Society for Research in Child Development, 59 (2–3, Serial No. 240), pp. 228–283. Cassidy, J., Kirsh, S. J., Scolton, K. L., & Parke, R. D. (1996). Attachment and representations of peer relationships. Developmental Psychology, 32, 892–904. Chorpita, B. F., & Barlow, D. H. (1998). The development of anxiety: The role of control in the early environment. Psychological Bulletin, 124, 3–21.

122

Lisa M. Diamond and Christopher P. Fagundes

Cicchetti, D., Ackerman, B. P., & Izard, C. E. (1995). Emotions and emotion regulation in developmental psychopathology. Development and Psychopathology, 7, 1–10. Cohn, J. F., & Tronick, E. Z. (1983). Three-month-old infants’ reaction to simulated maternal depression. Child Development, 54, 185–193. Colder, C. R., & Chassin, L. (1997). Affectivity and impulsivity: Temperament risk for adolescent alcohol involvement. Psychology of Addictive Behaviors, 11, 83–97. Coe, C. L., Rosenberg, L. T., & Levine, S. (1988). Immunological consequences of psychological disturbance and maternal loss in infancy. Advances in Infancy Research, 5, 97–134. Cole, P. M., Martin, S. E., & Dennis, T. A. (2004). Emotion regulation as a scientiﬁc construct: Methodological challenges and directions for child development research. Child Development, 75, 317–333. Collins, N. L. (1996). Working models of attachment: Implications for explanation, emotion, and behavior. Journal of Personality and Social Psychology, 71, 810–832. Collins, N. L., & Feeney, B. C. (2000). A safe haven: An attachment theory perspective on support seeking and caregiving in intimate relationships. Journal of Personality and Social Psychology, 78, 1053–1073. Conger, R. D., & Ge, X. (1999). Conﬂict and cohesion in parent–adolescent relations: Changes in emotional expression from early to midadolescence. In M. J. Cox & J. Brooks Gunn (Eds.), Conflict and cohesion in families: Causes and consequences (pp. 185–206). Mahwah, NJ: Lawrence Erlbaum Associates. Cook, W. L. (2000). Understanding attachment security in family context. Journal of Personality and Social Psychology, 78, 285–294. Cooper, M. L., Frone, M. R., Russell, M., & Mudar, P. (1995). Drinking to regulate positive and negative emotions: A motivational model of alcohol use. Journal of Personality and Social Psychology, 69, 990–1005. Cooper, M. L., Shaver, P. R., & Collins, N. L. (1998). Attachment styles, emotion regulation, and adjustment in adolescence. Journal of Personality and Social Psychology, 74, 1380–1397. Cooper, M. L., Albino, A., Orcutt, H. K., & Williams, N. (2004). Attachment styles and intrapersonal adjustment: A longitudinal study from adolescence into young adulthood. In J. A. Simpson & W. S. Rholes (Eds.), Adult attachment: New directions and emerging issues (pp. 438–466). New York: Guilford. Croiset, G., Heijnen, C. J., Van der Wal, W. E., De Boer, S. F., & De Wied, D. (1990). A role for the autonomic nervous system in modulating the immune response during mild emotional stimuli. Life Sciences, 46, 419–425. Crowell, J. A., Fraley, R. C., & Shaver, P. R. (1999). Measurement of individual differences in adolescent and adult attachment. In J. Cassidy & P. R. Shaver (Eds.), Handbook of attachment: Theory, research, and clinical applications (pp. 434–465). New York: Guilford. Cummings, E. M., & Davies, P. (1996). Emotional security as a regulatory process in normal development and the development of psychopathology. Development and Psychopathology, 8, 123–139. Davila, J., Karney, B. R., & Bradbury, T. N. (1999). Attachment change processes in the early years of marriage. Journal of Personality and Social Psychology, 76, 783–802. Davis, C. G., Nolen-Hoeksema, S., & Larson, J. (1998). Making sense of loss and beneﬁting from the experience: Two construals of meaning. Journal of Personality and Social Psychology, 75, 561–574. Deci, E. L., & Ryan, R. M. (2000). The ‘‘what’’ and ‘‘why’’ of goal pursuits: Human needs and the self-determination of behavior. Psychological Inquiry, 11, 227–268.

Developmental Perspectives

123

DeGangi, G. A., DiPietro, J. A., Greenspan, S. I., & Porges, S. W. (1991). Psychophysiological characteristics of the regulatory disordered infant. Infant Behavior and Development, 14, 37–50. Denham, S. A. (2006). Emotional competence: Implications for social functioning. In J. L. Luby (Ed.), Handbook of preschool mental health: Development, disorders, and treatment (pp. 23–44). New York: Guilford Press. Diamond, L. M. (2000). Passionate friendships among adolescent sexual-minority women. Journal of Research on Adolescence, 10, 191–209. Diamond, L. M. (2001). Contributions of psychophysiology to research on adult attachment: Review and recommendations. Personality and Social Psychology Review, 5, 276–295. Diamond, L. M., & Hicks, A. M. (2004). Psychobiological perspectives on attachment: Implications for health over the lifespan. In J. A. Simpson & W. S. Rholes (Eds.), Adult attachment: New directions and emerging issues (pp. 240–263). New York: Guilford. Diamond, L. M., & Hicks, A. M. (2005). Attachment style, current relationship security, and negative emotions: The mediating role of physiological regulation. Journal of Social and Personal Relationships, 22, 499–518. Diamond, L. M., Hicks, A. M., & Otter-Henderson, K. A. (2006). Physiological evidence for repressive coping among avoidantly attached adults. Journal of Social and Personal Relationships, 23, 205–229. DiTommaso, E., & Spinner, B. (1993). The development and initial validation of the social and emotional loneliness scale for adults (SELSA). Personality and Individual Differences, 14, 127–134. Dodge, K. A. (1991). Emotion and social information processing. In J. Garber & K. A. Dodge (Eds.), The development of emotion regulation and dysregulation (pp. 159–181). Cambridge: Cambridge University Press. Doyle, A. B., & Markiewicz, D. (2005). Parenting, marital conﬂict and adjustment from earlyto mid-adolescence: Mediated by adolescent attachment style? Journal of Youth and Adolescence, 34, 97–110. Eisenberg, N., & Fabes, R. A. (1992). Emotion, regulation, and the development of social competence. In M. S. Clark (Ed.), Emotion and social behavior (Vol. 14, pp. 119–150). Thousand Oaks, CA: Sage. Eisenberg, N., Fabes, R. A., Schaller, M., & Carlo, G. (1991). The relations of parental characteristics and practices to children’s vicarious emotional responding. Child Development, 62, 1393–1408. Eisenberg, N., Fabes, R. A., Guthrie, I. K., & Reiser, M. (2000). Dispositional emotionality and regulation: Their role in predicting quality of social functioning. Journal of Personality and Social Psychology, 78, 136–157. Eisenberg, N., Cumberland, A., Spinrad, T. L., Fabes, R. A., Shepard, S. A., Reiser, M., Murphy, B. C., Losoya, S. H., & Guthrie, I. K. (2001). The relations of regulation and emotionality to children’s externalizing and internalizing problem behavior. Child Development, 72, 1112–1134. Eisenberg, N., Spinrad, T. L., & Morris, A. S. (2002). Regulation, resiliency, and quality of social functioning. Self and Identity, 1, 121–128. El-Sheikh, M. (2005). Stability of respiratory sinus arrhythmia in children and young adolescents: A longitudinal examination. Developmental Psychobiology, 46, 66–74. Fabes, R. A., & Eisenberg, N. (1997). Regulatory control in adults’ stress-related responses to daily life events. Journal of Personality and Social Psychology, 73, 1107–1117. Fabes, R. A., Eisenberg, N., Karbon, M., Troyer, D., & Switzer, G. (1994). The relations of children’s emotion regulation to their vicarious emotional responses and comforting behaviors. Child Development, 65, 1678–1693.

124

Lisa M. Diamond and Christopher P. Fagundes

Fagot, B. I., & Kavanagh, K. (1990). The prediction of antisocial behavior from avoidant attachment classiﬁcation. Child Development, 61, 864–873. Feeney, B. C. (2007). The dependency paradox in close relationships: Accepting dependence promotes independence. Journal of Personality and Social Psychology, 92, 268–285. Feeney, J. A. (1999). Adult romantic attachment and couple relationships. In J. Cassidy & P. R. Shaver (Eds.), Handbook of attachment: Theory, research, and clinical applications (pp. 355–377). New York: Guilford. Feeney, J. A. (2000). Implications of attachment style for patterns of health and illness. Child: Care, Health and Development, 26, 277–288. Feeney, J. A., & Ryan, S. M. (1994). Attachment style and affect regulation: Relationships with health behavior and family experiences of illness in a student sample. Health Psychology, 13, 334–345. Feeney, J. A., Noller, P., & Callan, V. J. (1994). Attachment style, communication and satisfaction in the early years of marriage. In K. Bartholomew & D. Perlman (Eds.), Advances in personal relationships: A research annual (Vol. 5, pp. 269–308). London: Jessica Kingsley Publishers. Flinn, M. V., & England, B. G. (1995). Childhood stress and family environment. Current Anthropology, 36, 854–866. Fogel, A. (1992). Movement and communication in human infancy: The social dynamics of development. Human Movement Science, 11, 387–423. Folkman, S., & Moskowitz, J. T. (2000). Positive affect and the other side of coping. American Psychologist, 55, 647–654. Fox, N. A. (1994a). Deﬁnitions and concepts of emotion regulation: Introduction. In N. Fox (Ed.), The development of emotion regulation: Biological and behavioral considerations. Monographs of the Society for Research in Child Development, 59 (2–3, Serial No. 240), pp. 243–246. Fox, N.A. (1994b). The development of emotion regulation: Biological and behavioral considerations. Monographs of the Society for Research in Child Development, 59 (2–3, Serial No. 240). Fox, N. A. (1989). Psychophysiological correlates of emotional reactivity during the ﬁrst year of life. Developmental Psychology, 25, 495–504. Fox, N. A., & Calkins, S. D. (2003). The development of self-control of emotion: Intrinsic and extrinsic inﬂuences. Motivation and Emotion, 27, 7–26. Fraley, R. C., & Davis, K. E. (1997). Attachment formation and transfer in young adults’ close friendships and romantic relationships. Personal Relationships, 4, 131–144. Fraley, R. C., Waller, N. G., & Brennan, K. A. (2000). An item response theory analysis of self-report measures of adult attachment. Journal of Personality and Social Psychology, 78, 350–365. Fredrickson, B. L. (2001). The role of positive emotions in positive psychology: The broadenand-build theory of positive emotions. American Psychologist, 56, 218–226. Fredrickson, B. L., & Joiner, T. (2002). Positive emotions trigger upward spirals toward emotional well-being. Psychological Science, 13, 172–175. Fredrickson, B. L., & Levenson, R. W. (1998). Positive emotions speed recovery from the cardiovascular sequelae of negative emotions. Cognition and Emotion, 12, 191–220. Fredrickson, B. L., Mancuso, R. A., Branigan, C., & Tugade, M. M. (2000). The undoing effect of positive emotions. Motivation and Emotion, 24, 237–258. Frick, P. J., & Morris, A. S. (2004). Temperament and developmental pathways to conduct problems. Journal of Clinical Child and Adolescent Psychology, 33, 54–68. Fridja, N. H. (1986). The emotions. Cambridge, UK: Cambridge University Press.

Developmental Perspectives

125

Friedman, B. H., & Thayer, J. F. (1998). Autonomic balance revisited: Panic anxiety and heart rate variability. Journal of Psychosomatic Research, 44, 133–151. Fujita, F., Diener, E., & Sandvik, E. (1991). Gender differences in negative affect and well-being: The case for emotional intensity. Journal of Personality and Social Psychology, 61, 427–434. Furman, W., & Buhrmester, D. (1992). Age and sex differences in perceptions of networks of personal relationships. Child Development, 63, 103–115. Gable, S., & Isabella, R. A. (1992). Maternal contributions to infant regulation of arousal. Infant Behavior and Development, 15, 95–107. Gable, S. L., & Reis, H. T. (1999). Now and then, them and us, this and that: Studying relationships across time, partner, context, and person. Personal Relationships, 6, 415–432. Glaser, D. (2000). Child abuse and neglect and the brain: A review. Journal of Child Psychology and Psychiatry and Allied Disciplines, 41, 97–116. Glynn, L. M., Christenfeld, N., & Gerin, W. (2002). The role of rumination in recovery from reactivity: Cardiovascular consequences of emotional states. Psychosomatic Medicine, 64, 714–726. Gohm, C. L. (2003). Mood regulation and emotional intelligence: Individual differences. Journal of Personality and Social Psychology, 84, 594–607. Goldberg, S. (1997). Attachment and childhood behavior problems in normal, at-risk and clinical samples. New York: Guilford Press. Gottlieb, G. (1991). Experiential canalization of behavioral development: Theory. Developmental Psychology, 27, 4–13. Granger, D. A., Weisz, J. R., & Kauneckis, D. (1994). Neuroendocrine reactivity, internalizing behavior problems, and control-related cognitions in clinic-referred children and adolescents. Journal of Abnormal Psychology, 103, 267–276. Gray, J. A. (1995). A model of the limbic system and basal ganglia: Applications to anxiety and schizophrenia. In M. S. Gazzaniga (Ed.), The cognitive neurosciences (pp. 1165–1176). Cambridge, MA: MIT Press. Gross, J. J. (1998). Antecedent- and response-focused emotion regulation: Divergent consequences for experience, expression, and physiology. Journal of Personality and Social Psychology, 74, 224–237. Gross, J. J. (1999). Emotion regulation: Past, present, future. Cognition and Emotion, 13, 551–573. Gross, J. J., & Munoz, R. F. (1995). Emotion regulation and mental-health. Clinical Psychology: Science and Practice, 2, 151–164. Grossman, P., & Svebak, S. (1987). Respiratory sinus arrhythmia as an index of parasympathetic cardiac control during active coping. Psychophysiology, 24, 228–235. Grossman, P., Stemmler, G., & Meinhardt, E. (1990). Paced respiratory sinus arrhythmia as an index of cardiac parasympathetic tone during varying behavioral tasks. Psychophysiology, 27, 404–416. Grossman, P., Karemaker, J., & Wieling, W. (1991). Prediction of tonic parasympathetic cardiac control using respiratory sinus arrhythmia: The need for respiratory control. Psychophysiology, 28, 201–216. Grossman, P., Brinkman, A., & de Vries, J. (1992). Cardiac autonomic mechanisms associated with borderline hypertension under varying behavioral demands: Evidence for attenuated parasympathetic tone but not for enhanced beta-adrenergic activity. Psychophysiology, 29, 698–711. Grossmann, K. E., Grossmann, K., & Waters, E. (Eds). (2005). Attachment from infancy to adulthood: The major longitudinal studies. New York: Guilford Press.

126

Lisa M. Diamond and Christopher P. Fagundes

Gunnar, M. R. (2003). Integrating neuroscience and psychological approaches in the study of early experiences. In J. A. King, C. F. Ferris, & I. I. Lederhendler (Eds.), Roots of mental illness in children (pp. 238–247). New York: New York Academy of Sciences. Gunnar, M. R., & Donzella, B. (2002). Social regulation of cortisol levels in early human development. Psychoneuroendocrinology, 27, 199–220. Halberstadt, A. G., Cassidy, J., Stifter, C. A., Parke, R. D., & Fox, N. A. (1995). Self-expressiveness within the family context: Psychometric support for a new measure. Psychological Assessment, 7, 93–103. Halberstadt, A. G., Denham, S. A., & Dunsmore, J. C. (2001). Affective social competence. Social Development, 10, 79–119. Hamilton, C. E. (2000). Continuity and discontinuity of attachment from infancy through adolescence. Child Development, 71, 690–694. Hardy, D. F., Power, T. G., & Jaedicke, S. (1993). Examining the relation of parenting to children’s coping with everyday stress. Child Development, 64, 1829–1841. Hart, J., Gunnar, M. R., & Cicchetti, D. (1995). Salivary cortisol in maltreated children: Evidence of relations between neuroendocrine activity and social competence. Development and Psychopathology, 7, 11–26. Hauser, S. T., Gerber, E. B., & Allen, J. P. (1998). Ego development and attachment: Converging platforms for understanding close relationships. Mahwah, NJ: Lawrence Erlbaum Associates. Hazan, C., & Shaver, P. R. (1987). Romantic love conceptualized as an attachment process. Journal of Personality and Social Psychology, 52, 511–524. Hazan, C., & Zeifman, D. (1994). Sex and the psychological tether. In D. Perlman & K. Bartholomew (Eds.), Advances in personal relationships: A research annual (Vol. 5, pp. 151–177). London: Jessica Kingsley Publishers. Hazan, C., & Zeifman, D. (1999). Pair-bonds as attachments: Evaluating the evidence. In J. Cassidy & P. R. Shaver (Eds.), Handbook of attachment theory and research (pp. 336–354). New York: Guilford. Hazan, C., Gur-Yaish, N., & Campa, M. (2004). What does it mean to be attached? In J. A. Simpson & W. S. Rholes (Eds.), Adult attachment: New directions and emerging issues (pp. 55–85). New York: Guilford. Hertsgaard, L., Gunnar, M. R., Erickson, M. F., & Nachmias, M. (1995). Adrenocortical responses to the Strange Situation in infants with disorganized/disoriented attachment relationships. Child Development, 66, 1100–1106. Hessler, D. M., & Fainsilber Katz, L. (2007). Children’s emotion regulation: Self-report and physiological response to peer provocation. Developmental Psychology, 43, 27–38. Hofer, M. (1984). Relationships as regulators: A psychobiologic perspective on bereavement. Psychosomatic Medicine, 46, 183–197. Horsten, M., Ericson, M., Perski, A., Wamala, S. P., Schenck-Gustafsson, K., & Orth-Gome´r, K. (1999). Psychosocial factors and heart rate variability in healthy women. Psychosomatic Medicine, 61, 49–57. Irwin, M., Hauger, R., & Brown, M. (1992). Central corticotropin-releasing hormone activates the sympathetic nervous system and reduces immune function: Increased responsivity of the aged rat. Endocrinology, 131, 1047–1053. Isen, A. M. (1993). Positive affect and decision making. In M. Lewis & J. M. Haviland (Eds.), Handbook of emotions (pp. 261–277). New York: Guilford Press. Isen, A. M. (2000). Positive affect and decision making. In M. Lewis & J. M. Haviland-Jones (Eds.), Handbook of emotions (2nd ed., pp. 417–435). New York: Guilford Press.

Developmental Perspectives

127

Isen, A. M. (2002). A role for neuropsychology in understanding the facilitating inﬂuence of positive affect on social behavior and cognitive processes. In C. R. Snyder & S. J. Lopez (Eds.), Handbook of positive psychology (pp. 528–540). London: Oxford University Press. Isen, A. M. (2003). Positive affect as a source of human strength. In L. G. Aspinwall & U. M. Staudinger (Eds.), A psychology of human strengths: Fundamental questions and future directions for a positive psychology (pp. 179–195). Washington, D.C.: American Psychological Association. Isen, A. M., & Hastorf, A. H. (1982). Some perspectives on cognitive social psychology. In A. H. Hastorf & A. M. Isen (Eds.), Cognitive social psychology (pp. 1–31). New York: Elsevier North Holland. Jacobvitz, D., Curran, M., & Moller, N. (2002). Measurement of adult attachment: The place of self-report and interview methodologies. Attachment & Human Development, 4, 207–215. Keating, D. P. (1990). Adolescent thinking. In S. S. Feldman & G. R. Elliott (Eds.), At the threshold: The developing adolescent (pp. 54–89). Cambridge, MA: Harvard University Press. Kerns, K. A., & Barth, J. M. (1995). Attachment and play: Convergence across components of parent–child relationships and their relations to peer competence. Journal of Social and Personal Relationships, 12, 243–260. Kibler, J. L., Prosser, V. L., & Ma, M. (2004). Cardiovascular correlates of misconduct in children and adolescents. Journal of Psychophysiology, 18, 184–189. Kiecolt-Glaser, J. K., McGuire, L., Robles, T. F., & Glaser, R. (2002b). Psychoneuroimmunology: Psychological inﬂuences on immune function and health. Journal of Consulting and Clinical Psychology, 70, 537–547. Kim, K. J., Conger, R. D., Lorenz, F. O., & Elder, G. H., Jr. (2001). Parent-adolescent reciprocity in negative affect and its relation to early adult social development. Developmental Psychology, 37, 775–790. Kirkpatrick, L. A., & Davis, K. E. (1994). Attachment style, gender, and relationship stability: A longitudinal analysis. Journal of Personality and Social Psychology, 66, 502–512. Kirkpatrick, L. A., & Hazan, C. (1994). Attachment styles and close relationships: A four-year prospective study. Personal Relationships, 1, 123–142. Kirschbaum, C., Wolf, O. T., May, M., Wippich, W., & Hellhammer, D. H. (1996). Stress- and treatment-induced elevations of cortisol levels associated with impaired declarative memory in healthy adults. Life Sciences, 58, 1475–1483. Kliewer, W., Fearnow, M. D., & Miller, P. A. (1996). Coping socialization in middle childhood: Tests of maternal and paternal inﬂuences. Child Development, 67, 2339–2357. Kobak, R. R., & Ferenz Gillies, R. (1995). Emotion regulation and depressive symptoms during adolescence: A functionalist perspective. Development and Psychopathology, 7, 183–192. Kobak, R. R., & Sceery, A. (1988). Attachment in late adolescence: Working models, affect regulation, and representations of self and others. Child Development, 59, 135–146. Kobak, R. R., Sudler, N., & Gamble, W. (1991). Attachment and depressive symptoms during adolescence: A developmental pathways analysis. Development and Psychopathology, 3, 461–474. Kobak, R. R., Cole, H. E., Ferenz-Gillies, R., Fleming, W. S., & Gamble, W. (1993). Attachment and emotion regulation during mother-teen problem solving: A control theory analysis. Child Development, 64, 231–245. Kopp, C. B. (1989). Regulation of distress and negative emotions: A developmental view. Developmental Psychology, 25, 343–354.

128

Lisa M. Diamond and Christopher P. Fagundes

La Guardia, J. G., Ryan, R. M., Couchman, C. E., & Deci, E. L. (2000). Within-person variation in security of attachment: A self-determination theory perspective on attachment, need fulﬁllment, and well-being. Journal of Personality and Social Psychology, 79, 367–384. Lakey, B., McCabe, K. M., Fisicaro, S. A., & Drew, J. B. (1996). Environmental and personal determinants of support perceptions: Three generalizability studies. Journal of Personality and Social Psychology, 70, 1270–1280. Lane, R. D., Reiman, E. M., Bradley, M. M., Lang, P. J., Ahern, G. L., Davidson, R. J., & Schwartz, G. E. (1997). Neuroanatomical correlates of pleasant and unpleasant emotion. Neuropsychologia, 35, 1437–1444. Larsen, J. T., McGraw, A. P., & Cacioppo, J. T. (2001). Can people feel happy and sad at the same time? Journal of Personality and Social Psychology, 81, 684–696. Larsen, R. J. (2000). Toward a science of mood regulation. Psychological Inquiry, 11, 129–141. Larsen, R. J., Hemenover, S. H., Norris, C. J., & Cacioppo, J. T. (2003). Turning adversity to advantage: On the virtues of the coactivation of positive and negative emotions. In L. G. Aspinwall & U. M. Staudinger (Eds.), A psychology of human strengths: Fundamental questions and future directions for a positive psychology (pp. 211–225). Washington, D.C.: American Psychological Association. Larson, R. W., & Richards, M. H. (1994). Family emotions: Do young adolescents and their parents experience the same states? Journal of Research on Adolescence, 4, 567–583. Larson, R. W., Csikszentmihalyi, M., & Graef, R. (1980). Mood variability and the psychosocial adjustment of adolescents. Journal of Youth and Adolescence, 9, 469–490. Larson, R. W., Clore, G. L., & Wood, G. A. (1999). The emotions of romantic relationships: Do they wreak havoc on adolescents? In W. Furman, C. Feiring, & B. B. Brown (Eds.), Contemporary perspectives on adolescent romantic relationships (pp. 19–49). New York: Cambridge University Press. Laursen, B. (1996). Closeness and conﬂict in adolescent peer relationships: Interdependence with friends and romantic partners. In W. M. Bukowski, A. F. Newcomb, & W. W. Hartup (Eds.), The company they keep: Friendship in childhood and adolescence (pp. 186–210). New York: Cambridge University Press. Laursen, B., Coy, K. C., & Collins, W. A. (1998). Reconsidering changes in parent–child conﬂict across adolescence: A meta-analysis. Child Development, 69, 817–832. LeDoux, J. E. (1995). In search of an emotional system in the brain: Leaping from fear to emotion and consciousness. In M. S. Gazzaniga (Ed.), The cognitive neurosciences (pp. 1049–1062). Cambridge, MA: MIT Press. Loney, B. R., Butler, M. A., Lima, E. N., Counts, C. A., & Eckel, L. A. (2006). The relation between salivary cortisol, callous-unemotional traits, and conduct problems in an adolescent non-referred sample. Journal of Child Psychology and Psychiatry, 47, 30–36. Lupien, S., Lecours, A. R., Lussier, I., Schwartz, G., Nair, M. P., & Meaney, M. J. (1994). Basal cortisol levels and cognitive deﬁcits in human aging. Journal of Neuroscience, 14, 2893–2903. Lyons, K. S., & Sayer, A. G. (2005). Longitudinal dyad models in family research. Journal of Marriage and Family, 67, 1048–1060. Magai, C., Hunziker, J., Mesias, W., & Culver, L. C. (2000). Adult attachment styles and emotional biases. International Journal of Behavioral Development, 24, 301–309. Main, M., Kaplan, N., & Cassidy, J. (1985). Security in infancy, childhood, and adulthood: A move to the level of representation. Monographs of the Society for Research in Child Development, 50, 66–104. Markiewicz, D., Lawford, H., Doyle, A. B., & Haggart, N. (2006). Developmental differences in adolescents’ and young adults’ use of mothers, fathers, best friends, and romantic partners to fulﬁll attachment needs. Journal of Youth and Adolescence, 35, 127–140.

Developmental Perspectives

129

Marsiglio, W., Amato, P., Day, R. D., & Lamb, M. E. (2000). Scholarship on fatherhood in the 1990s and beyond. Journal of Marriage & the Family, 62, 1173–1191. Marvin, R. S., & Britner, P. A. (1999). Normative development: The ontogeny of attachment. In J. Cassidy & P. R. Shaver (Eds.), Handbook of attachment: Theory, research, and clinical applications (pp. 44–67). New York: Guilford. Masten, A. S. (2001). Ordinary magic: Resilience processes in development. The American Psychologist, 56, 227–238. McCormick, C. M., & Mathews, I. Z. (2007). HPA function in adolescence: Role of sex hormones in its regulation and the enduring consequences of exposure to stressors. Pharmacology, Biochemistry and Behavior, 86, 220–233. McEwen, B. S. (1995). Stressful experience, brain, and emotions: Developmental, genetic, and hormonal inﬂuences. In M. S. Gazzaniga (Ed.), The cognitive neurosciences (pp. 1117– 1135). Cambridge, MA: MIT Press. McEwen, B. S., & Stellar, E. (1993). Stress and the individual. Mechanisms leading to disease. Archives of Internal Medicine, 153, 2093–2101. McEwen, B. S., Angulo, J., Cameron, H., Chao, H. M., Daniels, D., Gannon, M. N., Gould, E., Mendelson, S., Sakai, R., & Spencer, R. (1992). Paradoxical effects of adrenal steroids on the brain: Protection versus degeneration. Biological Psychiatry, 31, 177–199. Mickelson, K. D., Kessler, R. C., & Shaver, P. R. (1997). Adult attachment in a nationally representative sample. Journal of Personality and Social Psychology, 73, 1092–1106. Mikulincer, M. (1998a). Adult attachment style and affect regulation: Strategic variations in self-appraisals. Journal of Personality and Social Psychology, 75, 420–435. Mikulincer, M. (1998b). Adult attachment style and individual differences in functional versus dysfunctional experiences of anger. Journal of Personality and Social Psychology, 74, 513– 524. Mikulincer, M., & Orbach, I. (1995). Attachment styles and repressive defensiveness: The accessibility and architecture of affective memories. Journal of Personality and Social Psychology, 68, 917–925. Mikulincer, M., & Shefﬁ, E. (2000). Adult attachment style and cognitive reactions to positive affect: A test of mental categorization and creative problem solving. Motivation and Emotion, 24, 149–174. Mikulincer, M., Shaver, P. R., & Pereg, D. (2003). Attachment theory and affect regulation: The dynamics, development, and cognitive consequences of attachment-related strategies. Motivation and Emotion, 27, 77–102. Miller, J. B., & Hoicowitz, T. (2004). Attachment contexts of adolescent friendship and romance. Journal of Adolescence, 27, 191–206. Miller, G. E., Chen, E., & Zhou, E. S. (2007). If it goes up, must it come down? Chronic stress and the hypothalamic-pituitary-adrenocortical axis in humans. Psychological Bulletin, 133, 25–45. Moos, R. H. (1988). Coping responses inventory manual. Palo Alto, CA: Stanford University and Department of Veterans Affairs Medical Centers. Moris, P., Meesters, C., & van den Berg, S. (2003). Internalizing and externalizing problems as correlates of self-reported attachment style and perceived parental rearing in normal adolescents. Journal of Child and Family Studies, 12, 171–183. Morris, A. S., Silk, J. S., Steinberg, L., Myers, S. S., & Robinson, L. R. (2007). The role of the family context in the development of emotion regulation. Social Development, 16, 361–388. Moss, H. B., Vanyukov, M., Yao, J. K., & Kirillova, G. P. (1999). Salivary cortisol responses in prepubertal boys: The effects of parental substance abuse and association with drug use behavior during adolescence. Biological Psychiatry, 45, 1293–1299.

130

Lisa M. Diamond and Christopher P. Fagundes

Munck, A., & Guyre, P. M. (1991). Glucocorticoids and immune function. In R. Ader, D. L. Falten, & N. Cohen (Eds.), Psychoneuroimmunology (2nd ed., pp. 447–474). New York: Academic Press. Nolen-Hoeksema, S., Parker, L. E., & Larson, J. (1994). Ruminative coping with depressed mood following loss. Journal of Personality and Social Psychology, 67, 92–104. Pandina, R. J., Johnson, V., & Labouvie, E. W. (1992). Affectivity: A central mechanism in the development of drug dependence. In M. Glantz & R. Pickens (Eds.), Vulnerability to drug abuse (pp. 179–209). Washington D.C.: American Psychological Association. Park, C. L., Cohen, L. H., & Murch, R. L. (1996). Assessment and prediction of stress-related growth. Journal of Personality, 64, 71–105. Patrick, D. L., Edwards, T. C., & Topolski, T. D. (2002). Adolescent quality of life, Part II: Initial validation of a new instrument. Journal of Adolescence, 25, 287–300. Phares, V., & Compas, B. E. (1992). The role of fathers in child and adolescent psychopathology: Make room for daddy. Psychological Bulletin, 111, 387–412. Pipp, S., & Harmon, R. J. (1987). Attachment as regulation: A commentary. Child Development, 58, 648–652. Porges, S. W. (1991). Vagal tone: An autonomic mediator of affect. In J. Garber & K. A. Dodge (Eds.), The development of emotion regulation and dysregulation (pp. 111–128). New York: Cambridge University Press. Porges, S. W. (1992). Autonomic regulation and attention. In B. A. Campbell, H. Hayne, & R. Richardson (Eds.), Attention and information processing in infants and adults (pp. 201–223). Hillsdale, NJ: Lawrence Erlbaum. Porges, S. W., Doussard-Roosevelt, J. A., & Maiti, A. K. (1994). Vagal tone and the physiological regulation of emotion. In N. Fox (Ed.), The development of emotion regulation: Biological and behavioral considerations. Monographs of the Society for Research in Child Development, 59 (2–3, Serial No. 240), pp. 167–186. Powers, S. I., Pietromonaco, P. R., Gunlicks, M., & Sayer, A. (2006). Dating couples’ attachment styles and patterns of cortisol reactivity and recovery in response to a relationship conﬂict. Journal of Personality and Social Psychology, 90, 613–628. Radloff, L. S. (1977). The CES-D Scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1, 385–401. Ramirez, J. M. (2003). Hormones and aggression in childhood and adolescence. Aggression and Violent Behavior, 8, 621–644. Reis, H. T. (2001). Relationship experiences and emotional well-being. In C. D. Ryff & B. H. Singer (Eds.), Emotion, social relationships, and health (pp. 57–85). New York: Oxford University Press. Reis, H. T., & Franks, P. (1994). The role of intimacy and social support in health outcomes: Two processes or one? Personal Relationships, 1, 185–197. Renken, B., Egeland, B., Marvinney, D., & Mangelsdorf, S. (1989). Early childhood antecedents of aggression and passive-withdrawal in early elementary school. Journal of Personality, 57, 257–281. Repetti, R. L., Taylor, S. E., & Seeman, T. E. (2002). Risky families: Family social environments and the mental and physical health of offspring. Psychological Bulletin, 128, 330–366. Rholes, W. S., Simpson, J. A., & Orina, M. M. (1999). Attachment and anger in an anxietyprovoking situation. Journal of Personality and Social Psychology, 76, 940–957. Richards, M. H., Crowe, P. A., Larson, R. W., & Swarr, A. (1998). Developmental patterns and gender differences in the experience of peer companionship during adolescence. Child Development, 69, 154–163. Roisman, G. I., Collins, W. A., Sroufe, L. A., & Egeland, B. (2005). Predictors of young adults’ representations of and behavior in their current romantic relationship: Prospective tests of the prototype hypothesis. Attachment & Human Development, 7, 105–121.

Developmental Perspectives

131

Rosenstein, D. S., & Horowitz, H. A. (1996). Adolescent attachment and psychopathology. Journal of Consulting and Clinical Psychology, 64, 244–253. Rothbart, M. K. (1991). Temperament: A developmental framework. In J. Strelau & A. Angleneitner (Eds.), Explorations in temperament: International perspectives on theory and measurement (pp. 61–74). New York: Plenum Press. Rothbart, M. K., & Derryberry, D. (1981). Development of individual differences in temperament. In M. E. Lamb & A. L. Brown (Eds.), Advances in developmental psychology (Vol. 1, pp. 37–86). Hillsdale, NJ: Earlbaum. Ryff, C. D. & Singer, B. H. (Eds). (2001). Emotions, social relationships, and health. New York: Oxford University Press. Ryff, C. D., Singer, B. H., Wing, E., & Love, G. D. (2001). Elective afﬁnities and uninvited agonies: Mapping emotion with signiﬁcant others onto health. In C. D. Ryff & B. H. Singer (Eds.), Emotions, social relationships, and health (pp. 133–174). New York: Oxford University Press. Saarni, C. (1992). Children’s emotional-expressive behaviors as regulators of others’ happy and sad emotional states. San Francisco, CA: Jossey-Bass. Salovey, P., & Mayer, J. D. (1990). Emotional intelligence. Imagination, Cognition, and Personality, 9, 185–211. Salovey, P., Mayer, J. D., Goldman, S. L., Turvey, C., & Palfai, T. P. (1995). Emotional attention, clarity, and repair: Exploring emotional intelligence using the trait meta-mood scale. In J. W. Pennebaker (Ed.), Emotion, disclosure, and health (pp. 125–154). Washington D.C.: American Psychological Association. Saul, J. P. (1990). Beat-to-beat variations of heart rate reﬂect modulation of cardiac autonomic outﬂow. News in Psychological Science, 5, 32–37. Savin-Williams, R. C., & Berndt, T. J. (1990). Friendship and peer relations. In S. S. Feldman & G. R. Elliott (Eds.), At the threshold: The developing adolescent (pp. 277–307). Cambridge, MA: Harvard University Press. Scarpa, A., & Raine, A. (1997). Psychophysiology of anger and violent behavior. Psychiatric Clinics of North America, 20, 375–394. Schaefer, E. S. (1965). Children’s reports of parental behavior: An inventory. Child Development, 36, 413–424. Schiefelbein, V. L., & Susman, E. J. (2006). Cortisol levels and longitudinal cortisol change as predictors of anxiety in adolescents. Journal of Early Adolescence, 26, 397–413. Schneiderman, N. (1983). Behavior, autonomic function and animal models of cardiovascular pathology. In T. M. Dembroski, T. H. Schmidt, & G. Blumchen (Eds.), Biobehavioral bases of coronary heart disease (pp. 304–351). New York: Karger. Schore, A. N. (1996a). Effects of a secure attachment relationship on right brain development, affect regulation, and infant mental health. Infant Mental Health Journal, 22, 269–276. Schore, A. N. (1996b). The experience-dependent maturation of a regulatory system in the orbital prefrontal cortex and the origin of developmental psychopathology. Development and Psychopathology, 8, 59–87. Schore, A. N. (2000). Attachment and the regulation of the right brain. Attachment and Human Development, 2, 23–47. Seeman, T. E. (2001). How do others get under our skin? Social relationships and health. In C. D. Ryff & B. H. Singer (Eds.), Emotion, social relationships, and health (pp. 189–209). New York: Oxford University Press. Senchak, M., & Leonard, K. E. (1992). Attachment styles and marital adjustment among newlywed couples. Journal of Social and Personal Relationships, 9, 51–64. Shaver, P. R., & Mikulincer, M. (2002). Attachment-related psychodynamics. Attachment & Human Development, 4, 133–161.

132

Lisa M. Diamond and Christopher P. Fagundes

Shaver, P. R., Hazan, C., & Bradshaw, D. (1988). Love as attachment: The integration of three behavioral systems. In R. J. Sternberg & M. L. Barnes (Eds.), The psychology of love (pp. 193–219). New Haven, CT.: Yale University Press. Shoal, G. D., Giancola, P. R., & Kirillova, G. P. (2003). Salivary cortisol, personality, and aggressive behavior in adolescent boys: A 5-year longitudinal study. Journal of the American Academy of Child & Adolescent Psychiatry, 42, 1101–1107. Siegel, D. J. (1999). The developing mind. New York: The Guilford Press. Siegel, D. J. (2001). Toward an interpersonal neurobiology of the developing mind: Attachment relationships, ‘‘mindsight,’’ and neural integration. Infant Mental Health Journal, 22, 67–94. Silk, J. S., Steinberg, L., & Morris, A. S. (2003). Adolescents’ emotion regulation in daily life: Links to depressive symptoms and problem behavior. Child Development, 74, 1869–1880. Simpson, J. A. (1990). Inﬂuence of attachment styles on romantic relationships. Journal of Personality and Social Psychology, 59, 971–980. Simpson, J. A., & Rholes, W. S. (1994). Stress and secure base relationships in adulthood. Advances in Personal Relationships, 5, 181–204. Simpson, J. A., & Rholes, W. S. (1998). Attachment in adulthood. In J. A. Simpson & W. S. Rholes (Eds.), Attachment theory and close relationships (pp. 3–21). New York: Guilford. Simpson, J. A., Rholes, W. S., & Nelligan, J. S. (1992). Support seeking and support giving within couples in an anxiety-provoking situation: The role of attachment styles. Journal of Personality and Social Psychology, 62, 434–446. Simpson, J. A., Rholes, W. S., & Phillips, D. (1996). Conﬂict in close relationships: An attachment perspective. Journal of Personality and Social Psychology, 71, 899–914. Simpson, J. A., Ickes, W., & Grich, J. (1999). When accuracy hurts: Reactions of anxiousambivalent dating partners to a relationship-threatening situation. Journal of Personality and Social Psychology, 76, 754–769. Snidman, N. (1989). Behavioral inhibition and sympathetic inﬂuence on the cardiovascular system. In J. S. Reznick (Ed.), Perspectives on behavioral inhibition (pp. 51–70). Chicago: University of Chicago Press. Soenens, B., Vansteenkiste, M., Lens, W., Luyckx, K., Goossens, L., Beyers, W., & Ryan, R. M. (2007). Conceptualizing parental autonomy support: Adolescent perceptions of promotion of independence versus promotion of volitional functioning. Developmental Psychology, 43, 633–646. Solomon, J., & George, C. (1999). The measurement of attachment security in infancy and childhood. In J. Cassidy & P. R. Shaver (Eds.), Handbook of attachment theory and research (pp. 287–318). New York: Guilford. Spangler, G., & Zimmermann, P. (1999). Attachment representation and emotion regulation in adolescents: A psychobiological perspective on internal working models. Attachment and Human Development, 1, 270–290. Spear, L. P. (2000). The adolescent brain and age-related behavioral manifestations. Neuroscience and Biobehavioral Reviews, 24, 417–463. Spear, J. F., Kronhaus, K. D., Moore, E. N., & Kline, R. P. (1979). The effect of brief vagal stimulation on the isolated rabbit sinus node. Circulation Research, 44, 75–88. Sroufe, L. A. (2005). Attachment and development: A prospective, longitudinal study from birth to adulthood. Attachment & Human Development, 7, 349–367. Sroufe, L. A., & Fleeson, J. (1986). Attachment and the construction of relationships. In W. W. Hartup & Z. Rubin (Eds.), Relationships and development (pp. 51–69). Hillsdale, NJ: Erlbaum. Sroufe, L. A., Egeland, B., Carlson, E., & Collins, W. A. (2005). Placing early attachment experiences in developmental context: The Minnesota longitudinal study. New York: Guilford Publications.

Developmental Perspectives

133

Stansbury, K., & Gunnar, M. R. (1994). Adrenocortical activity and emotion regulation. In N. Fox (Ed.), The development of emotion regulation: Biological and behavioral considerations. Monographs of The Society for Research in Child Development, 59 (2–3, Serial No. 240), pp. 108–134. Stanton, A. L., Danoff-Burg, S., Cameron, C. L., & Ellis, A. P. (1994). Coping through emotional approach: Problems of conceptualization and confounding. Journal of Personality and Social Psychology, 66, 350–362. Stanton, A. L., Kirk, S. B., Cameron, C. L., & Danoff-Burg, S. (2000). Coping through emotional approach: Scale construction and validation. Journal of Personality and Social Psychology, 78, 1150–1169. Steinberg, L., & Morris, A. S. (2001). Adolescent development. Annual Review of Psychology, 52, 83–110. Steinberg, L., & Silk, J. S. (2002). Parenting adolescents. Mahwah, NJ: Lawrence Erlbaum Associates. Steinberg, L., & Silverberg, S. B. (1986). The vicissitudes of autonomy in early adolescence. Child Development, 57, 841–851. Stifter, C. A., & Fox, N. A. (1990). Infant reactivity: Physiological correlates of newborn and 5-month temperament. Developmental Psychology, 26, 582–588. Taylor, S. E., Repetti, R. L., & Seeman, T. E. (1997). Health psychology: What is an unhealthy environment and how does it get under the skin? Annual Review of Psychology, 48, 411–447. Taylor, S. E., Dickerson, S. S., & Klein, L. C. (2002). Toward a biology of social support. In C. R. Snyder & S. J. Lopez (Eds.), Handbook of positive psychology (pp. 556–569). London: Oxford University Press. Thompson, R. (1994). Emotion regulation: A theme in search of deﬁnition. In N. Fox (Ed.), The development of emotion regulation: Biological and behavioral considerations. Monographs of the Society for Research in Child Development, 59 (2–3, Serial No. 240), pp. 225–252. Tobin, R. M., & Graziano, W. G. (2006). Development of regulatory processes through adolescence: A review of recent empirical studies. In D. K. Mroczek & T. D. Little (Eds.), Handbook of personality development (pp. 263–283). Mahwah, NJ: Lawrence Erlbaum Associates. Tomkins, S. S. (1984). Affect theory. In P. Ekman (Ed.), Emotion in the human face (2nd ed., pp. 353–395). New York: Cambridge University Press. Torquati, J. C., & Vazsonyi, A. T. (1999). Attachment as an organizational construct for affect, appraisals, and coping of late adolescent females. Journal of Youth and Adolescence, 28, 545–562. Trinke, S. J., & Bartholomew, K. (1997). Hierarchies of attachment relationships in young adulthood. Journal of Social and Personal Relationships, 14, 603–625. Uchino, B. N., Holt Lunstad, J., Uno, D., & Flinders, J. B. (2001). Heterogeneity in the social networks of young and older adults: Prediction of mental health and cardiovascular reactivity during acute stress. Journal of Behavioral Medicine, 24, 361–382. van Ijzendoorn, M. H., Schuengel, C., & Bakermans-Kranenburg, M. J. (1999). Disorganized attachment in early childhood: Meta-analysis of precursors, concomitants, and sequelae. Development and Psychopathology, 11, 225–249. Walker, E. F., Walder, D. J., & Reynolds, F. (2001). Developmental changes in cortisol secretion in normal and at-risk youth. Development and Psychopathology, 13, 721–732. Waters, E., Merrick, S., Treboux, D., Crowell, J., & Albersheim, L. (2000). Attachment security in infancy and early adulthood: A twenty-year longitudinal study. Child Development, 71, 678–683.

134

Lisa M. Diamond and Christopher P. Fagundes

Weinﬁeld, N. S., Sroufe, L. A., & Egeland, B. (2000). Attachment from infancy to early adulthood in a high-risk sample: Continuity, discontinuity, and their correlates. Child Development, 71, 695–702. Weiss, R. S. (1982). Attachment in adult life. In C. M. Parkes & J. Stevenson-Hinde (Eds.), The place of attachment in human behavior (pp. 171–184). New York: Basic Books. Williams, R. B., Jr. (1985). Neuroendocrine response patterns and stress: Biobehavioral mechanisms of disease. In R. B. Williams, Jr. (Ed.), Perspectives on behavioral medicine. Volume 2. Neuroendocrine control and behavior (pp. 71–101). Orlando, FL: Academic Press. Wills, T. A., Windle, M., & Cleary, S. D. (1998). Temperament and novelty seeking in adolescent substance use: Convergence of dimensions of temperament with constructs from Cloninger’s theory. Journal of Personality and Social Psychology, 74, 387–406. Youngblade, L. M., Park, K. A., & Belsky, J. (1993). Measurement of young children’s close friendship: A comparison of two independent assessment systems and their associations with attachment security. International Journal of Behavioral Development, 16, 563–587.

FUNCTION REVISITED: HOW INFANTS CONSTRUE FUNCTIONAL FEATURES IN THEIR REPRESENTATION OF OBJECTS

Lisa M. Oakesa and Kelly L. Madoleb a

b

CENTER FOR MIND AND BRAIN, UNIVERSITY OF CALIFORNIA, DAVIS, CA 95618, USA DEPARTMENT OF PSYCHOLOGY, WESTERN KENTUCKY UNIVERSITY, BOWLING GREEN, KY 42101-1030, USA

I. INTRODUCTION II. THE CONSTRUCT OF OBJECT FUNCTION A . DEFINING FUNCTION B . WHAT KIND OF PROPERTY IS FUNCTION? III. A NEW CONCEPTION OF FUNCTION A . FUNCTION AS AN EMERGENT FEATURE OF OBJECTS AND EVENTS B . INFANTS’ SENSITIVITY TO THE COMPONENTS OF FUNCTION IV. OUR RESEARCH ON INFANTS’ ATTENTION TO AND REPRESENTATION OF FUNCTION A . INFANTS’ SHIFTING ATTENTION TO FUNCTION AND APPEARANCE B . INFANTS’ DEVELOPING ATTENTION TO APPEARANCE AND FUNCTION C . INFANTS’ ATTENTION TO THE RELATION BETWEEN APPEARANCE, ACTION, AND OUTCOME D . OBJECT FUNCTION IN INFANTS’ CATEGORIZATION V. CONCLUSIONS ACKNOWLEDGEMENT REFERENCES

I. Introduction prior to operatory classiﬁcations based on y objective equivalences y there exists a mode of classiﬁcation based on the relationship between actions which are functional y (Piaget et al., 1977, p. 15) an object is ﬁrst identiﬁed as having important functional relations y and y perceptual analysis is derivative of the functional concept, not a priori essential to it. (Nelson, 1974, p. 284)

135 Advances in Child Development and Behavior R.V. Kail : Editor

Copyright r 2008 Elsevier B.V. All rights reserved.

136

Lisa M. Oakes and Kelly L. Madole Affordances relate the utility of things, events, and places to the needs of animals and their actions in fulﬁlling them y Affordances themselves are perceived and, in fact, are the essence of what we perceive. (Gibson, 1982, p. 60)

These quotations make clear that object function (or affordance, in Gibson’s terms) has long been believed to play a major role in infants’ conceptual development. Function has been argued to be central to some concepts (Nelson, 1972, 1973) and may be at the core of infants’ and young children’s conceptions of objects and artifacts (Keil, 1989). Functional commonalities can be more compelling in children’s forming of concepts and learning of words than other types of commonalities such as color or other aspects of perceptual similarity (Horst, Oakes, & Madole, 2005; Kemler Nelson et al., 2000b). Moreover, the importance of function may have an underlying neurophysiological basis. For example, apraxic patients have deﬁcits speciﬁc to their knowledge of object function, deﬁned as the intended use of those objects (Buxbaum & Saffran, 2002). It is somewhat surprising, given its importance, that researchers have not adopted a single, uniﬁed operational deﬁnition of object function. For example, in psychological research function sometimes refers to the goal of acting on an object (e.g., to sharpen pencils, Buxbaum & Saffran, 2002), sometimes refers to the characteristic actions of the objects themselves (e.g., an object swings when hung from a hook, Booth & Waxman, 2002), and sometimes refers to the consequences of acting in a particular way on an object (e.g., a squeaking sound is produced when an object is squeezed, Perone & Oakes, 2006). In other words, function is not uniformly deﬁned as why one acts on objects, how an object reacts to actions, or the consequence of actions on objects. For example, a shoe can protect the foot (why you use it), soar across the yard when thrown (the reaction when acted on), and make a loud noise (the consequence of banging it on the table). Should such different kinds of function be considered interchangeable in designing studies probing infants’ emerging understanding of function? This lack of a clear deﬁnition is perhaps even more surprising because function has often been considered a conceptual property of objects. For example, in her seminal work on children’s developing semantic knowledge, Nelson (1973, 1974, 1979) argued that children’s ﬁrst categories are based on function. That is, children have a conceptual basis for extending labels, rather than merely extending labels to objects that simply look alike. Perceptual similarities are used to extend labels only if they predict function—such as the ability of a bowl-shaped object to be put on the head like a hat. In subsequent work, function is at least partly equated with a deep, conceptual knowledge that goes beyond attention to appearance, physical structure, or idiosyncratic uses of an object. For example,

How Infants Construe Functional Features

137

according to Bloom (1996), the intended use of the designer of an object is the essence, or core, of artifact categories, even when an actor uses the object for a different purpose (for similar arguments see Casler & Kelemen, 2005; Keil, 1989; Matan & Carey, 2001). The results of several studies have been taken as evidence that even very young children’s understanding of function reﬂects a ‘‘design stance,’’ or the understanding that function is related to the intention of the designer (e.g., Jaswal, 2006). That is, functions are seen as intentional properties of objects, and evidence that infants recognize functions as intentional indicates that they are recognizing deep, conceptual properties of objects, as opposed to superﬁcial, perceptual properties. Similarly, arguments about the shape-bias in early language development often pit apparently superﬁcial shape similarities against deeper similarities in function. Researchers have examined whether shape itself is the salient feature—the assumption being that shape is a less sophisticated way of extending a label—or whether children consider shape and function together, which would suggest an understanding that structural features constrain functional features (Diesendruck, Hammer, & Catz, 2003a; Diesendruck, Markson, & Bloom, 2003b; Kemler Nelson et al., 2000a), and can be a cue to hidden functional features (Welder & Graham, 2001). As a result of this general conception of function, in infant studies function is often pitted against appearance in an effort to show the conceptual basis of infants’ categorization (e.g., Trauble & Pauen, 2008). However, we previously argued that the distinction between conceptual and perceptual categorization is not clear-cut (Madole & Oakes, 1999; Oakes & Madole, 2003). That is, it is not always clear that ‘‘conceptual’’ properties differ fundamentally from ‘‘perceptual’’ properties. Instead, properties may vary along a continuum in terms of how heavily they are speciﬁed by perceptual aspects (hue, contours, and so on) versus conceptual aspects (breathes, self-produced motion, and so on). Function, in fact, provides a particularly compelling example of a property that has both ‘‘conceptual’’ and ‘‘perceptual’’ aspects. Like many structural or appearance-based features, function can often be observed by looking at an object—perhaps as someone else interacts with it. Function, like many object characteristics, emerges from other properties or pieces of information available—such as whether the object is acted on, whether there is an outcome, the characteristics of the actor, the way the actor performs the action, and the individual’s past experience with the object. Although some researchers might interpret this complexity as evidence that function is conceptual in nature, in reality appearance-based features are characterized by similar complexity. Our perception of an object’s color, for example, is the combination of hue, saturation, and brightness. Moreover, an object’s

138

Lisa M. Oakes and Kelly L. Madole

perceived color changes with changes in the lighting conditions and as the object is moved or manipulated. Similarly, we recognize an object’s shape only by considering its boundaries, shading, how parts are organized, and so on. Like color, the apparent shape of an object changes as the object is seen from different views or as the object is manipulated. Thus, both appearance and function are complexly determined by multiple factors, and perceiving and representing either type of feature involves the integration of multiple sources of information. The distinction between obvious and nonobvious or conceptual and perceptual, therefore, is misleading, and leads to an incomplete understanding of how infants use many different features in their categorization. In this chapter, we examine the issue of object function and propose a new way of thinking about function as it relates to early conceptual development. Our view of function is similar, in many ways, to Barsalou and colleagues’ HIPE model of function (Barsalou, Sloman, & Chaigneau, 2005; Chaigneau, Barsalou, & Sloman, 2004). In this model, object functions are dynamically constructed using knowledge about how an object has been or was designed to be used (H), an actor’s intentions while interacting with the object (I), aspects of the physical environment that determine how an object can be used (P), and ﬁnally the event sequence (E) that involves the object’s behavior and any outcomes that occur during the interaction. As in that model, we do not see function as an independent, unitary property of objects. Instead, function is a mental model or construct that emerges as an actor interacts with an object and the state of the object changes over time. We did not think about function this way when we began studying the role of function in infants’ learning about, identifying, and categorization of novel objects nearly 20 years ago. Over the years our thinking about function has changed—from an intuitive conception of function (which we argue characterizes much of the work on object function in infancy) to a consideration of the features that comprise object function and how infants represent that collection of features. This chapter is the product of this nearly two decades of thinking about these issues. This chapter is divided into three major sections. In the ﬁrst section, we provide background information and an overview of the issues surrounding the study of object function. Here, we discuss how function has (and has not) been deﬁned in the literature as well as the nature of function as a property of objects. In the second section, we elaborate our conception of function as an emergent property of an object or event given other features or properties. In the context of this discussion we identify what we believe are the important features that typically comprise ‘‘function,’’ and we

How Infants Construe Functional Features

139

discuss evidence from developmental and cognitive science, as well as neuroscience, about how such features are represented. Finally, in the third section we describe our own program of work examining infants’ sensitivity to, representation of, and categorization using object function. This work began before much of the empirical and theoretical work described in the ﬁrst parts of the chapter was conducted. Thus, some of our original assumptions were naı¨ ve with respect to how function should be deﬁned or construed. However, this work provided an important starting point for our current understanding of object function and the future direction of our work in this area.

II. The Construct of Object Function It has been widely demonstrated that functional properties play a central role in object recognition and categorization (Keil, 1989; Rips, 1989). Although researchers have disagreed about whether young children’s categories depend on superﬁcial, ‘‘perceptual’’ similarities or deeper, ‘‘conceptual’’ functional similarities (Diesendruck et al., 2003a; Nelson, 1973), the features considered the ‘‘functions’’ of objects clearly have a powerful inﬂuence on how infants understand and learn about objects. For example, observing how objects are used (and perhaps the resulting effect) can increase infants’ attention to a dimension that predicts that feature. Wilcox and Chapa (2004), for example, showed that demonstrating the different functions of two objects that differed in color appeared to highlight the signiﬁcance of color: The function demonstration induced 7- and 9-month-old infants to use the colors of a different set of objects in a subsequent object individuation task; infant who did not see the function demonstration did not use color to individuate objects. Booth and Waxman (2002) observed a similar effect when testing 14-month-old infants’ attention to the surface similarity of a collection of objects. Infants only seemed to recognize those commonalities when the objects’ functions were demonstrated during learning. Function may also be important in infants’ mapping of names onto objects. Generally, children use object shape to extend the names of novel objects (e.g., Landau, Smith, & Jones, 1988). However, when the function of objects is demonstrated, children as young as 2 years of age will generalize the names of those objects to other new objects with the same function (Kemler Nelson et al., 2000b). Thus, ample evidence across many different kinds of studies indicates that functional properties play a central role in the identiﬁcation and categorization of objects.

140

Lisa M. Oakes and Kelly L. Madole

A. DEFINING FUNCTION Despite its signiﬁcance in young children’s (and adults’) object concepts, function has rarely been clearly, uniformly, or explicitly deﬁned. Consider the following deﬁnitions that come from research on infant conceptual development, adult language, and neuropsychological work with aphasic patients. Function is what an object is used for (Kellenbach, Brett, & Patterson, 2003, p. 30) the action an object affords and is speciﬁcally ﬁtted for (Wilcox & Chapa, 2004, p. 279) the use of a thing (Booth, 2006, p. 146) the purpose for which a particular object is conventionally used. (Regier, Carlson, & Corrigan, 2005, p. 191) capacities of an object to act or be acted upon in a speciﬁc manner y (Prasada, 2005, p. 206)

This list shows that there is no single, clear deﬁnition of function used in the literature. Moreover, many studies claim to manipulate function or probe knowledge of object function without providing any deﬁnition at all (e.g., Buxbaum & Saffran, 2002; Trauble & Pauen, 2008). Often, function is deﬁned intuitively—researchers create stimulus objects that have a particular ‘‘function,’’ but that function is based on the researchers’ intuition of what a function is. Some researchers simply state that function is the ‘‘use’’ of the object without clearly indicating whether it is the intended use by the creator, how the object is typically used, or the possible uses of the object (e.g., Booth, 2006). In part, function may not be explicitly deﬁned because researchers have a belief that object function is selfevident—we recognize it when we see it. An additional problem arises because deﬁnitions of function vary depending on the researcher’s goals in studying function. For some, functions are the uses intended by the creator of the object, and thus the primary research goal is to determine when children adopt this ‘‘design stance’’ (Jaswal, 2006; Matan & Carey, 2001). How structural properties constrain what can be done with an object is sometimes important in the functions used (Kemler Nelson, 1999), although it is not always clear whether those structural constraints are imposed by the potential actions by the individual (and thus would change with development) or constraints imposed by the design of the maker, or both. For others, functions are actions on objects that produce outcomes (e.g., Wilcox & Chapa, 2004), and can be arbitrarily related to the object structure or the intention of the

How Infants Construe Functional Features

141

designer. Clearly, all these features, along with many others (the physical limitations of the actor, the consequence of acting on the object, how the object has been used in the past, and so on) can be considered aspects of function. Such differing deﬁnitions may not reﬂect the same underlying construct. Features or properties that are labeled ‘‘object function’’ in different studies may not actually reﬂect the same type of feature or property, and thus they may not be processed in the same way. For example, Trauble and Pauen (2008) operationalized function by creating complex novel objects with parts that allowed them to be used with another device to produce some outcome—for example, a protrusion on the object could be inserted in an opening of an apparatus to produce a sound. Is such a function equivalent to squeezing an object to produce a squeak (Horst et al., 2005), using a scoop to pour salt (Wilcox & Chapa, 2004), or hanging a bag by its handle so it would swing (Booth & Waxman, 2002)? Although such functions are treated interchangeably, the fact that researchers have adopted different criteria for considering a feature as part of the object’s function makes it difﬁcult to compare the results of their work. For example, Trauble and Pauen (2008) and Madole, Oakes, and Cohen (1993) both used object-examining tasks to investigate infants’ attention to object function. However, Trauble and Pauen (2008) found that 11-month-old infants were sensitive to object function, whereas Madole, Oakes, and Cohen found that 14-month-old infants, but not 10-month-old infants, were sensitive to function. Such discrepant ﬁndings may reﬂect differences in how the researchers deﬁned function (and instantiated function in their stimuli). The lack of consensus in what constitutes an object’s function, limits comparisons across investigations and weakens converging evidence about the importance of function. Despite the lack of a single, agreed-upon operational deﬁnition of object function, most conceptions of function include the same type of features, at least in the context of infant conceptual development. Functions generally involve an action and an outcome. An object’s function is that it bangs on a peg and makes a sound (Wilcox & Chapa, 2004), swings to hit chimes, making them ring (Booth, 2006), or clicks when rolled (Horst et al., 2005; Perone & Oakes, 2006). Functions also appear to include actions performed by a human agent. In virtually all studies examining infants’ attention to function, human agents (or at least human hands) demonstrate object functions such as pouring salt (Wilcox & Chapa, 2004) or pushing an object which then clicks (Horst et al., 2005; Perone & Oakes, 2006). Researchers differ, however, in how they emphasize these properties. For example, for some, functions require that actions on objects produce an outcome. Scooping actions result in the pouring of salt. Swinging an object

142

Lisa M. Oakes and Kelly L. Madole

causes chimes to ring. Squeezing an object causes a squeaking sound to be heard. Moreover, the presence of an outcome may have particular signiﬁcance when learning about objects. Wilcox and Chapa (2004), for example, found that although demonstrating the different functions of differently colored objects enhanced attention to color, demonstrating ‘‘non-functional’’ actions (i.e., moving the object without any obvious outcome) did not enhance infants’ responding to color. The power of function to enhance attention to predictive features lies, at least partly, in the presence of an outcome. However, most instantiations of function in the infant literature conﬂate the action and the outcome. This is a problem because we still know relatively little about how these different aspects of function are perceived by infants—that is, we do not know whether infants represent the action and outcome as a uniﬁed, single feature of function. Moreover, conﬂating action and outcome does not acknowledge the contribution of each dimension in infants’ attention to, processing of, and representation of the events. Consider the outcomes in the previous examples: salt moves from one location (the scoop) to another (the table) during a few seconds of the event, a chime rings for a short time when the object strikes it, squeaking is heard intermittently during the action on the object. These outcomes are intermittent, dynamic, and often involve a sound. Thus, the outcomes may attract infants’ attention and their representation of function may be a result of their attention to the outcome. For example, the difference between the functional and non-functional manipulation observed by Wilcox and Chapa’s (2004) may not reﬂect the fact that one condition involved function and the other did not; rather, the difference may reﬂect the fact that one condition involved an outcome and the other did not, and the effects may solely have been due to infants’ attending to the outcomes. There is no evidence that infants attended to the action (or associated the action with a particular outcome) in either context. Another functional feature emphasized by researchers is the presence of an actor. Properties of objects that could be construed as relevant to function—for example, their characteristic movement trajectories—but that do not involve a human agent typically are not described in terms of function. Rakison (2004; Rakison & Poulin-Dubois, 2002), for example, does not use the label ‘‘function’’ when studying infants’ representations of the distinctive movement trajectories of objects. Booth (2000) made her actions ‘‘non-functional’’ by having them occur in the absence of an actor. Clearly, therefore, researchers assume that functions involve the relation between an actor and an object. Interestingly, infants’ learning of nonfunctional features—such as movement trajectories—does parallel the development of infants’ learning of functions in some ways. Although

How Infants Construe Functional Features

143

infants, like adults, may assume that functions are actions performed on an object by an agent, this hypothesis has not been adequately tested. We have not yet systematically studied whether the presence or absence of an agent (human or non-human) changes infants’ understanding of a collection of actions that might be described as the object’s function. Note, however, that in the context of animate objects function is meaningful even in the absence of an agent. For example, Rakison and Cohen (1999) referred to the functions of the legs of cows (walking), and observed that in the second year infants interacted with objects with legs in functionally appropriate ways (i.e., they made them hop or walk). This example illustrates the many uses of the term function. When referring to an inanimate object, the function may require actions performed by an agent. But, when referring to parts of an animate object, a more teleological referent may be appropriate. That is, for animate objects, the functions of parts may refer to what those parts allow the animate being to do—or what those parts are for (Kelemen et al., 2003). A ﬁnal aspect of function that is critical according to many theorists is the intention of the creator (Bloom, 1996; Casler & Kelemen, 2007). Signiﬁcant research has been aimed at attempting to understand when children adopt this view of function. For example, Casler and Kelemen (2005) observed that children as young as 2 years of age can learn after only one episode what a tool is ‘‘for.’’ They suggest that this ﬁnding reveals that very young children are biased to categorize tools by their intentional design, and have the foundation of the design stance (but see Truxaw et al., 2006, for an alternative view). These features—actions that produce outcomes, the intentional action of a human actor—are important for the researchers’ deﬁnitions of function. We know little about how infants characterize function. For example, infants might well consider self-produced movement, agent-produced movements, and agent-produced movements leading to an outcome all as ‘‘functional.’’ Although we know that older children are sensitive to the causal relation between functional and perceptual features (Kemler Nelson et al., 2000b) and to the conventional and intended uses of objects (Diesendruck et al., 2003b; Kemler Nelson, Herron, & Holt, 2003) little is known about what features are important for infants’ construal of a feature (or set of features) as functional. A truly general deﬁnition of function, in the context of conceptual development, may be impossible. Still, we need to be more precise about the use of the term ‘‘function’’ for at least two reasons. First, different ﬁndings with regard to the use of functional properties in identifying and categorizing objects may stem, in part, from differing deﬁnitions of function. Resolving these different results may be possible only through

144

Lisa M. Oakes and Kelly L. Madole

careful clariﬁcation of the relevant terms. Second, attention to functional properties is likely a gradually developing achievement that proceeds from attention to simpler features. Understanding how these different features come together to produce an understanding of function may be critical to fully understanding conceptual development.

B. WHAT KIND OF PROPERTY IS FUNCTION? A potentially more serious issue than a lack of a deﬁnition for function is with the fundamental way that function has typically been conceptualized. Function has, at least implicitly, been characterized as a property of objects. In other words, function is considered to be inherent in objects and is one of many separable properties or features that together allow us to categorize objects. This way of thinking about function in the study of infant cognition can be traced to the seminal work of Katherine Nelson (1972, 1973, 1974). According to Nelson, what an individual can do with an object or what objects themselves do is at the core of children’s concepts of objects; the surface feature properties that help deﬁne the category are constrained by this function. For example, a child’s ﬁrst category of ‘‘hat’’ denotes objects (of a general shape) that can be put on one’s head. Similarly, Mervis (1985), in describing one child’s horn category, argued that although the child used form to infer the function of the items, all the items in the horn category had the same function—they could be blown. Interestingly, Nelson (1974) argued that early concept formation is based on an unanalyzed representation of a whole object, rather than on individual features or attributes. However, subsequent research often treated function as though it is, in fact, an individual feature or attribute that can be manipulated in much the same way that one manipulates a feature like color or shape. One approach has been to consider the appearance and function as two separate features that can be crossed to produce novel objects. In a classic study, Gentner (1978), for example, presented children with two objects, each with a distinct function and a distinct appearance. Children’s label extensions were tested by presenting them with a hybrid object—an object with the appearance of one object but the function of the other. In a different design that similarly considers surface features and function as properties of objects, Landau, Smith and Jones (1998) taught children the names of objects and then asked them to extend those names to new objects that either were the same as the familiar object in terms of shape or a material-based function. This design similarly pits the two types of properties against one another, treating them as somewhat equivalently good object features. Indeed, in these and other

How Infants Construe Functional Features

145

studies of infants’ attention to object function, the researchers’ goal was to determine whether function or some alternate object property was the driving factor in categorizing or naming an object. This conceptualization of function derives, in part, from a particular view of object concepts. Concepts are often referred to as the mental representation of things—for example, a category of objects (Murphy, 2002). This deﬁnition is useful for pointing out that concepts are (in a sense) ‘‘in the head’’ and are not the same as the ‘‘real’’ object. For example, one’s concept of dog is not the same as the category of real dogs that exist in the world. Traditionally, theorists have viewed these mental representations as stable and symbolic (e.g., Mandler, 2004). This relatively static view of concepts has led researchers to view the features that comprise those concepts as stable as well. It is, therefore, not surprising that much work on object function has focused on function as an intrinsic property of objects that has its own correspondingly stable mental representation. For example, if our concept of fork is characterized by a stable, symbolic mental representation, then our representation of this object’s function, to eat with, is also stable and symbolic. The notion that function is a separable feature of objects that can be manipulated arose from this tradition. However, subsequent research and thinking has revealed that function— like many object features—is not a static, separable feature of objects. Instead, function, and indeed objects concepts more generally, are dynamic and changing depending on the context.

III. A New Conception of Function A. FUNCTION AS AN EMERGENT FEATURE OF OBJECTS AND EVENTS We believe that the prevalent way of thinking about function is misleading. The very idea that function is a unitary feature that is inherent in a particular object, or set of objects, may have contributed to the imprecision in how function is instantiated in the literature. It is not at all obvious that function is always (or ever) a feature of objects in the same way as is color or shape. That is, although color, shape, and function all may be multiply determined, only function requires the interaction between the object and some actor—either the creator of the object (who presumably had an intended use of the object) or an individual currently interacting with the object (presumably for some purpose). In fact, function may never be fundamentally intrinsic to an object. Instead function is intrinsic to the relation between an object and an actor. Thus, function may

146

Lisa M. Oakes and Kelly L. Madole

be emergent in a different way than are other object properties that are attended to, encoded, and used as a basis of categorization. Our view of function has many similarities to Barsalou and colleagues’ (Barsalou et al., 2005; Chaigneau et al., 2004) HIPE model. Recall that in this model different pieces of knowledge about an object’s use, the actors intentions, the physical limitations of an object’s use, and the sequence of events when using an object are joined together in a causal model that determines one’s ‘‘functional sense’’ about a particular object. Thus, function is a relational construct that involves all of the aspects we described in the previous section. Unlike the views described earlier, however, Barsalou does not argue that function is a single property that reﬂects any one of those features—for example, it is not solely the designer’s intended use or the uses that are afforded by the physical structure. Rather, function emerges from many aspects of the object, the agent, and the context. According to this perspective, therefore, construing function as a property of objects that can be manipulated is misleading—what can be manipulated is a person’s experience with an object, their knowledge of the intentions of the designer, their understanding of how it can be acted on, and so on. This conception of function derives from the tradition of embodied cognition, which has a very different notion of concept from the view described earlier. Rather than focusing on concepts as mental states that refer to things outside the mind, the focus is on mechanisms that produce conceptual processing (Barsalou et al., 2003). For example, Smith (2005) argues for a dynamic systems view of cognition in which concepts are not ‘‘constants in the head,’’ but rather emerge in real time controlled by a number of variables. Smith points out that concepts are not constant, but change with the context, our knowledge, goals, and the like. Relative stability emerges, according to this view, because of the stability of such factors, not because of an underlying stable concept. Thus, no features are stable—even features such as color and shape depend on the viewing conditions and one’s goal and motivation at the moment. Our position is that infants’ concept of function is not an independent, unitary property of an object that can be manipulated (although we have deﬁned function in that way in the past), but emerges from the interaction of many factors and is stable to the extent that the context highlights the same features of the object, the context, and the interaction between the agent and the object. Although the HIPE theory is not developmental in focus, it provides a starting point for a theory of how attention to function, as an emergent property, might develop, particularly during infancy when conceptual understanding is emerging. That is, under the view that function is a property of an object (presumably the use intended by the designer), the

How Infants Construe Functional Features

147

function of the object does not change (it is to sharpen pencils, provide light, etc.) unless the object undergoes radical changes in structure. Indeed, researchers have asked whether children believe the function of objects change with changes in how the object is used (e.g., using a teapot as a vase) or changes in superﬁcial features (e.g., changing the color) (Gutheil et al., 2004; Matan & Carey, 2001). However, if function is property of the relation between an actor and an object, the function can change if the actor undergoes critical change (such as developing new motor, cognitive, or linguistic skills) even if the object remains the same over time. For example, a wooden block may have the function of banging before developing the manual dexterity to stack objects, but the function of stacking once such dexterity has developed. Consider some of the factors that can impact infants’ interactions with objects, potentially changing the function of those objects: (1) infants’ previous experience with that and similar objects, (2) infants’ ability to recall and apply that previous experience, (3) infants’ ability to act on the objects, (4) the salience of different features of the objects and the actor, (5) infants’ ability or inability to connect different parts of the event (e.g., the action and the outcome), and (6) contextual factors that may reduce or increase the information-processing demand placed on the infant. This list is not exhaustive, but serves as a starting point for illustrating the fact that function is, to a large extent, a moving target. No single feature that we can indisputably call function will be developmentally primary; instead, how infants conceptualize function—and whether the function they attend to, perceive, and encode corresponds to an adult-deﬁned function—will be determined by features of the context and the infants’ developmental level (Madole & Oakes, 1999; Oakes & Madole, 2000, 2003). Nelson’s (1979) classic account of object function reveals a sensitivity to the fact that function is not stable and unchanging. She described four ways function could be deﬁned: (1) actions on things, (2) the independent activity of a thing, (3) the reaction of a thing to an action on it, and (4) the use, or utility of a thing. Note that this classic view of function separates the key elements of the factors in the HIPE model: actions performed on objects, how objects react, either on their own or in reaction to those actions, and the intention of the actor or creator of the object. Unlike Barsalou, however, Nelson describes these as different ways of deﬁning function, not as elements of any function. These deﬁnitions of function appear to represent a rough progression in terms of the cognitive skills required. To recognize and represent the actions on things and the independent activity of things (1 and 2 above) requires only that one recognize and represent observable properties. An infant can directly observe whether an object is squeezed or whether it jumps. Indeed,

148

Lisa M. Oakes and Kelly L. Madole

by 4 or 5 months infants have these skills—they detect changes in action and how objects behave (for a review, see Kellman & Arterberry, 1998). Thus, from a young age infants have the ability to detect the most primitive aspects of function. However, if such features do form the foundation of a primitive recognition of function, this represents a very immature understanding of function relative to the way such properties have been operationalized in studies with older children and adults. The third deﬁnition—the reaction of a thing to an action on it—requires integrating two distinct events. For example, learning that an object squeaks when an actor squeezes it requires that the infant detect and encode the squeezing, squeaking, and the contingency between them. The ability to detect and learn such associations seems to emerge later than the ability to attend to and encode the individual features (Madole et al., 1993; Rakison, 2004; Younger & Cohen, 1986). Thus, we would expect that associating actions performed on objects and the outcomes of those actions would emerge relatively late in infancy. Finally, the fourth deﬁnition—function as the use or utility of a thing— seems to require abilities acquired beyond infancy. Speciﬁcally, if this aspect of function refers to the intended use of a thing—either intended by the designer or the actor—it seems unlikely that infants have access to this information. Infants can observe an object being used, but they cannot be told why the object was made or that, although we are currently using the object as a hammer, it is actually a shoe. As described earlier, a large number of investigators have been examining the developmental origins of the design stance, or the belief that artifacts are deﬁned centrally by the use intended by the designer (Bloom, 1996; Matan & Carey, 2001). Although some evidence indicates that by age 2 children may be sensitive to this kind of information (Casler & Kelemen, 2005), infants may not use such information to determine the functions of objects. Nonetheless, infants may have a primitive understanding of the intended use of objects. Even if they do not have access to information about the intention of the designer or the actors’ goals, infants may be able to differentiate intentional from accidental actions (Carpenter, Call, & Tomasello, 2005). In addition, infants almost certainly do have access to information about objects’ typical uses. Because the designer’s intended use is often highly confounded with how objects are typically used, infants’ recognition of the typical use of objects may bootstrap an eventual understanding of the intended use. Thus, there may be a rough developmental progression from function as ‘‘identifying actions on an object’’ or ‘‘recognizing action’’ to function as ‘‘understanding how an object is intended to be used.’’ This proposed developmental progression, however, highlights a clear gap in how the latter form of function can be derived from the former. This gap may be

How Infants Construe Functional Features

149

at least partially ﬁlled by considering the third notion of function: The reaction of a thing to an action upon it. This idea of function brings together the ﬁrst two deﬁnitions via an emerging understanding of the causal relations that join them. Moreover, it might provide the groundwork for an understanding of the conventional or intended use of an object. Although the remote control could be used for banging on the table, the fact that an interesting reaction occurs when the buttons are pressed might provide a basis for understanding that the buttons exist for just this purpose. Thus, we propose the following conceptualization of how attention to and representation of function emerges over infancy and early childhood. Because we argue that the perception of object function actually involves integrating several different aspects of events and contexts, we believe that there is no single point at which infants recognize function as a property of objects. Rather, across development infants come to recognize and represent new kinds of features of objects and events, and these features comprise their conception of function. Function, therefore, is not (by this view) a single, coherent property of objects, but rather function is an emergent property of objects, events, or situations given the presence of several component features. Obviously, this conceptualization makes it difﬁcult to ask questions such as ‘‘is function more or less salient with development’’ and ‘‘when do infants ﬁrst categorize on the basis of function.’’ However, we argue that by considering the shifting ways in which infants might construe object function, we can get a deeper understanding how their developmentally appropriate understanding of function is incorporated into their object representations.

B. INFANTS’ SENSITIVITY TO THE COMPONENTS OF FUNCTION An important step in understanding how infants conceptualize object function is to consider whether infants are sensitive to the components of function described earlier, as well as when infants integrate such pieces of information. Although studies have examined infants’ sensitivity to features of events that may contribute to their understanding of function, little work has been directed at addressing such questions in the context of infants’ understanding of object function itself. To further our understanding of infants’ developing conceptions of object function, in this section we review what we know from the literature about infants’ developing sensitivity to aspects of events that relate to object function. We use Barsalou and colleagues’ (Barsalou et al., 2005; Chaigneau et al., 2004) HIPE model

150

Lisa M. Oakes and Kelly L. Madole

as a framework for thinking about the primary aspects of function. In particular, we examine whether infants are sensitive to (1) the history of objects uses, (2) the intentions of the actor or designer, (3) the physical features that constrain objects’ uses, and (4) the event sequences in which the objects participate. Currently, our understanding of infants’ sensitivity to each of these aspects is incomplete—particularly as it relates to infants’ perception of function. However, by identifying gaps in the literature, we also identify important avenues for future research. 1.

2.

History of how objects have been used. Our knowledge about infants’ understanding of artifact history is limited. Certainly, infants learn about actions performed on objects from observing the history of their uses. They imitate actions they have seen performed on objects (e.g., Meltzoff, 1988) and they dishabituate when they see objects acted on in new ways (Perone & Oakes, 2006). But the nature of this behavior and its relation to infants’ underlying knowledge about objects is unclear. Infants’ imitative ability as well as their ability to recognize when an object is acted on in a novel way might simply reﬂect the ability to form associations between an action and some other property of the object. We do not know whether infants expect that how an object has been used historically is how the object should be used, or whether they expect that objects can be acted on in multiple ways. Moreover, for infants, the historical use of objects might be irrelevant to considering the object’s function. Indeed, several studies have shown that, with development, an object’s history or the creator’s intent becomes more important in children’s judgments about an object’s function (German & Johnson, 2002; Matan & Carey, 2001). Thus, although we know that infants’ attend to and recognize the history of an object’s use, we do not know how such information is integrated into their representation of the object’s function. Intended use by the designer or the actor. Infants may have a primitive understanding of the ‘‘use’’ or ‘‘utility’’ of an object. Perceiving an actor’s behavior as intentional toward an object (even if one does not know what those intentions are) or expecting that actions produce outcomes (even if one does not have expectations about what outcomes should be produced) can be thought of as the foundations of understanding the objects have uses. For example, Buresh and Woodward (2007) reported that infants recognize a single actor’s behavior (consistently reaching toward a particular object on several trials) and use this information to anticipate that actor’s future behavior, but not the future behavior of another actor. They

How Infants Construe Functional Features

3.

151

interpreted this ﬁnding as suggesting that by 13 months, infants restrict goals to speciﬁc actors. Other ﬁndings have been taken as evidence that even younger infants are sensitive to the intentions of an actor (e.g., Song, Baillargeon, & Fisher, 2005). When imitating the actions of an adult model, infants in the second year perform actions that produce the intended goal of an actor rather than exactly mimicking his or her precise actions (Carpenter et al., 2005; Gergely, Bekkering, & Kiraldy, 2002). Although this evidence may reﬂect sensitivity to the goals of the actor, how infants use such information in their conceptualization of function is uncertain. Certainly, performing an action that reproduces the agent’s goal (if not his or her actions) reveals that infants must have associated that goal with the object (otherwise, they would not have been able to reproduce the goal). However, this pattern may reﬂect infants’ selective attention to the result of the action, rather than to the details of the action itself. Infants do seem to have some of the skills required for at least a primitive understanding the intensions of how objects are used—perhaps an expectation for how a particular actor will behave based on past experience or perhaps selective attention to the end result of an action (as opposed to detailed understanding of the action itself ). What is unknown, however, is when infants integrate these expectations or perceptions into their conceptions of the functions of objects (e.g., it can illuminate, it can roll). It is one thing to be sensitive to actions as intentional; it is quite another to incorporate the knowledge into understanding the intentional use of an object. How physical structure constrains the use of an object. Clearly, infants detect and encode the physical structure of objects—they detect changes in features such as object shape, color, and pattern (see Kellman & Arterberry, 1998, for a review). However, we know little about how infants’ sensitivity to such features inﬂuences their understanding of object function. Indeed, researchers often assume that infants are insensitive to such constraints, presenting infants with arbitrary structure–function combinations. In many studies, objects with identical structures but that differ in some superﬁcial property (such as color) perform different functions. Thus, a nonstructural feature (in this case color) is predictive of function, and the structural constraints on the functions of the objects are the same. In such cases, structural information is not useful for determining the functions of the objects. For example, in Wilcox and Chapa’s (2004) study, the structure that afforded banging or pouring was identical in the two objects; the surface feature that predicted function was color.

152

4.

Lisa M. Oakes and Kelly L. Madole

Color is not a feature of objects that would constrain the ability of that object to bang or pour. Similarly, in our own studies, the same objects can be shaken, rolled, inverted, squeezed (Horst et al., 2005; Madole et al., 1993; Perone & Oakes, 2006). In our studies the objects have the structural support to allow any function—they have wheels, they can (presumably) have ratting objects inside, and so on. A sophisticated conceptualization of function, however, may require noting the correspondence between the structural feature that affords particular actions or uses and those actions or uses. Because humans can perform so many different actions on objects, acting on and recognizing actions on objects requires linking speciﬁc actions with speciﬁc objects. Forming such representations is likely difﬁcult because they involve the integration of information processed by both the dorsal and ventral streams (Buxbaum, Kyle, & Menon, 2005; Chao & Martin, 2000; Wolk, Coslett, & Glosser, 2005), two visual streams that in primates process different information. The ventral stream is used for recognizing and identifying objects, and the dorsal stream is used for recognizing and controlling actions (Goodale & Milner, 1992). Recognizing that particular objects can be acted on in speciﬁc ways, therefore, may require integration of information processed by these two different streams. We know very little about how infants represent these different kinds of information, or how they represent information processed by the two streams. However, at least in the context of short term or working memory, there has been considerable interest in this kind of integration, which is assumed to develop between 6 and 12 months of age (Kaldy & Leslie, 2005; Oakes, Ross-Sheehy, & Luck, 2006). Unlike the kinds of contexts in which infants’ attention to function is typically assessed, in short-term or working-memory tasks infants are given only brief exposures to stimuli and their ability to remember features (or combinations of features) over the very short term (sometimes less than 1 s) is assessed. Thus, although integration on this timescale certainly contributes to how infants encode features of complex objects and events, the kind of integration of information processed by the dorsal and ventral streams in such tasks may or may not be central to the kind of integration required for acting on and perceiving actions on objects. The event sequence. We perhaps know the most about the skills infants posses for representing this last component. Infants develop the ability to link actions and outcomes (the components of the event sequence)

How Infants Construe Functional Features

153

in the ﬁrst year. At 6 months, infants perceive causal relations of launching events (Leslie, 1984). Later in infancy, spatiotemporal cues help infants parse causal chains (Cohen et al., 1999). By 9 months, infants segment human action into meaningful components related to actions and goals (e.g., picking up a towel from the ﬂoor) (Baldwin et al., 2001). Thus, infants have the prerequisite ability to link action and outcomes when observing demonstrations of object function and as described previously, infants might respond differently to actions that produce an outcome than to those that do not (Wilcox & Chapa, 2004). Understanding how infants integrate actions and the outcomes is important because researchers often conﬂate these, assuming that a function is an action that produces an outcome, such as squeezing an object to make it squeak, inserting a part to make a sound, or swinging an object to make chimes ring (Booth, 2006; Horst et al., 2005; Trauble & Pauen, 2008). Barsalou and colleagues’ (Barsalou et al., 2005; Chaigneau et al., 2004) conception of the event sequence also reﬂects this assumption—in their model, the action and outcome are parts of a single component, the event sequence. However, actions and outcomes may in fact be distinct features of function. Speciﬁcally, the human adult brain represents action (the characteristic ways of moving an object, such as the circular motion used when operating the pencil sharpener) separately from the goal or the outcome of acting on an objects (e.g., using a pencil sharpener to sharpen a pencil) (e.g., Kellenbach et al., 2003). For example, Buxbaum and Saffran (2002) found that apraxic patients had deﬁcits in their knowledge of how objects were manipulated, but not their knowledge of the intended function (i.e., the consequence of that manipulation). In a PET study of normal adults, Kellenbach et al. (2003) found brain regions that responded speciﬁcally when identifying the actions performed on objects but not the functions of those objects. Moreover, action, and not function, is central to how people represent manipulable objects (e.g., Yoon & Humphreys, 2005). In summary, infants clearly develop sensitivity to many of the components involved in function, and thus they have the tools required for conceptualizing function, at least in a primitive way. In the following section, we review our work on infants’ sensitivity to object function. This work provides a foundation for understanding how infants’ conceptualize function, and how the components discussed previously inﬂuence infants’ responding to features adults would label as an object’s ‘‘function.’’

154

Lisa M. Oakes and Kelly L. Madole

IV. Our Research on Infants’ Attention to and Representation of Function Our work examining infants’ attention to, representation of, and categorization based on object function began almost two decades ago. We were motivated by an interest in infants’ use of ‘‘non-obvious’’ features in categorization (for reviews of these issues and our perspective, see Madole & Oakes, 1999; Oakes & Madole, 1999, 2000, 2003). When we designed our ﬁrst studies examining these issues in the late 1980s, much of the work described in previous sections had not yet been conducted. Subsequent work has reﬁned our thinking about function. Not surprisingly, therefore, in our early work function was intuitively, and often imprecisely, deﬁned and how function in one study is related to function in other studies is not always clear. However, despite these limitations, our program of research illustrates how we can begin to go beyond our initial question—when do infants categorize using function—and answer questions about how infants conceptualize object function. Speciﬁcally, we can ask how function is related to other properties, what kinds of functions infants detect and represent, and the relative importance of function versus other types of features when identifying and categorizing objects. Moreover, we believe this work reveals that function is similar to other properties that are often regarded as more perceptual. Function is not always the most salient feature; whether infants seem to treat function as more important than other features depends on the context. No single functional feature seems to have a special status; instead the particular feature or combination of features to which infants attend is determined by multiple factors including an infant’s age, how the stimuli are presented, and whether items are presented in isolation or in the context of other items. Function does not necessarily reﬂect the infants’ deep, theoretical understanding of the object or the connections among objects—any more than do features such as perceptual similarity, shape, and the like. Of course, infants’ developing understanding of the world—particularly of physical mechanisms of causality—will constrain their understanding of object function (as well as other properties such as shape). Thus, infants’ use of object function is not unrelated to their emerging conceptual understanding of objects, categories, and causality. However, early attention to object function may not be sufﬁcient evidence that infants have a deep, conceptual understanding of manipulable objects or artifacts. In line with this ‘‘constructivist’’ view, our research has focused on how infants’ attention to and understanding of function changes with increasing information-processing

How Infants Construe Functional Features

155

skills, motor abilities, ‘‘theories’’ (i.e., expectations for how objects are constructed), history with objects, and so on.

A. INFANTS’ SHIFTING ATTENTION TO FUNCTION AND APPEARANCE When we began our studies, a large body of research had established that infants perceive and represent the surface features of objects. Less was known about infants’ perception and representation of object function. Thus, our ﬁrst question was simply ‘‘Can infants learn both the surface features and functions of objects?’’ Initially, we assumed, on the basis of a general understanding of infants’ cognitive and perceptual development, that function would be more difﬁcult to perceive because of greater demands on the information-processing system. More challenging stimuli (e.g., images with more elements) take more time for infants to learn, and are more difﬁcult for younger infants to perceive and remember (Cohen, 1988). Surface features are relatively static and unchanging; the color, presence of particular parts, and overall shape, for example, are available most of the time that objects are visible, although some actions on objects may temporarily transform the object’s shape (e.g., squeezing) or obscure some features momentarily (e.g., grabbing an object). Functional features present greater demands on infants’ perceptual, attentional, and memory processes—for example, they involve an action that unfolds over time and an outcome that occurs only at speciﬁc points during the action. Thus, accurately perceiving those features requires tracking them over changes in space and time—perhaps recognizing commonalities across transformations, changes in lighting conditions, and occlusion. Selectively attending to relevant features, and ignoring irrelevant ones, may be particularly difﬁcult in contexts involving changes in luminance and abrupt onsets, features that effectively capture attention. Linking actions and outcomes requires detecting the spatial and temporal contingency between different parts of events and integrating that information. Thus, we predicted that, in general, infants would attend to surface features earlier in development than they would attend to functional features. Our ﬁrst study revealed exactly this developmental pattern (Madole et al., 1993). We deﬁned function as an action that can be performed on the object resulting in some outcome—shaking resulted in rattling and pushing resulted in the wheels rolling. We employed an object-examining task in which the actual objects are presented to the infant and the infant is allowed to

156

Lisa M. Oakes and Kelly L. Madole

manipulate and explore the object for a ﬁxed period of time. Each trial began with the experimenter demonstrating the function of the toy (i.e., either rolling or shaking it), and then the infant was given the toy to explore for 30 s. Thus, the function was only available to infants during the ﬁrst few seconds of the trial unless they reproduced that function themselves during that trial. The design of this experiment is presented in Figure 1. During an initial pretest, infants received trials with the two items to be presented during test. This pretest provided a baseline for infants’ preference for one object over the other. Next, infants were presented with one of those items on each of eight familiarization trials. Finally, the two test items were presented once again. We asked whether infants’ preference for the ‘‘novel’’ item (i.e., the item that was not presented during familiarization) changed from pretest to test. Infants were randomly assigned to one of the three conditions: In the appearance novel condition the two objects differed in appearance but had

Condition: Preset Phase (2 trials, 30 s in duration. 1 object presented on each trial)

Familiarization Phase (8 trials, 30 s in duration, 1 object presented on each trial)

Appearance Novel

rolls when pushed

rolls when pushed

Function Novel

rolls when rattles when shaken pushed

Appearance & Function Novel

rolls when rattles when shaken pushed

rolls when pushed rolls when pushed rolls when pushed rolls when pushed rolls when pushed rolls when pushed rolls when pushed rolls when pushed

Familiar

Novel

Familiar

Novel

Familiar

Novel

Test Phase (2 trials, 30 s in duration, 1 object presented on each trial) rolls when rolls when pushed pushed

rolls when rattles when shaken pushed

rolls when rattles when pushed shaken

Fig. 1. The stimuli and design of Madole et al. (1993), Experiment 1. An object-examining task was used. Infants were given objects one at a time on 30-trials. The session was divided into a pretest (2 trials with the novel and familiar items), familiarization (8 trials with the familiar item), test (2 trials with the novel and familiar items). Infants were tested in one of three conditions: an Appearance Novel condition in which the two items differed only in appearance, a Function Novel condition in which the two items differed only in appearance, and an Appearance and Function Novel condition in which the two items differed in both appearance and function.

How Infants Construe Functional Features

157

the same function (e.g., the tall yellow and red round objects both rolled when pushed), in the function novel condition the two objects had the same appearance but performed different functions (one rattled when shaken and the other’s wheels rolled when it was pushed), and in the appearance and function novel condition the two objects differed in both appearance and function (e.g., a tall yellow object that rolled when pushed and a round red object that rattled when shaken). The preference scores are in Figure 2. For the purposes of presentation here, novelty preference scores are shown: For both the pretest and test, we calculated infants’ preference for the novel item by dividing their attention (as measured by the duration of examining) to that item by their attention to both the novel and familiar items. Thus, a novelty preference score of .50 represents equal attention to both items, and a score above .50 represents greater attention to the novel item. All infants looked about equally to the two items during the pretest (the white bars). After familiarization (the black bars), in contrast, every group except one had novelty preference scores that were well above chance during test, and they clearly increased their novelty preference from pretest to test. The one group who failed to show this novelty preference was the 10-month-old infants who were shown two items that differed only in

Fig. 2. Novelty preference scores derived from the examining durations during pretest and test from Madole et al. (1993), Experiment 1 by condition. Novelty preference scores greater than chance (.50) indicate that infants preferred the novel item over the familiar item.

158

Lisa M. Oakes and Kelly L. Madole

function. These infants had novelty preference scores near chance both before and after familiarization. Familiarization trials with a rattling object, for example, did not induce a preference for rolling during test, if those two objects had the same appearance. We therefore concluded that younger infants were more attentive to the static appearance features, and older infants were attentive to both those features and the more dynamic functional features, consistent with our prediction that informationprocessing demands underlie infants’ selective attention to appearance or function (Madole & Oakes, 2005; Madole et al., 1993). However, subsequent research and theorizing leads to different predictions for how infants should allocate their attentional resources to functional versus surface features of objects. Speciﬁcally, low-level perceptual features of the stimuli associated with function may result in function having priority over appearance. For example, auditory features may be selectively attended over visual ones (Robinson & Sloutsky, 2004). Moving stimuli are often preferred over static stimuli (Shaddy & Colombo, 2004). Stimuli that appear abruptly may recruit attention more effectively than stimuli that are constantly available or that appear in other ways (Yantis & Jonides, 1990). Such results lead to the prediction that infants should prefer and selectively attend to dynamic, changing features that have an auditory component (such as function) over static, unchanging features that are solely visual (such as appearance). Indeed, increasing evidence shows such selective attention for relatively dynamic over relatively static features. Bahrick, Gogate, and Ruiz (2002), for example, found that 5½-month-old infants learned the actions performed by an actor (e.g., brushing her hair) but failed to remember the details of her face or the object she used in those actions. For these infants, therefore, the dynamic actions were more salient than the relatively static surface features of the actress’s face or the object she used. When presented with competing static and dynamic information, young infants apparently selectively attend to dynamic features, at least under these circumstances. We observed this pattern for infants’ attention to appearance versus function in a visual-habituation task in which, in contrast to our objectexamining task described earlier, infants are shown movies of events in which the function is demonstrated several times during the course of a trial and they simply sit and watch those movies. Using such a task, we examined infants’ attention to appearance and to function, deﬁned as an action performed on an object and a resulting outcome. The stimuli were always movies in which a hand acted on an object and the outcome of that action was a sound (see Figure 3). In each experiment, infants were familiarized with one event (e.g., a hand rolling a yellow object

How Infants Construe Functional Features

159

Fig. 3. Stimuli and design for Perone et al. (in press) and Horst et al. (2005). Infants were familiarized with a single movie in which a hand reaches for, grasps, and acts on an object, and a sound is produced. Infants then were tested with two new events, one that differs from the familiar in appearance and another that differs from the familiar in function. On each trial, infants’ looking time to the movie was assessed.

accompanied by clicking) and then were tested with a change in appearance (e.g., a hand rolling a purple object accompanied by clicking), a change in function (e.g., a hand pulling the top of the yellow object accompanied by whistling), and changes in both (e.g., a hand inverting a pink object accompanied by a mooing sound). Perone et al. (in press) observed that 6- to 7-month-old infants were more attentive to the function than the appearance in these events. The results of two experiments using the design just described are presented in Figure 4 (the 6- and 7-month-old infants in that ﬁgure). We have plotted infants’ increase in looking to tests that involve a change in appearance or a change in function, relative to their looking to a familiar item. Speciﬁcally, we subtracted infants’ looking during a trial with the familiar item from their

Lisa M. Oakes and Kelly L. Madole

160

18 16

Dishabituation Score (in s)

14 12 10 8 6 4 2 0 6 months

7 months

10 months

-2 Appearance Change

Function Change

Fig. 4. Infants’ dishabituation score (looking time to the novel item-looking time to the familiar item) for tests involving a change in appearance or a change in function for (A) 6-month-old infants observed by Perone et al. (in press), (B) 7-month-old infants observed by Perone et al. (in press), and (C) 10-month-old infants observed by Horst et al. (2005). Error bars represent 95% confidence intervals.

during each test trial with a novel item; if infants increase their interest to a novel stimulus, this dishabituation score will be greater than zero. In general, infants in this age range robustly responded when the function changed, but they responded only weakly (or not at all) when just appearance changed. Thus, the dynamic functional features were more salient than the static appearance features. Horst et al. (2005) tested 10-month-old infants in essentially the same experiment. The data are also presented in Figure 4. In contrast to the 6- to 7-month old infants, these older infants attended to and represented both function and appearance. They signiﬁcantly increased their looking to both a change in function and a change in appearance. These results, therefore, reveal the opposite developmental trajectory from that we observed in our previous study. Whereas Madole et al. (1993) reported that infants responded ﬁrst to static appearance features, the subsequent studies suggested that dynamic functional features have developmental priority over the static appearance features. Moreover, the 10-month-old infants observed by Madole et al. failed to attend to function, whereas the 10-month-old infants observed by Horst et al. (2005) robustly

How Infants Construe Functional Features

161

attended to function. Clearly, these two sets of studies revealed different patterns and suggest that function and appearance were differentially effective at recruiting infants’ attention in these two experimental contexts. Why did we observe that infants found appearance more salient than function when using object examining, but observed the opposite pattern when using a visual-habituation task? Object examining, in which infants actively manipulate and explore objects, may tap different cognitive processes than does visual habituation and familiarization, and different developmental trajectories reveal that the two types of processes develop independently. Indeed, this is the view argued by Mandler (2004). However, our ﬁndings are the opposite from what one would predict from Mandler’s reasoning. She argues that object examining elicits attention to more conceptual, non-obvious features, whereas visual tasks elicit responding to appearance features. In this case, attention to appearance should dominate in the visual task but not the object-examining task, and yet this is exactly the opposite of what we observed. We believe that the apparent discrepancies reﬂect the fact that function is an emergent feature as described in Section III. Contextual factors, the infant’s past experience, and characteristics of the different properties all contribute to which features are most salient at a given point in development. Recall that dynamic features (such as hand moving and grabbing an object or an intermittently heard sound) might be more salient than static features (such as the color of the object) due to low-level perceptual factors such as movement, ﬂicker, and abrupt onset. In some contexts, function will have these properties. In our visual tasks, infants see a 7 s event sequence in which a hand is in motion most of the time and the object changes position and may change shape momentarily. Thus, movement, ﬂicker, and abrupt onset characterize the action—and its associated outcome. Functional features in this task, therefore, may be highly salient. In object examining, in contrast, the function is only demonstrated at the start of the trial. For the majority of the trial, the infant is free to explore and manipulate the object, and reproduce the function if he or she chooses. Thus, in this context, the surface features may be more salient because the infant can look at the object from a wider variety of angles and the function may be only minimally available (because the infant does not reproduce it). Object-examining and visual-habituation tasks also differ in their general information-processing demands. We have argued elsewhere that infants may respond differently in visual tasks than in object-examining tasks because the former places fewer demands on the infants’ informationprocessing resources (Oakes & Madole, 2003; see also Younger & Furrer, 2003). In object examining, infants have access to much more

162

Lisa M. Oakes and Kelly L. Madole

information—both relevant and irrelevant. In the study by Madole et al. (1993), for example, infants could view the object from multiple perspectives, gaining information about what each side, the top, and bottom looked like. They had access to what effect shaking and rolling had on the object (if they chose to produce both actions). They had information about the object’s texture, weight, taste, and so on. Clearly this is much more information than that available in the studies by Perone et al. (in press) or Horst et al. (2005). In these studies, infants could see only the front of the object. On each trial they had information about the effect of only one action produced on the object (the one demonstrated by the agent). They did not have other views of the object, or information about the taste, texture, or weight of the object. Thus, the differences in attention to appearance and function may be related to differences in how much information is available to the infant and infants’ inability to selectively attend to relevant information. Indeed, as we discuss subsequently, Horst et al. observed that 10-month-old infants attended to function over appearance in a more demanding visual-habituation context. The effects of the information-processing demands and low-level visual and auditory characteristics of the stimuli may not be independent of one another. Our view of function is that such factors should together determine how infants perceive the events and the objects in them. Indeed, variations in the information-processing load likely alter the effect of low-level sensory characteristics of the stimuli (e.g., loudness, luminance, movement). For adults, distractors are less likely to capture attention when the central task is demanding (e.g., Lavie & Cox, 1997). Similar interactions have been observed for control of infants’ attention (Oakes, Tellinghuisen, & Tjebkes, 2000). Thus, to the extent that such processes are relevant for how infants deal with competing stimulus properties—such as appearance versus function—we predict that the effect of characteristics of function (e.g., ﬂicker, motion) would be determined, in part, by the informationprocessing demand (or ‘‘load’’) of the task in general. Thus, it may be misleading to suggest that, developmentally, infants always initially attend to function or always initially attend to the surface features. Which features are attended to earlier in development may depend, critically, on the context. In some contexts, low-level features may draw attention to some aspects of the event (such as the function). Do the varying results described here suggest that it is hopeless to try to derive some underlying principles about the development of attention to features like appearance and function? We are not this pessimistic. Despite the fact that we observed different developmental trajectories in the two tasks (and with different sets of stimuli), we can still attempt to understand the mechanisms of developmental change in infants’ attention to such

How Infants Construe Functional Features

163

features. In fact, the two developmental trajectories we have observed are not very different at the broadest level. In both contexts, we observed that younger infants attended to and remembered a smaller subset of the information available than did older infants. One explanation for this ﬁnding is that because of more limited abilities to encode information in visual short-term memory (Ross-Sheehy, Oakes, & Luck, 2003) and control visual attention (Oakes, Kannass, & Shaddy, 2002), younger infants attended to only some of the features present and focused their attention and memory processing on those features. As infants become able to hold more information (perhaps for longer) in short-term memory and more effectively control their attention, they apparently become able to attend to and encode more features and a wider variety of features. The difference between the two experimental contexts is whether the youngest infants attended to appearance or function. Of course, at this point an informationprocessing explanation for these changes is speculative, but consistent with what we know about changes in infants’ encoding abilities, attention, and memory. Additional studies are needed to conﬁrm whether developmental changes in infants’ attention to appearance and function are, in fact, the result of more changes in such information-processing abilities. We can therefore begin to ask what factors inﬂuence infants’ ability to attend to more features in one or the other context. It has become increasingly apparent that changes in motor development inﬂuence infants’ perception of the surface features of objects in visual task. Needham (2000), for example, observed that 4-month-old infants who were more active during object exploration also were more sensitive to the surface features that deﬁned object boundaries in a visual task. This relation is not surprising given the signiﬁcance of object boundaries for effectively reaching for and grasping objects. Such ﬁndings do not show that object exploration induces changes in visual perception, but they do show that perception and action are related, and provide the background for asking how changes in infants’ ability to act on objects might bring about changes in their perception of appearance and function. Therefore, infants’ ability to attend to and encode object appearance in addition to object function in visual habituation is likely inﬂuenced by changes in motor skill. In a study with 6-month-old infants, Perone et al. (in press) found that not only were infants more responsive to changes in function than to changes in appearance, but also that infants’ emerging reaching and grasping skills were related to their responding to a change in appearance. In this experiment, we tested infants in an experiment with the design in Figure 3, and we assessed their skill at reaching for and retrieving objects (see Figure 5). We found that attention to a change in appearance, but not to a change in function, was related to infants’ skillful grasping and

164

Lisa M. Oakes and Kelly L. Madole

Fig. 5. An infant in the object exploration task used by Perone et al. (in press). Infants were seated with their caregiver; four toys were arranged around their feet.

object exploration in this task. In particular, infants who successfully grasped toys more frequently tended to look longer when appearance changed at test. Moreover, this relation held even when we controlled for infants’ sitting ability, age, and their general interest in novelty. Obviously, this work is a ﬁrst step in understanding how infants’ attention to the features of the objects in this context develops. But, this step is critically important because it suggests a mechanism (the emergence of new motor skills) that may account for developmental changes in attention to appearance and function. Clearly there are intricate and complex relations between how infants integrate information about action and the surface features of objects. The results of the studies described in this section underscore the fact that questions such as ‘‘when do infants perceive function?’’ or ‘‘is function or appearance more salient?’’ are not straightforward. Observing discrepancies like that described here led to our rethinking of function. Clearly, from our ﬁndings, function does not seem to be a single feature of objects that can be

How Infants Construe Functional Features

165

manipulated equivalently in different contexts—it is not simply what an object does or the reaction of an object to some action. Whether infants selectively attended to function or appearance seemed to be complexly determined by how those features were presented, the salience of different aspects of the features, and the information-processing demands of the task.

B. INFANTS’ DEVELOPING ATTENTION TO APPEARANCE AND FUNCTION It is important to keep in mind, however, that even though young infants may represent appearance and function separately, they are not independent. One explanation for Perone et al.’s (in press) ﬁnding that infants with more advanced motor skills were more attentive to object appearance is that these infants have begun the process of learning that appearance and function are related. Structural properties of objects are a component of function. Adults know that the structural properties of objects constrain the functions—or afford the actions that can be performed. Objects with wheels can be rolled. Compressible objects can be squeezed. Objects with beads (or rice or beans) in them rattle when shaken. Moreover, for adults, the visual perception of manipulable artifacts (such as tools) engages brain areas thought to be important in the representation of action on those objects (Martin, 2007). Thus, representing appearance and function together is an important step in developing a mature, sophisticated conception of function. Integrating appearance and function, however, is critically related to integrating what information with how information—or information thought to be processed by the ventral and dorsal visual streams (Goodale & Milner, 1992). Integrating this kind of information—or even attending to both kinds of information simultaneously—may be particularly difﬁcult for infants (Mareschal & Johnson, 2003). Thus, it would not be surprising to observe a protracted development of infants’ integrated representation of function and appearance. One of our experiments indeed revealed a relatively late emerging recognition of the relation between object appearance and object function (Madole et al., 1993). We familiarized 10-, 14-, and 18-month-old infants with two objects in an object-examining task. The objects had different appearances and functions—for example, a tall, yellow object that rolled when pushed and a red round object that rattled when shaken. Thus, function, or the object-action correspondence, was correlated with appearance. We asked whether infants learned the combination of a

Lisa M. Oakes and Kelly L. Madole

166

particular function and a particular appearance—for example, that yellow objects roll and red objects rattle. If so, during test infants should prefer a red object that rolled to a red object that rattled. Indeed, this is exactly what we observed in 18-month-old infants (see Figure 6). At this age, infants apparently could associate appearance with function (e.g., red objects rattle when you shake them) in object examining. Fourteen-month-old infants, in contrast, did not associate the appearance and function. They did not respond to a novel combination of the appearance and function, responding instead to novel individual features. They learned that objects could be red or yellow and could shake or roll, but not that red objects rolled and yellow objects rattled. Thus, attention to the combination of function and appearance—at least in object examining—emerges relatively late in infancy. This understanding likely does not emerge full-blown at this age. In fact, the developmental achievements necessary for understanding the relation between appearance and function almost certainly begin to emerge much earlier, as infants begin exploring objects. Skillful object exploration requires (and may even

20 18

Examining time (in s)

16 14 12 10 8 6 4

14 months

10 months Correlated

Uncorrelated

18 months Novel

Fig. 6. Examining times (in seconds) to the correlated, uncorrelated, and completely novel tests by age for Madole et al. (1993), Experiment 2. The correlated item was one of the two items presented during familiarization. The uncorrelated item was a new combination of the familiar function and appearance. The novel item had a new appearance and function. Error bars represent 71 SE.

How Infants Construe Functional Features

167

induce) the recognition of a connection between the surface features of objects and how the object can be acted upon. With development, infants use visual information to adjust their manual actions on objects (Corbetta, Thelen, & Johnson, 2000). They adjust their exploratory activity in response to the material and texture of objects (Bourgeois et al., 2005). Note however, that the appearance–function relation in the objects used here was arbitrary. Nothing about the structure of the objects immediately revealed their functions. Thus, there may be a delay before infants apply their knowledge about the relation between appearance and function to objects in which that relation is less transparent. The lack of transparency however does not mean these relations lack ecological validity: Many real objects are characterized by a similar arbitrariness in terms of the relation between structure and function. One might have to learn, for example, that the silver key starts the car, but the gold key opens the ofﬁce, because the actual mechanism underlying this function is unobservable. An important achievement, then, is learning which appearance-based features are most predictive of function.

C. INFANTS’ ATTENTION TO THE RELATION BETWEEN APPEARANCE, ACTION, AND OUTCOME Through the course of conducting this research, we realized that neither appearance nor function is a simple, unitary feature of objects. As our understanding of function has become more reﬁned, our approach to considering the relation between object appearance and object function has also become more reﬁned. In fact, the appearance of an object as a whole might be considerably less important than the appearance of relevant parts of that object. In a set of experiments that further clariﬁed the nature of infants’ attention to the association between appearance and function, Madole and Cohen (1995) asked whether 14- and 18-month-old infants’ could associate the function with the appearance of only a part of the object, and whether some associations are more easily learned than others. Using a visual-habituation procedure, we familiarized infants with two events in which the appearances of an object’s parts were correlated with different functions. In one condition, the appearance of the part predicted whether or not it was functional—for example, small, black wheels rolled but large, red wheels did not roll. Both 14- and 18-month-old infants dishabituated to a new object in which the previously non-functional part now worked (see the meaningful panel in Figure 7). In the example just given, the large red wheels now rolled. In another condition the appearance of one part was associated with whether or not the other part was

Lisa M. Oakes and Kelly L. Madole

168

30

25

Looking Time (in s)

20

15

10

5

0 14 months 18 months Meaningful Correlated

14 months

18 months Arbitary

Uncorrelated

Novel

Fig. 7. Looking times (in seconds) to the correlated, uncorrelated, and novel tests for 14- and 18-month-old infants habituated to the arbitrary and meaningful correlations in Madole and Cohen (1995). In the arbitrary condition, the appearance of one part was correlated with the function of a different part. In the meaningful condition, the appearance of a part was correlated with the function of that part. Error bars represent 71 SE.

functional. So, for example, if the protrusion on the top of the object was a green tree, the wheels rolled—regardless of the appearance of those wheels. Interestingly, although 14-month-old infants also learned this relation, 18-month-old infants did not (see the arbitrary panel in Figure 7). Only the 14-month-old infants dishabituated to the uncorrelated event that violated the association presented in habituation. Thus, they learned that the appearance of one part was predictive of whether another part functioned. What had developed between 14 and 18 months? During this time, at least for these kinds of objects, infants apparently become increasingly sensitive to the way in which appearance and function are actually related in real objects. Their learning of new associations is constrained by their developing theories of how objects are constructed or work, their understanding of action at a distance (i.e., that causes and outcomes are usually spatially contiguous), the extraction of statistical regularities they have observed about objects, or some combination of these three. Thus, development proceeds beyond simply associating functions or actions with

How Infants Construe Functional Features

169

object appearance to understanding how function or actions and appearances are related. But how do infants’ conceptualize function? The studies described up to this point laid the groundwork for demonstrating that infants are sensitive to the object property that adults would label ‘‘function,’’ and that they correlate such function with appearance. Now, if function is a complex, multiply determined property, it is important to identify exactly which of the features that might comprise function are detected and represented, as well how those features are combined. When infants represented function in our studies, did they represent a single, coherent feature—such as ‘‘squeaks when squeezed?’’ Our results can be equally well explained by infants’ attention to a single component—for example, the same patterns would have emerged if infants attended to just the outcome. Even if infants do learn several components relevant to function, a full understanding of how infants conceive of function requires that we establish how they combine those components. Subsequently, we examined infants’ attention to the actions and outcomes relevant to function. Obviously, infants may be sensitive to other components of function such as the intentions of an actor or designer, how the object has been historically used, or how the object’s structure constrains the function. However, central to any demonstration of function is an action on an object and some reaction (in our case the object apparently emitting a sound). Perhaps before one can detect intentional actions, learn patterns of how objects are used, and so on, one must be able to represent actions and outcomes, and link these features to other components. In other words, action and outcome may be components of the most primitive conceptions of object function. Indeed, according to Nelson (1974) infants’ ﬁrst concepts are focused on objects that can be acted on and that engage in dynamic changes (such as rolling). To test whether 10-month-old infants responded to changes in both action and sound, Perone and Oakes (2006) habituated infants to a single event and then tested their responding when just the sound or the action changed. Using essentially the same design as depicted in Figure 3, we habituated infants to a single event and then tested them with events that differed from the familiarization event only in sound or only in action. For example, infants habituated to the object that clicked when it rolled, might receive tests with an object that mooed when it rolled and an object that clicked when it was squeezed. Thus, if infants attended to and represented both the action and the sound they would dishabituate to both of these tests. As shown in Figure 8, infants did indeed respond to both changes. We replicated this ﬁnding in a second experiment when infants were habituated to four different object appearances, but a constant sound and action.

Lisa M. Oakes and Kelly L. Madole

170

22

10-month-old Infants

7-month-old infants

Dishabituation Score (in s)

17

12

7

2 One Object

Four Objects

Sound Alone

Action Alone

Action Change

One-dimension

Two-dimensions

-3 Sound Change

Fig. 8. Dishabituation scores (in seconds) to (A) a change in sound and action by 10-month-old infants in Perone and Oakes (2006), who were habituated to a single object or four different objects, and (B) a change in sound and a change in action by 7-month-old infants in Perone & Oakes (2008). Error bars represent 95% confidence intervals.

Again, 10-month-old infants dishabituated if just the sound changed or if just the action changed. Together, these results show that 10-month-old infants attend to and represent both of these components of function. We also examined 7-month-old infants’ responding to these changes using a different experimental design (Perone & Oakes, 2008). In these experiments, infants were again habituated to a single event and then tested with two new events. However, whether infants responded to a change in action or sound was tested between-subjects. Infants received a test event that differed in only one property (e.g., sound) and a second test event that differed in two properties (e.g., action and appearance). Comparing these two tests allowed us to test the alternative possibility that infants at this age generally ﬁnd multiple changes more compelling than fewer changes. One group of infants received a test event that had only a change in sound and one that had changes in both action and appearance; the other group received a test event that had only a change in action and one that had changes in both sound and appearance. Seven-month-old infants dishabituated when just the sound or just the action changed, as well as when sound or action changed with appearance (see Figure 8). Thus, 7-month-old

How Infants Construe Functional Features

171

infants responded to changes in either feature, and did not respond more when two features changed. Together, all these results show that both 7and 10-month-old infants represented action and sound. At least by 7 months, therefore, infants learn about two critical components of function—actions and outcomes. Of course, these features in isolation are not functions—as described earlier, action without an outcome is not considered sufﬁcient to be function (Wilcox & Chapa, 2004), and for older children and adults functions are, in part, actions constrained by the physical structure of objects (Kemler Nelson et al., 2000a). Thus, in the context of function, components such as action and outcome are important in combination with other components that help to deﬁne the relation between them. It is not immediately obvious how infants might combine the various components of function. Researchers’ intuitive deﬁnitions of function suggest that binding the action and outcome is critical. Functional demonstrations or objects are those in which an action produces an outcome (Welder & Graham, 2001; Wilcox & Chapa, 2004), and functions are made more salient by increasing the intensity of their outcomes (Booth, 2006). Clearly for researchers, then, actions and outcomes are bound. Thus, an important achievement for infants would be the development of this same intuition about function—that is, that actions produce a particular outcome, and function is acting in a particular way in order to produce a speciﬁc outcome—and they should represent the association between actions and their outcomes. Indeed, by 7 months—and certainly by 10 months—infants have many of the skills necessary to link the action and the outcome. Our events are characterized by a clear contingency and evidence of a causal mechanism. The outcome is spatiotemporally contiguous with the action. The hand makes contact with the object and as it acts the sound is heard. Infants are sensitive to exactly these cues, at least for physical causality—that is, when one object makes contact with another object, launching it into motion (Leslie, 1984; Oakes, 1994; Oakes & Cohen, 1990). Thus, infants have the skills required to bind action and outcome. Moreover, such causal connections may be particularly compelling, and as a result infants ﬁnd the binding of action and outcome salient. As described earlier, however, the human adult brain represents separately actions on objects and the results or goals of those actions (e.g., Kellenbach et al., 2003), and action, not function, appears to be central to how people represent manipulable objects (e.g., Yoon & Humphreys, 2005). Despite researchers’ intuitions, therefore, the binding of action and sound may not be the most salient to infants. Instead, actions may be more central to infants’ object representations—at least when

172

Lisa M. Oakes and Kelly L. Madole

encoding and remembering manipulable objects. Indeed, Nelson (1972, 1973, 1974) argued that children’s earliest concepts are based on similarities in the actions that they can perform on the objects (e.g., all balls can be rolled), and children must learn the features of objects that predict those actions (e.g., round shape predicts the ability to roll). Therefore, the most salient binding may be between object appearance (e.g., color, shape) and the actions that can be performed on objects. Madole and Cohen (1995) found that 14-month-old infants form such associations, although they are not constrained to link the appearances of speciﬁc parts with the functions of those parts—suggesting that they do not have an adult understanding of how the physical structure allows or inhibits some actions. Linking actions and appearances may be difﬁcult, however, because such links involve integrating information processed by both the ventral and dorsal streams (Buxbaum et al., 2005; Chao & Martin, 2000; Wolk et al., 2005). Mareschal and Johnson (2003) reported evidence suggesting that at 4 months dorsal and ventral processing are quite distinct. In short term or working memory, integration of such information occurs between 6 and 8 months (Kaldy & Leslie, 2005; Oakes et al., 2006). When object function is demonstrated, actions occur brieﬂy and intermittently. In the studies described here, the resulting sound also occurred brieﬂy and intermittently. Limitations in short-term or working-memory ability may contribute to infants younger than 8 months having difﬁculty linking such features with appearance. Moreover, recall that appearance is less salient for young infants than are the dynamic features of sound and action. Thus, despite the importance of the links between action and appearance, sensitivity to these combinations of components may not emerge until relatively late in infancy. Finally, the combination of the outcome and the appearance may be the most salient. In the real world, the particular outcomes produced by acting on objects are determined, in part, by the structure of those objects. Objects with wheels can roll; objects without wheels cannot roll. Infants’ history with different types of objects, therefore, may cause them to attend to such combinations. Although, as just mentioned, because appearance is generally less salient than the outcomes used here, these combinations may not be the ﬁrst learned. Thus, there are reasons to suspect that infants may attend to any one of the possible associations between object appearance, actions on objects, and outcomes of actions. Perone and Oakes (2006) tested among these three possible patterns in 10-month-old infants by familiarizing infants with two different events like those depicted in Figure 3. One feature was the same in both events—for example, the objects had the same appearance in both events, or the same action was performed in both events. The other two

How Infants Construe Functional Features

173

features were associated and were different in the two events. If the two objects looked the same, then the action and sound were associated (e.g., infants might see a purple object that squeaked when it was squeezed on half of the trials and that clicked when it was rolled on the other trials). In this way, infants could learn either the four independent features— rolling, squeezing, squeaking, and clicking—or two sets of bound features—rolling/clicking and squeezing/squeaking. Infants then were tested with a switched or uncorrelated event—an event that had all the familiar features, but the features were recombined. For example, squeezing would produce clicking. We tested 10-month-old infants in one of the three conditions—one that tested their learning of the association between sound and action, one that tested their learning of the association between appearance and action, or one that tested their learning of the association between appearance and sound. As can be seen in Figure 9, only infants familiarized with events in which the action and appearance were associated (e.g., purple objects are squeezed, yellow objects are rolled) looked longer to the uncorrelated (or switched) event than to the correlated (or familiar) event. The other infants all looked equivalently to the correlated event and the uncorrelated event. Thus, 10-month-old infants only learned the association between action and appearance. Apparently at this age infants do not perceive function as a single feature unifying action and outcome. Putting these results in the framework of Barsalou and colleagues’ (Barsalou et al., 2005; Chaigneau et al., 2004) model, infants at this age appear most sensitive to the P aspect of the HIPE model of function—they learned how the physical structure was associated with the actions performed on the object. Although in our stimuli this association was arbitrary (all of the objects could be rolled, squeezed, inverted) these 10-month-old infants may have already formed a bias to attend to this association from their experience in real life. That is, they had learned that the shape, size, and other features of objects determine what actions can be performed on them. In Gibson’s (1988) terms, appearance is related to the affordances of the objects. Thus, we have begun to build an understanding of infants’ perception and representation of object function as multiply determined by the factors described earlier. Although our work has uncovered some of the ways infants combine the information available in these events, several important questions remain. First, the effect of the outcome on infants’ learning of function is unclear. Recall that actions without outcomes, or with less salient outcomes, are not processed the same as are actions with clear, salient outcomes (Booth, 2006; Wilcox & Chapa, 2004). However, we found 10-month-old infants fail to learn the association between the action and

Lisa M. Oakes and Kelly L. Madole

174

30

Looking Time (in s)

25

20

15

10

5

0 Sound/Action Combined Familiar

Sound/Appearance Combined Uncorrelated

Action/ Appearance Combined Novel

Fig. 9. Looking time to the familiar, uncorrelated, and novel events by 10-month-old infants in Perone and Oakes (2006), by condition. In the sound/action combined condition, infants were familiarized with two events that instantiated a constant relation between the sound and the action (e.g., squeezing produced squeaking). In the sound/appearance combined condition, infants were familiarized with two events that instantiated a constant relation between sound and appearance (e.g., purple objects squeaked). In the action/appearance combined condition, infants were familiarized with two events that instantiated a constant relation between action and appearance (e.g., purple objects are squeezed). Error bars represent 71 SE.

the outcome, instead focusing on the action and appearance. Despite the fact that infants fail to learn that particular outcomes are associated with particular actions, it may be critically important that the actions actually produced outcomes. That is, the effect of the action may have facilitated infants’ learning of the association between the action performed on the object and the appearance of that object. For example, although models of category structure such as the causal status hypothesis (Ahn & Luhmann, 2005) argue for the priority of causes in categorization, properties may not be perceived as causes unless they are linked with an outcome. Thus, although infants do not link the action with the outcome, the fact that the actions produce outcomes may be key to their learning of object function. In addition, questions remain about how attention to combinations of features develops. Madole and Cohen (1995) observed that younger infants

How Infants Construe Functional Features

175

responded to a broader range of associations among features than did older infants. If this pattern is general to different aspects of linking functional features, we may observe that younger infants detect and learn a broader range of feature bindings—for example, the binding between sound and action as well as the binding between action and appearance—and that only with experience with the real objects do they restrict their attention to only the action/appearance bindings. Indeed, unlike older children, our 10-month-old infants were sensitive to any structure–action combination (not just those in which the physical structure constrains what one can do with an object). Thus, even for their attention to the appearance/action links, infants may initially be sensitive to a broad range of associations— even those that adults ﬁnd arbitrary—and only gradually come to recognize only the plausible associations, much like pattern observed by Madole and Cohen. Alternatively, infants’ developing understanding of the links between the components of function may take a different path. They may become increasingly sensitive to the associations among the features in these events and the action/appearance association is simply the ﬁrst one learned. Older infants may show sensitivity to other relations that younger infants do not—such as the association between particular actions and particular outcomes. Only by testing older and younger infants can we answer these questions. In summary, infants clearly do not perceive the functions used here as single, uniﬁed features of objects. Rather, they attend to the components and combine those components in non-obvious ways. Although we are only just beginning to understand the development of infants’ attention to and integration of the components of object function, the work we have discussed here is enlightening. This work will serve as the foundation of future work examining how infants’ perception and representation of function changes with increased sensitivity to the components of function and their ability to integrate and coordinate these difference pieces of information.

D. OBJECT FUNCTION IN INFANTS’ CATEGORIZATION Note that we have wandered quite far from the original focus of the study of function in the developmental literature. Nelson (1973, 1974) was concerned with children’s conceptual development as it was related to language acquisition. She was interested in understanding which concepts were ﬁrst learned and labeled. Her empirical work examined young children’s vocabularies to establish the foundation of their ﬁrst words.

176

Lisa M. Oakes and Kelly L. Madole

Mervis (1985) in her classic study used a functional deﬁnition for a nonverbal category. She argued that examining the assumptions infants make about the function of an object (from the actions they attempt to perform on that object) provided insight into the development of categorical knowledge. None of the work we have described addresses these questions. Instead, our work has primarily focused on how infants identify and represent individual objects or learn the correlations or associations among features when familiarized with only two objects. These are critically important questions and inform us about infants’ conceptual development. However, we also need to establish how infants use these features to form categories. Because we have focused on how function is construed, we have spent less time understanding how infants use commonalities in function as compared to appearance to group objects or events. Obviously, a complete understanding of how function forms the basis of a category of objects depends on a fuller understanding of how infants conceptualize function itself. However, we have begun to examine how infants use commonalities in appearance and function to form categories. Horst et al. (2005) habituated 10-month-old infants with a series of events like those in Figure 3. Unlike the previous experiments described here, infants were familiarized with four different events that depicted a category of objects. For half of the infants, the objects all had the same appearance—for example, a purple, round object—but each object had a different function. On one trial, for example, the hand squeezed the object and it squeaked and on a different trial the hand inverted the object and it mooed. Infants in this condition saw the hand engaging in four different actions toward objects with the same appearance (presented one at a time on different trials), and each action was associated with a unique outcome (sound). In this context, infants could learn the single common appearance, the four different functions, or both. For the other infants, the common feature across trials was the function. Infants saw objects with four different appearances—the purple object, a multi-colored pyramid shaped object, and so on—but the same function was demonstrated on each object (e.g., squeezing that produced a squeaking sound). In this context, infants could learn the single common function, four different appearances, or both. Infants’ responding during test revealed that when familiarized with four different items—whether those items shared a common function or a common appearance—the function was the most salient feature (see Figure 10). Infants who were familiarized with four different appearances and one function dishabituated when shown a new event involving a familiar

How Infants Construe Functional Features

177

35 30

Looking time (in s)

25 20 15 10 5 0 Appearance Constant Completely Familiar

Familiar Appearance/Novel Function

Function Constant Novel Appearance/Familiar Function

Fig. 10. Looking time (in seconds) to the novel appearance and novel function test by 10-monthold infants in Horst et al. (2005), who were habituated to four items that had the same appearance but different functions (Appearance Constant condition) or four items that had the same function but different appearances (Function Constant condition). Error bars represent 71 SE.

appearance and a new function. Infants who were familiarized with four different functions and a single appearance dishabituated when shown a new event involving a familiar appearance and a new function. Of course, we do not know whether infants perceive and conceptualize function in the same way in this context as they do when they are habituated to only one or two items. The representations we uncovered when we presented infants with only one or two events during familiarization may not generalize to the increased information-processing demands of seeing four different events during familiarization. For example, although 10-month-old infants have little difﬁculty attending to and encoding appearance when familiarized with a single event, they do not attend to and encode appearance as readily when familiarized with four different events. Therefore, in this categorization context, infants might detect other types of combinations of features and the salience of sound or action change. Obviously, our future research in this area will be informed by our ﬁndings on how infants perceive and conceptualize function more generally.

178

Lisa M. Oakes and Kelly L. Madole

V. Conclusions We have studied object function for nearly 20 years. During this time our thinking has gradually shifted from considering function as a single, coherent feature that can be manipulated (or treated as a single property of an object) to considering function as multiply determined and emerging from the conﬂuence of many different factors. This shift is inspired, in part, by the movement in cognitive science away from thinking about concepts as static and symbolic, to the recognition of concepts as dynamic and embodied. Research on how infants conceptualize the particular function of an object provides an excellent context for exploring this new view of concepts. Infants’ understanding of function in a given context may involve their recognition of the ways the object has been used, the intentions of the actor, the physical constraints of the object, and the relation between the action and an outcome, similar to Barsalou and colleagues’ (Barsalou et al., 2005; Chaigneau et al., 2004) view of adults’ understanding of function. This view of infants’ developing understanding of object function has much in common with contemporary dynamic systems views of cognitive development (Smith, 2005, forthcoming; Smith & Thelen, 2003). As in that approach, we argue that the conceptualization of the function of the object is created in the moment through the interactions of multiple factors. Consider the function of a spoon for an infant. At one point in time, the function of a spoon is for mouthing, for example, if the infant has sore gums, the spoon is cool to the touch, is sufﬁciently round to provide a comforting surface, and the infant can bring the spoon to her mouth. At another point in time, the spoon’s function is to make noise with, if the infant is seated at a hard surface, the spoon is rigid and has good resonance qualities, and the infant has the motor coordination to bang the spoon. At still another moment the spoon’s function is to transfer food to one’s mouth, if the child has developed the ﬁne motor skills to grasp the spoon properly, is interested in the available food, and the spoon’s structure can provide the support needed. Thus, the spoon does not have a single, unchanging function that is part of the individual’s stable concept for ‘‘spoon.’’ Rather, the concept and the function of the spoon is emergent given these contextual factors, and these concepts and functions might then inﬂuence the actor’s subsequent identiﬁcation and categorization of the object. We do not mean to suggest that the spoon has no conventional function, or that the designer’s intention is irrelevant. We do propose, however, that those factors work along with other contextual and developmental factors to create the object’s function at any given moment. The work we have presented illustrates how we developed this perspective. Our view of function derived from the traditional views of

How Infants Construe Functional Features

179

function described in the ﬁrst section. These views have been useful for understanding the signiﬁcant role object function plays in infant cognitive processing. However, further understanding of function requires a different approach. By examining the component aspects of function—and how infants represent both those component features and the links between them—we provide a foundation for deeper understanding of the role function plays in infants’ understanding of the world, for example, by disentangling their representations of the actions performed on objects and their outcomes. Our thinking about function also has been shaped by our perspective on conceptual development in infancy more generally. Function has been promoted as a non-obvious or deep-level conceptual property that members of a category share (Keil, 1989). Indeed, our initial work in this area was motivated, in part, by a desire to understand how infants integrate nonobvious features into their categorization of objects. However, the conceptual ranking of function has been elevated to the ‘‘essential’’ status for artifact categories (Bloom, 1996). Infants’ functional actions on objects—for example, making animals drink from cups—have been used as evidence that their grouping of objects is not merely perceptual, but contains conceptual information (Mandler & McDonough, 1998). We propose, in contrast, that infants’ representation of function is not qualitatively different from their representation of other, more obvious features. Perceiving and representing function involves attending to and combining the components described earlier. Thus, although accurately representing function aids in making inferences about features of objects that are not immediately visible, recognition of such features may not involve different processes from those that are used when recognizing other types of features of objects. The concept of object function is fuzzy, and is not easily characterized by a set of necessary and sufﬁcient features. We hope that research programs like ours will provide a deeper understanding into how the human mind conceptualizes the collection of properties that together comprise function. Moreover, by studying the development of this conceptualization we gain understanding not only of how to think about function in general, but also how infants’ understanding of object function changes. An adult conception of the function of an object may incorporate an understanding of the intentions of the actor or object’s creator, the causal relation between the action and the outcome, the affordance of the surface features of the object to allow the action, and the mechanism hidden in the object that produces the outcome. The infant conception of the function of an object, in contrast, may incorporate only some of these features. An unresolved issue, of course, is whether such a primitive understanding of function actually

180

Lisa M. Oakes and Kelly L. Madole

constitutes object function (although this may be a philosophical rather than a psychological question). Although some might interpret our perspective as begging the question as to whether infants actually represent object function, we think the study of infant conceptual development should move beyond this question. Instead of trying to identify the point at which infants have an adult conception of function, it will be more fruitful to explore the features infants do represent, and the broad range of developmental changes that allow infants to represent new features of objects in the world around them. Speciﬁcally, although clearly beyond the scope of this paper, we believe that the components of functions (and their associations) are learned by detecting regularities in how objects are structured, how people interact with objects, how objects respond to such interactions, and so on. Indeed, results like those reported by Madole and Cohen (1995) show the powerful inﬂuence of such statistical regularities on infants’ changing conceptualization of function. Our future work will be aimed at a deeper understanding of how such a mechanism can induce changes in infants’ perception of, attention to, and encoding of function. In summary, we have attempted to bring clarity to a feature of objects that is often not clearly deﬁned, and we are making progress toward understanding how infants think about a feature that is fundamentally important to how a large class of objects is deﬁned. By examining infants’ changing understanding of object function, we gain a deeper and more general insight into infants’ conceptual development.

Acknowledgement Preparation of this chapter and much of the work described was made possible by grants HD49840, HD49143, and MH64020 awarded to LMO. We wish to thank Sammy Perone for extremely helpful comments on this chapter, and for insightful and lively discussions of these issues.

REFERENCES Ahn, W., & Luhmann, C. C. (2005). Demystifying theory-based categorization. In L. Gershkoff-Stowe & D. H. Rakison (Eds.), Building object categories in developmental time. Carnegie Mellon symposium on cognition (pp. 277–300). Mahwah, NJ: Lawrence Erlbaum Associates. Bahrick, L. E., Gogate, L. J., & Ruiz, I. (2002). Attention and memory for faces and actions in infancy: The salience of actions over faces in dynamic events. Child Development, 73, 1629–1643.

How Infants Construe Functional Features

181

Baldwin, D. A., Baird, J. A., Saylor, M. M., & Clark, M. A. (2001). Infants parse dynamic action. Child Development, 72, 708–717. Barsalou, L. W., Simmons, W. K., Barbey, A. K., & Wilson, C. D. (2003). Grounding conceptual knowledge in modality-speciﬁc systems. Trends in Cognitive Sciences, 7, 84–91. Barsalou, L. W., Sloman, S., & Chaigneau, S. (2005). The HIPE theory of function. In L. Carlson & E. van der Zee (Eds.), Functional features in language and space: Insights from perception, categorization, and development (pp. 131–148). Oxford: Oxford. Bloom, P. (1996). Intention, history, and artifact concepts. Cognition, 60, 1–29. Booth, A. E. (2000). The facilitative effect of agent-produced motions on categorization in infancy. Infant Behavior and Development, 23, 153–174. Booth, A. E. (2006). Object function and categorization in infancy: Two mechanisms of facilitation. Infancy, 77, 145–169. Booth, A. E., & Waxman, S. (2002). Object names and object functions serve as cues to categories for infants. Developmental Psychology, 38, 948–957. Bourgeois, K., Khawar, A. W., Neal, S. A., & Lockman, J. J. (2005). Infant manual exploration of objects, surfaces, and their interrelations. Infancy, 8, 233–252. Buresh, J. S., & Woodward, A. L. (2007). Infants track action goals within and across agents. Cognition, 104, 287–314. Buxbaum, L. J., & Saffran, E. M. (2002). Knowledge of object manipulation and object function: Dissociations in apraxic and nonapraxic subjects. Brain and Language, 82, 179–199. Buxbaum, L. J., Kyle, K. M., & Menon, R. (2005). On beyond mirror neurons: Internal representations subserving imitation and recognition of skilled object-related actions in humans. Cognitive Brain Research, 25, 226–239. Carpenter, M., Call, J., & Tomasello, M. (2005). Twelve- and 18-month-olds copy actions in terms of goals. Developmental Science, 8, F13–F20. Casler, K., & Kelemen, D. (2005). Young children’s rapid learning about artifacts. Developmental Science, 8, 472–480. Casler, K., & Kelemen, D. (2007). Reasoning about artifacts at 24 months: The developing teleo-functional stance. Cognition, 103, 120–130. Chaigneau, S. E., Barsalou, L. W., & Sloman, S. A. (2004). Assessing the causal structure of function. Journal of Experimental Psychology-General, 133, 601–625. Chao, L. L., & Martin, A. (2000). Representation of manipulable, man-made objects in the dorsal stream. NeuroImage, 12, 478–484. Cohen, L. B. (1988). An information processing approach to infant cognitive development. In L. Weiskranz (Ed.), Thought without language (pp. 211–228). New York: Clarendon Press. Cohen, L. B., Rundell, L. J., Spellman, B. A., & Cashon, C. H. (1999). Infants’ perception of causal chains. Psychological Science, 10, 412–418. Corbetta, D., Thelen, E., & Johnson, K. (2000). Motor constraints on the development of perception-action matching in infant reaching. Infant Behavior and Development, 23, 351–374. Diesendruck, G., Hammer, R., & Catz, O. (2003a). Mapping the similarity space of children and adults’ artifact categories. Cognitive Development, 18, 217–231. Diesendruck, G., Markson, L., & Bloom, P. (2003b). Children’s reliance on creator’s intent in extending names for artifacts. Psychological Science, 14, 164–168. Gentner, D. (1978). What looks like a jiggy but acts like a zimbo: A study of early word meaning using artiﬁcial objects. Papers and Reports on Language Development, 15, 1–6. Gergely, G., Bekkering, H., & Kiraldy, I. (2002). Rational imitation in preverbal infant. Nature, 415, 755. German, T. P., & Johnson, S. C. (2002). Function and the origins of the design stance. Journal of Cognition and Development, 3, 279–300.

182

Lisa M. Oakes and Kelly L. Madole

Gibson, E. J. (1982). The concept of affordance in development: The renascence of functionalism. In W. A. Collins (Ed.), The concept of development: The Minnesota Symposium on Child Psychology (Vol. 15, pp. 55–81). Hillsdale, NJ: Erlbaum. Gibson, E. J. (1988). Exploratory behavior in the development of perceiving, acting, and the acquiring of knowledge. In M. R. Rosenzweig & L. W. Porter (Eds.), Annual Review of Psychology (Vol. 39, pp. 1–41). Palo Alto, CA: Annual Reviews. Goodale, M. A., & Milner, D. A. (1992). Separate visual pathways for perception and action. Trends in Neuroscience, 15, 20–25. Gutheil, G., Bloom, P., Valderrama, N., & Freedman, R. (2004). The role of historical intuitions in children’s and adults’ naming of artifacts. Cognition, 91, 23–42. Horst, J. S., Oakes, L. M., & Madole, K. L. (2005). What does it look like and what can it do? Category structure inﬂuences how infants categorize. Child Development, 76, 614–631. Jaswal, V. K. (2006). Preschoolers favor the creator’s label when reasoning about an artifact’s function. Cognition, 99, B83–B92. Kaldy, Z., & Leslie, A. M. (2005). A memory span of one? Object identiﬁcation in 6.5-monthold infants. Cognition, 97, 153–177. Keil, F. C. (1989). Concepts, kinds, and cognitive development. Cambridge, MA: MIT Press. Kelemen, D., Widdowson, D., Posner, T., Brown, A. L., & Casler, K. (2003). Teleo-functional constraints on preschool children’s reasoning about living things. Developmental Science, 6, 329–345. Kellenbach, M. L., Brett, M., & Patterson, K. (2003). Actions speak louder than functions: The importance of manipulability and action in tool representation. Journal of Cognitive Neuroscience, 15, 30–46. Kellman, P. J., & Arterberry, M. E. (1998). The cradle of knowledge: Development of perception in infancy. Cambridge, MA: MIT Press. Kemler Nelson, D. G. (1999). Attention to functional properties in toddlers’ naming and problem-solving. Cognitive Development, 14, 77–100. Kemler Nelson, D. G., Frankenﬁeld, A., Morris, C., & Blair, E. (2000a). Young children’s use of functional information to categorize artifacts: Three factors that matter. Cognition, 77, 133–168. Kemler Nelson, D. G., Russell, R., Duke, N., & Jones, K. (2000b). Two-year-olds will name artifacts by their functions. Child Development, 71, 1271–1288. Kemler Nelson, D. G., Herron, L., & Holt, M. B. (2003). The sources of young children’s name innovations for novel artifacts. Journal of Child Language, 30, 823–843. Landau, B., Smith, L. B., & Jones, S. S. (1988). The importance of shape in early lexical learning. Cognitive Development, 3, 299–321. Landau, B., Smith, L. B., & Jones, S. S. (1998). Object shape, object function, and object name. Journal of Memory and Language, 38, 1–27. Lavie, N., & Cox, S. (1997). On the efﬁciency of visual selective attention: Efﬁcient visual search leads to inefﬁcient distractor rejection. Psychological Science, 8, 395–398. Leslie, A. M. (1984). Spatiotemporal contiguity and perception of causality in infants. Perception, 13, 287–305. Madole, K. L., & Cohen, L. B. (1995). The role of object parts in infants’ attention to formfunction correlations. Developmental Psychology, 31, 317–332. Madole, K. L., & Oakes, L. M. (1999). Making sense of infant categorization: Stable processes and changing representations. Developmental Review, 19, 263–296. Madole, K. L., & Oakes, L. M. (2005). Infants’ attention to and use of functional properties in categorization. In L. Carlson & E. van der Zee (Eds.), Functional features in language and space: Insights from perception, categorization, and development (pp. 275–292). Oxford: Oxford University Press.

How Infants Construe Functional Features

183

Madole, K. L., Oakes, L. M., & Cohen, L. B. (1993). Developmental changes in infants’ attention to function and form-function correlations. Cognitive Development, 8, 189–209. Mandler, J. M. (2004). The foundations of mind: Origins of conceptual thought. New York: Oxford University Press. Mandler, J. M., & McDonough, L. (1998). Studies in inductive inference in infancy. Cognitive Psychology, 23, 60–96. Mareschal, D., & Johnson, M. H. (2003). The ‘‘what’’ and ‘‘where’’ of object representations in infancy. Cognition, 88, 259–276. Martin, A. (2007). The representation of object concepts in the brain. Annual Review of Psychology, 58, 25–45. Matan, A., & Carey, S. (2001). Developmental changes within the core of artifact concepts. Cognition, 78, 1–26. Meltzoff, A. N. (1988). Infant imitation and memory: Nine-month-olds in immediate and deferred tests. Child Development, 59, 217–225. Mervis, C. B. (1985). On the existence of prelinguistic categories—a case-study. Infant Behavior and Development, 8, 293–300. Murphy, G. L. (2002). The big book of concepts. Cambridge, MA: MIT Press. Needham, A. (2000). Improvements in object exploration skills may facilitate the development of object segregation in early infancy. Journal of Cognition and Development, 1, 131–156. Nelson, K. (1972). The relation of form recognition to concept development. Child Development, 43, 67–74. Nelson, K. (1973). Some evidence for cognitive primacy of categorization and its functional basis. Merrill-Palmer Quarterly, 19, 21–39. Nelson, K. (1974). Concept, word, and sentence: Interrelations in acquisition and development. Psychological Review, 81, 267–285. Nelson, K. (1979). Explorations in the development of a functional semantic system. In W. A. Collins (Ed.), The Minnesota Symposium on Child Psychology: Children’s language and communication (Vol. 12, pp. 47–81). Hillsdale, NJ: Erlbaum. Oakes, L. M. (1994). The development of infants’ use of continuity cues in their perception of causality. Developmental Psychology, 30, 748–756. Oakes, L. M., & Cohen, L. B. (1990). Infant perception of a causal event. Cognitive Development, 5, 193–207. Oakes, L. M., & Madole, K. L. (1999). From seeing to thinking: Reply to Mandler. Developmental Review, 19, 307–318. Oakes, L. M., & Madole, K. L. (2000). The future of infant categorization research: A processoriented approach. Child Development, 71, 119–126. Oakes, L. M., & Madole, K. L. (2003). Principles of developmental change in infants’ category formation. In D. H. Rakison & L. M. Oakes (Eds.), Early category and concept development: Making sense of the blooming, buzzing confusion (pp. 132–158). New York: Oxford University Press. Oakes, L. M., Tellinghuisen, D. J., & Tjebkes, T. L. (2000). Competition for infants’ attention: The interactive inﬂuence of attentional state and stimulus characteristics. Infancy, 1, 347–361. Oakes, L. M., Kannass, K. N., & Shaddy, D. J. (2002). Developmental changes in endogenous control of attention: The role of target familiarity on infants’ distraction latency. Child Development, 73, 1644–1655. Oakes, L. M., Ross-Sheehy, S., & Luck, S. J. (2006). Rapid development of binding in visual short-term memory. Psychological Science, 17, 781–787. Perone, S., & Oakes, L. M. (2006). It clicks when it is rolled and squeaks when it is squeezed: What 10-month-old infants learn about object function. Child Development, 77, 1608–1622.

184

Lisa M. Oakes and Kelly L. Madole

Perone, S., Madole, K. L., Ross-Sheehy, S., Carey, M., & Oakes, L. M. (in press). The relation between infants activity with objects and attention to object appearance. Developmental Psychology. Perone, S., & Oakes, L. M. (2008). Seven-month-old infants’ attention to the components of function. Unpublished manuscript. Piaget, J., Grize, J. B., Szeminsk, A., & Bang, V. (1977). Epistomology and psychology of functions. Boston: D. Reidel Publishing Company. Prasada, S. (2005). Being near the ceramic, but not near the mug: On the role of construal in spatial language. In L. Carlson & E. van der Zee (Eds.), Functional features in language and space: Insights from perception, categorization, and development (pp. 205–234). Oxford: Oxford University Press. Rakison, D. H. (2004). Infants’ sensitivity to correlations between static and dynamic features in a category context. Journal of Experimental Child Psychology, 89, 1–30. Rakison, D. H., & Cohen, L. B. (1999). Infants’ use of functional parts in basic-like categorization. Developmental Science, 2, 423–432. Rakison, D. H., & Poulin-Dubois, D. (2002). You go this way and I’ll go that way: Developmental changes in infants’ detection of correlations among static and dynamic features in motion events. Child Development, 73, 682–699. Regier, T., Carlson, L., & Corrigan, B. (2005). Attention in spatial language: Bridging geometry and function. In L. Carlson & E. van der Zee (Eds.), Functional features in language and space: Insights from perception, categorization, and development (pp. 191–204). Oxford: Oxford University Press. Rips, L. J. (1989). Similarity, typicality, and categorization. In S. Vosniadou & A. Ortony (Eds.), Similarity and analogical reasoning (pp. 21–59). New York, NY: Cambridge University Press. Robinson, C. W., & Sloutsky, V. M. (2004). Auditory dominance and its change in the course of development. Child Development, 75, 1387–1401. Ross-Sheehy, S., Oakes, L. M., & Luck, S. J. (2003). The development of visual short-term memory capacity in infants. Child Development, 74, 1807–1822. Shaddy, D. J., & Colombo, J. (2004). Developmental changes in infant attention to dynamic and static stimuli. Infancy, 5, 355–365. Smith, L. B. (2005). Cognition as a dynamic system: Principles from embodiment. Developmental Review, 25, 278–298. Smith, L. B. (forthcoming). Dynamic systems, sensory-motor processes, and the origins of stability and ﬂexibility. In J. P. Spencer, M. S. Thomas, & J. L. McClelland (Eds.), Toward a new grand theory of development? Connectionism and dynamic systems theory re-considered. New York, NY: Oxford University Press. Smith, L. B., & Thelen, E. (2003). Development as a dynamic system. Trends in Cognitive Sciences, 7, 343–348. Song, H., Baillargeon, R., & Fisher, C. (2005). Can infants attribute to an agent a disposition to perform a particular action? Cognition, 98, B45–B55. Trauble, D., & Pauen, S. (2008). The role of functional information for infant categorization. Cognition, 105, 362–379. Truxaw, D., Krasnow, M. M., Woods, C., & German, T. P. (2006). Conditions under which function information attenuates name extension via shape. Psychological Science, 17, 367–371. Welder, A. N., & Graham, S. A. (2001). The inﬂuences of shape similarity and shared labels on infants’ inductive inferences about nonobvious object properties. Child Development, 72, 1653–1673. Wilcox, T., & Chapa, C. (2004). Priming infants to attend to color and pattern information in an individuation task. Cognition, 90, 265–302.

How Infants Construe Functional Features

185

Wolk, D. A., Coslett, H., & Glosser, G. (2005). The role of sensory-motor information in object recognition: Evidence from category-speciﬁc visual agnosia. Brain and Language, 94, 131–146. Yantis, S., & Jonides, J. (1990). Abrupt visual onset and selective attention: Voluntary versus automatic allocation. Journal of Experimental Psychology: Human Perception and Performance, 16, 121–134. Yoon, E. Y., & Humphreys, G. W. (2005). Direct and indirect effects of action on object classiﬁcation. Memory and Cognition, 33, 1131–1146. Younger, B. A., & Cohen, L. B. (1986). Developmental changes in infants’ perception of correlations among attributes. Child Development, 57, 803–815. Younger, B. A., & Furrer, S. (2003). A comparison of visual familiarization and objectexamining measures of categorization in 9-month-old infants. Infancy, 4, 327–348.

This page intentionally left blank

TRANSACTIONAL FAMILY DYNAMICS: A NEW FRAMEWORK FOR CONCEPTUALIZING FAMILY INFLUENCE PROCESSES

Alice C. Schermerhorn and E. Mark Cummings DEPARTMENT OF PSYCHOLOGY, UNIVERSITY OF NOTRE DAME, NOTRE DAME, IN 46556, USA

I. INTRODUCTION II. SETTING THE STAGE A. HISTORICAL OVERVIEW B. THEORETICAL BASES FOR TRANSACTIONAL FAMILY DYNAMICS III. TRANSACTIONAL FAMILY DYNAMICS: AN EMERGING THEME A. WHY IS A TRANSACTIONAL FAMILY DYNAMICS MODEL NEEDED? B. WHAT KIND OF APPROACH IS NEEDED? C. THE HIERARCHICALLY ORGANIZED SYSTEMS OF TRANSACTIONAL FAMILY DYNAMICS D. A COMPREHENSIVE MODEL OF TRANSACTIONAL FAMILY DYNAMICS IV. MAPPING EMPIRICAL WORK ONTO A TRANSACTIONAL FAMILY DYNAMICS FRAMEWORK A. TRANSACTIONAL DYNAMICS OF THE PARENT–CHILD RELATIONSHIP B. TRANSACTIONAL DYNAMICS BETWEEN PARENT–CHILD AND INTERPARENTAL RELATIONSHIPS C. TRANSACTIONAL DYNAMICS OF INTERPARENTAL RELATIONSHIPS AND CHILDREN D. FAMILY-WIDE TRANSACTIONAL DYNAMICS V. DISCUSSION A. AN AGENDA FOR FUTURE RESEARCH: SOME HYPOTHESES ABOUT TRANSACTIONAL FAMILY DYNAMICS B. CONCLUSIONS ACKNOWLEDGEMENT REFERENCES

I. Introduction people are not just onlooking hosts of internal mechanisms orchestrated by environmental events. They are agents of experiences rather than simply undergoers of experiences. (Bandura, 2001, p. 4)

187 Advances in Child Development and Behavior R.V. Kail : Editor

Copyright r 2008 Elsevier B.V. All rights reserved.

188

Alice C. Schermerhorn and E. Mark Cummings

In research on family inﬂuence processes, there is a growing dissatisfaction with existing models that assume unidirectional pathways and underemphasize dynamic processes. However, aside from acknowledging the problem, few systematic proposals have been advanced for more sophisticated ways of thinking about these pathways of inﬂuence. Addressing this gap, we propose transactional family dynamics as a new way of understanding family inﬂuence processes. Transactional family dynamics refers to the myriad ways in which family members inﬂuence one another, that is, mutual inﬂuence processes within families. For example, these processes may include complex patterns of inﬂuence in interparental, father–child, mother–child, and sibling relationships. Notably, our interest is in transactional processes—not unidirectional processes—that is, inﬂuence processes continuously moving in both directions over time. Our interest in transactional family dynamics began with a review of the literature on child effects on families (Cummings & Schermerhorn, 2003) and empirical tests of child effects on marital conﬂict (Schermerhorn, Cummings, & Davies, 2005; Schermerhorn et al., 2007; Schermerhorn, Chow, & Cummings, 2007). We were intrigued to ﬁnd that children’s responses to interparental conﬂict predicted change in interparental conﬂict itself—either increases or decreases, depending on the nature of the child’s response. Expanding our focus, we also found transactional links between interparental and parent–child relationships (i.e., mother–child and father–child) over time (Schermerhorn, Cummings, & Davies, 2008). We were impressed by the extent of the evidence for the transactional nature of these processes and by the multiple pathways of inﬂuence between multiple family members and relationships. These ﬁndings, and emerging results from other laboratories, prompted us to think about the need for a new framework for conceptualizing the multitude of family inﬂuence processes. That is, rather than focusing narrowly on just one pathway (e.g., children’s inﬂuence on marital conﬂict), we wanted our model to encompass the many pathways, and to integrate emerging empirical work suggesting the importance of multiple pathways of inﬂuence. The notion of transactional family dynamics refers to inﬂuence processes among multiple family relationships, including the inﬂuence of individual family members on family relationships, the inﬂuence of family relationships on one another, and family-wide inﬂuences. The transactional family dynamics framework also includes the reverse direction of effects. For example, with regard to the inﬂuence of individuals on family relationships, one would also be concerned with the inﬂuence of family relationships on individual family members. The aim of our approach is to provide a framework for representing these processes across multiple family relationships.

Transactional Family Dynamics

189

These processes unfold in ‘‘real time,’’ or moment-by-moment interactions, as well as ‘‘developmental time,’’ or long periods of time. Such processes encompass behaviors intended to inﬂuence other family members, but also include family members’ unintentional inﬂuence on one another. Thus, another key point of our approach is that dynamic processes of inﬂuence operate at multiple levels of analysis, including varying lengths of time, or time scales. Some time ago, Thelen and Ulrich (1991) called for investigators to develop dynamic accounts of behavior at many levels of analysis. Consistent with that message, our aim is also to describe and identify the nested, multiply caused phenomenon of family inﬂuence. Although there are increasing calls for broader conceptualizations of families (Cox & Paley, 1997; Jenkins et al., 2005a), increasing the complexity of the study of families presents theoretical and practical challenges that remain to be addressed. Currently, much of the literature on families at least implicitly reﬂects a narrow conceptualization of families, for example, assessing only one direction of inﬂuence, or focusing on only one or a couple of family members. The narrow focus also presents a problem for the clinician, by endorsing therapies that may be ill-suited for real families because of failure to consider important directions of inﬂuence. Moreover, a gap in the study of family inﬂuence processes is the lack of an overarching theoretical framework to unite and integrate research concerning multiple directions of inﬂuence. Thus, the development of a transactional family dynamics framework was motivated by the urgent need advocated by many in the discipline to move toward models that embrace the complexity of family relationships. In this context, it is important to consider the factors that contribute to the inherent complexity of mutual family inﬂuence processes. First, families have a hierarchical organization, with individuals nested within dyads and triads, which are nested within families. Figure 1a depicts this hierarchical organization. Second, families may include multiple family dyads and triads, and therefore, a multitude of inﬂuence pathways among them. Third, family inﬂuence processes unfold in real time interactions, as well as in the context of processes that may extend over longer periods of time (e.g., the development of emotional bonds or attachments; Bowlby, 1973). That is, time is hierarchically organized, with smaller time scales nested within increasingly long time scales, with varying possible lengths of time for potentially critical changes to occur (Cole & Maxwell, 2003). Fourth, there are different conceptualizations of inﬂuence and change in family relationships, including change from one time point to the next and overall patterns of change. Our framework recognizes and attempts to accommodate these complicating factors, for example, by classifying ﬁndings by family relationship and by conceptualization of inﬂuence and change.

Alice C. Schermerhorn and E. Mark Cummings

190 (a)

Child(ren)

Mother

Parent-child Relationships

Father

Interparental Relationship

Children and their Siblings Increasing Complexity

Parent-child and Interparental Relationships

Interparental Relationships and Children

Sibling and Parent-child Relationships

Family-wide (b)

Duration 1A

Duration 1B

Duration 2A

Duration 2B Duration 3A

Duration 3B

Duration 4A

Duration 4B Duration 5A

Duration 5B

Fig. 1. The hierarchical organization of family influence processes: (a) in terms of family members and family relationships; and (b) in terms of time scales.

We begin with a brief historical overview of some of the theoretical inﬂuences that have laid the foundations for this emerging approach to understanding family inﬂuence processes. We then describe transactional family dynamics at a theoretical level, providing a set of organizing concepts and principles. Next, selected research consistent with this framework is reviewed, toward showing how these seemingly disparate directions in research ﬁt together, and underscoring the utility of the transactional family dynamics model. We conclude by highlighting some possible directions and hypotheses for future research. The aim of this chapter is to articulate and advance a model of transactional family dynamics as a framework for conceptualizing and studying family inﬂuence processes. Allen et al. (2006) developed the metaphor of a dance for the concept of family inﬂuence processes. Arguing that the critical issue is not one of seeking to understand causality, but rather of seeking to understand the dance itself, Allen and colleagues pointed out that the dance might be lead by more than one family member. Moreover, our ability to distinguish

Transactional Family Dynamics

191

cause and effect is very limited, as even the most sophisticated research designs and data analytic methods cannot prove that our causal models are correct, but rather, can only prove that they are incorrect. Moreover, Allen et al. argued, questions about causality are not even the right questions to be asking. That is, ‘‘Observing a dance doesn’t tell us who’s in the lead, and knowing who’s in the lead doesn’t tell us who decided what dance to do, or what music to play’’ (Allen et al., 2006). Instead, we should be asking questions about the dance itself—the complex process of leading and following, giving and taking. We elaborate this metaphor by pointing out that the metaphor is not mechanistic—there is no automatic correspondence between action and reaction (although there is certainly a strong relationship between the two). Thus, this metaphor allows for individuality—it affords space for interpretation of events, personality, and agency. That is, the metaphor allows for interpreting the music through dance.

II. Setting the Stage A. HISTORICAL OVERVIEW Traditionally, the commonly held view of family relationships was unidirectional—the direction of inﬂuence was believed to be parent-tochild, with scant consideration of child effects on parents. Then, beginning in the late 1960s, Richard Q. Bell (1968, 1971, 1979) called attention to the reverse direction of effects, namely, child-to-parent inﬂuence processes. The late 1960s through the early 1980s saw an upsurge of research aimed at distinguishing child-to-parent effects from parent-to-child effects, and the view that children inﬂuence their parents gained acceptance (see Kahn & Antonucci, 1980; Powers et al., 1983; Sameroff, 1975a, 1975b). However, in the subsequent period, relatively few researchers examined bidirectional processes (see Dunn, Hinde & Stevenson-Hinde, and Kuczynski for exceptions), perhaps partly because of methodological and statistical limitations (Lytton, 1982) and even decreased interest in the topic (Lytton, 1990b). Moreover, although some (e.g., Engfer, 1988; Lytton, 1982) had called for examination of the whole family, suggesting that children, siblings, marriages, and parenting might all be related as inﬂuences, relatively few studies adopted or advanced a family-wide model of transactional dynamics. Gradually, however, conceptualizations of relatively complex patterns of family inﬂuence gained acceptance. A resurgence of theory and research relevant to transactional family dynamics began in the late 1990s, including increasing examination of additional family relationships, such as sibling

192

Alice C. Schermerhorn and E. Mark Cummings

relationships, and links between pairs of family relationships, such as between marital and parent–child relationships. Furthermore, statistical approaches began to emerge that enabled the testing of increasingly complex and sophisticated models of family inﬂuence, addressing a longstanding barrier to progress on these notions. In addition, several decades of research on multiple areas of family functioning have yielded a broader and deeper knowledge of families that enhances understanding of transactional family processes. For example, signiﬁcant advances have been made in the study of parent–child attachment, parenting, parental psychopathology, marital conﬂict, and child functioning in the context of families (see Connell & Goodman, 2002; Cummings & Davies, 2002; Davies & Cummings, 1994; Gray & Steinberg, 1999; Grych & Fincham, 1990; Thompson & Raikes, 2003). With these empirical and methodological advances, there is now potential for substantial progress in the study of transactional family dynamics. That is, it is now possible to return productively to the innovative questions raised several decades ago about child effects and the accompanying ideas. These ideas may well have been ahead of their time in the early 1980s, but are very appropriate and valuable for emerging directions in family research at this time. New questions for this approach are also informed by the theoretical and empirical advances in family research that began in the 1990s. Thus, the time is ripe to move forward with new advances in what we term transactional family dynamics.

B. THEORETICAL BASES FOR TRANSACTIONAL FAMILY DYNAMICS Several major theories contribute to our conceptualization of transactional family dynamics (i.e., mutual inﬂuence processes within families; Table I). From these major theories, we have drawn a number of themes that inform our framework. First, multiple family members and family relationships inﬂuence one another continuously over time. Second, families are organized hierarchically, with individuals nested within family dyads (e.g., marriages) and triads (e.g., mother–father–child), which are, in turn, nested within families. At the same time, inﬂuence processes are also viewed as ‘‘circular,’’ with a continuous cycle of mutual inﬂuence in which action, reaction, and further reaction occur constantly (Granic, 2000). Third, time is also hierarchically organized, with moments nested within hours, which are nested within days, which are nested within weeks, which are nested within months, and so on. Importantly, processes unfolding over different time scales are qualitatively different from one another. That is, although inﬂuence processes in different time scales have much in common with one

Transactional Family Dynamics

193

Table I Transactional Family Dynamics: Theoretical Foundations Reference

Summary of contribution

Family relationship

Bandura (2001, 2006) Bell (1968, 1971, 1979) Bogartz (1994)

Notions of human agency Child effects Merits and weaknesses of dynamic systems theory Interacting time scales Reciprocity and interconnectedness of multiple family relationships Bioecological model of human development

Individual Parent–child Child developmental processes Parent–child Parent–child

Developmental psychopathology Developmental perspective on parent–child interactions Social Relations Model and Actor–Partner Interdependence Model Family systems theory

Person–environment

Bowlby (1973) Bretherton (1985)

Bronfenbrenner (1979, 1986, 1988, 2005), Bronfenbrenner & Morris (1998) Cicchetti (2006), Cicchetti et al. (1988) Collins & Madsen (2003) Cook (2003), Cook & Kenny (2005), Kashy & Kenny (2000) Cox & Paley (1997, 2003) Cummings & Schermerhorn (2003)

Developmental perspective on children’s inﬂuence on family relationships

Emery (1982) Granic (2000, 2005), Granic et al. (2003), Granic & Hollenstein (2003, 2006), Granic & Patterson (2006) Hinde & Stevenson-Hinde (1987) Kelley et al. (1983)

Child effects on marriage Dynamic systems theory applied to child antisocial behavior and reciprocity in family relationships Child development in context of social networks Mutual inﬂuence processes, distinguished between emotion, thought, and behavior Bidirectionality and child agency

Kuczynski & Hildebrandt (1997), Kuczynski et al. (1997), Kuczynski & Parkin (2007), Lollis & Kuczynski (1997) Lewis (2000, 2002, 2004), Lewis et al. (1999), Howe & Lewis (2005)

Time scales of developmental processes, dynamic systems models of development

Lytton (1982, 2000)

Parent- and child-effects

Parent–child, person– environment

Parent–child Parent–child

Multiple family relationships Multiple family relationships, especially child and interparental Child and interparental Parent–child

Family, teacher, peer Close relationships

Parent–child

Utility of dynamic systems (DS) approaches to explain development Parent–child

194

Alice C. Schermerhorn and E. Mark Cummings Table I (Continued )

Reference

Summary of contribution

Family relationship

Maccoby (1984) McHale & Fivaz-Depeursinge (1999), McHale et al. (2003) Minuchin (1985)

Mutual inﬂuence processes Family-wide concepts, family alliances, child effects Family systems theory

Newson & Newson (1976)

Child’s developing inﬂuence on others Coercive family processes, bidirectionality and antisocial behavior Socialization and interactional processes during adolescence Transactional models of child development, intervention research Child effects on marriage

Parent–child Parent–child, coparenting Multiple family relationships Parent–child

Patterson (1982), Patterson, DeBaryshe, & Ramsey (1989), Patterson & Fisher (2002) Powers et al. (1983)

Sameroff (1975a, 1975b, 1995), Sameroff & Fiese (2000), Sameroff & MacKenzie (2003) Sanders, Nicholson, & Floyd (1997) Scarr & McCartney (1983), McCartney (2003) Schaffer (1999)

Smith (2005), Smith & Thelen (2003), Thelen & Smith (1994, 1998), Thelen and Ulrich (1991)

Selection and creation of one’s environment Bidirectionality of relationships and child development Dynamic systems principles and developmental psychology

Parent–child

Parent–child

Parent–child, person– environment Child and interparental Person–environment Parent–child

Child motor and cognitive development

another, they also have important differences, which we discuss later. Fourth, multiple relevant conceptualizations of inﬂuence and change are posited. More speciﬁcally, we distinguish between inﬂuence processes involving association or contingency between family members, change from one time point to the next, and overall patterns of change. We return to these themes throughout this chapter, as they contribute substantially to the transactional family dynamics framework. In the rest of this section, we provide an overview of research and theoretical directions that provide a foundation for our approach, highlighting their relevance to our framework. 1. Child Effects Our conceptualization of transactional family dynamics originated with our interest in child effects (Cummings & Schermerhorn, 2003).

Transactional Family Dynamics

195

Bell (1968, 1971, 1979) was the ﬁrst to develop a well-articulated call to study effects of children on parents. A long-standing assumption in socialization research had been that the child’s effects on the parent–child system could be attributed to biological characteristics of the child. Bell argued cogently for recognition of the importance of child effects in their own right, independent of the issue of biology. Building on these ideas, Lytton (1982) suggested that the existing literature may well hide child effects, because of its reliance on cross-sectional designs and correlational analyses, as well as its handling of data in ways that precluded the examination of children’s inﬂuence on parents. Subsequently, other investigators also emphasized bidirectional models of inﬂuence on children’s socialization (Dunn, 1997; Kuczynski, Marshall, & Schell, 1997; Maccoby, 1984; Powers et al., 1983), and the bidirectional and multifaceted development of parent–child interactions (Collins & Madsen, 2003). Moreover, child effects operate from the moment an infant is born, and children’s behavior serves a homeostatic function, regulating the behavior of other family members (McHale, Kavanaugh, & Berkman, 2003). Children are not passive recipients of parenting, but rather, active participants in parent–child relationships (Cole, 2003; Emery et al., 1983; Maccoby, 1984; Stifter, 2003). Parenting practices and child functioning are a product of both parent and child characteristics and behavior (Lytton 1990a; Patterson & Fisher, 2002), and maternal responding to young children’s misbehavior depends in part on the type of misbehavior (Grusec & Kuczynski, 1980). Thus, the parent–child relationship can be described as reciprocal, involving mutual inﬂuence between parent and child (Bretherton, 1985). Furthermore, parent–child interactions occur in a wide range of contexts (e.g., play, caregiving, teaching), and parent–child interactions in one context may affect interactions in another context (Lollis & Kuczynski, 1997). Relatedly, child effects occur, not solely within the mother–child relationship (a primary focus of earlier research), but also within father– child relationships, and children inﬂuence their siblings’ relationships with their parents (McHale et al., 2003). Moreover, children’s inﬂuence extends to the marital relationship (Cummings & Schermerhorn, 2003). Children’s behavioral dysregulation during marital conﬂict may reﬂect ‘‘taking on a symptom’’ (Emery, 1982), intended to distract parents from marital difﬁculties. In contrast, children’s hostility in the context of interparental hostility may escalate coercive family processes (Patterson, 1982), thereby promoting increased marital discord over time. Notably, agency and bidirectionality are to be distinguished from one another. Bidirectional effects include any behavioral, psychological, or biological processes that alter relations between two people, but are not necessarily self-initiated or intentional. However, Kuczynski and colleagues

196

Alice C. Schermerhorn and E. Mark Cummings

have called for understanding children as agents in the family (Kuczynski et al., 1997; Kuczynski & Hildebrandt, 1997). Agency has been described as intentional inﬂuence on one’s functioning and life circumstances (Bandura, 2006) and as exercising the ability to engage in intentional behavior, choose methods of inﬂuencing others, reﬂect on behavior, interpret communications, and make assertions (Kuczynski & Parkin, 2007). Therefore, agentic effects are a subset of bidirectional effects. That is, the concept of agency makes stronger suppositions about the individual’s role, including underlying motivations, organization, and plans. Bandura (2006) identiﬁed four core properties of agency: (a) developing an action plan; (b) setting goals and anticipating likely outcomes; (c) acting on one’s intentions; and (d) evaluating those actions. In the context of family relationships, we have deﬁned children’s agency as their behaviors that are designed to inﬂuence family members (Cummings & Schermerhorn, 2003). Kuczynski and Parkin (2007, p. 261) wrote, ‘‘A challenge for the future is to develop models that consider parents and children interacting simultaneously as agents and adapting to each other’s agency during interactions.’’

2. Dialectical Models Dialectical models have also informed our thinking about family inﬂuence processes and the hierarchical organization of families. Kuczynski and Parkin (2007) characterized dialectical models as reﬂecting intentionality and portraying the individual as active, rather than reactive. One key concept within dialectics is the unity of opposites; that is, the notion that the individual must be recognized as a part of a whole, and that in order to understand the individual, one must examine interrelations between part and whole (Kuczynski & Parkin, 2007). We draw on this notion of the unity of opposites in explicating the hierarchical organization of families, which consist of relationships, which consist of individuals. Also important is the notion of contradiction, or the role of opposing elements in producing quantitative and qualitative change; this process of contradiction and change is a pervasive part of family life. For example, contradiction may result from differences between husbands’ and wives’ parenting values, compounded by an opposing child value of security and family cohesion. Out of these opposing elements, which are nested within a larger family system, emerges change—ideally, a synthesis, resolving the differing parental values, ending the conﬂict, and thereby restoring children’s sense of family security. Thus, in the context of family inﬂuence processes, the unity of opposites and contradiction may work together to produce a synthesis, reﬂecting change within the family.

Transactional Family Dynamics

197

3. Bronfenbrenner’s Ecological Model of Human Development Urie Bronfenbrenner’s work has considerably inﬂuenced our thinking about transactional family relationships, particularly in terms of notions of hierarchical organization. His ecological theory of human development (1979, 1986) provides a model of the mutual development of the individual and the multiple, nested environments within which the individual functions (micro-, meso-, exo-, macrosystems) over the life span (chronosystem). Of particular relevance to our work, Bronfenbrenner (1979) emphasized the role of reciprocity in human interactions. In later formulations of his bioecological theory of human development, Bronfenbrenner (1988, 2005) placed particular emphasis on four broad and interrelated components of human development: (a) the developing person’s characteristics; (b) interaction between the person and environment; (c) environmental contexts ranging from proximal to distal in relation to the person; and (d) the progression of time. Thus, Bronfenbrenner’s ecological model provides a useful foundation for conceptualizing transactional family dynamics because it emphasizes the hierarchical organization of systems. We apply this notion of hierarchical organization to the context of families, and more speciﬁcally, family inﬂuence processes.

4. Individual–Environment Interaction Because our framework involves mutual inﬂuence of the individual on others in the family and multiple pathways of inﬂuence, notions of interactions between the person and the environment (or family) are important to our model. Kelley et al. (1983) linked mutual inﬂuence processes in close relationships with events in the environment and highlighted the interactive roles of emotion, cognition, and behavior in mutual inﬂuence processes. McCartney and Scarr also presented revolutionary ideas about individual–environment interaction (McCartney, 2003; Scarr & McCartney, 1983). In particular, their notion of niche-picking, or the individual’s selection and creation of environments that provide a good ﬁt to the individual, is closely related to our views of mutual inﬂuence among family members. That is, we see family members’ inﬂuence on one another as part of the process of shaping one’s environment—changing the family environment. Advancing notions of individual–environment interaction, Sameroff (1975a, 1975b, 1995) called for moving beyond examining static characteristics of the person and the environment. He suggested that researchers should instead examine the dynamic, continual transactions between the person and the environment. Sameroff argued that development is not solely a result of either characteristics of the person or environment, but instead results from the process by which these characteristics develop

198

Alice C. Schermerhorn and E. Mark Cummings

through their mutual inﬂuence over time. We think this conceptualization of transactions is critically important; consequently, we use the word transactional in our framework.

5. Developmental Psychopathology The ﬁeld of developmental psychopathology has also contributed to the theoretical underpinnings of this framework, because of its emphasis on the hierarchical organization of developing systems. Stemming from the ﬁeld of developmental psychopathology, an organizational perspective on human development emphasizes viewing, not just discrete domains of development, but rather, the overall organization of development across domains, including interrelations among domains (Cicchetti, Toth, & Bush, 1988; The Carolina Consortium on Human Development, 1996). Thus, the individual is viewed holistically (Cicchetti, 2006), and interactions between genes, neurobiology, psychology, and social functioning are viewed as critical in determining behavior (Cicchetti et al., 1988). Moreover, the organism is regarded as fully integrated, such that lower-level events, such as cellular functioning, can inﬂuence higher-level events, such as thought and emotion, and vice versa (Cicchetti, 2006). By extension, then, even higher-level events, like the mutual inﬂuence of family members and families, are an important focus of developmental psychopathology. Relatedly, we view time as hierarchically organized, with shorter time scales nested within longer time scales; moment-by-moment inﬂuence processes contribute to long-term inﬂuence processes—as well as the reverse—and both long and short time scales uniquely contribute to the whole of family experience. Thus, our notions of transactional family dynamics reﬂect circularity in patterns of interaction and inﬂuence, as well as the hierarchical organization of families and time.

6. Dynamic Systems Theory Dynamic systems theory has also inﬂuenced our thinking about the hierarchical organization of families and of time, as well as our conceptualizations of change. In particular, dynamic systems principles are well suited to examining complex questions about the interrelatedness of the whole and its parts (Bogartz, 1994; Smith, 2005), and thus, provide an ideal framework for research on family inﬂuence processes (Granic, 2000; O’Brien, 2005). Thus, we draw on dynamic systems principles in addressing the hierarchical organization of families, with multiple individuals and relationships nested within families, and the hierarchical organization of time, with multiple time scales nested within one another.

Transactional Family Dynamics

199

Dynamic systems theory addresses the process of change and development, rather than developmental outcomes; in dynamic systems terms, there is no end point of development (Thelen & Ulrich, 1991). Moreover, with its central focus on change and change in the rate of change, dynamic systems theory points to questions about both (a) change from one time point to the next; and (b) overall patterns of change. Chief among the contributions of dynamic systems theory is a set of concepts facilitating examination of overall patterns of change. Such patterns include stabilization, destabilization, and self-regulation. In a ground-breaking application of dynamic systems theory to the ﬁeld of developmental psychology, Thelen and Ulrich (1991) described motor development as the process of repeated cycles of stabilizing and destabilizing behavior patterns. In terms of social development, relationships may develop partly as a function of stabilizing and destabilizing behavior patterns of family members. For example, when parents repeatedly respond sensitively, their infants develop stable views of their parents as dependable. Moreover, family relationships may be self-regulating, with tendencies to return to baseline levels of functioning. As an illustration, a mother and her adolescent might have a fairly close relationship, but there may be periods of more or less closeness; that is, the system may oscillate back and forth past its baseline level of closeness. Thus, dynamic systems principles and methods afford opportunities to deepen conceptualization and empirically based knowledge of family inﬂuence processes. However, dynamic systems methods rely on mathematics-intensive procedures, and relatively little research has utilized this approach.

7. Social Relationships as a Context for Development We also draw on the notion of social relationships as contexts for development; that is, notions of others’ inﬂuence on one’s change and development. Just as Bronfenbrenner and others have outlined models of hierarchical organization of the environment, Hinde and Stevenson-Hinde (1987) conceptualized children’s development within social relationships in terms of (a) links between the child and the social interactions in which they participate; and (b) links between social interactions and the relationships within which they are nested. Hinde and Stevenson-Hinde also emphasized the history of interactions and relationship functioning as contributors to subsequent interactions and relationship functioning. Moreover, all of these processes are conceptualized as inﬂuencing, and being inﬂuenced by, children’s interactions and relationships with others in their social networks, whose interactions and relationships are, in turn, inﬂuenced by other people with their own social relationships. Thus, this

200

Alice C. Schermerhorn and E. Mark Cummings

conceptualization of social relationships as hierarchically organized ﬁts with our transactional family dynamics framework, with interactions nested within dyads and triads, who in turn, are nested within families. 8. Family Systems Theory Family systems theory emphasizes the interdependent nature of subsystems within families (Cox & Paley, 1997; Minuchin, 1985), conceptualizing families as organized wholes (Cox & Paley, 2003). These notions gave rise to our views of families as hierarchically organized, consisting of multiple family members and relationships. Families are capable of both self-regulation and self-reorganization (Cox & Paley, 1997). Self-regulation involves stabilizing interaction patterns; for example, there may be rapid changes in family conﬂict followed by self-regulation back to the family’s typical low levels of conﬂict. Self-reorganization refers to adaptation to the environment. For example, a downturn in the economy may cause a father to lose his job, which may prompt the family to reorganize itself around new roles, such as the mother becoming the primary source of income. Similarly, Bretherton (1985) discussed links between children’s internal representations of multiple family relationships, and McHale and FivazDepeursinge (1999) called for an examination of families as wholes, rather than as a group of individuals or dyads. Moreover, they described the notion of a family’s personality as the family’s tendency toward certain emotions and behaviors. For example, one family may have a warm and expressive personality, whereas another family may tend toward a cold, detached personality. Thus, these notions of families as hierarchically organized wholes with their own personalities, and of multiple pathways of inﬂuence play an integral part in our conceptualization of transactional family dynamics. 9. Parent and Child Development Family inﬂuence processes depend, in part, on child and parent development. That is, the relationship between two family members is a developing one, with each member of the relationship affecting the other member over time. Maccoby (1984) discussed at length the effect of child development on bidirectionality. Maccoby highlighted the role of such developmental factors as physical growth, language development, conceptions of others, and autonomy in children’s interactions with their parents. As they develop, children become better able to communicate with family members and become increasingly aware of others’ points of view, as well as becoming more skilled at portraying themselves favorably (Newson &

Transactional Family Dynamics

201

Newson, 1976). Children also become more skillful in their approach to noncompliance with parental requests (Kuczynski & Kochanska, 1990). Moreover, children show increasing emotion during conﬂict with mothers and siblings during the second year of life, and they show increasing understanding of their family members and of ways to comfort their siblings (Dunn & Munn, 1985). These changes enable children to better coordinate their own activities with those of other family members. At the same time, parents also develop and change in many ways, developments that are, themselves, important to the changing nature of family inﬂuence processes. For example, parents adjust their parenting and disciplinary styles to match their children’s development (Kuczynski et al., 1987). That is, concurrent with changes in children’s interactions with family members, parents respond to their children’s cognitive development by using increasingly verbal instructions and explanations in place of physical demonstrations, and by making more sophisticated verbal responses to their children’s requests (Maccoby, 1984). With maturity, children are more likely to be inﬂuenced by their parents’ petitions to their sense of fairness, and their parents respond to this change by decreasing their emphasis on reward and punishment. Furthermore, older children’s greater understanding of mutual obligations means that, as children get older, their parents are more effectively able to discipline by revoking their children’s privileges. Older children can also be inﬂuenced by their parents’ emphasis on what other people will think of their behavior. Although we have discussed development here primarily in terms of the parent–child relationship, the same principles apply to other family relationships. Moreover, parents also develop as individuals, independent of their development as parents; that is, their development as adults, outside the realm of the family, likely also contributes to the dynamics of family inﬂuence processes (Sarah J. Schoppe-Sullivan, personal communication, July 26, 2007). Thus, the interacting effects of all family members’ development contribute in important ways to family inﬂuence processes.

III. Transactional Family Dynamics: An Emerging Theme A. WHY IS A TRANSACTIONAL FAMILY DYNAMICS MODEL NEEDED? A transactional family dynamics model addresses a gap in conceptualizing family inﬂuence processes, consistent with the complexity of families. That is, we propose a model of multiple family members and family relationships nested within families, connected via multiple pathways of

202

Alice C. Schermerhorn and E. Mark Cummings

inﬂuence. Moreover, these inﬂuence processes unfold over the course of multiple, nested time scales, each of which contributes uniquely to development. Lastly, inﬂuence and change can be conceptualized in terms of association and contingency of family members’ behavior, observationto-observation change, and change in the overall pattern of inﬂuence. The transactional family dynamics approach can help inform and increase the accuracy of conceptualizations of other domains of family research. That is, in order to develop a fuller understanding of family processes, it is important to test hypotheses in ways that are as precise as possible, including framing investigations to capture the complexity of families. Failure to account for these inﬂuence processes may create distortions throughout the research process, from selecting research methods to conducting data analyses (misspeciﬁcation of statistical models) to interpreting results. Although the issue of bidirectionality is commonly acknowledged at a conceptual level, even when the data necessary to assess bidirectional effects are available, these processes are often overlooked in statistical analyses. Moreover, a theoretical framework is needed to unify the work of many different investigators and to provide a framework addressing questions about how all of this work ﬁts together. Together, the research and theories reviewed support a theoretical perspective that is a useful model for examining how family members and family relationships inﬂuence one another over time. In a subsequent section, we show how existing work ﬁts within this framework.

B. WHAT KIND OF APPROACH IS NEEDED? A model is needed that emphasizes the hierarchical organization of family members (family members are nested within family relationships, which are nested within families) and of time scales (shorter time scales are nested within longer time scales). Moreover, our model facilitates distinguishing between inﬂuence in the form of association and contingency, change from one time point to the next, and the overall pattern of change, as well as examining pathways of inﬂuence between multiple family dyads and triads. That is, it is important to account for the complex pathways between multiple family relationships, including the circular directions of inﬂuence that underlie transactional processes, each of which inﬂuences other family relationships at the same time that they, themselves, are changing. Researchers and theorists such as Bell, Bronfenbrenner, Hinde, and Lytton were already pointing to complex research questions such as these in

Transactional Family Dynamics

203

the 1980s, but did not yet have the methodological or statistical tools to actually answer them. For example, Lytton (1982, p. 273) wrote extensively about practical problems in studying family process, including the presence of multiple family members, limitations of statistical software, consideration of the broader ecology within which the family exists, and concluded The state-of-the-art analysis in this area prevents us from looking simultaneously at the interaction of all these systems over time in any rigorous fashion. We will, I think, have to conﬁne ourselves to investigating one or two aspects at one attempt y While this may seem a piecemeal approach, it enables us to make progress in testing different theoretical models by methods of manageable complexity.

Since that time, advances in methodological and statistical tools have given rise to the possibility for substantial advances to be made now in the study of family inﬂuence processes. Moreover, the work of various researchers reﬂects considerable progress in studying aspects of family inﬂuence processes. Their work has great utility for developing integrated models of family inﬂuence, increasing the accuracy with which they reﬂect family life and family functioning. Thus, we are now ready to return to the complex questions about family relationships that have been laid out for us—a legacy we have inherited from these leaders in the ﬁeld. C. THE HIERARCHICALLY ORGANIZED SYSTEMS OF TRANSACTIONAL FAMILY DYNAMICS At the core of our model of transactional family dynamics is the notion that behavior has multiple, hierarchically ordered causes, with events at lower levels of the hierarchy inﬂuencing, and being inﬂuenced by, events at higher levels of the hierarchy. For example, family inﬂuence processes begin with the actions of a single family member. At the next level of the hierarchy, family inﬂuence processes involve multiple family members, who are nested within dyads, acting and interacting with one another. In turn, multiple family dyads and triads, which are nested within families, inﬂuence one another via multiple pathways (see Figure 1a). Moreover, time is hierarchically organized, with real time processes nested within increasingly longer time scales. An additional factor creating complexity for the study of family inﬂuence processes is the multiple conceptualizations of change and inﬂuence. To impose some order on this complexity, we classify inﬂuence processes in terms of three different conceptualizations of inﬂuence: Systems A, B, and C. System A focuses on associations between two family members (or between two family relationships) reﬂecting mutual inﬂuence of both people

204

Alice C. Schermerhorn and E. Mark Cummings

on each other, without a change in either one. System B involves change processes within families from one time point to the next, again in terms of family inﬂuences. In particular, we consider how change unfolds over different time scales. System C focuses on change in the rate of change, that is, the overall pattern of change. These systems apply equally well to the various levels of the family hierarchy, pathways of family inﬂuence, and time scales. There is no great signiﬁcance to the names we have assigned to these three systems; they are simply used as a short-hand to refer to the corresponding conceptualizations of change and inﬂuence. However, the research questions that map onto one system are qualitatively different from those that map onto another system. Figure 2 depicts the relations between Systems A, B, and C, as well as the components of each. System A shows interconnected characteristics, behavior, and developmental status, symbolizing their inﬂuence on one another. System B reﬂects change from one time point to the next, and refers to different time scales. System C depicts the dynamics of the system as a whole, escalating and de-escalating over time, as symbolized by the dashed, wavy arrows. Several terms from dynamic systems theory are also illustrated in System C; they are explained in a later section. As shown in the ﬁgure, Systems A and B are subsumed within System C and contribute much to it. The portion of the ﬁgure that illustrates Systems A and B represents a cross-section of System C, a snapshot in time of the overall pattern of change. Notably, arrows symbolize inﬂuence of processes in different time scales on one another, as well as inﬂuence of different systems on one another. Circles are used to symbolize the continuity of inﬂuence processes and to emphasize the process, as opposed to an endpoint or outcome. Although the systems can be distinguished from one another in terms of their statistical meanings, more useful for the present purpose is that the systems are also distinguished from one another conceptually, by virtue of their relationship to change. That is, research questions can be classiﬁed in terms of the way inﬂuence and change are conceptualized, and these classiﬁcations map directly onto the systems we outline. In System A, research questions focus on inﬂuence and association in the absence of change. We can think of this as the level or amount of inﬂuence at one time point. Thus, the focus is one family member’s inﬂuence on another; that is, one family member elicits a particular response from another, but that response does not necessarily reﬂect a change from an earlier point in time. In System B, inﬂuence and change are conceptualized in terms of change from one time point to the next. Reﬂecting a different, more topographical perspective—a bird’s eye view—System C deﬁnes inﬂuence and change in terms of overall patterns of inﬂuence and change; that is, the rhythms of a relationship.

Attractor 1

Phase shift

Duration 2

Duration 1

Fig. 2. Conceptual figure depicting relationships between Systems A, B, and C.

System C

Attractor 2

Duration 3

Developmental Status

Characteristics Behavior

System A

System B

Duration 4

Duration 5

Transactional Family Dynamics 205

206

Alice C. Schermerhorn and E. Mark Cummings

These conceptual distinctions reﬂect qualitative differences. The systems, thus, give rise to distinct research questions, and therefore offer unique ways of thinking about family inﬂuence processes. Together, these three systems allow us to address the complexity of family inﬂuence processes along multiple dimensions previously outlined, including nested time scales, the hierarchical organization of family relations, differing notions of inﬂuence and change (which correspond to the systems), and pathways of inﬂuence among multiple family dyads and triads.

1. System A System A pertains to family inﬂuence processes without change—that is, the amount of inﬂuence in a relationship at any point in time. Thus, we are interested in qualities of individual family members, such as personal characteristics and values, developmental status, and behavior as they inﬂuence the behavior of other family members. Examples of personal characteristics and values include age, pubertal status, physical size, gender, temperament, adjustment problems, political views, moral values, and other beliefs. System A does not directly involve change processes, but it does involve qualities that inﬂuence family members, either at one moment in time or over the course of time—that is, inﬂuence without change in inﬂuence. For example, a father might encourage his son to pursue sports and his daughter to play the piano; in this example, the father is treating his children differently because of their gender. Notice that this example does not involve change in the amount of inﬂuence, but rather, it reﬂects the inﬂuence of a personal characteristic in eliciting the father’s behavior. The father’s behavior is caused, at least in part, by personal characteristics of his children. As another example, a mother may allow her son to stay out late than her daughter, because her son is reserved and cautious, whereas her daughter is impulsive. Thus, one family member’s behavior may be contingent on characteristics of another family member. Importantly, although this system does not reﬂect change, it can reﬂect causal processes, as the above examples illustrate (notably, an experimental design is needed to demonstrate causation), as well as the inﬂuence of family members on one another (e.g., granting requests, reciprocal responding). Much of the empirical literature on family inﬂuence processes involves these sorts of response processes (Table II). As outlined brieﬂy already, many individual characteristics are important to transactional family dynamics. Child behavior problems are one example (Jenkins et al., 2005b). Moreover, temperament can contribute to inﬂuence processes; for example, a baby who is cheerful and outgoing probably makes her parents feel happy.

Table II Transactional Family Dynamics: Empirical Contributions Study

Sample (age) and design

Transactional results

Cross-sectional; 15-min observations of each mother with (a) her own child, (b) a child without CD, and (c) a child with CD

A

Compared with nonconduct disordered (NCD) children, conduct disordered (CD) children elicited more requests and negative responses from both mothers of CD and mothers of NCD children.

Cross-sectional; questionnaire, observational, vocabulary test

A

Child-speciﬁc differential parenting was linked with concurrent adjustment problems for all siblings in 2 of the 3 datasets.

3 waves, each spaced 1 year apart; questionnaire data; used structural equation model (SEM) with autoregressive controls for some constructs

B

Maternal monitoring predicted later child externalizing, but child externalizing did not predict later monitoring. Child temperament predicted later parent–child relationship quality, which was linked with concurrent monitoring.

207

System

Transactional Family Dynamics

Transactional dynamics of the parent–child relationship Anderson, Lytton, & 32 mother–child dyads; Romney (1986) half of the children met Diagnostic and Statistical Manual (DSM) III criteria for Conduct Disorder (CD), half of the children did not have disorders Boyle et al. (2004) 3 datasets, consisting of the following: 2,128 4–16-year-olds (894 families); 7,392 4–11-year-olds (3,376 families); 2,876 3–14-year-olds (1,218 families); mothers and teachers Brody (2003) 156 African American children in singleparent households, as well as their mothers and teachers; average child age at Time 1=11 years

Methodologies

208

Table II (Continued ) Sample (age) and design

Methodologies

System

Transactional results

Brophy & Dunn (2002)

30 families of ‘‘hard to manage’’ children (conduct disordered and/or hyperactive) and 26 families of control children; child ages were 3.6–4.6 years at Time 1

2 waves, spaced 18 months apart; intelligence testing, observational

A (tests of links from Time 1 to Time 2 were not reported)

Brunk & Henggeler (1984)

32 mothers; ranging in age from 25 to 48 years

Cross-sectional, experimental design: 2 10-year-old male confederates displayed either anxiouswithdrawn (AW) or conduct-problem (CP) behavior

A

Clark et al. (2000)

108 mothers and their infants (infants averaged 9 months of age at Time 1)

2 waves spaced 4–6 months apart; observational, and questionnaire data

A

Compared with control children, in mother–child dyads with hard to manage children: (a) at Time 1, mother–child conversational turn-taking was lower and maternal negative control was higher; (b) at Time 2, mothers showed more negative control and less positive control. In the AW condition, participants showed more rewarding and helping behavior than in the CP condition; in the CP condition, participants showed more discipline, commands, and ignoring than in the AW condition. Child negative emotionality predicted subsequent maternal power assertion, but not maternal responsiveness; child emotionality interacted with maternal personality in links with subsequent maternal power assertion.

Alice C. Schermerhorn and E. Mark Cummings

Study

85 mother–child pairs; children were preschoolers at Time 1

2 waves, spaced approximately 2 years apart; questionnaire, observational data

B

Cook & Kenny (2005)

203 mother–adolescent dyads

B

Covell & Abramovitch (1987)

123 children (ages 5–15 years) and 54 of their parents

2 waves spaced 1 year apart; questionnaire data; used autoregressive controls Cross-sectional; structured story-based interview

Covell & Miles (1992)

Study 1: 120 4–9-year-old children; Study 2: 180 4–12-year-old children and their parents

Cross-sectional; structured story-based interview

Eisenberg et al. (1999)

94 children (preschool/kindergarten-age at Time 1) and their mothers and fathers

5 waves, spaced 6 months to 2 years apart; questionnaire data; used SEM with autoregressive controls for all variables.

B (because the data reﬂect participants’ beliefs about change) A

B

Within an interaction, mothers reciprocated their daughters’ positive emotional expressions more than they reciprocated negative emotional expressions. Mothers’ reciprocation of children’s expression of anger predicted increases in child externalizing problems. Associations between mother and child attachment security were bidirectional. Children reported that they could change maternal mood; their mothers conﬁrmed their perception.

209

Children indicated they would be more likely to directly intervene when they caused parental anger than when work-related problems or interparental conﬂict caused parental anger. Reciprocal links were found between some (but not all) time points for: (a) child anger, hostility, irritation and parental punitive and distressed reactions; and (b) child regulation and parental punitive (but not distressed) reactions.

Transactional Family Dynamics

Cole et al. (2003)

210

Table II (Continued ) Sample (age) and design

Methodologies

System

Transactional results

Granic et al. (2003)

149 children (9–10 years old at Time 1) and their parents

C

Behavioral variability during family interactions peaked at 13–14 years of age.

Granic & Lamey (2002)

36 8–12-year-old boys with clinical levels of externalizing symptoms and their mothers

5 waves, spaced 2 years apart; observational data from a problemsolving task; state space grids Time-series (10-min problem-solving task); questionnaire and observational; state space grids

C

Grundy (2007)

133 mothers and their ﬁrstborn children, who were 4th graders at Time 1

5 waves spaced 1 year apart; questionnaire data

B

Grusec & Kuczynski (1980)

20 mothers of 4–5-yearolds; 20 mothers of 6.5–8-year-olds

Cross-sectional; structured observational task

A

The mother–child interactions of children with externalizing symptoms were characterized by a permissive pattern, whereas the interactions of children with both externalizing and internalizing symptoms changed from a permissive pattern to one that was mutually hostile, following a cue to resolve a conﬂict. Maternal monitoring-related knowledge and preadolescent children’s competent behavior predicted increases in one another over time. Maternal response to scenario involving child misbehavior depended partly on type of misbehavior.

Alice C. Schermerhorn and E. Mark Cummings

Study

48 parents of at least 1 4–7-year-old

Cross-sectional; structured interviews in home or lab

B (because the data reﬂect participants’ discourse about change)

Hollenstein et al. (2004)

240 kindergarten children and their parents; families were from a low-income neighborhood 494 adolescent girls

2 waves during kindergarten; timeseries parent–child observation, questionnaire data 2 waves, spaced 1 year apart; questionnaire, structured interview, and physical growth data; used autoregressive controls for all constructs

C

1,077 14-year-old children and their parents

2 waves, spaced 2 years apart; questionnaire data

Huh et al. (2006)

Kerr & Stattin (2003a)

B

B

Parents reported that the parent– child relationship is most strengthened by their efforts at companionship and by their child’s compliance and efforts at companionship; the parent– child relationship is most weakened by parental overuse of authority and child noncompliance. Child internalizing and externalizing problems were linked with rigidity in parent– child interactions. Adolescent externalizing symptoms predicted subsequent decreases in parental support and control; adolescent substance abuse and parental control predicted subsequent decreases in one another. Child delinquency predicted subsequent parenting, but ﬁndings regarding the reverse direction of effects were mixed, with some evidence that delinquency predicts decreases in parental monitoring.

Transactional Family Dynamics

Harach & Kuczynski (2005)

211

Table II (Continued ) Methodologies

System

Transactional results

Kochanska & Kuczynski (1991)

50 5-year-old children and their mothers, 26 of whom had histories of depression and 24 of whom did not

Cross-sectional, timeseries parent–child observation

A

Lewis et al. (1999)

8 mother–infant dyads

C

Lytton (1979)

136 boys (age 25–35 months), and their mothers and fathers (when available)

2 3-week waves (at infant ages 10–12 weeks and 26–28 weeks); observations of separations and reunions; used state space grids Cross-sectional; timeseries data from home observations

Depressed mothers’ autonomygranting was dependent on both child cooperation and maternal mood; moreover, mothers reciprocated their children’s behavior more than children reciprocated their mothers’ behavior. Examined distress intensity and attention to mother; found that baseline levels had substantial stability and inﬂuence on behavior.

Lytton (1982)

136 boys (92 twins, 44 singletons; age range 25–35 months) and their parents

3 waves (Time 2 was 1 week after Time 1; Time 3 was approximately 7 years after Time 1); in-home interviews and observations; used SEM with autoregressive controls for some constructs

A (although the data were sequential, the study examined contingency and association, not change) B

Children complied with their mothers more when their fathers were present. Parents most frequently exhibited no response to either child compliance or noncompliance, but fathers responded signiﬁcantly less than mothers to child compliance. Child effects on attachment were more prominent in the shortterm, but parent effects on compliance were more prominent in the short-, medium-, and long-terms; parents treated their children differently as a function of (largely genetically based) differences in child behavior.

Alice C. Schermerhorn and E. Mark Cummings

Sample (age) and design

212

Study

94 mothers of 3–6-yearold children

Cross-sectional; questionnaires, home interviews, and responses to vignettes

A

Masche et al. (2006)

Sample 1: 1,339 adolescents (average Time 1 age=13.7); Sample 2: 1,343 adolescents (average Time 1 age=15.8) and their parents

4 waves, spaced 1 year apart; questionnaire; used autoregressive controls for all constructs

B

Shearer et al. (2005)

170 mothers and 159 fathers (average age of their ﬁrstborn child was 16.3 years at Time 2)

2 waves, spaced 4 years apart; questionnaire and interview

B

Steinberg (1981)

27 mother–father–son triads; sons were 11–14 years old at Time 1

3 waves, spaced 6 months apart; questionnaire and observational; computed change scores for each dependent variable

B

213

Maternal emotion regulation varied as a function of child emotion type (sadness, anger, fear), but not as a function of child temperament. Adolescents’ externalizing behavior (delinquency, problems at home) predicted less parental behavioral control and positive behavior and more aversive parental behavior, which predicted increases in adolescent externalizing behavior. Maternal ratings of mother– adolescent relations were most positive following increases in parent–child acceptance and decreases in conﬂict. Parents reported granting their children more autonomy as a function of their child’s development. During the period of adolescence prior to the pubertal apex, sons and mothers interrupted each other more and explained themselves less, and sons deferred to their mothers less; after the pubertal apex, mothers interrupted their sons less. Across ages 11–14, fathers’ interruptions of their sons increased and sons showed more deference toward their fathers.

Transactional Family Dynamics

Martini et al. (2004)

214

Table II (Continued ) Sample (age) and design

Methodologies

System

Transactional results

Stice & Barrera (1995)

441 10–16-year-olds (half of whom were children of alcoholics) and their parents

2 waves, spaced 1 year apart; interview; used autoregressive controls for all constructs

B

Trickett & Kuczynski (1986)

40 families with a child age 4–11 years, half of whom had at least 1 abusive parent

Cross-sectional; daily reports (5 days) of child misbehavior and parental responding

A

Tucker et al. (2003)

188 families with 2 adolescent siblings (ﬁrstborns’ average age was 15 years; secondborns’ average age was 12.5 years) 1,222 1–36-month-old children and their mothers

Cross-sectional; home interviews of each family member

A

Adolescent externalizing problems and substance use predicted decreases in parental control and support, whereas parental control and support predicted decreases in substance use, but not externalizing. Abusive parents most frequently responded to child misbehavior with punishment, whereas for nonabusive parents, choice of discipline depended on misbehavior type. Parents treat their children differently as a function of child personal characteristics.

5 waves (ages 1-, 6-, 15-, 24-, and 36-months) home interviews and questionnaires; used autoregressive controls for all constructs

B

Warren et al. (2006)

Whereas maternal depression predicts increases in the duration of infant awakening, longer durations of infant awakening predict decreases in maternal depression.

Alice C. Schermerhorn and E. Mark Cummings

Study

Frosch et al. (2000)

53 families with a preschool-age child (average age=3 years)

Howes & Markman (1989)

20 families with children ages 1–3 years

Owen & Cox (1997)

38 families (mothers, fathers, and infants)

2 waves, spaced 2.5 years apart; Q-sort, observational, and questionnaire data 2 waves: premarriage and post-birth (3–5 years later); questionnaires, Q-sort 3 waves (prenatal, 3 months of age, 1 year of age); observational, interview, strange situation

B

A

A

A

A

A

Child positive affect predicted increases in supportive coparenting, but coparenting did not predict change in positive affect. For families with low marital adjustment, children of secure mothers were more secure than children of insecure mothers. Marital functioning and the mother–child relationship both predicted one another, and maternal personality predicted functioning in both relationships. Links between concurrent and subsequent marital functioning, parent–child security, and parenting. Premarital assessment of interparental relationship quality was correlated with mother–child security at 1–3 years of age. Marital conﬂict predicted insecurity in mother- and father–child relationships.

Transactional Family Dynamics

Transactional dynamics between parent–child and interparental relationships Davis (2007) 59 families with a 2 waves, spaced 1 year preschool child apart; mother–father– child interaction; used autoregressive controls for all constructs Eiden et al. (1995) 45 mothers and their Cross-sectional; mother– 16–62-month-old child interaction, children Attachment Q-set, Adult Attachment Interview, questionnaire Engfer (1988) 36 families with a 5 waves (at child birth, 4-month-old baby and at 4, 8, 18, and 43 month of age); questionnaire and observational

215

Table II (Continued ) Methodologies

System

Transactional results

Schermerhorn et al. (2008)

232 children (kindergarten-age at Time 1), their mothers, and fathers

B

Found links between children’s emotional security about marital conﬂict, mother–child, and father–child relationships.

Schoppe et al. (2001)

57 families with a 3-yearold at Time 1

3 waves, each spaced 1 year apart; story-stem completion task; used SEM with autoregressive controls for each construct 2 waves, spaced 1 year apart; triadic (mother– father–child) observation and questionnaires

A

Schoppe-Sullivan et al. (2004)

46 families with a 6-month-old at Time 1

B

Schoppe-Sullivan et al. (2007)

283 families with a child age 8–16 years at Time 1

2 waves, spaced approximately 2.5 years apart; dyadic (parent–child) and triadic (mother–father– child) observations, questionnaires; used autoregressive controls for all constructs 3 waves, spaced 1 year apart; questionnaire data; used autoregressive controls for some constructs

For families with low positive affect, supportive coparenting predicted low child externalizing problems, whereas undermining coparenting predicted more externalizing in families with more negative affect. Coparenting in early childhood predicts subsequent marital behavior, but marital behavior during early childhood does not predict subsequent coparenting.

B

Parenting (behavioral control, autonomy-granting, and warmth) mediated relations between marital conﬂict and child adjustment; with autoregressive controls of child adjustment, behavioral control continued to mediate relations between marital conﬂict and children’s internalizing symptoms.

Alice C. Schermerhorn and E. Mark Cummings

Sample (age) and design

216

Study

Talbot & McHale (2004)

50 families with a 12-month-old child

Cross-sectional; questionnaire, observational (including triadic interaction)

Marital quality and aspects of parenting predicted concurrent coparenting harmony and negativity.

A

Destructive marital conﬂict tactics were linked with children’s insecure emotional responding; constructive marital conﬂict tactics were linked with children’s secure emotional responding. Diary reports of destructive marital conﬂict tactics and negative marital emotions were associated with children’s aggressive responding to analog depictions of marital conﬂict; children’s aggressive responses to marital conﬂict were linked with externalizing problems. Parents’ negative emotions and destructive conﬂict tactics were associated with children’s insecure responding; parents’ positive emotions and constructive conﬂict tactics were associated with children’s secure responding.

Transactional dynamics of interparental relationships and children Cummings et al. (2003) 116 families with a child Cross-sectional; aged 8–16 years questionnaires, home diary reports completed on each of 15 days; analyses included cross-reporter tests Cummings et al. (2004) 108 families with a child Cross-sectional; aged 8–16 years questionnaires, home diary reports completed on each of 15 days, and children’s responses to videotaped depictions of marital conﬂict episodes

A

Cummings et al. (2002)

A

51 couples with a child aged 4–11 years

Cross-sectional; home diary reports completed on each of 6 days

Transactional Family Dynamics

A

217

218

Table II (Continued ) Sample (age) and design

Methodologies

System

Transactional results

Cummings et al. (2006)

Study 1: 226 children, ages 9–18 and their parents; Study 2: 232 children, ages 5–7 and their parents

B

Children’s emotional insecurity about the marital relationship linked destructive marital conﬂict and increases in children’s internalizing and externalizing problems.

Davies et al. (2002)

285 children, ages 11–13, and their parents

Study 1: 2 waves, spaced 2 years apart, used questionnaire data; Study 2: 3 waves, each spaced 1 year apart, used questionnaire and observational data; SEM was used in both studies, with autoregressive controls in Study 2. 2 waves spaced 2 years apart; questionnaire; used SEM with autoregressive controls

B

Jenkins et al. (2005)

296 children from 127 families; average child age was 5 years at Time 1; siblings ranged in age from 6 to 17 years; parents and teachers; sample included stepfamilies

Children’s emotional insecurity about the marital relationship linked destructive marital conﬂict with increases in children’s internalizing and externalizing problems. Child externalizing problems predicted increases in marital conﬂict.

2 waves, spaced 2 years apart; questionnaire; used autoregressive controls for marital conﬂict

B

Alice C. Schermerhorn and E. Mark Cummings

Study

113 married or cohabiting couples, with a child in the age range of 8–16 years (average age=11)

Cross-sectional; home diary reports completed on each of 15 days, questionnaire; used dynamic systems modeling

C

Schermerhorn et al. (2005)

115 children (kindergarten-age at Time 1), their mothers, and fathers

B

Schermerhorn et al. (2007)

232 couples with a kindergarten-age child at Time 1

2 waves, spaced 1 year apart; story-stem completion task; used SEM with autoregressive control for marital conﬂict 3 waves, each spaced 1 year apart; questionnaire and observational data; used SEM with autoregressive controls for marital conﬂict

4-wave design, 1 year between waves; home interviews; autoregressive controls were included for some constructs

B

Longitudinal links between maternal psychological functioning and parenting, younger and older sibling competence, and younger sibling self-regulation.

219

Family-wide transactional dynamics Brody et al. (2003) 152 single-mother families with at least 2 children and the children’s teachers; average child ages were approximately 12 years for ﬁrstborns and 9 years for second-borns

B

Child agentic behavior predicted less destructive, more constructive, more resolved marital conﬂict; child negativity predicted more destructive, less constructive, more unresolved marital conﬂict; from one marital conﬂict to the next, husbands’ behavior changes rapidly; wives’ conﬂict resolution slows down this rapid change in their husbands’ behavior. Destructive marital conﬂict predicted more child negative emotionality, which predicted more child perceived agency; perceived agency predicted increases in marital conﬂict. Destructive marital conﬂict predicted more negative emotionality; emotionality predicted more subsequent agentic behavior and behavioral dysregulation, which predicted change in marital conﬂict.

Transactional Family Dynamics

Schermerhorn et al. (2007)

220

Table II

Study

Sample (age) and design

Methodologies

System

Transactional results

Bumpus et al. (2001)

194 families with an 8th– 10th grade ﬁrstborn child and at least 1 child 1–4 years younger 3,681 sibling pairs, approximately 4–7 years of age

Cross-sectional; home interviews and nightly telephone calls (7 nights for children; 3 nights for parents) 2 waves (14 weeks before birth of second child and 4 years later); questionnaire

A

91 2-parent families with 2 or more children; at Time 1, older siblings were 4th/5th graders and younger siblings were Z6 years old 37 families with an 8–11month-old child at Time 1

4 waves, each spaced 6 months apart; homebased questionnaires, interviews

A

2 waves, spaced 3 years apart; observational and questionnaire data

A

Parental autonomy-granting varied as a function of child birth-order, gender, and girls’ menarcheal status, as well as maternal gender role attitudes. Signiﬁcant correlations were found between marital, parent–child, and sibling relationships in positivity and negativity. Both concurrently and 1 year subsequently, sibling warmth is linked with (a) children’s satisfaction in parent–child relationships; and (b) marital satisfaction and love. Marital functioning and family processes in infancy predicted child internalizing and externalizing problems 3 years later.

Dunn et al. (1999)

McGuire et al. (1996)

McHale & Rasmussen (1998)

A

Alice C. Schermerhorn and E. Mark Cummings

(Continued )

Rasbash et al. (2007)

Vuchinich et al. (1988)

52 families with 1–6 children, aged 2–22 years

Cross-sectional; observations of positive and negative emotional expression in each family dyad (sibling, marital, mother–child, father– child); used autoregressive controls for all constructs

B

Greater reciprocity was present in sibling and in marital dyads than in parent–child dyads; genetic effects were responsible for substantial variability at the individual level; and positivity emerged as an individual-speciﬁc variable, whereas negativity emerged as a relationshipspeciﬁc variable.

Cross-sectional; questionnaire

A

Cross-sectional; observations of 39 dinners with 2 parents and 1–6 children ages 3–22 years, 17 dinners with mother and 1–3 children ages 2–6 years

A

Correlations were found between (a) parent–child and sibling conﬂict; (b) parent–child and marital conﬂict; (c) marital and sibling conﬂict. During conﬂict within a family dyad, a third family member intervenes 1/3 of the time, with preferred strategies including mediation (parents) and distraction (children).

Transactional Family Dynamics

Rinaldi & Howe (2003)

657 families with both parents and 2 children (10–18 years old, not more than 4 years apart); sibling pairs in nondivorced families were: 92 monozygotic twin, 94 dizygotic twin, and 90 full (nontwin), and in stepfamilies were: 171 full, 104 half, and 124 unrelated 60 families with a 5th/6th grade child and the child’s closest-in-age sibling

221

222

Alice C. Schermerhorn and E. Mark Cummings

Having discussed System A primarily in terms of individual family members, we now note that System A concepts also apply to family relationships and to families as a whole. That is, families and family relationships have characteristics, values, and developmental status of their own. For example, family relationships can be described as supportive or as having an even distribution of power; whole families can be described as warm, emotionally expressive, secure, and open. Thus, the explanations and description we have provided for System A apply not only at the level of individual family members, but also at the level of family relationships and families as a whole.

2. System B System B revolves around change from one time point to the next. Change unfolds over nested time scales, with events occurring on shorter time scales inﬂuencing events occurring on longer time scales of years. Consistent with that, dynamic systems theory emphasizes this nested, interdependent nature of time, and we refer to this unfolding process in System B. Individuals’ behavior during family interactions, on a time scale of seconds, are nested within family inﬂuence processes that are on much longer time scales. Thus, one important way in which transactional family dynamics are hierarchically organized is in terms of the nested time scales in which they unfold. Thus, System B involves these time-linked inﬂuence processes. Notably, as with System A, the concept of System B applies, not only at the level of individual family members, but also at the level of family relationships and whole families. Several researchers have distinguished between micro processes unfolding over short periods of time, and macro processes unfolding over long periods of time (Bandura, 2001; Lewis, 2002; Lytton, 1982; Patterson, 1997). There have also been calls to examine relations between different time scales (Granic & Patterson, 2006; Lewis, 2002; Thelen, 1995; van Gelder & Port, 1995). Highlighting the unique contributions of different time scales to development, Smith (2005) emphasized the contributions of ‘‘real time’’ processes to developmental change, and Lytton argued that cause–effect relations might differ as a function of timeframe. In the context of inﬂuence processes, we consider what might transpire over different time scales, and we suggest issues to consider in terms of the inﬂuence of various time scales on each other. Figure 3 illustrates these notions, applied to the arena of children’s inﬂuence on marital conﬂict. We use labels to refer to processes that unfold over different time scales (Durations 1–5; see Figure 1b). As with the systems, the labels we apply to the time scales are arbitrary and are used only to simplify our discussion. Notably, we deﬁne episodes as any

Day 2

Beginning of Year 1

Day 3

… Day 731

Beginning of Year 3

Effects of marital conflict, child involvement in conflict, and child adjustment at Year 1 on those same constructs at Year 3 (Duration 5)

Fig. 3. Transactional influence between marital conflict and children across different time scales.

Effects of marital conflict and child involvement in conflict and adjustment across timescales

Child effects on conflict resolution within conflicts (Duration 1 and 2)

Day 1

Child effects on conflict resolution between conflicts (Duration 3 and 4)

Transactional Family Dynamics 223

224

Alice C. Schermerhorn and E. Mark Cummings

interactions between family members, including disagreements, casual conversations, and nonverbal communications. The label ‘‘Duration 1’’ refers to the most intense exchanges during episodes between family members, for example, the most heated period of an interparental disagreement. These exchanges occur in real time—over seconds or minutes. As an example of Duration 1 processes, if one family member speaks defensively, cries, or is silent, that likely inﬂuences the behavior of another family member at the next moment. Thus, one important inﬂuence process involves the most intense exchange between family members—the epicenter of an episode. Duration 2 consists of inﬂuence processes spanning the entire episode, not just the most intense portion of the episode. Thus, Duration 2 includes Duration 1 and the rest of the episode as well. That is, this time scale involves all of the inﬂuence processes that occur throughout an episode. Duration 3 includes the proximal events that lead up to an episode, as well as the period during which a family member is actively thinking about an interaction or episode, but the episode has come to an end (although feelings stemming from the interaction might remain intense). Duration 4 is the time frame during which a family member periodically recalls aspects of an interaction or episode, in between periods of thinking about other things. Duration 5 is the period during which a family member is inﬂuenced at a broader, more global level by the episode or interaction. For example, a father–child disagreement over an adolescent’s breaking curfew could create lasting ‘‘coolness’’ between parent and child. If the offense is repeated, the parents’ response is likely to solidify, resulting in further punishment and further distrust, this time sustained over a longer period of time. We think that examining multiple time scales offers the possibility for developing new insights about inﬂuence processes. The hierarchical organization of time scales is depicted in Figure 1b; ‘‘A’’ and ‘‘B’’ are used to denote different individual interactions or episodes; for example, ‘‘Duration 1A’’ and ‘‘Duration 2A’’ refer to the ﬁrst two Durations of Episode A, whereas ‘‘Duration 1A’’ and ‘‘Duration 1B’’ refer to the ﬁrst Durations of Episode A and Episode B. We would suggest that the contributions of these time scales to development cannot be disentangled. Arrows indicate the ﬂow from real time to long-term processes. For example, processes unfolding during the period of the most intense exchange (Duration 1) inﬂuence the rest of the episode (Duration 2), which is largely limited by much longer-term processes (Duration 5). Furthermore, we propose that the most intense exchange during one interaction inﬂuences the most intense exchange during the next interaction (e.g., the inﬂuence of Duration 1A on Duration 1B). For example, a child’s tearfulness during a marital disagreement might prompt her parents to

Transactional Family Dynamics

225

handle their next disagreement behind closed doors. Notably, this process is not limited to temporally adjacent episodes, but includes episodes separated by other, intervening episodes. We also suggest cross-inﬂuence among the time scales (e.g., the inﬂuence of Duration 1A on Duration 4B). For example, a family member may tend to remember a heated exchange for several days. Moreover, Duration 4 processes might also feed back to inﬂuence both Duration 1 and Duration 5. For example, thinking for several days about an earlier disagreement might provoke further episodes of intense conﬂict and might lead to more negative attributions about the relationship over time. Moreover, the most intense exchange might reﬂect the inﬂuence of a speciﬁc aspect of a previous interaction on a speciﬁc aspect of a subsequent interaction. For example, the wife’s raising tangential grievances during one conﬂict episode (Duration 2A) may prompt the husband to do the same the next time (Duration 2B). (For simplicity of presentation, in Figure 1b we included only a small subset of all of the possible arrows depicting inﬂuence processes.) Thus, one family member’s behavior in previous interactions may inﬂuence other family members’ expectations of future behavior. These expectations, in turn, inﬂuence family members’ behavior in subsequent interactions (Bowlby, 1973). Thus, various time scales interact and produce further development. In addition to the inﬂuence of various time scales on one another, we propose qualitative differences between processes unfolding at different time scales, and that different time scales have unique implications for family inﬂuence processes. In particular, we suggest that real time processes involve speciﬁc, concrete behaviors, whereas longer-term processes reﬂect more global processes, such as ﬁrmly held beliefs. Another possibility may be that more emotion is elicited at shorter time scales (Durations 1 and 2) than at longer ones (especially Duration 5); that is, emotion may arise more frequently during interactions in real time than in developmental time. Moreover, inﬂuence processes during shorter time scales may reﬂect more automated cognition, and inﬂuence processes during longer time scales may reﬂect more controlled cognition (see Klaczynski & Daniel’s (2005) description of experiential and analytic reasoning systems), especially reﬂecting the development of attributions about a family member or relationship. However, this notion is purely speculative, as we know of only one study that has investigated this issue. Lytton (1982) found that parental discipline may appear as a response to child behavior in the short-term (Duration 1), but in the long-term (Duration 5), it may be inﬂuenced more by parental values and beliefs, both of which are integrated into a parent’s approach to parenting. Notably, in terms of methodological considerations, observations are needed from a range of time scales—each is equally important to our gaining insight into the functioning of these processes.

226

Alice C. Schermerhorn and E. Mark Cummings

That is, we need measures from many time scales in order to advance current knowledge of how transactional family dynamics work. As noted by Cole and Maxwell (2003), a key issue for future research is to identify appropriate time frames over which various causal processes unfold; identifying such time frames remains an empirical question for many areas of family research. 3. System C System C involves change in the rate of change, or the overall pattern of change, in family inﬂuence processes. As with Systems A and B, System C applies to individual family members, family relationships, and whole families. System C is largely concerned with how relationships cycle. For example, a relationship may experience cycles of escalation, involving periods of time with more than the usual number of disagreements, followed by periods of relative cohesion and peace. Thus, this system involves the rhythms of a relationship, with ﬂuctuations in conﬂict and harmony. This section focuses on principles from dynamic systems theory and is somewhat technical. However, we provide deﬁnitions of terms and illustrate points with hypothetical examples pertaining to family inﬂuences processes; thus, we think that this material, although technical, is helpful in clarifying what we mean by System C, as well as the explanatory potential of looking at family inﬂuence processes through this lens. In the language of dynamic systems theory, attractor refers to the baseline level of the phenomenon, the natural tendency, or stable, recurrent patterns of interaction (see Figure 2). Thus, attractors can be conceptualized as a relationship’s natural behavioral tendency in terms of inﬂuence processes. In terms of transactional family dynamics, the attractor would be a family’s baseline family inﬂuence processes. Using an example from our previous work, the attractor would be the baseline level of marital conﬂict (whatever that is for a given couple), with a corresponding baseline level of child behavioral responding to marital conﬂict. Moreover, there may be different attractors for individual family members and for relationships. Drawing on dynamic systems principles, control variables are events that disturb a relationship, moving it away from the attractor state. Thus, an increase in the frequency and destructiveness of marital conﬂict might lead to an increase in a child’s efforts to resolve marital conﬂict, followed by a change in marital functioning, in turn resulting in a change in child behavior. Change in System C can be predicted by any of a large number of factors. Examples include normative developmental processes as previously described, major conﬂicts or other threats to relationship stability (e.g., an extra-marital affair), new friendships, work-related stress (distress, eustress), functioning of relationships with extended family.

Transactional Family Dynamics

227

Consistent with these notions, Granic (2000) described the parent–child relationship as self-organizing and discussed the roles of each person in guiding the other back toward baseline levels of behavior (both individual and dyadic baseline levels). For example, if a child is upset, a parent may try to soothe the child to get him back to his normal calm state. In terms of stability, family relationships typically have a baseline, or attractor, for inﬂuence processes speciﬁc to that relationship; relationships tend to return themselves to their baseline, a process known as self-stabilization or self-organization. Several other patterns of change are also relevant to family inﬂuence processes. Acceleration refers to an increase in the rate of change and deceleration refers to a decrease in the rate of change (Bisconti, Bergeman, & Boker, 2004). If the rate of change in a family inﬂuence process increases, the dyad will move back and forth past baseline more and more rapidly. For example, compared with the husband, the wife may have a more rapid increase in marital dissatisfaction in response to a major marital disagreement. Alternatively, reﬂecting deceleration, a spouse’s mood might change increasingly slowly over time. The amplitude, or level, of the behavior can also ﬂuctuate. An escalating pattern would involve increases in the amplitude of the behavior—for example, increases in the husband’s use of destructive marital conﬂict tactics over time. The level can also oscillate back and forth past the baseline. For example, in the aftermath of a marital disagreement, partners may feel very angry one moment, much less angry the next moment, and more angry again after that. Over time, these swings in the amount of anger may decrease, and partners may settle back out at their baseline level of anger. Such a pattern of change would reﬂect damping. Moreover, a damping of one partner’s anger might help bring about a damping of the other partner’s anger. Thelen and Ulrich (1991) provide a useful guide for conducting research that is consistent with dynamic systems principles. Their ﬁrst step is to identify what they refer to as the collective variable, or index of the change process. In terms of family inﬂuence processes, the collective variable is a behavior in real time, such as talking, ignoring, and misbehaving, that indexes the overall pattern of change in the inﬂuence process (e.g., escalating, damping). Development occurs when disturbances introduced by some aspect of the family relationship (e.g., cognitive development in one family member, action by another member of the family, marked physical growth) cause the relationship to shift from one attractor to another. This type of shift is development. In dynamic systems terms, this is referred to as a phase shift (see Figure 2); such periods represent changes in the overall pattern of inﬂuence. Because they involve change, phase shifts present opportunities to

228

Alice C. Schermerhorn and E. Mark Cummings

learn about the dynamics of the relationship. It is during these phase shifts that novel behavioral forms can emerge. Thus, new forms are not necessarily produced by the environment beyond the dyad, triad, or family, but rather, can emerge from within. Control variables are factors that are responsible for these changes; thus, an understanding of development requires identiﬁcation of the variables that lead to phase shifts in attractor states. Thus, an important task for researchers is to identify the points of transition where loss of stability occurs (i.e., points at which a relationship is unstable, and therefore, more open to change). In terms of transactional family dynamics, possible transition points include periods of growth in child cognitive development, adolescent pubertal development, and divorce. During these periods, there is the potential for substantial, sustained change in family inﬂuence processes. For example, one question may be whether children’s efforts to inﬂuence the marital relationship ﬂuctuate more when marital conﬂict increases. If so, then marital conﬂict is a potential control variable for that child behavior. The next step would be to manipulate the hypothesized control variables to test whether that produces a phase shift (Thelen & Ulrich, 1991). Intervention programs and experiments offer ethical ways of examining differences between experimental and control groups as a function of the hypothesized control variables. D. A COMPREHENSIVE MODEL OF TRANSACTIONAL FAMILY DYNAMICS Thus, transactional family dynamics provides a framework for organizing and integrating information about family inﬂuence processes. One hope is that the theoretical notions we describe here may serve as a catalyst prompting others to outline their own (possibly quite different) theoretical notions regarding the family inﬂuence processes we describe as transactional family dynamics. By bringing together the work of many scholars of transactional family dynamics, particularly those focused on different family relationships, different time scales, and different conceptualizations of inﬂuence and change, we hope the eventual result will be the development of a comprehensive theory of transactional family dynamics.

IV. Mapping Empirical Work onto a Transactional Family Dynamics Framework We now provide an overview of selected empirical work that is relevant to transactional family dynamics, demonstrating how this work ﬁts into our

Transactional Family Dynamics

229

framework (see Table II). We organize the studies according to the family relationships they address, in order of increasing complexity (see Figure 1a) and according to the systems they reﬂect (see Figure 2), and when possible, we provide at least one example of each system for each family relationship. Notably, relatively few groups have conducted work examining change in the rate of change (System C), perhaps because of the newness of the necessary statistical (dynamic systems modeling) and graphical (Gridware; Lamey et al., 2004) approaches. In the next sections of the chapter, we highlight a subset of these studies. Our goal in this section is to show how selected studies ﬁt within a transactional family dynamics framework, rather than to provide an exhaustive review of the evidence for and against transactional processes, and to highlight gaps in what is known about these processes. Although we show in Table II how inﬂuence processes in other family relationships (children and their siblings, sibling and parent–child relationships) ﬁt within our framework, due to space limits our discussion in the text focuses on transactional inﬂuence in (a) the parent–child relationship, (b) the interparental relationship, (c) links between the parent– child and interparental relationships, (d) links between the interparental relationship and children, and (e) family-wide processes. In cataloging the studies by system in Table II, the passage of time was a requirement for classiﬁcation in Systems B and C. Because cross-sectional studies cannot examine change, their ﬁndings are most consistent with static views of characteristics and behavior, and thus, they typically reﬂect System A (but see Covell & Abramovitch (1987) for an exception). Moreover, because we have deﬁned Systems B and C to reﬂect change, studies that do not examine change, even if they include longitudinal data, are not consistent with either Systems B or C. That is, studies testing associations between one construct at one time point and another construct at a later time point without testing for change in the second construct (e.g., via autoregressive controls, growth curve modeling) were classiﬁed in System A. In addition, some studies reﬂect more than one system. Because fewer studies have examined System C than System B, and fewer have examined System B than System A, in Table II we indicate System C when possible, followed by System B when possible.

A. TRANSACTIONAL DYNAMICS OF THE PARENT–CHILD RELATIONSHIP 1. System A Our focus in this section is on studies examining the inﬂuence of individual characteristics, behavior, and developmental status on parents

230

Alice C. Schermerhorn and E. Mark Cummings

and children (see Table II). Substantial evidence suggests that parents treat their children differently on the basis of personal characteristics of the child. For example, parents’ time involvement, affection, and disciplinary practices vary as a function of child gender, and privilege-granting and chore assignments vary as a function of child age and birth order (Tucker, McHale, & Crouter, 2003). Several studies have examined links between parents and children in terms of psychosocial functioning (see Table II). Among depressed mothers, maternal autonomy granting is a function of both child behavior and maternal negative mood (Kochanska & Kuczynski, 1991). Interestingly, mothers reciprocate their children’s behavior more than children reciprocate their mothers’. Examining children’s adjustment problems as a characteristic eliciting differential responding from parents, Boyle et al. (2004) found links (in two out of three studies) between child-speciﬁc differential maternal parenting and adjustment problems across siblings. In addition, in mother– child dyads with conduct disordered or hyperactive children, mothers use more negative control and less positive control, and mother–child conversational turn-taking is diminished, compared with mother–child dyads with nondisordered children (Brophy & Dunn, 2002). Moreover, when interacting with anxious-withdrawn children, adults exhibit more effort toward eliciting responses from the child; when interacting with conduct-disordered children, adults exhibit more effort toward restricting the child’s behavior (Brunk & Henggeler, 1984). Providing further insight into this process, conduct-disordered children elicit more negative responses and more requests from mothers, compared with nonconduct-disordered children (Anderson et al., 1986). This ﬁnding holds equally for mothers who are themselves parents of a conduct-disordered child and for mothers who are not parents of a conduct-disordered child. These ﬁndings support the notion that the direction of effects between conduct disorders and parenting may be child-to-parent as much as parent-to-child. With regard to children’s temperament, maternal emotion regulation varies as a function of children’s emotion (sadness, anger, fear), but not as a function of other dimensions of children’s temperament (Martini, Root, & Jenkins, 2004). Moreover, child negative emotionality predicts subsequent maternal power assertion, and for mothers low in perspective-taking or high in extraversion, child negative emotionality is linked with more power assertion (Clark, Kochanska, & Ready, 2000). Thus, the mother–child relationship appears to be more closely related to child emotionality and mother personality than to other dimensions of child temperament. Several studies have examined the contingency of parents’ and children’s behavior, that is, the degree to which one family member’s behavior is contingent on another’s (see Table II). Interestingly, children are more

Transactional Family Dynamics

231

likely to try to reduce parental anger when they are the cause of the anger than when the anger is caused by difﬁculties at work or interparental conﬂict (Covell & Miles, 1992). With regard to child compliance, young children comply more with their mothers when their fathers are present (Lytton, 1979). Surprisingly, the most frequent parental response to both child compliance and noncompliance is actually a complete lack of response from the parent, and fathers respond even less than mothers do to compliance. Elucidating these contingent processes in violent families, abusive parents tend to respond to all child misbehavior with punishment, whereas for nonabusive parents, the discipline strategy depends on the type of misbehavior (Trickett & Kuczynski, 1986).

2. System B In this section, we discuss transactional parent–child inﬂuence processes in terms of change from one time point to the next. Several studies have examined change as a function of parent–child interactions (see Table II). During the preschool years, mothers reciprocate their daughters’ (but not their sons’) positive emotional expressions more than they reciprocate their expressions of anger (Cole, Teti, & Zahn-Waxler, 2003). However, when mothers do reciprocate their children’s angry expressions, that predicts increases in children’s externalizing problems. Examining these links from preschool through age 12, children’s anger predicts increases in both punitive and distressed parental reactions, which predict increases in children’s anger (Eisenberg et al., 1999). In contrast, children’s self-regulation predicts decreases in punitive (but not distressed) parental reactions, and both punitive and distressed reactions predict decreases in self-regulation. Family interaction patterns change somewhat over the course of adolescence. During the period of adolescence prior to the pubertal apex, sons and mothers interrupt each other more and explain themselves less, and sons defer to their mothers less; after the pubertal apex, however, mothers interrupt their sons less (Steinberg, 1981). In contrast, across puberty, fathers’ interruptions of their sons increase and sons show more deference toward their fathers. In terms of parent–child relationship quality, mothers describe their relationships with their adolescents as most positive following increases in parent–child acceptance and decreases in parent–child conﬂict (Shearer, Crouter, & McHale, 2005). Moreover, parents grant their children more autonomy as their children develop. Relatedly, parents perceive that the parent–child relationship is most strengthened by their children’s compliance and by both parent and child efforts at companionship, whereas parental overuse of authority and child noncompliance are most

232

Alice C. Schermerhorn and E. Mark Cummings

detrimental to the parent–child relationship (Harach & Kuczynski, 2005). Interestingly, children perceive that they are capable of changing their mothers’ moods, endorsing gift-giving and verbal strategies for improving maternal mood, and their mothers agree that their children are able to change their moods (Covell & Abramovitch, 1987). Links between child adjustment problems (externalizing, delinquency) and parenting support notions of transactional processes (see Table II). Highlighting the transactional nature of these links, adolescent externalizing problems predict less parental behavioral control and positive behavior and more aversive parental behavior, which in turn predict more adolescent externalizing problems (Masche, Stattin, & Kerr, 2006). Relatedly, preadolescent competent behavior and maternal monitoring-relevant knowledge are reciprocally linked, with both constructs predicting increases in one another (Grundy, Gondoli, & Blodgett Salaﬁa, 2007). Moreover, parental control and support and adolescent substance use predict decreases in one another over time (Huh et al., 2006; Stice & Barrera, 1995). Some studies have found stronger support for child effects or parent effects, however (see Table II). For example, some work suggests that child delinquency prompts changes in parental monitoring, rather than the reverse direction, with some evidence that monitoring may actually decrease in the face of delinquency (Kerr & Stattin, 2003a). Moreover, adolescent externalizing symptoms predict decreases in parental support and control, whereas these dimensions of parenting do not predict changes in externalizing (Huh et al., 2006; Stice & Barrera, 1995). In contrast, whereas maternal monitoring predicts subsequent child externalizing problems, externalizing problems do not predict subsequent maternal monitoring (Brody, 2003). Debate regarding the direction of effects is available in Brody (2003), Capaldi (2003), and Kerr and Stattin (2003a, 2003b). Mother attachment security and adolescent attachment security predict increases in one another (Cook & Kenny, 2005). Toward the goal of teasing apart parental and children’s inﬂuence across time scales in the domain of attachment security, Lytton (1982) found evidence that child effects outweighed parent effects in the short-term, and neither child nor parent is more inﬂuential on the other’s attachment in the medium- or long-terms. In contrast, his work suggested that in the domain of discipline and compliance, parent effects outweigh child effects in the short-, medium-, and long-terms. With regard to examination of transactional links between parental psychopathology and children’s functioning or behavior in terms of change from one time point to the next, one study found that maternal depression predicts increases in the duration of infant awakening (Warren et al., 2006). Interestingly, longer durations of infant awakening predict decreases in

Transactional Family Dynamics

233

maternal depression. Thus, maternal depression and infant awake time are reciprocally linked, but in opposing directions. Notably, as can be seen from Table II, most of the studies in the parent–child domain for System B utilized data from relatively long time scales (1 or more years between assessments), with a few studies drawing on medium-term time scales (several months), and one study examining somewhat shorter time scales. 3. System C We now turn our attention to studies examining inﬂuence processes in terms of the overall patterns of change in parent–child relationships (see Table II). Examining infants’ distress intensity and attention to their mothers, Lewis, Lamey, and Douglas (1999) found that baseline levels of these constructs had substantial stability and inﬂuence on their behavior. During kindergarten-age children’s interactions with their parents, rigidity—that is, a lack of ﬂexible adaptability—is linked with externalizing and internalizing problems (Hollenstein et al., 2004). During adolescence, variability and instability in behavioral responding during parent–child interactions peaks at around age 13–14 years (Granic et al., 2003). In terms of links with adjustment problems, externalizing children’s interactions with their mothers are characterized by a permissive pattern, whereas the mother–child interactions of children with both externalizing and internalizing symptoms change from a permissive pattern to one that is mutually hostile (Granic & Lamey, 2002). In terms of time scales, as can be seen from Table II, all of the studies in System C used short units of time. 4. Summary There has been a considerable amount of work examining the inﬂuence of child characteristics and contingencies, and change from one time point to the next, and several studies have examined overall patterns of change in parent–child inﬂuence processes. However, as can be seen from the studies we reviewed (see Table II), much of this literature is based on studies with mothers, rather than including both parents. Therefore, we know much less about these transactional processes in the father–child relationship. In terms of overall patterns of change, work in this area has utilized data drawn from real time observations, but to our knowledge, no studies have examined patterns of change in terms of longer time scales. Thus, there are several gaps in our knowledge of transactional processes in the parent–child relationship. Moreover, little is known about positive or adaptive functioning and parental mental health. Thus, one question might involve examining whether children’s positive behavior (e.g., prosocial behavior, helping with household tasks) predicts decreases in parental psychopathology.

234

Alice C. Schermerhorn and E. Mark Cummings

B. TRANSACTIONAL DYNAMICS BETWEEN PARENT–CHILD AND INTERPARENTAL RELATIONSHIPS 1. System A Several studies have addressed associations between aspects of the parent–child and interparental relationships (see Table II). Interparental hostility is linked with low levels of mother–child attachment security several years later, and mother- and father–child security are associated with concurrent low levels of interparental conﬂict and high levels of interparental positivity (Frosch, Mangelsdorf, & McHale, 2000). Linking these constructs to maternal security, in families with low levels of marital adjustment, children of secure mothers are more securely attached to their mothers, compared with children of insecure mothers (Eiden, Teti, & Corns, 1995). In fact, for children of secure mothers, mother–child attachment security is unrelated to marital adjustment. In contrast, even premarital relationship quality is linked with mother–child attachment at 1–3 years of age, as is concurrent marital functioning (Howes & Markman, 1989), and pre-birth marital conﬂict predicts less secure mother- and father– child relations at 12–15 months of age (Owen & Cox, 1997). Coparenting, or joint parenting by adults in a family, is another important domain of family life. For families with low levels of positive affect, supportive coparenting predicts low levels of child externalizing problems, and in families with high levels of negative affect, undermining coparenting predicts externalizing problems (Schoppe, Mangelsdorf, & Frosch, 2001). Moreover, the combination of undermining coparenting and maladaptive family structure (e.g., triangulation) predicts higher levels of externalizing problems. Furthermore, marital quality and parental ﬂexibility and self-control have been jointly linked with concurrent coparenting harmony and negativity (Talbot & McHale, 2004). In addition, Engfer (1988) found evidence of four mechanisms linking the parent–child and marital relationships: (a) marital functioning inﬂuences the mother–child relationship, (b) the mother attempts to compensate for an unsatisfactory marriage by fulﬁlling her love and intimacy needs from the parent–child relationship, (c) the stresses of childcare inﬂuence the marital relationship, and (d) maternal personality inﬂuences both the marital and parent–child relationships.

2. System B A number of studies have also examined transactional inﬂuence between marital and parent–child relationships in terms of change from one time point to the next. For example, parental behavioral control mediated the link

Transactional Family Dynamics

235

between marital conﬂict and change in children’s internalizing problems (Schoppe-Sullivan, Schermerhorn, & Cummings, 2008); moreover, parental behavioral control, autonomy-granting, and warmth mediated relations between marital conﬂict and child internalizing and externalizing. Relatedly, children’s positive affect predicts increases in supportive coparenting, but coparenting does not predict change in children’s affect (Davis, 2007). Moreover, although coparenting predicts change in marital behavior, marital behavior does not predict change in coparenting (Schoppe-Sullivan et al., 2004). In addition, children’s secure internal representations of father–child relations predict increases in their representations of emotional security about marital function and representations of security about both the mother- and father–child relationships (Schermerhorn et al., 2008). Moreover, secure representations of the mother–child relationship predict increases in the security of father–child representations. Notably, all of the System B studies used fairly long time scales (1 year or longer between waves), and we know of no System C studies linking the parent–child and interparental relationships (i.e., studies examining change in the rate of change). 3. Summary Considerable research has examined links between the interparental and parent–child relationships in terms of associations between characteristics and functioning and in terms of change from one time point to the next. However, links between characteristics of children (e.g., age and gender) and change in interparental and parent–child relationships have been understudied, and many questions about coparenting remain unanswered. Moreover, examination of overall patterns of change (System C) in links between the interparental and parent–child relationships remains another gap in the literature. This type of work, drawing largely on dynamic systems theory and methods, is relatively new in family research, and thus, there are relatively few examples. Notably, dynamic systems modeling involves complicated statistical procedures. At the same time, this approach holds tremendous promise for contributing to the development of richer understanding of family inﬂuence processes.

C. TRANSACTIONAL DYNAMICS OF INTERPARENTAL RELATIONSHIPS AND CHILDREN 1. System A Building on existing questionnaire- and laboratory-based work, a new direction in studies examining links between marital conﬂict and child

236

Alice C. Schermerhorn and E. Mark Cummings

functioning is the use of diary methods, which provide a more ecologically valid test of these links (Cummings & Davies, 1994). Based on diary methods, destructive marital conﬂict tactics have been linked with children’s emotional insecurity and adjustment problems (Cummings, Goeke-Morey, & Papp, 2003) and aggression (Cummings, Goeke-Morey, & Papp, 2004), whereas constructive marital conﬂict tactics have been linked with children’s emotional security and lower levels of aggression. Moreover, both negative marital emotions and destructive marital conﬂict tactics are linked with children’s emotionally insecure responses, and positive marital emotions and constructive marital conﬂict tactics are linked with children’s secure responses (Cummings et al., 2002).

2. System B In terms of processes linking marital conﬂict and change in child functioning, children’s emotional insecurity about marital relations serves as an explanatory mechanism. In kindergarten-age children, children’s emotional insecurity links destructive marital conﬂict with increases in children’s internalizing and externalizing problems (Cummings et al., 2006). Moreover, in a sample of preadolescent children, children’s emotional insecurity about marital conﬂict, but not their cognitions about marital conﬂict, served to link destructive marital conﬂict with increases in children’s internalizing and externalizing problems (Davies et al., 2002). Notably, few studies have examined the ways in which children contribute to change in the marital relationship. However, in a landmark study, Jenkins et al. (2005b) found that in families with high levels of child externalizing problems, externalizing problems predicted increases in marital conﬂict. In contrast, preadolescents’ competent behavior does not predict subsequent marital conﬂict (Grundy et al., 2007). Subsequent research advanced this area of work further by examining links between children’s patterns of responding during marital conﬂict and subsequent marital conﬂict. Speciﬁcally, we have conducted several studies examining children’s intentional inﬂuence on change in marital conﬂict. We found that destructive marital conﬂict predicted more child negative emotionality, which related to greater perceptions of agency, agentic behavior, and behavioral dysregulation (Schermerhorn et al., 2005; Schermerhorn et al., 2007). Perceived agency and agentic behavior, in turn, were associated with subsequent decreases in destructive marital conﬂict, whereas behavioral dysregulation was linked with subsequent increases in destructive marital conﬂict. Person-oriented analyses of agentic and dysregulated responses indicated distinct clusters of children (low behavioral, agentic, high behavioral), and cluster membership was linked

Transactional Family Dynamics

237

with individual differences in marital and psychosocial functioning (Schermerhorn et al., 2007). Notably, as can be seen in Table II, all of these studies used long time scales, with one or more years between waves. Extending our examination of agency and behavioral dysregulation to shorter-term processes, we utilized diary data to assess children’s inﬂuence on marital conﬂict within conﬂict episodes, using dynamic systems modeling (Duration 2; Schermerhorn et al., 2007). We found that agentic behavior predicted less destructive conﬂict and more constructive and resolved conﬂict, and that dysregulated and negative child behavior predicted more destructive conﬂict and less constructive and resolved conﬂict. Thus, the results of our examination of shorter-term processes were consistent with our ﬁndings examining these processes over longer periods of time. 3. System C In our study of children’s inﬂuence over shorter time scales (i.e., Schermerhorn et al., 2007), we also examined dynamic processes between husbands’ and wives’ conﬂict behavior, using dynamic systems modeling to examine each spouses’ change around their baselines. We found that husbands’ behavior during each conﬂict was inﬂuenced by their own behavior during the immediately preceding conﬂict, such that their behavior changed considerably from one conﬂict to the next, tending to oscillate back and forth past baseline levels (Durations 3–4). This is, husbands who had high levels of negativity during one conﬂict tended to have low levels of negativity in the next conﬂict. Interestingly, husbands’ conﬂict resolution during one disagreement was inﬂuenced by their wives’ resolution during the preceding disagreement, meaning that husbands’ behavior in one disagreement was similar to their wives’ behavior from the preceding disagreement. This work thus represents an effort to examine overall patterns of change. That is, we examined husbands’ and wives’ inﬂuence on their own, and each others’, patterns of change. However, we know of no other studies directed toward these goals in the context of transactional links between interparental relationships and children. 4. Summary Many key questions about transactional links between marital and child functioning have been addressed, both in terms of key characteristics of marital conﬂict and child adjustment, and in terms of change processes. However, a critical gap in the literature involves examination of overall patterns of change in links between marital conﬂict and children (System C). Moreover, although previous work has examined both day-to-day time

238

Alice C. Schermerhorn and E. Mark Cummings

scales and yearly time scales, very little work has examined other time scales, and thus, questions regarding inﬂuence processes unfolding over these time scales remain unanswered.

D. FAMILY-WIDE TRANSACTIONAL DYNAMICS 1. System A A number of researchers have examined family-wide transactional family dynamics at the level of inﬂuence processes and associations between characteristics (see Table II). First, differences in parental-autonomy granting have been linked to several parent and child characteristics. For example, ﬁrstborns are granted more autonomy than second-borns, especially in older girl–younger boy dyads (Bumpus, Crouter, & McHale, 2001). Furthermore, ﬁrstborn girls are granted less autonomy in families with more traditional gender role attitudes, compared with families having less traditional attitudes. Moreover, in families with less traditional maternal gender role attitudes, postmenarcheal girls are granted more autonomy than either postmenarcheal girls whose mothers have more traditional gender role attitudes or premenarcheal girls. With regard to the marital relationship, longitudinal links have been found between marital hostility and affection, parent–child negativity, and sibling negativity and positivity (Dunn et al., 1999). In addition, marital functioning and family-wide hostility, harmony, and parenting discrepancies during infancy have been linked with child internalizing and externalizing problems 3 years later (McHale & Rasmussen, 1998). In addition, both concurrently and 1 year later, sibling warmth is linked with both children’s satisfaction in their relationships with their parents and with marital satisfaction and love (McGuire, McHale, & Updegraff, 1996). In terms of handling of conﬂict, links have been found between (a) parent–child and sibling conﬂict; (b) parent–child and marital conﬂict; and (c) marital and sibling conﬂict (Rinaldi & Howe, 2003). Moreover, research suggests that conﬂict affects multiple members of the family, extending to include family members not initially involved in the dispute. That is, when family dyads have conﬂict, a third family member intervenes approximately 1/3 of the time. However, when third parties do get involved, they are less likely to respond to conﬂicts with a conﬂictual response themselves. Parents-as-third-parties most frequently use mediation (especially mothers) and power-invoking (especially fathers) strategies (Vuchinich, Emery, & Cassidy, 1988). In contrast, distraction is the strategy of choice for children, followed by mediation as a second choice.

Transactional Family Dynamics

239

2. System B With regard to whole-family processes and change from one time point to the next, previous work has addressed a variety of interrelated constructs. For example, older sibling competence predicts improvement in maternal psychological functioning, which is linked with maternal warmth toward younger siblings, which in turn, predicts subsequent younger sibling self-regulation (Brody et al., 2003). In addition, younger sibling self-regulation is predicted by prior levels of older sibling competence and younger sibling self-regulation predicts subsequent younger sibling competence. Interestingly, sibling dyads show the most reciprocity of both positive and negative emotional expression of any family dyad (Rasbash et al., 2007). That is, one sibling’s positivity predicts an increase in the other sibling’s positivity, and one sibling’s negativity predicts an increase in the other sibling’s negativity. Notably, father–child dyads show the least reciprocity, particularly for fathers’ emotional expressions. Interestingly, positive emotional expressions are reciprocated less frequently than negative emotional expressions across all family dyads. In conducting this investigation, Rasbash et al. (2007) used an innovative extension of Kashy and Kenny’s (2000) Actor–Partner Interdependence Model (APIM), applied to family data. Using APIM, the researcher simultaneously models actor effects, which reﬂect the prediction of a person’s current behavior based on that same person’s past behavior, and partner effects, or the inﬂuence of the other person on one’s own behavior (Cook & Kenny, 2005). Thus, Rasbash et al.’s work makes a major contribution to the conceptualization of families, by distinguishing actor, partner, and relationship effects.

3. Summary Thus, these ﬁndings are consistent with notions of family-wide inﬂuence processes. However, relatively little work has examined the inﬂuence of families’ characteristics on individual family members. The relatively few family-wide studies examining change need to be supplemented with further examination of these links and further examination of multiple time scales, although notably, System B studies have used both short and long time scales (see Table II). Moreover, we know of no studies examining overall patterns of change in whole-family transactional dynamics. Nonetheless, given the difﬁculties inherent in studying family-wide inﬂuence processes, we recognize the work that has been done in this area as a remarkable contribution to the literature, and we highlight gaps toward the goal of stimulating further research in this area.

240

Alice C. Schermerhorn and E. Mark Cummings

V. Discussion Given this evidence regarding the transactional nature of family inﬂuence processes across dyads, triads, and whole families, we believe that we can further advance this area of work by viewing it through the lens of the transactional family dynamics perspective. This framework provides a way of cataloging what is known about these processes and highlighting gaps in our knowledge (see Table II). We also endeavor to provide a model of family inﬂuence processes that is realistic in terms of nested time scales, the hierarchical organization of family relationships, addressing association and contingency, time-point to time-point change, overall patterns of change, and the complexity of pathways between multiple family dyads and triads. Work in this area is not only important for advancing theory and research, but it may also facilitate more effective clinical work with families through better understanding of transactional family processes that may underlie the development and maintenance of mental health problems in families.

A. AN AGENDA FOR FUTURE RESEARCH: SOME HYPOTHESES ABOUT TRANSACTIONAL FAMILY DYNAMICS This theoretical framework highlights important gaps, and points to signiﬁcant goals for future research pertinent to developing a more complete understanding of transactional family processes. One set of issues involves time scales. That is, we have speculated about differences between processes unfolding over different time scales, but empirical work is needed to test these notions. One possibility is that there are similarities between different time scales. Thus, one might speculate that events occurring on a time scale of seconds are a microcosm of what happens on a time scale of hours, days, weeks, months, and years. One basis for this prediction is that the actors have similar intent, goals, motivations, and relationships with each other, regardless of the time frame, and thus, the processes may be similar in form. For example, real time inﬂuence processes during a marital conﬂict (Durations 1 and 2) may involve a lot of anger, whereas inﬂuence processes over the course of several days (Durations 4 and 5) may involve high levels of negative spousal attributions. Thus, the processes may reﬂect important differences in terms of emotional experiences (more prominent in real-time) and cognitive experiences (more prominent over longer time scales). At the same time, these processes can be seen as similar in terms of their likely effects on individuals and relationships. In any case, similarities and differences in processes over different time scales are in urgent need of further study. Moreover, an alternative possibility is that processes operate

Transactional Family Dynamics

241

differently in different time scales, in which case studying potential predictors of those differences, or of differences between families or family relationships, could contribute important new insights to the literature. Another interesting question may involve the inﬂuence of processes in one time scale on processes in another time scale. Furthermore, the coming together of multiple interdependent time scales may support the emergence of new patterns of inﬂuence. Finally, we think that distinguishing between time scales offers an interesting way of thinking about transactional family dynamics, suggesting an important new category of research questions. Second, we suggest the need for further study of how processes in one system inﬂuence processes in other systems. That is, characteristics of individual family members or family relationships, such as gender or temperament, might inﬂuence one another (System A). That inﬂuence process might cause one family member to change his way of relating to another family member. For example, perhaps after many years of a wife’s battle with depression, her husband may feel depressed himself, and begin behaving more negatively to his wife than before (System B). However, a subsequent change in another family relationship, such as a resolution of a long-standing disagreement between siblings, may cause the father to decide to behave more positively toward his wife. Over time, this shift from negative to positive behavior may solidify, resulting in a new pattern of behavior, and eventually leading to a sustained shift in behavior patterns (System C). A third set of hypotheses involves intentionality—to what degree are inﬂuence processes intentional? For example, if a child tries to resolve interparental conﬂict, is it the child’s efforts that bring about the decrease in conﬂict, or is it the child’s involvement that signals to the parents that the child is distressed, causing the parents to resolve the conﬂict out of concern for the child? Additionally, in order to be capable of engaging in some forms of agentic behavior, at least a minimal level of development in certain domains (e.g., cognitive, emotional) must be reached. For example, children may not be able to conceptualize and enact certain types of mediation in parental disputes until they achieve relatively advanced levels of cognitive functioning. Thus, further research should investigate links between developmental processes and the emergence of children’s agency, as well as identifying the domains of development that contribute to the appearance of agency (Sarah J. Schoppe-Sullivan, personal communication, July 26, 2007). Thus, there are critical questions to be addressed regarding the degree to which inﬂuence processes are attributable to intentional vs. unintentional behavior. That is, to what extent do family members act as agents of change in their families? A fourth direction for future research involves separating out the complex web of factors that contribute to, and are part and parcel of, transactional

242

Alice C. Schermerhorn and E. Mark Cummings

family dynamics. Experimental methods, following the example of Brunk and Henggeler (1984) and others (see also Cummings, 1995), and intervention research have the potential to facilitate progress toward this goal, because of the control afforded the researcher over various stimuli. That is, experimental control over independent variables provides an opportunity to learn more about dependent variables. Another approach would be to examine inﬂuence processes during phase transitions. Phase transitions to consider would include adolescence, milestones in cognitive development, divorce, school entry, job loss, and a death in the family. B. CONCLUSIONS Many researchers and theorists have called for greater attention to whole-family processes. As a call to action on this issue, recognizing the complexity of families—with nested individuals, dyads, and triads—Cox and Paley argued that (1997, p. 260): although a number of researchers y have emphasized the importance of data collected at multiple levels (e.g., individual, dyadic, whole family), it is rare for family research to include measurement that reﬂects all levels of the family. Even when researchers purport to have done so, the measurement often is not faithful to the level of analysis that is intended.

Further highlighting the complexity of transactional family dynamics, particularly with regard to nested time scales, Kuczynski and Parkin (2007) emphasized that instances of inﬂuence within families are not isolated events, but rather, each instance represents one thread interwoven into the fabric of family life and family experience, the whole of which produces continuous change. Thus, our goal is to, not only study the thread, but also to study—and come to understand—the fabric itself.

Acknowledgement We thank Paul Schermerhorn for his contributions to the conceptualization of this work, and Sarah J. Schoppe-Sullivan and Scott E. Maxwell for their helpful comments on an earlier version of this manuscript.

REFERENCES Allen, J. P., Antonishak, J., McFarland, F. C., & Meyer, J. R. (2006). Parents inﬂuencing peers and/or peers inﬂuencing parents? Transactional model of autonomy development from age

Transactional Family Dynamics

243

13 to 18. In C. Kouros (Chair), Adolescents’ influence in the family: Sibling, marital, and parent–child relationships. Paper presented at the biennial meetings of the Society for Research on Adolescence, San Francisco, CA. Anderson, K. E., Lytton, H., & Romney, D. M. (1986). Mothers’ interactions with normal and conduct-disordered boys: Who affects whom? Developmental Psychology, 22, 604–609. Bandura, A. (2001). Social cognitive theory: An agentic perspective. Annual Review of Psychology, 52, 1–26. Bandura, A. (2006). Toward a psychology of human agency. Perspectives on Psychological Science, 1, 164–180. Bell, R. Q. (1968). A reinterpretation of the direction of effects in studies of socialization. Psychological Review, 75, 81–95. Bell, R. Q. (1971). Stimulus control of parent or caretaker behavior by offspring. Developmental Psychology, 4, 63–72. Bell, R. Q. (1979). Parent, child, and reciprocal inﬂuences. American Psychologist, 34, 821–826. Bisconti, T. L., Bergeman, C. S., & Boker, S. M. (2004). Emotional well-being in recently bereaved widows: A dynamical systems approach. Journal of Gerontology: Psychological Sciences, 59B, P158–P167. Bogartz, R. S. (1994). The future of dynamic systems models in developmental psychology in the light of the past. Journal of Experimental Child Psychology, 58, 289–319. Bowlby, J. (1973). Attachment and loss. Vol. 2. Separation. New York: Basic Books. Boyle, M. H., Jenkins, J. M., Georgiades, K., Cairney, J., Duku, E., & Racine, Y. (2004). Differential-maternal parenting behavior: Estimating within- and between-family effects on children. Child Development, 75, 1457–1476. Bretherton, I. (1985). Attachment theory: Retrospect and prospect. In I. Bretherton & E. Waters (Eds.), Growing points in attachment theory and research. Monographs of the Society for Research in Child Development, 50(1-2, Serial No. 209), 3–35. Brody, G. H. (2003). Parental monitoring: Action and reaction. In A. C. Crouter & A. Booth (Eds.), Children’s influence on family dynamics: The neglected side of family relationships (pp. 163–169). Mahwah, NJ: Lawrence Erlbaum Associates, Publishers. Brody, G. H., Kim, S., Murry, V. M., & Brown, A. C. (2003). Longitudinal direct and indirect pathways linking older sibling competence to the development of younger sibling competence. Developmental Psychology, 39, 618–628. Bronfenbrenner, U. (1979). The ecology of human development: Experiments by nature and design. Cambridge, MA: Harvard University Press. Bronfenbrenner, U. (1986). Ecology of the family as a context for human development: Research perspectives. Developmental Psychology, 22, 723–742. Bronfenbrenner, U. (1988). Interacting systems in human development. Research paradigms: Present and future. In N. Bolger, A. Caspi, G. Downey, & M. Moorehouse (Eds.), Persons in context: Developmental processes (pp. 25–49). New York: Cambridge University Press. Bronfenbrenner, U. (2005). Making human beings human: Bioecological perspectives on human development. Thousand Oaks, CA: Sage Publications Ltd. Bronfenbrenner, U., & Morris, P. A. (1998). The ecology of developmental processes. In W. Damon & R. M. Lerner (Eds.), Handbook of child psychology, Volume 1: Theoretical models of human development (5th ed., pp. 993–1028). Hoboken, NJ: Wiley. Brophy, M., & Dunn, J. (2002). What did mummy say? Dyadic interactions between young ‘‘hard to manage’’ children and their mothers. Journal of Abnormal Child Psychology, 30, 103–112. Brunk, M. A., & Henggeler, S. W. (1984). Child inﬂuences on adult controls: An experimental investigation. Developmental Psychology, 20, 1074–1081.

244

Alice C. Schermerhorn and E. Mark Cummings

Bumpus, M. F., Crouter, A. C., & McHale, S. M. (2001). Parental autonomy granting during adolescence: Exploring gender differences in context. Developmental Psychology, 37, 163–173. Capaldi, D. M. (2003). Parental monitoring: A person–environment interaction perspective on this key parenting skill. In A. C. Crouter & A. Booth (Eds.), Children’s influence on family dynamics: The neglected side of family relationships (pp. 171–179). Mahwah, NJ: Lawrence Erlbaum Associates, Publishers. Cicchetti, D. (2006). Development and psychopathology. In D. Cicchetti & D. J. Cohen (Eds.), Developmental psychopathology, Volume 1: Theory and method (2nd ed., pp. 1–23). Hoboken, NJ: Wiley. Cicchetti, D., Toth, S., & Bush, M. (1988). Developmental psychopathology and incompetence in childhood: Suggestions for intervention. In B. Lahey & A. Kazdin (Eds.), Advances in clinical child psychology (pp. 1–71). New York: Plenum Press. Clark, L. A., Kochanska, G., & Ready, R. (2000). Mothers’ personality and its interaction with child temperament as predictors of parenting behavior. Journal of Personality and Social Psychology, 79, 274–285. Cole, P. M. (2003). The developmental course from child effects to child effectiveness. In A. C. Crouter & A. Booth (Eds.), Children’s influence on family dynamics: The neglected side of family relationships (pp. 109–118). Mahwah, NJ: Lawrence Erlbaum Associates, Publishers. Cole, D. A., & Maxwell, S. E. (2003). Testing mediational models with longitudinal data: Questions and tips in the use of structural equation modeling. Journal of Abnormal Psychology, 112, 558–577. Cole, P. M., Teti, L. O., & Zahn-Waxler, C. (2003). Mutual emotion regulation and the stability of conduct problems between preschool and early school age. Development and Psychopathology, 15, 1–18. Collins, W. A., & Madsen, S. D. (2003). Developmental change in parenting interactions. In L. Kuczynski (Ed.), Handbook of dynamics in parent–child relations (pp. 49–66). Thousand Oaks, CA: Sage. Connell, A. M., & Goodman, S. H. (2002). The association between psychopathology in fathers versus mothers and children’s internalizing and externalizing behavior problems: A meta-analysis. Psychological Bulletin, 128, 746–773. Cook, W. L. (2003). Quantitative methods for deductive (theory-testing) research on parent– child dynamics. In L. Kuczynski (Ed.), Handbook of dynamics in parent–child relations (pp. 347–372). Thousand Oaks, CA: Sage. Cook, W. L., & Kenny, D. A. (2005). The actor–partner interdependence model: A model of bidirectional effects in developmental studies. International Journal of Behavioral Development, 29, 101–109. Covell, K., & Abramovitch, R. (1987). Understanding emotion in the family: Children’s and parents’ attributions of happiness, sadness, and anger. Child Development, 58, 985–991. Covell, K., & Miles, B. (1992). Children’s beliefs about strategies to reduce parental anger. Child Development, 63, 381–390. Cox, M. J., & Paley, B. (1997). Families as systems. Annual Review of Psychology, 48, 243–267. Cox, M. J., & Paley, B. (2003). Understanding families as systems. Current Directions in Psychological Science, 12, 193–196. Cummings, E. M. (1995). Usefulness of experiments for the study of the family. Journal of Family Psychology, 9, 175–185. Cummings, E. M., & Davies, P. T. (1994). Children and marital conflict: The impact of family dispute and resolution. New York and London: The Guilford Press. Cummings, E. M., & Davies, P. T. (2002). Effects of marital conﬂict on children: Recent advances and emerging themes in process-oriented research. Journal of Child Psychology and Psychiatry, 43, 31–63.

Transactional Family Dynamics

245

Cummings, E. M., & Schermerhorn, A. C. (2003). A developmental perspective on children as agents in the family. In L. Kuczynski (Ed.), Handbook of dynamics of parent-child relations (pp. 91–108). Thousand Oaks, CA: Sage. Cummings, E. M., Goeke-Morey, M. C., Papp, L. M., & Dukewich, T. L. (2002). Children’s responses to mothers’ and fathers’ emotionality and tactics in marital conﬂict in the home. Journal of Family Psychology, 16, 478–492. Cummings, E. M., Goeke-Morey, M. C., & Papp, L. M. (2003). Children’s responses to everyday marital conﬂict tactics in the home. Child Development, 74, 1918–1929. Cummings, E. M., Goeke-Morey, M. C., & Papp, L. M. (2004). Everyday marital conﬂict and child aggression. Journal of Abnormal Child Psychology, 32, 191–202. Cummings, E. M., Schermerhorn, A. C., Davies, P. T., Goeke-Morey, M. C., & Cummings, J. S. (2006). Interparental discord and child adjustment: Prospective investigations of emotional security as an explanatory mechanism. Child Development, 77, 132–152. Davies, P. T., & Cummings, E. M. (1994). Marital conﬂict and child adjustment: An emotional security hypothesis. Psychological Bulletin, 116, 387–411. Davies, P. T., Harold, G. T., Goeke-Morey, M. C., & Cummings, E. M. (2002). Child emotional security and interparental conﬂict. Monographs of the Society for Research in Child Development, 67(3, Serial No. 270), vii–115. Davis, E. (2007). Exploring potential effects of child behavior on the coparenting relationship. Unpublished master’s thesis, The Ohio State University. Dunn, J. (1997). Commentary: Lessons from the study of bi-directional effects. Journal of Social and Personal Relationships, 14, 565–573. Dunn, J., & Munn, P. (1985). Becoming a family member: Family conﬂict and the development of social understanding in the second year. Child Development, 56, 480–492. Dunn, J., Deater-Deckard, K., Pickering, K., Golding, J., & the ALSPAC Study Team. (1999). Siblings, parents, and partners: Family relationships within a longitudinal community study. Journal of Child Psychology and Psychiatry, 40, 1025–1037. Eiden, R. D., Teti, D. M., & Corns, K. M. (1995). Maternal working models of attachment, marital adjustment, and the parent–child relationship. Child Development, 66, 1504–1518. Eisenberg, N., Fabes, R. A., Shepard, S. A., Guthrie, I. K., Murphy, B. C., & Reiser, M. (1999). Parental reactions to children’s negative emotions: Longitudinal relations to quality of children’s social functioning. Child Development, 70, 513–534. Emery, R. E. (1982). Interparental conﬂict and the children of discord and divorce. Psychological Bulletin, 92, 310–330. Emery, R. E., Binkoff, J. A., Houts, A. C., & Carr, E. G. (1983). Children as independent variables: Some clinical implications of child-effects. Behavior Therapy, 14, 398–412. Engfer, A. (1988). The interrelatedness of marriage and the mother–child relationship. In R. A. Hinde & J. Stevenson-Hinde (Eds.), Relationships within families: Mutual influences (pp. 104–118). Oxford: Clarendon Press. Frosch, C. A., Mangelsdorf, S. C., & McHale, J. L. (2000). Marital behavior and the security of preschooler–parent attachment relationships. Journal of Family Psychology, 14, 144–161. Granic, I. (2000). The self-organization of parent–child relations: Beyond bidirectional models. In M. D. Lewis & I. Granic (Eds.), Emotion, development, and self-organization: Dynamic systems approaches to emotional development (pp. 267–297). New York: Cambridge University Press. Granic, I. (2005). Timing is everything: Developmental psychopathology from a dynamic systems perspective. Developmental Review, 25, 386–407. Granic, I., & Hollenstein, T. (2003). Dynamic systems methods for models of developmental psychopathology. Development and Psychopathology, 15, 641–669.

246

Alice C. Schermerhorn and E. Mark Cummings

Granic, I., & Hollenstein, T. (2006). A survey of dynamic systems methods for developmental psychopathology. In D. Cicchetti & D. J. Cohen (Eds.), Developmental psychopathology, Vol 1: Theory and method (2nd ed., pp. 889–930). Hoboken, NJ: Wiley. Granic, I., & Lamey, A. V. (2002). Combining dynamic systems and multivariate analyses to compare the mother–child interactions of externalizing subtypes. Journal of Abnormal Child Psychology, 30, 265–283. Granic, I., & Patterson, G. R. (2006). Toward a comprehensive model of antisocial development: A dynamic systems approach. Psychological Review, 113, 101–131. Granic, I., Dishion, T. J., & Hollenstein, T. (2003). The family ecology of adolescence: A dynamic systems perspective on normative development. In G. R. Adams & M. D. Berzonsky (Eds.), Blackwell handbook of adolescence (pp. 60–91). Malden, MA: Blackwell Publishing. Granic, I., Hollenstein, T., Dishion, T. J., & Patterson, G. R. (2003). Longitudinal analysis of ﬂexibility and reorganization in early adolescence: A dynamic systems study of family interactions. Developmental Psychology, 39, 606–617. Gray, M. R., & Steinberg, L. (1999). Unpacking authoritative parenting: Reassessing a multidimensional construct. Journal of Marriage and the Family, 61, 574–587. Grundy, A. M., Gondoli, D. M., & Blodgett Salaﬁa, E. H. (2007). Marital conﬂict and preadolescent behavioral competence: Maternal knowledge as a longitudinal mediator. Journal of Family Psychology, 21, 675–682. Grusec, J. E., & Kuczynski, L. (1980). Direction of effect in socialization: A comparison of the parent’s versus the child’s behavior as determinants of disciplinary techniques. Developmental Psychology, 16, 1–9. Grych, J. H., & Fincham, F. D. (1990). Marital conﬂict and children’s adjustment: A cognitivecontextual framework. Psychological Bulletin, 108, 267–290. Harach, L. D., & Kuczynski, L. J. (2005). Construction and maintenance of parent–child relationships: Bidirectional contributions from the perspective of parents. Infant and Child Development, 14, 327–343. Hinde, R. A., & Stevenson-Hinde, J. (1987). Interpersonal relationships and child development. Developmental Review, 7, 1–21. Hollenstein, T., Granic, I., Stoolmiller, M., & Snyder, J. (2004). Rigidity in parent–child interactions and the development of externalizing and internalizing behavior in early childhood. Journal of Abnormal Child Psychology, 32, 595–607. Howe, M. L., & Lewis, M. D. (2005). The importance of dynamic systems approaches for understanding development. Developmental Review, 25, 247–251. Howes, P., & Markman, H. J. (1989). Marital quality and child functioning: A longitudinal investigation. Child Development, 60, 1044–1051. Huh, D., Tristan, J., Wade, E., & Stice, E. (2006). Does problem behavior elicit poor parenting? A prospective study of adolescent girls. Journal of Adolescent Research, 21, 185–204. Jenkins, J., Simpson, A., Dunn, J., Rasbash, J., & O’Connor, T. G. (2005a). Mutual inﬂuence of marital conﬂict and children’s behavior problems: Shared and nonshared family risks. Child Development, 76, 24–39. Jenkins, J. M., Dunn, J., O’Connor, T. G., Rasbash, J., & Behnke, P. (2005b). Change in maternal perception of sibling negativity: Within- and between-family inﬂuences. Journal of Family Psychology, 19, 533–541. Kahn, R. L., & Antonucci, T. C. (1980). Convoys over the life course: Attachment, roles, and social support. In P. Baltes & O. G. Brim, Jr. (Eds.), Life-span development and behavior (Vol. 3, pp. 253–286). New York: Academic Press. Kashy, D. A., & Kenny, D. A. (2000). The analysis of data from dyads and groups. In H. T. Reis & C. M. Judd (Eds.), Handbook of research methods in social and personality psychology (pp. 451–477). New York: Cambridge University Press.

Transactional Family Dynamics

247

Kelley, H. H., Berscheid, E., Christensen, A., Harvey, J. H., Huston, T. L., Levinger, G., McClintock, E., Peplau, L. A., & Peterson, D. R. (1983). Close relationships. New York: W.H. Freeman & Company. Kerr, M., & Stattin, H. (2003a). Parenting of adolescents: Action or reaction? In A. Booth & A. C. Crouter (Eds.), Children’s influence on family dynamics: The neglected side of family relationships (pp. 121–151). Mahwah, NJ: Lawrence Erlbaum Associates. Kerr, M., & Stattin, H. (2003b). Straw men, untested assumptions, and bi-directional models: A response to Capaldi and Brody. In A. Booth & A. C. Crouter (Eds.), Children’s influence on family dynamics: The neglected side of family relationships (pp. 181–187). Mahwah, NJ: Lawrence Erlbaum Associates. Klaczynski, P. A., & Daniel, D. B. (2005). Individual differences in conditional reasoning: A dual-process account. Thinking and Reasoning, 11, 305–325. Kochanska, G., & Kuczynski, L. (1991). Maternal autonomy granting: Predictors of normal and depressed mothers’ compliance and noncompliance with the requests of ﬁve-year-olds. Child Development, 62, 1449–1459. Kuczynski, L., & Hildebrandt, N. (1997). Models of conformity and resistance in socialization theory. In J. E. Grusec & L. Kuczynski (Eds.), Parenting and children’s internalization of values: A handbook of contemporary theory (pp. 227–256). New York: Wiley. Kuczynski, L., & Kochanska, G. (1990). Development of children’s noncompliance strategies from toddlerhood to age 5. Developmental Psychology, 26, 398–408. Kuczynski, L., & Parkin, C. M. (2007). Agency and bidirectionality in socialization: Interactions, transactions, and relational dialectics. In J. E. Grusec & P. D. Hastings (Eds.), Handbook of socialization: Theory and research (pp. 259–283). New York: Guilford. Kuczynski, L., Kochanska, G., Radke-Yarrow, M., & Girnius-Brown, O. (1987). A developmental interpretation of young children’s noncompliance. Developmental Psychology, 23, 799–806. Kuczynski, L., Marshall, S., & Schell, K. (1997). Value socialization in a bidirectional context. In J. E. Grusec & L. Kuczynski (Eds.), Parenting and children’s internalization of values: A handbook of contemporary theory (pp. 23–50). New York: Wiley. Lamey, A., Hollenstein, T., Lewis, M. D., & Granic, I. (2004). GridWare (Version 1.1). [Computer software]. http://www.statespacegrids.org Lewis, M. D. (2000). The promise of dynamic systems approaches for an integrated account of human development. Child Development, 71, 36–43. Lewis, M. D. (2002). Interacting time scales in personality (and cognitive) development: Intention, emotions, and emergent forms. In N. Granott & J. Parziale (Eds.), Microdevelopment: Transition processes in development and learning (pp. 183–212). New York: Cambridge University Press. Lewis, M. D. (2004). Trouble ahead: Predicting antisocial trajectories with dynamic systems concepts and methods. Journal of Abnormal Child Psychology, 32, 665–671. Lewis, M. D., Lamey, A. V., & Douglas, L. (1999). A new dynamic systems method for the analysis of early socioemotional development. Developmental Science, 2, 457–475. Lollis, S., & Kuczynski, L. (1997). Beyond one hand clapping: Seeing bidirectionality in parent–child relations. Journal of Social and Personal Relationships, 14, 441–461. Lytton, H. (1979). Disciplinary encounters between young boys and their mothers and fathers: Is there a contingency system? Developmental Psychology, 15, 256–268. Lytton, H. (1982). Two-way inﬂuence processes between parents and child—when, where and how? Canadian Journal of Behavioural Science, 14, 259–275. Lytton, H. (1990a). Child and parent effects in boys’ conduct disorder: A reinterpretation. Developmental Psychology, 26, 683–697.

248

Alice C. Schermerhorn and E. Mark Cummings

Lytton, H. (1990b). Child effects—still unwelcome? Response to Dodge and Wahler. Developmental Psychology, 26, 705–709. Lytton, H. (2000). Toward a model of family-environmental and child-biological inﬂuences on development. Developmental Review, 20, 150–179. Maccoby, E. E. (1984). Socialization and developmental change. Child Development, 55, 317–328. Martini, T. S., Root, C. A., & Jenkins, J. M. (2004). Low and middle income mothers’ regulation of negative emotion: Effects of children’s temperament and situational emotional responses. Social Development, 13, 515–530. Masche, J. G., Stattin, H., & Kerr, M. (2006). What do parents do when faced with adolescent problem behaviors? Parents’ and youths’ perspectives. In J. G. Masche (Chair), Stepping up or backing down? Parents’ reactions to adolescents’ problem behaviors. Paper presented at the meeting of the Society for Research on Adolescence, San Francisco, CA. McCartney, K. (2003). On the meaning of models: A signal amidst the noise. In A. C. Crouter & A. Booth (Eds.), Children’s influence on family dynamics (pp. 27–36). Mahwah, NJ: Lawrence Erlbaum Associates. McGuire, S., McHale, S. M., & Updegraff, K. (1996). Children’s perceptions of the sibling relationship in middle childhood: Connections within and between family relationships. Personal Relationships, 3, 229–239. McHale, J. P., & Fivaz-Depeursinge, E. (1999). Understanding triadic and family group interactions during infancy and toddlerhood. Clinical Child and Family Psychology Review, 2, 107–127. McHale, J. P., & Rasmussen, J. L. (1998). Coparental and family group-level dynamics during infancy: Early family precursors of child and family functioning during preschool. Development and Psychopathology, 10, 39–59. McHale, J. P., Kavanaugh, K. C., & Berkman, J. M. (2003). Sensitivity to infants’ cues: As much a mandate for researchers as for parents. In A. C. Crouter & A. Booth (Eds.), Children’s influence on family dynamics (pp. 91–108). Mahwah, NJ: Lawrence Erlbaum Associates. Minuchin, P. (1985). Families and individual development: Provocations from the ﬁeld of family therapy. Child Development, 56, 289–302. Newson, J., & Newson, E. (1976). Seven years old in the home environment. New York: Wiley. Owen, M. T., & Cox, M. J. (1997). Marital conﬂict and the development of infant–parent attachment relationships. Journal of Family Psychology, 11, 152–164. O’Brien, M. (2005). Studying individual and family development: Linking theory and research. Journal of Marriage and Family, 67, 880–890. Patterson, G. R. (1982). Coercive family process. Eugene, OR: Castalia. Patterson, G. R. (1997). Performance models for parenting: A social interactional perspective. In J. E. Grusec & L. Kuczynski (Eds.), Parenting and children’s internalization of values: A handbook of contemporary theory (pp. 193–226). New York: Wiley. Patterson, G. R., & Fisher, P. A. (2002). Recent developments in our understanding of parenting: Bidirectional effects, causal models, and the search for parsimony. In M. H. Bornstein (Ed.), Handbook of parenting, Vol. 5: Practical issues in parenting (pp. 59–88). Mahwah, NJ: Lawrence Erlbaum Associates. Patterson, G. R., DeBaryshe, B. D., & Ramsey, E. (1989). A developmental perspective on antisocial behavior. American Psychologist, 44, 329–335. Powers, S. I., Hauser, S. T., Schwartz, J. M., Noam, G. G., & Jacobson, A. M. (1983). Adolescent ego development and family interaction: A structural-developmental perspective. In J. D. Grotevant & C. R. Cooper (Eds.), Adolescent development in the family: New directions for child development (Vol. 22). San Francisco: Jossey-Bass.

Transactional Family Dynamics

249

Rasbash, J., O’Connor, T. G., Jenkins, J., & Reiss, D. (2007). Deconstructing relationships in families: Contrasting individual and relationship hypotheses in a genetically-informative family study. Unpublished manuscript. Rinaldi, C. M., & Howe, N. (2003). Perceptions of constructive and destructive conﬂict within and across family subsystems. Infant and Child Development, 12, 441–459. Sameroff, A. (1975a). Transactional models in early social relations. Human Development, 18, 65–79. Sameroff, A. (1975b). Early inﬂuences on development: Fact or fancy? Merrill-Palmer Quarterly, 21, 267–294. Sameroff, A. J. (1995). General systems theories and developmental psychopathology. In D. Cicchetti & D. J. Cohen (Eds.), Developmental psychopathology, Vol. 1: Theory and methods (pp. 659–695). Oxford: Wiley. Sameroff, A. J., & Fiese, B. H. (2000). Transactional regulation: The developmental ecology of early intervention. In J. P. Shonkoff & S. J. Meisels (Eds.), Handbook of early childhood intervention (2nd ed., pp. 135–159). New York: Cambridge University Press. Sameroff, A. J., & MacKenzie, M. J. (2003). Research strategies for capturing transactional models of development: The limits of the possible. Development and Psychopathology, 15, 613–640. Sanders, M. R., Nicholson, J. M., & Floyd, F. J. (1997). Couples’ relationships and children. In W. K. Halford & H. J. Markman (Eds.), Clinical handbook of marriage and couples interventions. New York: Wiley. Scarr, S., & McCartney, K. (1983). How people make their own environments: A theory of genotype - environment effects. Child Development, 54, 424–435. Schaffer, H. R. (1999). Understanding socialization: From unidirectional to bidirectional conceptions. In M. Bennett (Ed.), Developmental psychology: Achievements and prospects (pp. 272–288). Philadelphia: Psychology Press. Schermerhorn, A. C., Cummings, E. M., & Davies, P. T. (2005). Children’s perceived agency in the context of marital conﬂict: Relations with marital conﬂict over time. Merrill-Palmer Quarterly, 51, 121–144. Schermerhorn, A. C., Chow, S-M., & Cummings, E. M. (2007). Dynamics of family influence processes during interparental conflict. Manuscript in preparation. Schermerhorn, A. C., Cummings, E. M., & Davies, P. T. (2008). Children’s representations of multiple family relationships: Organizational structure and development in early childhood. Journal of Family Psychology, 22, 89–101. Schermerhorn, A. C., Cummings, E. M., DeCarlo, C. A., & Davies, P. T. (2007). Children’s inﬂuence in the marital relationship. Journal of Family Psychology, 21, 259–269. Schoppe, S. J., Mangelsdorf, S. C., & Frosch, C. A. (2001). Coparenting, family process, and family structure: Implications for preschoolers’ externalizing behavior problems. Journal of Family Psychology, 15, 526–545. Schoppe-Sullivan, S. J., Mangelsdorf, S. C., Frosch, C. A., & McHale, J. (2004). Associations between coparenting and marital behavior from infancy to the preschool years. Journal of Family Psychology, 18, 194–207. Schoppe-Sullivan, S. J., Schermerhorn, A. C., & Cummings, E. M. (2007). Marital conﬂict and children’s adjustment: Evaluation of the parenting process model. Journal of Marriage and Family, 69, 1118–1134. Shearer, C. L., Crouter, A. C., & McHale, S. M. (2005). Parents’ perceptions of changes in mother–child and father–child relationships during adolescence. Journal of Adolescent Research, 20, 662–684. Smith, L. B. (2005). Cognition as a dynamic system: Principles from embodiment. Developmental Review, 25, 278–298.

250

Alice C. Schermerhorn and E. Mark Cummings

Smith, L. B., & Thelen, E. (2003). Development as a dynamic system. Trends in Cognitive Sciences, 7, 343–348. Steinberg, L. D. (1981). Transformations in family relations at puberty. Developmental Psychology, 17, 833–840. Stice, E., & Barrera, M. (1995). A longitudinal examination of the reciprocal relations between perceived parenting and adolescents’ substance use and externalizing behaviors. Developmental Psychology, 31, 322–334. Stifter, C. A. (2003). Child effects on the family: An example of the extreme case and a question of methodology. In A. C. Crouter & A. Booth (Eds.), Children’s influence on family dynamics: The neglected side of family relationships (pp. 79–89). Mahwah, NJ: Lawrence Erlbaum Associates. Talbot, J. A., & McHale, J. P. (2004). Individual parental adjustment moderates the relationship between marital and coparenting quality. Journal of Adult Development, 11, 191–205. The Carolina Consortium on Human Development. (1996). Developmental science: A collaborative statement. In R. B. Cairns, G. H. Elder, & E. J. Costello (Eds.), Developmental science: Cambridge studies in social and emotional development (pp. 1–6). New York: Cambridge University Press. Thelen, E. (1995). Time-scale dynamics and the development of an embodied cognition. In R. F. Port & T. van Gelder (Eds.), Mind as motion: Explorations in the dynamics of cognition (pp. 69–100). Cambridge, MA: The MIT Press. Thelen, E., & Smith, L. B. (1994). A dynamic systems approach to the development of cognition and action. Cambridge, MA: MIT. Thelen, E., & Smith, L. B. (1998). Dynamic systems theories. In W. Damon & R. M. Lerner (Eds.), Handbook of child psychology. Volume 1: Theoretical models of human development (5th ed., pp. 563–634). Hoboken, NJ: Wiley. Thelen, E., & Ulrich, B. D. (1991). Hidden skills: A dynamic systems analysis of treadmill stepping during the ﬁrst year. Monographs of the Society for Research in Child Development, 56, [Serial No 223], v–103. Thompson, R. A., & Raikes, H. A. (2003). Toward the next quarter-century: Conceptual and methodological challenges for attachment theory. Development and Psychopathology, 15, 691–718. Trickett, P. K., & Kuczynski, L. (1986). Children’s misbehaviors and parental discipline strategies in abusive and nonabusive families. Developmental Psychology, 22, 115–123. Tucker, C. J., McHale, S. M., & Crouter, A. C. (2003). Dimensions of mothers’ and fathers’ differential treatment of siblings: Links with adolescents’ sex-typed personal qualities. Family Relations, 52, 82–89. Van Gelder, T., & Port, R. F. (1995). It’s about time: An overview of the dynamical approach to cognition. In R. F. Port & T. van Gelder (Eds.), Mind as motion: Explorations in the dynamics of cognition (pp. 1–44). Cambridge, MA: The MIT Press. Vuchinich, S., Emery, R. E., & Cassidy, J. (1988). Family members as third parties in dyadic family conﬂict: Strategies, alliances, and outcomes. Child Development, 59, 1293–1302. Warren, S. L., Howe, G., Simmens, S. J., & Dahl, R. E. (2006). Maternal depressive symptoms and child sleep: Models of mutual inﬂuence over time. Development and Psychopathology, 18, 1–16.

THE DEVELOPMENT OF RATIONAL THOUGHT: A TAXONOMY OF HEURISTICS AND BIASES

Keith E. Stanovicha, Maggie E. Toplakb, and Richard F. Westc a

DEPARTMENT OF HUMAN DEVELOPMENT AND APPLIED PSYCHOLOGY, UNIVERSITY OF TORONTO, TORONTO M5S 1V6, CANADA b DEPARTMENT OF PSYCHOLOGY, YORK UNIVERSITY, TORONTO, ONTARIO M3J 1P3, CANADA c DEPARTMENT OF GRADUATE PSYCHOLOGY, JAMES MADISON UNIVERSITY, HARRISONBURG, VIRGINIA 22807, USA

I. II.

INTRODUCTION EXPERIMENTALLY TRACTABLE DEFINITIONS OF RATIONAL THOUGHT

III.

DUAL-PROCESS MODELS OF COGNITION

IV.

A PRELIMINARY TAXONOMY OF RATIONAL THINKING ERRORS

V. VI.

VII.

CLASSIFYING HEURISTICS AND BIASES EXEMPLAR DEVELOPMENTAL STUDIES IN THE DIFFERENT CATEGORIES OF THE TAXONOMY A . DEFAULT TO TYPE 1 PROCESSING: VIVIDNESS EFFECTS B . FOCAL BIAS: FRAMING EFFECTS C . OVERRIDE FAILURE: DENOMINATOR NEGLECT D . OVERRIDE FAILURE: BELIEF BIAS E . MINDWARE GAPS F . HYBRID REASONING PROBLEMS CONCLUSION: SPECIFICITY AND GENERALITY IN THE DEVELOPMENT OF RATIONAL THOUGHT

REFERENCES

I. Introduction The most well-known indicators of cognitive functioning, intelligence and cognitive ability tests, do not assess a critical aspect of thinking—the ability to think rationally. To think rationally means adopting appropriate goals, taking the appropriate action given one’s goals and beliefs, and holding beliefs that are commensurate with available evidence. Standard intelligence tests do not assess such functions (Perkins, 1995, 2002;

251 Advances in Child Development and Behavior R.V. Kail : Editor

Copyright r 2008 Elsevier B.V. All rights reserved.

252

Keith E. Stanovich et al.

Stanovich, 2002, 2008c; Sternberg, 2003, 2006). For example, although intelligence tests do assess the ability to focus on an immediate goal in the face of distraction, they do not assess at all whether a person has the tendency to develop goals that are rational in the ﬁrst place. Likewise, intelligence tests are good measures of how well a person can hold beliefs in short-term memory and manipulate those beliefs, but they do not assess at all whether a person has the tendency to form beliefs rationally when presented with evidence. And again, similarly, intelligence tests are good measures of how efﬁciently a person processes information that has been provided, but they do not at all assess whether the person is a critical assessor of information as it is gathered in the natural environment. Variation in intelligence has been one of the most studied topics in psychology for many decades (Deary, 2001; Geary, 2005; Lubinski, 2004), and the development of the cognitive abilities related to intelligence is likewise a central topic in developmental science (Anderson, 2005; Bjorklund, 2004; Kail, 2000). In contrast, variation in rational thought among adults has only recently been the focus of research (Bruine de Bruin, Parker, & Fischhoff, 2007; Stanovich & West, 1998b, 2000). Similarly, the empirical literature on the development of rational thinking is still relatively sparse (see Byrnes, 1998; Klaczynski, 2001a; Kokis et al., 2002; Reyna, Lloyd, & Brainerd, 2003). In this chapter, we summarize some of the existing work on the development of rational thought and we attempt to provoke more such work by providing a framework for classifying rational thinking errors.

II. Experimentally Tractable Deﬁnitions of Rational Thought Cognitive scientists recognize two types of rationality: instrumental and epistemic. The simplest deﬁnition of instrumental rationality is: Behaving in the world so that you get exactly what you most want, given the resources (physical and mental) available to you. Somewhat more technically we could characterize instrumental rationality as the optimization of the individual’s goal fulﬁllment. Economists and cognitive scientists have reﬁned the notion of optimization of goal fulﬁllment into the technical notion of expected utility. The model of rational judgment used by decision scientists is one in which a person chooses options based on the option which has the largest expected utility (see Dawes, 1998; Hastie & Dawes, 2001; Wu, Zhang, & Gonzalez, 2004). The other aspect of rationality studied by cognitive scientists is termed epistemic rationality. This aspect of rationality concerns how well beliefs map onto the actual structure of the world. Epistemic rationality is

Development of Rational Thought

253

sometimes called theoretical rationality or evidential rationality (see Audi, 1993, 2001; Foley, 1987; Harman, 1995; Manktelow, 2004; Over, 2004). Instrumental and epistemic rationality are related. The aspect of beliefs that enter into instrumental calculations (i.e., tacit calculations) are the probabilities of states of affairs in the world. One of the fundamental advances in the history of modern decision science was the demonstration that if people’s preferences follow certain patterns (the so-called axioms of choice—things like transitivity and freedom from certain kinds of context effects) then they are behaving as if they are maximizing utility—they are acting to get what they most want (Edwards, 1954; Jeffrey, 1983; Luce & Raiffa, 1957; Savage, 1954; von Neumann & Morgenstern, 1944). This is what makes people’s degrees of rationality measurable by the experimental methods of cognitive science. Although it is difﬁcult to assess utility directly, it is much easier to assess whether one of the axioms of rational choice is being violated. This has been the logic of the seminal heuristics and biases research program inaugurated in the much-cited studies of Kahneman and Tversky (1972, 1973, 1979) (Tversky & Kahneman, 1974, 1981, 1983, 1986). Researchers in the heuristics and biases tradition have demonstrated in a host of empirical studies that people violate many of the strictures of rationality and that the magnitude of these violations can be measured experimentally. For example, people display conﬁrmation bias, test hypotheses inefﬁciently, display preference inconsistencies, do not properly calibrate degrees of belief, overproject their own opinions onto others, combine probabilities incoherently, and allow prior knowledge to become implicated in deductive reasoning (for summaries of the large literature, see Baron, 2000; Evans, 1989, 2007; Gilovich, Grifﬁn, & Kahneman, 2002; Kahneman & Tversky, 2000; Shaﬁr & LeBoeuf, 2002; Stanovich, 1999, 2004). These violations of rational strictures have spawned the biases that have been conjectured to explain the irrational thinking: base-rate neglect, framing effects, representativeness biases, anchoring biases, availability bias, outcome bias, and vividness effects, to name just a few. Degrees of rationality can be assessed in terms of the number and severity of such cognitive biases that individuals display. Failure to display a bias becomes a measure of rational thought. The thinking errors and biases that have been demonstrated in the judgment and decision-making literature have proliferated to the extent that a taxonomy of these systematic error types is badly needed. We have developed a taxonomy for the adult literature (Stanovich, 2008a, 2008b; Toplak et al., 2007), and in the remainder of this chapter will demonstrate its applicability to the sparse but growing literature on the development of rational thought. The taxonomy builds on the concepts of dual-process

254

Keith E. Stanovich et al.

theories of cognition. Because such theories have been discussed previously in Advances (see Klaczynski, 2004), we shall provide only a brief introduction in Section III before moving on to introduce the taxonomy.

III. Dual-Process Models of Cognition Virtually all attempts to classify heuristics and biases tasks end up utilizing a dual-process framework because most of the tasks in the heuristics and biases literature were deliberately designed to pit a heuristically triggered response against a normative response generated by the analytic system. As Kahneman (2000) notes, ‘‘Tversky and I always thought of the heuristics and biases approach as a two-process theory’’ (p. 682). Since Kahneman and Tversky launched the heuristics and biases approach in the 1970s, a wealth of evidence has accumulated in support of the dual-process approach. Evidence from cognitive neuroscience and cognitive psychology converges on the conclusion that mental functioning can be characterized by two different types of cognition having somewhat different functions and different strengths and weaknesses (Brainerd & Reyna, 2001; Evans, 2003, 2008a, 2008b; Evans & Over, 1996, 2004; Kahneman & Frederick, 2002, 2005; Metcalfe & Mischel, 1999; Sloman, 1996, 2002; Stanovich, 1999). There are many such theories (over 20 dual-process theories are presented by Stanovich, 2004) and they have some subtle differences, but they are similar in that all distinguish autonomous from nonautonomous processing. The two types of processing were termed systems in earlier writings, but theorists have been moving toward more atheoretical characterizations so we shall follow Evans (2008b) in using the terms Type 1 and Type 2 processing. The deﬁning feature of Type 1 processing is its autonomy. Type 1 processes are termed autonomous because: (1) their execution is rapid, (2) their execution is mandatory when the triggering stimuli are encountered, (3) they do not put a heavy load on central processing capacity (i.e., they do not require conscious attention), (4) they do not depend on input from high-level control systems, and (5) they can operate in parallel without interfering with each other or with Type 2 processing. Type 1 processing would include: behavioral regulation by the emotions; the encapsulated modules for solving speciﬁc adaptive problems that have been posited by evolutionary psychologists; processes of implicit learning; and the automatic ﬁring of overlearned associations (see Evans, 2007, 2008a; Stanovich, 2004).

Development of Rational Thought

255

Type 2 processing contrasts with Type 1 processing on each of the critical properties that deﬁne the latter. Type 2 processing is relatively slow and computationally expensive—it is the focus of our awareness. Many Type 1 processes can operate at once in parallel, but only one (or a very few) Type 2 thoughts can be executing at once—Type 2 processing is thus serial processing. Type 2 processing is often language based. One of the most critical functions of Type 2 processing is to override Type 1 processing. All of the different kinds of Type 1 processing (processes of emotional regulation, Darwinian modules, associative and implicit learning processes) can produce responses that are irrational in a particular context if not overridden. In order to override Type 1 processing, Type 2 processing must display at least two (possibly related) capabilities. One is the capability of interrupting Type 1 processing and suppressing its response tendencies. Type 2 processing thus involves inhibitory mechanisms of the type that have been the focus of recent work on executive functioning (Hasher, Lustig, & Zacks, 2007; Miyake et al., 2000; Zelazo, 2004). But the ability to suppress Type 1 processing gets the job only half done. Suppressing one response is not helpful unless a better response is available to substitute for it. Where do these better responses come from? One answer is that they come from processes of hypothetical reasoning and cognitive simulation that are a unique aspect of Type 2 processing (Evans, 2007; Kahneman & Tversky, 1982; Nichols & Stich, 2003). When we reason hypothetically, we create temporary models of the world and test out actions (or alternative causes) in that simulated world. To reason hypothetically we must, however, have one critical cognitive capability— the ability to distinguish our representations of the real world from representations of imaginary situations. For example, when considering an alternative goal state different from the one we currently have, we must be able to represent our current goal and the alternative goal and to keep straight which is which. Likewise, we need to be able to differentiate the representation of an action about to be taken from representations of potential alternative actions we are considering. But the latter must not infect the former while the mental simulation is being carried out. Thus, in a much-cited article, Leslie (1987) modeled pretense by positing a so-called secondary representation (see Perner, 1991) that was a copy of the primary representation but that was decoupled from the world so that it could be manipulated—that is, be a mechanism for simulation. The important issue for our purposes is that decoupling secondary representations from the world and then maintaining the decoupling while simulation is carried out is a Type 2 processing operation. It is computationally taxing and greatly restricts the ability to do any other Type 2 operation. In fact, decoupling

Keith E. Stanovich et al.

256

operations might well be a major contributor to a distinctive Type 2 property—its seriality. Cognitive decoupling must take place when an individual engages in a simulation of alternative worlds in order to solve a problem. Problemsolving tasks that necessitate fully disjunctive reasoning (see Johnson-Laird, 2006; Shaﬁr, 1994) provide examples of the situations that require fully decoupled simulation. Fully disjunctive reasoning involves considering all possible states of the world when deciding among options or when choosing a problem solution in a reasoning task. Consider the following problem, taken from the work of Levesque (1986, 1989) and studied by our research group (see Toplak & Stanovich, 2002): Jack is looking at Anne but Anne is looking at George. Jack is married but George is not. Is a married person looking at an unmarried person? (A) Yes

(B) No

(C) Cannot be determined

The vast majority of people answer (C) (cannot be determined) when in fact the correct answer to this problem is (A) (yes). To answer correctly, both possibilities for Anne’s marital status (married and unmarried) must be considered to determine whether a conclusion can be drawn. If Anne is married, then the answer is ‘‘Yes’’ because she would be looking at George who is unmarried. If Anne is not married, then the answer is still ‘‘Yes’’ because Jack, who is married, would be looking at Anne. Considering all the possibilities (the fully disjunctive reasoning strategy) reveals that a married person is looking at an unmarried person whether Anne is married or not. The fact that the problem does not reveal whether Anne is married suggests to people that nothing can be determined. Many people make the easiest (incorrect) inference from the information given and do not proceed with the more difﬁcult (but correct) inference that follows from fully disjunctive reasoning. Not all Type 2 processing represents fully explicit cognitive simulation, however. Or, to put it another way: all hypothetical thinking involves Type 2 processing (Evans & Over, 2004), but not all Type 2 processing involves hypothetical thinking. What has been termed serial associative cognition (Stanovich, 2008a) represents this latter category. It can be understood by considering a discussion of the four-card selection task in a theoretical paper on dual-processes by Evans (2006) (see also Evans & Over, 2004). In Wason’s (1966) four-card selection task the participant is told the following: Each of the boxes below represents a card lying on a table. Each one of the cards has a letter on one side and a number on the other side. Here is a rule: If a card has a

Development of Rational Thought

257

vowel on its letter side, then it has an even number on its number side. As you can see, two of the cards are letter-side up, and two of the cards are number-side up. Your task is to decide which card or cards must be turned over in order to ﬁnd out whether the rule is true or false. Indicate which cards must be turned over.

The participant chooses from four cards labeled K, A, 8, 5 (corresponding to not-P, P, Q, and not-Q). The correct answer is to pick the A and the 5 (P and not-Q), but the most common answer is to pick the A and 8 (P and Q)—the so-called matching response. Evans (2006) points out that the previous emphasis on the matching bias evident in the task (Evans, 1972, 1998, 2002; Evans & Lynch, 1973) might have led some investigators to infer that the analytic system is not actively engaged in the task. In fact, matching bias might be viewed as just one of several such suggestions in the literature that much thinking during the task is Type 1 processing (see Hardman, 1998; Margolis, 1987; Stanovich & West, 1998a; Tweney & Yachanin, 1985). In contrast, however, Evans (2006) presents evidence indicating that Type 2 processing may be going on during the task—even on the part of the majority who do not give the normatively correct response but instead give the PQ response. First, in discussing the card inspection paradigm (Evans, 1996) that he pioneered (see also Ball et al., 2003; Lucas & Ball, 2005; Roberts & Newton, 2001), Evans (2006) notes that although participants look disproportionately at the cards they will choose (the ﬁnding leading to the inference that heuristic processes were determining the responses), the lengthy amount of time they spend on those cards suggests that analytic thought (Type 2 processing) is occurring (if only to generate justiﬁcation for the heuristically triggered choices). Secondly, in verbal protocol studies, participants can justify their responses (indeed, can rationalize any set of responses they are told are correct; see Evans & Wason, 1976) with arguments that sometimes refer to the hidden side of cards chosen. In fact, Type 2 processing is occurring in this task, but it is not full-blown cognitive simulation of alternative world models. It is thinking of a shallower type—serial associative cognition. Serial associative cognition is not rapid and parallel like Type 1 processes, but is nonetheless rather inﬂexibly locked into an associative mode that takes as its starting point a model of the world that is given to the subject. In the inspection paradigm, individuals are justifying heuristically chosen responses (P and Q for the standard form of the problem), and the heuristically chosen responses are driven by the model given to the participant by the rule. Likewise, Evans and Over (2004) note that in the studies of verbal protocols, when participants made an incorrect choice, they referred to the hidden sides of the cards they are going to pick, but referred only to veriﬁcation when they did so. Thus, the evidence suggests that people

258

Keith E. Stanovich et al.

accept the rule as given, assume it is true, and simply describe how they would go about verifying it. The fact that they refer to hidden sides does not mean that they have constructed any alternative model of the situation beyond what was given to them by the experimenter and their own assumption that the rule is true. They then reason from this single focal model—systematically generating associations from this focal model but never constructing another model of the situation. This is why the central characteristic of serial associative cognition is that it displays a focal bias. One way in which to contextualize the idea of focal bias is as the second stage in a framework for thinking about human information processing that dates to the mid-1970s—the idea of humans as cognitive misers (Dawes, 1976; Taylor, 1981; Tversky & Kahneman, 1974). There are in fact two aspects of cognitive miserliness. Dual-process theory has heretofore highlighted only Rule 1 of the cognitive miser: default to Type 1 processing whenever possible. But defaulting to Type 1 processing is not always possible—particularly in novel situations where there are no stimuli available to domain-speciﬁc evolutionary modules. Type 2 processing procedures will be necessary, but a cognitive miser default is operating even there. Rule 2 of the cognitive miser is that, when Type 1 processing will not yield a solution: default to serial associative cognition with a focal bias (not fully decoupled cognitive simulation). The notion of a focal bias conjoins several closely related ideas in the literature—Evans, Over, and Handley’s (2003) singularity principle, Johnson-Laird’s (1999, 2005) principle of truth, focusing (Legrenzi, Girotto, & Johnson-Laird, 1993), the effect/effort issues discussed by Sperber, Cara, and Girotto (1995), and ﬁnally the focalism (Wilson et al., 2000) and belief acceptance (Gilbert, 1991) issues that have been prominent in the social psychological literature. Our notion of focal bias conjoins many of these ideas under the overarching theme that they all have in common—that humans will ﬁnd any way they can to ease the cognitive load and process less information. Focal bias combines all of these tendencies into the basic idea that the information processor is strongly disposed to deal only with the most easily constructed cognitive model. So the focal model that will dominate processing—the only model that serial associative cognition deals with—is the most easily constructed model. The most easily constructed model: tends to represent only one state of affairs; it accepts what is directly presented and models what is presented as true; it ignores moderating factors—probably because taking account of those factors would necessitate modeling several alternative worlds and this is just what a focal processing allows us to avoid. And ﬁnally, given the voluminous literature in cognitive science on belief bias and the informal reasoning literature on myside bias, the easiest models to represent clearly

Development of Rational Thought

259

appear to be those closest to what a person already believes and has modeled previously (e.g., Evans & Feeney, 2004; Stanovich & West, 2007). Thus, serial associative cognition is deﬁned by its reliance on a single focal model that triggers all subsequent thought. Framing effects (an exemplar task is presented later in this chapter), for instance, are a clear example of serial associative cognition with a focal bias. As Kahneman (2003) notes, ‘‘the basic principle of framing is the passive acceptance of the formulation given’’ (p. 703). The frame presented to the subject is taken as focal, and all subsequent thought derives from it rather than from alternative framings because the latter would necessitate more computationally expensive simulation operations.

IV. A Preliminary Taxonomy of Rational Thinking Errors To this brief overview of the architectural assumptions of dual-process theory, we need to add one reminder before proceeding to outline our taxonomy. That reminder is about the importance of knowledge bases. An aspect of dual-process theory that has been relatively neglected is that the override process is not simply procedural but instead utilizes content—that is, it uses declarative knowledge and strategic rules (linguistically coded strategies) to transform a decoupled representation. In the previous dualprocess literature, override has been treated as a somewhat disembodied process. The knowledge bases and strategies that are brought to bear on the secondary representations during the simulation process have been given little attention. Thus it is important to remember that Type 2 processes access not only knowledge structures but, importantly, accesses the person’s opinions, beliefs, and reﬂectively acquired goal structure. Also accessed are micro-strategies for cognitive operations and production system rules for sequencing behaviors and thoughts. Likewise Type 1 processing implicates not only encapsulated knowledge bases from evolutionary adaptations, but also information that has become tightly compiled and available to the Type 1 processing of the autonomous mind due to overlearning and practice. The rules, procedures, and strategies that can be retrieved and used to transform decoupled representations have been referred to as mindware (Perkins, 1995). If one is going to trump a Type 1 response with conﬂicting information or a learned rule, one must have previously learned the information or the rule. If, in fact, the relevant mindware is not available because it has not been learned, then the cause of the thinking error will be missing mindware rather than override failure. This distinction in fact represents the beginning of a taxonomy of the causes of cognitive failure

Keith E. Stanovich et al.

260

related to rational behavior that we have used to organize the heuristics and biases literature (Stanovich, 2008a, 2008b), and to classify various practical problems of rational thinking—for example, to understand the thinking problems of pathological gamblers (Toplak et al., 2007). Figure 1 presents an initial attempt at a taxonomy of rational thinking problems. Presented at the top of Figure 1 are the two defaults of the cognitive miser listed in order of relative cognitive engagement. The characteristic presented ﬁrst is defaulting to the response options primed by Type 1 processing. It represents the shallowest kind of processing because Categories of Cognitive Failure Default toType 1 Processing Cognitive Miserliness Serial Associative Cognition with a Focal Bias

Override Failure

Failure of Sustained Decoupling

Probability Knowledge Mindware Gap

Importance of Alternative Hypotheses Many Domain-Specific Knowledge Structures Lay Psychological Theory Evaluation-Disabling Strategies

Contaminated Mindware

“Self” Encourages Egocentric Processing Many Domain-Specific Knowledge Structures

Fig. 1. A preliminary taxonomy of thinking errors.

Development of Rational Thought

261

no Type 2 processing is done at all. The second type of processing tendency of the cognitive miser is to engage in serial associative cognition with a focal bias. This characteristic represents a tendency to over-economize during Type 2 processing—speciﬁcally, to fail to engage in the full-blown simulation of alternative worlds or to engage in fully disjunctive reasoning. The third category of rational thinking problems represented is that of override failure. Here, unlike in the ﬁrst two cases, Type 2 cognitive decoupling is engaged. Inhibitory Type 2 processes try to take the Type 1 processing ofﬂine, but they fail. So in override failure, cognitive decoupling does take place, but it fails to suppress Type 1 processing. Portrayed next in Figure 1 are categories of cognitive failure that are related to mindware problems. Mindware problems are divided into mindware gaps and contaminated mindware. When an override of Type 1 processing is necessary but the mindware necessary for a substitute response is not available, then we have a case of a mindware gap. Although mindware gaps may lead to sub-optimal reasoning, the next category in the taxonomy is designed to draw attention to the fact that not all mindware is helpful—either to goal attainment or to epistemic accuracy. In fact, some acquired mindware can be the direct cause of irrational actions that thwart our goals. Such effects thus deﬁne another category in the taxonomy of cognitive failures: contaminated mindware. Turning ﬁrst to the category of mindware gaps, the curved rectangles in the ﬁgure are meant to represent missing knowledge bases. We have not represented an exhaustive set of knowledge partitionings—to the contrary, we have represented a minimal sampling of a potentially large set of coherent knowledge bases in the domains of probabilistic reasoning, causal reasoning, logic, and scientiﬁc thinking, the absence of which could result in irrational thought or behavior. The two represented are mindware categories that have been implicated in research in the heuristics and biases tradition: missing knowledge about probability and probabilistic reasoning strategies; and ignoring alternative hypotheses when evaluating hypotheses. These are just a few of many mindware gaps that have been suggested in the literature on behavioral decision making. There are many others, and the box labeled ‘‘Many Domain-Speciﬁc Knowledge Structures’’ indicates this. Finally, at the bottom of Figure 1 is the category of contaminated mindware. The curved rectangles represent problematic knowledge and strategies. They do not represent an exhaustive partitioning (the mindwarerelated categories are too diverse for that), but instead represent some of the mechanisms that have received some discussion in the literature. First is a subcategory of contaminated mindware that is much discussed in the literature—mindware that contains evaluation-disabling properties.

262

Keith E. Stanovich et al.

Some of the evaluation-disabling properties that help keep some mindware lodged in their hosts are: the promise of punishment if the mindware is questioned; the promise of rewards for unquestioning faith in the mindware; or the thwarting of evaluation attempts by rendering the mindware unfalsiﬁable (Dawkins, 1993; Dennett, 2006; Lynch, 1996; Stanovich, 2004). The second subcategory of contaminated mindware that has been discussed by several theorists is a concept of ‘‘self ’’ that serves to encourage egocentric thinking (Blackmore, 1999; Dennett, 1991, 1995). The self, according to these theorists, is a mechanism that fosters one characteristic of focal bias: that we tend to build models of the world from a single myside perspective. The egocentrism of the self was of course evolutionarily adaptive. Nonetheless, it is sometimes nonoptimal in a technological environment different from the environment of evolutionary adaptation. For example, myside processing makes difﬁcult such modern demands as: unbiasedness; sanctioning of nepotism; and discouragement of familial, racial, and religious discrimination. Finally, the last subcategory of contaminated mindware pictured in Figure 1 is meant to represent what is actually a whole set of categories: mindware representing speciﬁc categories of information or maladaptive culturally conditioned beliefs. As with the mindware gap category, there may be a large number of misinformation-ﬁlled or contaminated mindware that would support irrational thought and behavior. Lay psychological theory (or, folk theory) refers to the theories that people have about their own minds and is represented as both contaminated mindware and a mindware gap in Figure 1. Mindware gaps in this domain would be represented by the many things about our own minds that we do not know; for example, how quickly we will adapt to both fortunate and unfortunate events (Gilbert, 2006). Other things we think we know about our own minds are wrong. These misconceptions represent contaminated mindware. An example would be the folk belief that we accurately know our own minds. This contaminated mindware accounts for the incorrect belief that we always know the causes of our own actions (Nisbett & Wilson, 1977) and thinking that although others display myside and other thinking biases, we ourselves have special immunity from the very same biases (Ehrlinger, Gilovich, & Ross, 2005; Pronin, 2006). Finally, note the curved, double-headed arrow in Figure 1 indicating an important relation between the override failure category and the mindware gap category. In case of override failure, an attempt must be made to trump a response primed by Type 1 processing with alternative information or a learned rule. For an error to be classiﬁed as an override failure, one must have previously learned the alternative information or an alternative rule

Development of Rational Thought

263

different from the Type 1 response. If, in fact, the relevant mindware is not available because it has not been learned—or at least not learned to the requisite level to sustain override—then we have a case of a mindware gap rather than override failure. Note one interesting implication of the relation between override failure and mindware gaps—the fewer gaps one has, the more likely that an error may be attributable to override failure. Errors made by someone with considerable mindware installed are more likely to be due to override failure than to mindware gaps. Of course, the two categories trade off in a continuous manner with a fuzzy boundary between them. A well-learned rule not appropriately applied is a case of override failure. As the rule is less and less well instantiated, at some point it is so poorly compiled that it is not a candidate to override the Type 1 response and thus the processing error becomes a mindware gap.

V. Classifying Heuristics and Biases Table I classiﬁes many of the thinking errors discussed in the heuristics and biases literature in terms of the taxonomy in Figure 1. For example, the three Xs in the ﬁrst column signify three phenomena that represent defaulting to Type 1 processing: vividness effects, affect substitution, and impulsively associative thinking. Defaulting to the most vivid stimulus is a common way that the cognitive miser avoids Type 2 processing (Nisbett & Ross, 1980). Likewise defaulting to affective valence is often used in situations with emotional salience. And affect substitution is a speciﬁc form of a more generic trick of the cognitive miser, attribute substitution (Kahneman & Frederick, 2002)—substituting for a hard question an easier one that requires only Type 1 processing. Previously, we discussed a disjunctive reasoning problem from the work of Levesque (1986, 1989): ‘‘Jack is looking at Anne but Anne is looking at George.’’ Failure on problems of this type is an example of the intellectual laziness termed impulsively associative thinking (Stanovich, 2008a, 2008b). Here, participants look for any simple association that will prevent them from having to engage in Type 2 thought (in this case associating Anne’s unknown status with the response ‘‘cannot be determined’’). The second category of thinking error presented in Table I is overreliance on serial associative cognition with a focal bias (the most easily constructed model). This error often occurs in novel situations where some Type 2 processing is necessary. Framing effects are the example here (‘‘the basic principle of framing is the passive acceptance of the formulation given,’’ p. 703, Kahneman, 2003). The frame presented to the subject is

264

Table I A Basic Taxonomy of Thinking Errors Tasks, effects, and processing styles

The cognitive miser

Default to Type 1 processing

Failure of sustained decoupling

Mindware gaps (MG)

Probability knowledge

Alternative thinking

MG and CM

Contaminated mindware (CM)

Lay psychological theory

Evaluation disabling strategies

Self and egocentric processing

X X

Keith E. Stanovich et al.

Vividness effects Attribute substitution Impulsively associative thinking Framing effects Denominator neglect Belief bias Self-control problems Conjunction errors Noncausal baserates Bias blind spot Four-card selection Myside processing Affective forecasting errors Conﬁrmation bias

Focal bias

Override failure

X

X X X X X X X X X X

X

X

X

X X X

Development of Rational Thought

265

taken as focal, and all subsequent thought derives from it rather than from alternative framings because the latter would require more thought. Pure override failure—the third category of thinking errors presented in Table I—is illustrated by three effects: belief bias effects (e.g., De Neys, 2006; Dias, Roazzi, & Harris, 2005; Evans, 2007; Evans, Barston, & Pollard, 1983; Evans & Curtis-Holmes, 2005; Evans & Feeney, 2004; Markovits & Nantel, 1989), denominator neglect (Denes-Raj & Epstein, 1994; Kirkpatrick & Epstein, 1992; Pacini & Epstein, 1999) and self-control problems such as the inability to delay gratiﬁcation (Ainslie, 2001; Baumeister & Vohs, 2007; Loewenstein, Read, & Baumeister, 2003; Mischel, Shoda, & Rodriguez, 1989; Rachlin, 2000). (Belief bias and denominator neglect are discussed below). Table I also portrays two examples of mindware gaps related to missing probability knowledge: noncausal base-rate usage and conjunction errors (Kahneman & Tversky, 1972, 1973; Tversky & Kahneman, 1983). Listed next is the bias blind spot—the fact that people view other people as more biased than themselves (Pronin, 2006). The bias blind spot is thought to arise because people have incorrect lay psychological theories. They think, incorrectly, that biased thinking on their part would be detectable by conscious introspection. In fact, most social and cognitive biases operate unconsciously. Several of the remaining tasks illustrated in Table I represent irrational thought problems that are hybrids. That is, they are co-determined by several different cognitive difﬁculties. For example, we speculate that problems with the Wason four-card selection task (Evans, 2007; Wason, 1966, 1968) are multiply determined. People may have trouble with that task because they have not well instantiated the mindware of alternative thinking—the learned rule of the value in thinking of the false situation or thinking about a hypothesis other than the one you have. Alternatively, people might have trouble with the task because of a focal bias: they focus on the single model given in the rule (e.g., ‘‘a vowel must have even number on its other side’’) and do all of their reasoning from only this assumption without ﬂeshing out other possibilities. Table I represents both of these possibilities. Another thinking error with multiple determinants is myside processing (Baron, 1995; Klaczynski & Lavallee, 2005; Perkins, 1985, 1995; Stanovich & West, 2007). Excessive myside thinking is no doubt fostered by contaminated mindware—our notion of ‘‘self ’’ makes us egocentrically think that the world revolves around ourselves. But a form of focal bias may be contributing to that error as well—the bias to base processing on the mental model that is the easiest to construct. What easier model is there to construct than a model based on our own previous beliefs and

266

Keith E. Stanovich et al.

experiences? Such a focal bias is different from the egocentric mindware of the self. The focal bias is not egocentric in the motivational sense that we want to build our self esteem or sense of self worth. The focal bias is simply concerned with conserving computational capacity and it does so in most cases by encouraging reliance on a model from a myside perspective. Both motivationally driven ‘‘self ’’ mindware and computationally driven focal biases might be contributing to myside processing, making it another multiply determined bias. Errors in affective forecasting are likewise multiply determined (see Table I). Affective forecasting refers to our ability to predict what will make us happy in the future. People are surprisingly poor at affective forecasting (Gilbert, 2006; Hsee & Hastie, 2006; Kahneman, Diener, & Schwarz, 1999; Kahneman et al., 2006). We often make choices that reduce our happiness because we ﬁnd it hard to predict what will make us happy. For instance, people underestimate how quickly they will adapt to both fortunate and unfortunate events. One reason that people overestimate how unhappy they will be after a negative event is that they have something missing from their lay psychological theories—the personal theories they use to explain their own behavior. They fail to take into account the rationalization and emotion-dampening protective thought they will engage in after the negative event (‘‘I really didn’t want the job anyway,’’ ‘‘colleagues told me he was biased against older employees’’). People’s lay theories of their own psychology do not give enough weight to these factors and thus they fail to predict how much their own psychological mechanisms will damp down any unhappiness about the negative event. Another even more important source of affective forecasting errors is focal bias. Researchers in the affective forecasting literature have theorized speciﬁcally about focalism interfering with hedonic predictions (‘‘predictors pay too much attention to the central event and overlook context events,’’ p. 31, Hsee & Hastie, 2006). For example, a sports fan overestimates how happy the victory of the home team will make him two days after the event. When making the prediction, he ﬁxates on the salient focal event—winning the game—simulates the emotion he will feel in response to the event, and projects that same emotion two days into the future. What does not enter into his model—because such models are not easy to construct in imagination (hence too effortful for the cognitive miser)—is the myriad of other events that will be happening two days after his game and that will then impinge on his happiness in various ways (most of these other events will not be as happiness-inducing as was winning the game). Finally, Table I simply notes with one example that there may be even more complex effects in the heuristics and biases literature. The confusing concept of conﬁrmation bias (Evans, 1989, 2007; Klayman & Ha, 1987;

Development of Rational Thought

267

Nickerson, 1998) may be an example. Depending on how it is deﬁned, it could result from focal bias or from a failure to have instantiated the thinking mindware that prompts a consideration of alternative hypotheses. In more motivationally based accounts however, conﬁrmation bias might arise because of evaluation-disabling strategies embodied in contaminated mindware.

VI. Exemplar Developmental Studies in the Different Categories of the Taxonomy Table I is not an exhaustive listing of heuristics and biases tasks, but it does give a ﬂavor for how some much-cited effects and biases ﬁt into the framework (for other attempts to classify rational thinking errors, see Arkes, 1991; Harvey, 2007; Larrick, 2004; McFadden, 1999; Reyna et al., 2003; Stanovich, 2008b). However, the complexity of even this partial list will help to explain why we earlier characterized the literature on the developmental of rational thinking ability as sparse. Although there have been some initial developmental studies on most of the tasks in the heuristics and biases literature, none of the biases in the list has been the subject of intense investigation. In short, the literature is spread widely, but it is thin. In this section, we discuss exemplar developmental studies from several of the categories in the taxonomy. A. DEFAULT TO TYPE 1 PROCESSING: VIVIDNESS EFFECTS One of the most common Type 1 processing defaults of the cognitive miser is the tendency to default to vivid presentations of information and to avoid nonsalient numerical presentations of evidence. In the heuristics and biases literature, a typical problem would require the participant to make an inductive inference in a simulation of a real-life decision. The information relevant to the decision is conﬂicting and is of two different types. One type of evidence is statistical: either probabilistic or aggregate base-rate information that favors one of the bipolar decisions. The other evidence is a concrete case or vivid personal experience that points in the opposite direction. The classic Volvo versus Saab item (see p. 285 of Fong, Krantz, & Nisbett, 1986) provides an example. In this problem, a couple is deciding to buy one of two otherwise equal cars. Consumer surveys, statistics on repair records, and polls of experts favor the Volvo over the Saab. However, a friend reports experiencing a severe mechanical problem with the Volvo he owns. The participant is asked to provide advice to the

Keith E. Stanovich et al.

268

couple. Preference for the Volvo indicates a tendency to rely on the largesample information in spite of salient personal testimony. A preference for the Saab indicates reliance on the personal testimony over the opinion of experts and the large-sample information. Kokis et al. (2002) adapted several problems such as this for children. For example, one problem went as follows: Erica wants to go to a baseball game to try to catch a ﬂy ball. She calls the main ofﬁce and learns that almost all ﬂy balls have been caught in section 43. Just before she chooses her seat, she learns that her friend Jimmy caught 2 ﬂy balls last week sitting in section 10. Which section is most likely to give Erica the best chance to catch a ﬂy ball? (a) (b) (c) (d)

Deﬁnitely section 43 Probably section 43 Probably section 10 Deﬁnitely section 10

Selection of option (a) or (b) indicates the use of the aggregate base-rate information. Selection of options (c) or (d) indicates that the child is using the vivid information from a friend (that happens to be of lower diagnosticity). Kokis et al. (2002) found a signiﬁcant developmental trend whereby 13–14-year-olds displayed signiﬁcantly less reliance on the vivid personal information than did a group of 10–11-year-olds. Jacobs and Potenza (1991) found an analogous signiﬁcant developmental trend in the so-called object condition of their experiment where the problems were similar to those used by Kokis et al. (2002). However, in the so-called social condition of the Jacobs and Potenza study, the developmental trend was in the opposite direction—more reliance on the vivid information and less reliance on the more diagnostic statistical information by the older children. A consideration of the nature of the social problems reveals why this was the case. Here is an example of a social problem: In Juanita’s class, 10 girls are trying out to be cheerleaders and 20 are trying out for the band. Juanita is very popular and very pretty. She is always telling jokes and loves to be around people. Do you think Juanita is trying out to be a cheerleader or for the band?

Here, the statistical information points in the direction of band but the personal information points in the direction of cheerleader. But to understand the diagnosticity of the indicant information in this problem one must have knowledge of a social stereotype (that popular girls are drawn more to cheerleading than to band). Knowledge of this stereotype might well increase with age and thus be less available to the younger children. In short, the indicant information is less available to the younger children in the social condition.

Development of Rational Thought

269

Thus, the performance in the Jacobs and Potenza (1991) study is less inconsistent with the ﬁndings of Kokis et al. (2002) than may be apparent on the surface. The same is true of the social problems in a study by Davidson (1995). Finally, the developmental trend in Kokis et al. and in the object condition of Jacobs and Potenza (1991) is consistent with studies of individual differences within an age group. Reliance on vivid individuating information is negatively correlated with cognitive ability (Kokis et al., 2002; Stanovich & West, 1998b, 1998c).

B. FOCAL BIAS: FRAMING EFFECTS The second default of the cognitive miser is to default to serial associative cognition with a focal bias. Framing effects represent the classic example of this default in the heuristics and biases literature. For example, in discussing the mechanisms causing framing effects, Kahneman has stated that ‘‘the basic principle of framing is the passive acceptance of the formulation given’’ (2003, p. 703). The frame presented to the subject is taken as focal, and all subsequent thought derives from it rather than from alternative framings because the latter would require more thought. One of the most compelling framing demonstrations is from the early work of Tversky and Kahneman (1981): Decision 1. Imagine that the U.S. is preparing for the outbreak of an unusual disease, which is expected to kill 600 people. Two alternative programs to combat the disease have been proposed. Assume that the exact scientiﬁc estimates of the consequences of the programs are as follows: If Program A is adopted, 200 people will be saved. If Program B is adopted, there is a one-third probability that 600 people will be saved and a two-thirds probability that no people will be saved. Which of the two programs would you favor, Program A or Program B?

Most people when given this problem prefer Program A—the one that saves 200 lives for sure. There is nothing wrong with this choice taken alone. However, inconsistent responses to another problem deﬁne a framing effect: Decision 2. Imagine that the U.S. is preparing for the outbreak of an unusual disease, which is expected to kill 600 people. Two alternative programs to combat the disease have been proposed. Assume that the exact scientiﬁc estimates of the consequences of the programs are as follows: If Program C is adopted, 400 people will die. If Program D is adopted, there is a one-third probability that nobody will die and a two-thirds probability that 600 people will die. Which of the two programs would you favor, Program C or Program D?

Most people when presented with Decision 2 prefer Program D. Thus, across the two problems, the most popular choices are Program A and

270

Keith E. Stanovich et al.

Program D. The problem here is that Decision 1 and Decision 2 are really the same decision—they are merely redescriptions of the same situation. Program A and C are the same. That 400 will die in Program C implies that 200 will be saved—precisely the same number saved (200) in Program A. Likewise, the two-thirds chance that 600 will die in Program D is the same two-thirds chance that 600 will die (‘‘no people will be saved’’) in Program B. Many people show inconsistent preferences—their choice switches depending on the phrasing of the question. This is an example of a problem with very transparent equivalence. When presented with both versions of the problem together, most people agree that the problems are identical and that the alternative phrasing should not have made a difference. Such a lack of so-called descriptive invariance is a very fundamental violation of some of the simplest strictures of rational thought (see Tversky & Kahneman, 1981, 1986). A theory of why these framing effects occur was presented in the prospect theory of Kahneman and Tversky (1979) which contains the key assumption that the utility function is steeper (in the negative direction) for losses than for gains. This explains why people tend to be more risk averse for gains than for losses. The literature on framing effects in adults is vast (see Kahneman & Tversky, 1984, 2000; Kuhberger, 1998; Levin et al., 2002; Maule & Villejoubert, 2007). However, the developmental literature is quite small. Obviously, the complexity of the problems has to be vastly reduced and made appropriate for children. Outcomes in developmental studies become small prizes that the children receive instead of the imaginary deaths or real money that is used in adult studies. Several investigators have creatively adapted framing paradigms for children, but the results of these experiments have not converged. Levin and colleagues (Levin & Hart, 2003; Levin et al., 2007) found no developmental trend for framing effects. Children (6–8-year-olds) were more risk averse for gains than for losses in the manner that prospect theory predicts, but the magnitude of the framing effects that they displayed was the same as that found for adults. The results of Levin and colleagues were not completely convergent with those in a study by Reyna and Ellis (1994). The data patterns in the latter study were complex, however, and highly variable over the ages studied. Framing interacted with level of risk and magnitude of reward at certain ages. Brieﬂy though, preschoolers’ responses were consistent across frames—they displayed no framing effect. Second graders displayed a small reverse framing effect—they were more risk averse for losses. Fifthgraders displayed a small reverse framing effect at the highest reward magnitude and the standard framing effect (more risk seeking for losses) only for lowest reward magnitude.

Development of Rational Thought

271

The results of Levin and colleagues and those of Reyna and Ellis (1994) are consistent in one important respect however—neither study found a monotonically decreasing framing effect with age. This lack of developmental trend converges with studies of individual differences within an age group. Framing effects show very small (sometimes zero) correlations with cognitive ability (Bruine de Bruin et al., 2007; Parker & Fischhoff, 2005; Stanovich & West, 1999, 2008b; Toplak & Stanovich, 2002).

C. OVERRIDE FAILURE: DENOMINATOR NEGLECT One of several phenomena in the heuristics and biases literature that illustrates the failure of sustained decoupling (see Table I) is the phenomenon of denominator neglect. Epstein and colleagues (Denes-Raj & Epstein, 1994; Kirkpatrick & Epstein, 1992; Pacini & Epstein, 1999) demonstrated that it can result in a startling failure of rational judgment. Adults in several of his experiments were presented with two bowls of jelly beans. In the ﬁrst were nine white jelly beans and one red jelly bean. In the second were 92 white jelly beans and 8 red jelly beans. A random draw was to be made from one of the two bowls and if the red jelly bean was picked, the participant would receive a dollar. The participant could choose which bowl to draw from. Although the two bowls clearly represent a 10% and an 8% chance of winning a dollar, a number of subjects chose the 100 bean bowl, thus reducing their chance of winning. The majority did pick the 10% bowl, but a healthy minority (from 30 to 40% of the participants) picked the 8% bowl. Although most of these participants in the minority were aware that the large bowl was statistically a worse bet, that bowl also contained more enticing winning beans—the 8 red ones. In short, the tendency to respond to the absolute number of winners, for these participants, trumped the formal rule (pick the one with the best percentage of reds) that they knew was the better choice. Kokis et al. (2002) found no signiﬁcant trend for denominator neglect to decrease across their two age groups of 10–11-year-olds and 13–14-year-olds. However, Kokis et al. (2002) did ﬁnd that cognitive ability was negatively correlated with denominator neglect to a signiﬁcant degree. Kokis et al. used a paradigm very similar to that of Epstein and colleagues, but Klaczynski (2001b) altered the paradigm in an interesting way, by presenting options with equal probabilities and a response option that reﬂected the equivalent status of the two options. He had participants select from three alternatives: (a) a jar with 1 winning ticket out of 10; (b) a jar with 10 winning tickets out of 100; and (c) that it would not matter which jar was picked. Picking alternative (b) would indicate denominator neglect. Alternative (c) is the normatively correct response. Only a minority of

Keith E. Stanovich et al.

272

participants chose the normatively correct response, but the normatively correct response did increase across three groups: early adolescents, middle adolescents, and young adults. In contrast, the majority response— denominator neglect—did not show a developmental trend. D. OVERRIDE FAILURE: BELIEF BIAS Clearer results have been obtained with another phenomenon caused by the failure of sustained decoupling—belief bias (see Table I). Consider the following syllogism: Premise 1. All living things need water Premise 2. Roses need water Therefore, Roses are living things Judge the conclusion as either logically valid or invalid.

Premise 1 says that all living things need water, not that all things that need water are living things. So, just because roses need water, it does not follow from Premise 1 that they are living things. However, consider the following syllogism with exactly the same structure: Premise 1. All insects need oxygen Premise 2. Mice need oxygen Therefore, Mice are insects

In both problems, prior knowledge about the nature of the world (that roses are living things and that mice are not insects) is becoming implicated in a type of judgment that is supposed to be independent of content: judgments of logical validity. In the rose problem, prior knowledge was interfering, and in the mice problem prior knowledge was facilitative. Belief bias occurs when judgments of the believability of the conclusion interfere with judgments of logical validity. Using syllogistic reasoning problems suitably modiﬁed for children, Kokis et al. (2002) found a signiﬁcant developmental trend whereby 13–14-year-olds displayed signiﬁcantly less belief bias than did a group of 10–11-year-olds. The developmental trend in Kokis et al. is consistent with studies of individual differences within an age group, in which belief bias is negatively correlated with cognitive ability (De Neys, 2006; Handley et al., 2004; Sa´, West, & Stanovich, 1999; Stanovich & West, 1998b).

Development of Rational Thought

273

E. MINDWARE GAPS Mindware gaps are common in the domain of probability knowledge and judgment. Many notable heuristics and biases cluster in that domain (see the classic studies by Kahneman & Tversky, 1972, 1973). Conjunction effects represent rational thinking errors that arise because of mindware gaps in the domain of probability. Consider another problem that is famous in the literature of cognitive psychology, the so-called Linda problem (Tversky & Kahneman, 1983). Linda is 31-years-old, single, outspoken, and very bright. She majored in philosophy. As a student, she was deeply concerned with issues of discrimination and social justice, and also participated in anti-nuclear demonstrations. Please rank the following statements by their probability, using 1 for the most probable and 8 for the least probable. a. b. c. d. e. f. g. h.

Linda is a teacher in an elementary school ____ Linda works in a bookstore and takes Yoga classes ____ Linda is active in the feminist movement ____ Linda is a psychiatric social worker ____ Linda is a member of the League of Women Voters ____ Linda is a bank teller ____ Linda is an insurance salesperson ____ Linda is a bank teller and is active in the feminist movement ____

Most people make what is called a ‘‘conjunction error’’ on this problem. Because alternative h (Linda is a bank teller and is active in the feminist movement) is the conjunction of alternatives c and f, the probability of h cannot be higher than that of either c (Linda is active in the feminist movement) or f (Linda is a bank teller). All feminist bank tellers are also bank tellers, so h cannot be more probable than f—yet often over 80% of the adults in studies rate alternative h as more probable than f, thus displaying a conjunction error. Davidson (1995) reported the counterintuitive ﬁnding of a developmental trend of increasing conjunction errors with age. However, the stimuli in this study had the same problem as those in the social condition of the Jacobs and Potenza (1991) study discussed above—they depended on the knowledge of a stereotype that might increase with age. For example, one item went as follows: Mrs. Hill is not in the best health and she has to wear glasses to see. Her hair is gray and she has wrinkles. She walks kind of hunched over. Do you think that Mrs. Hill is: a waitress in a local restaurant; an old person who has grandchildren and a waitress at a local restaurant; etc.

274

Keith E. Stanovich et al.

Knowledge of the stereotype might lead children to make a conjunction error, but the knowledge of the stereotype is undoubtedly more extensive among sixth graders (Davidson’s oldest group) than among second graders (Davidson’s youngest group). Another way to state the problem with the Davidson (1995) stimuli is to say that the potential of the stimuli to trigger a representativeness judgment is confounded with age. Such stimuli are ﬁne for certain questions but they are not appropriate stimuli to use when addressing the generic question of whether the propensity to commit a conjunction error with age increases or decreases. Klaczynski (2001a) used stimuli that were less confounded in this manner and found no developmental trend in the number of conjunction errors from early adolescence (mean age 12.4 years) to middle adolescence (16.3 years). A developmental trend in favor of more optimal reasoning by older than younger adolescents has been found in other domains of probabilistic reasoning, including the gambler’s fallacy and the law of large numbers (Klaczynski, 2000, 2001a, 2001b; Klaczynski & Narasimham, 1998).

F. HYBRID REASONING PROBLEMS As Table I illustrates, several thinking errors in the heuristics and biases literature are multiply determined. In Wason’s four-card selection task, at the very least incorrect responses are due to focal bias and to the failure to understand the importance of thinking about alternative hypotheses. There may be even more sources of error on this problem (Evans, 2007; Klauer, Stahl, & Erdfelder, 2007; Osman & Laming, 2001). Developmental and individual differences studies of the task are consistent, however. Superior performance on the task is associated with development (Klaczynski, 2001a; Overton, Byrnes, & O’Brien, 1985) and with cognitive ability (DeShon et al., 1998; Stanovich & West, 1998a; Toplak & Stanovich, 2002; Valentine, 1975). As a ﬁnal example of a multiply determined cognitive bias, consider myside bias: People evaluate evidence, generate evidence, and test hypotheses in a manner biased toward their own opinions (Baron, 1991, 1995; Kuhn, 1991; Perkins, 1985; Perkins, Farady, & Bushey, 1991; Stanovich & West, 2007; Toplak & Stanovich, 2003). Myside bias derives from focal bias but also from mindware that structures our knowledge of the world from an egocentric perspective. Developmental trends in myside processing have been studied by Klaczynski and colleagues (1997; Klaczynski & Gordon, 1996; Klaczynski, Gordon, & Fauth, 1997; Klaczynski & Lavallee, 2005; Klaczynski &

Development of Rational Thought

275

Narasimham, 1998). They presented participants with ﬂawed hypothetical experiments that led to either opinion-consistent or opinion-inconsistent conclusions and evaluated the quality of the reasoning used when the participants critiqued the ﬂaws in the experiments. In these experiments, myside bias effects were also evident—participants found more ﬂaws when the experiment’s conclusions were inconsistent than when they were consistent with their opinions and beliefs. However, there was no developmental trend for myside bias to increase or decrease—at least for the early, middle, and late adolescent groups that were the focus of most of the Klaczynski lab’s research. The lack of a developmental trend in myside bias in the Klaczynski group’s studies is consistent with studies of individual differences within an age group. Across a variety of myside paradigms, there has been little evidence that myside bias is associated with cognitive ability (Klaczynski & Lavallee, 2005; Klaczynski & Robinson, 2000; Macpherson & Stanovich, 2007; Sa´ et al., 2005; Stanovich & West, 2007, 2008a; Toplak & Stanovich, 2003).

VII. Conclusion: Speciﬁcity and Generality in the Development of Rational Thought The overall conclusion we derive from our survey of results from our own laboratory and others is that the development of rational thought has been inadequately speciﬁed. Collectively, these biases (and many others in the heuristics and biases literature not covered here) deﬁne departures from rationality and hence indirectly index rationality itself. The minimal conclusion to be drawn from the body of developmental work taken as a whole is that children unequivocally do show every one of the biases that have been displayed in the adult literature: vividness effects, framing effects, denominator neglect, belief bias, conjunction errors, myside bias, etc. Beyond this minimal conclusion, however, few conclusions about development appear to generalize across the various domains and biases. First, none of the areas covered in this review have been the subject of enough research with convergent outcomes to establish developmental trends with conﬁdence. This caveat aside, the suggestive trends that are in the literature vary considerably across the various cognitive biases. Across the age ranges studied, there appear to be developmental increases in the avoidance of belief bias, analytic responding in the selection task, and reliance on statistical information in the face of conﬂicting personal

276

Keith E. Stanovich et al.

testimonials. All of these trends are in the direction of increasingly rational thought, at least according to accepted normative standards. These ﬁndings contrast, however, with the trends apparent in developmental studies of myside bias and framing effects. Neither of these biases was attenuated by development. The lack of developmental decreases in these two biases is interestingly convergent with ﬁndings that neither bias displays much of a correlation with intelligence (Stanovich & West, 2007, 2008b). These aspects of rational thought are thus quite independent of cognitive ability, reinforcing the fundamental point with which we started this chapter: measures of cognitive ability, as traditionally deﬁned, fail to assess degrees of rationality. In the case of denominator neglect and conjunction errors, the literature on developmental trends is simply too inconsistent and sparse to warrant any conclusions at this point. These two effects highlight how thin the literature is on children’s rational thought. One cannot amalgamate all of the studies reviewed here (as well as others not reviewed) as evidence on a single issue—this would misleadingly suggest that the literature is more extensive than it is. Rational thought spans many domains—it encompasses many different thinking dispositions and knowledge domains, each of which has been investigated separately in the adult literature. For example, there is an enormous literature on conjunction effects (Fisk, 2004), on framing (Kahneman & Tversky, 2000), on base-rate usage (Koehler, 1996), and on every one of the myriad effects in the heuristics and biases literature. A parallel effort in each of these domains will be necessary in order to fully understand the development of rational thought. Understanding the development of rationality will clearly be a tall order. It will be worth the effort, however, not just for scientiﬁc reasons. Assumptions about the nature and development of rationality are implicated in judgments of legal responsibility. Reyna and Farley (2006) have recently emphasized how background assumptions about adolescent rationality frame efforts to change adolescent risk behavior. For example, theories that stress adolescent feelings of invulnerability serve to avoid the attribution of irrationality to adolescents who engage in high-risk behavior. If these adolescents have strong feelings of invulnerability, or if they drastically underestimate the probabilities of negative outcomes, then a consequentialist calculation might well make engaging in high-risk behaviors rational for them. An alternative approach, one supported by the research in the heuristics and biases tradition, would relax the rationality assumption, and conclude instead that some of these adolescent behaviors violate standard rational strictures (Stanovich, 2004, 2008c). Such a stance ﬁnds additional motivation from an observation that Reyna and Farley (2006)

Development of Rational Thought

277

discuss: many adolescents making poor choices are alienated from the choices they make. As Reyna and Farley put it ‘‘people who take unhealthy risks often agree that their behavior is irrational, on sober reﬂection, but they gave in to temptation or were not thinking at the time of the decision, and are worse off for having done so’’ (p. 35). Instead of the economics-like assumption of adolescents as coherent rational actors, dual-process theories of the type we have discussed highlight the image of a decision maker in conﬂict. This comports well with the fact that many adolescents with behavior problems will indeed verbally reject their own behavior. Such a philosophical reorientation could, as Reyna and Farley (2006) demonstrate, have profound implications for how we interpret many ﬁndings in the area of adolescent decision making.

REFERENCES Ainslie, G. (2001). Breakdown of will. Cambridge: Cambridge University Press. Anderson, M. (2005). Marrying intelligence and cognition: A developmental view. In R. J. Sternberg & J. E. Pretz (Eds.), Cognition and intelligence (pp. 268–287). New York: Cambridge University Press. Arkes, H. R. (1991). Costs and beneﬁts of judgment errors: Implications for debiasing. Psychological Bulletin, 110, 486–498. Audi, R. (1993). The structure of justification. Cambridge: Cambridge University Press. Audi, R. (2001). The architecture of reason: The structure and substance of rationality. Oxford: Oxford University Press. Ball, L. J., Lucas, E. J., Miles, J., & Gale, A. (2003). Inspection times and the selection task: What do eye-movements reveal about relevance effects. Quarterly Journal of Experimental Psychology, 56A, 1053–1077. Baron, J. (1991). Beliefs about thinking. In J. Voss, D. Perkins, & J. Segal (Eds.), Informal reasoning and education (pp. 169–186). Hillsdale, NJ: Erlbaum. Baron, J. (1995). Myside bias in thinking about abortion. Thinking and Reasoning, 1, 221–235. Baron, J. (2000). Thinking and deciding (3rd ed.). Cambridge, MA: Cambridge University Press. Baumeister, R. F. & Vohs, K. D. (Eds.) (2007). Handbook of self-regulation: Research theory, and applications. New York: Guilford Press. Bjorklund, D. F. (2004). Children’s thinking: Cognitive development and individual differences. Belmont, CA: Wadsworth. Blackmore, S. (1999). The meme machine. New York: Oxford University Press. Brainerd, C. J., & Reyna, V. F. (2001). Fuzzy-trace theory: Dual processes in memory, reasoning, and cognitive neuroscience. In H. W. Reese & R. Kail (Eds.), Advances in child development and behavior (Vol. 28, pp. 41–100). San Diego: Academic Press. Bruine de Bruin, W., Parker, A. M., & Fischhoff, B. (2007). Individual differences in adult decision-making competence. Journal of Personality and Social Psychology, 92, 938–956. Byrnes, J. P. (1998). The nature and development of decision making. Mahwah, NJ: Erlbaum. Davidson, D. (1995). The representativeness heuristic and the conjunction fallacy effect in children’s decision making. Merrill-Palmer Quarterly, 41, 328–346. Dawes, R. M. (1976). Shallow psychology. In J. S. Carroll & J. W. Payne (Eds.), Cognition and social behavior (pp. 3–11). Hillsdale, NJ: Erlbaum.

278

Keith E. Stanovich et al.

Dawes, R. M. (1998). Behavioral decision making and judgment. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), The handbook of social psychology (Vol. 1, pp. 497–548). Boston: McGrawHill. Dawkins, R. (1993). Viruses of the mind. In B. Dahlbom (Ed.), Dennett and his critics (pp. 13–27). Cambridge, MA: Blackwell. Deary, I. J. (2001). Intelligence: A very short introduction. Oxford: Oxford University Press. De Neys, W. (2006). Dual processing in reasoning—two systems but one reasoner. Psychological Science, 17, 428–433. Denes-Raj, V., & Epstein, S. (1994). Conﬂict between intuitive and rational processing: When people behave against their better judgment. Journal of Personality and Social Psychology, 66, 819–829. Dennett, D. C. (1991). Consciousness explained. Boston: Little Brown. Dennett, D. C. (1995). Darwin’s dangerous idea: Evolution and the meanings of life. New York: Simon & Schuster. Dennett, D. C. (2006). From typo to thinko: When evolution graduated to semantic norms. In S. C. Levinson & P. Jaisson (Eds.), Evolution and culture (pp. 133–145). Cambridge, MA: MIT Press. DeShon, R. P., Smith, M. R., Chan, D., & Schmitt, N. (1998). Can racial differences in cognitive test performance be reduced by presenting problems in a social context? Journal of Applied Psychology, 83, 438–451. Dias, M., Roazzi, A., & Harris, P. L. (2005). Reasoning from unfamiliar premises: A study with unschooled adults. Psychological Science, 16, 550–554. Edwards, W. (1954). The theory of decision making. Psychological Bulletin, 51, 380–417. Ehrlinger, J., Gilovich, T., & Ross, L. (2005). Peering into the bias blind spot: People’s assessments of bias in themselves and others. Personality and Social Psychology Bulletin, 31, 680–692. Evans, J. St. B. T. (1972). Interpretation and matching bias in a reasoning task. Quarterly Journal of Experimental Psychology, 24, 193–199. Evans, J. St. B. T. (1989). Bias in human reasoning: Causes and consequences. Hove, UK: Lawrence Erlbaum Associates. Evans, J. St. B. T. (1996). Deciding before you think: Relevance and reasoning in the selection task. British Journal of Psychology, 87, 223–240. Evans, J. St. B. T. (1998). Matching bias in conditional reasoning: Do we understand it after 25 years? Thinking and Reasoning, 4, 45–82. Evans, J. St. B. T. (2002). Matching bias and set sizes: A discussion of Yama (2001). Thinking and Reasoning, 8, 153–163. Evans, J. St. B. T. (2003). In two minds: Dual-process accounts of reasoning. Trends in Cognitive Sciences, 7, 454–459. Evans, J. St. B. T. (2006). The heuristic-analytic theory of reasoning: Extension and evaluation. Psychonomic Bulletin and Review, 13, 378–395. Evans, J. St. B. T. (2007). Hypothetical thinking: Dual processes in reasoning and judgment. New York: Psychology Press. Evans, J. St. B. T. (2008a). Dual-processing accounts of reasoning, judgment and social cognition. Annual Review of Psychology, 59, 255–278. Evans, J. St. B. T. (2008b). How many dual-process theories do we need? One, two, or many? In J. Evans & K. Frankish (Eds.), In two minds: Dual processes and beyond. Oxford: Oxford University Press. Evans, J. St. B. T., & Curtis-Holmes, J. (2005). Rapid responding increases belief bias: Evidence for the dual-process theory of reasoning. Thinking and Reasoning, 11, 382–389.

Development of Rational Thought

279

Evans, J. St. B. T., & Feeney, A. (2004). The role of prior belief in reasoning. In J. P. Leighton & R. J. Sternberg (Eds.), The nature of reasoning (pp. 78–102). Cambridge: Cambridge University Press. Evans, J. St. B. T., & Lynch, J. S. (1973). Matching bias in the selection task. British Journal of Psychology, 64, 391–397. Evans, J. St. B. T., & Over, D. E. (1996). Rationality and reasoning. Hove, England: Psychology Press. Evans, J. St. B. T., & Over, D. E. (2004). If. Oxford: Oxford University Press. Evans, J. St. B. T., & Wason, P. C. (1976). Rationalization in a reasoning task. British Journal of Psychology, 67, 479–486. Evans, J. St. B. T., Barston, J., & Pollard, P. (1983). On the conﬂict between logic and belief in syllogistic reasoning. Memory and Cognition, 11, 295–306. Evans, J. St. B. T., Over, D. E., & Handley, S. J. (2003). A theory of hypothetical thinking. In D. Hardman & L. Maachi (Eds.), Thinking: Psychological perspectives on reasoning (pp. 3–22). Chicester: Wiley. Fisk, J. E. (2004). Conjunction fallacy. In R. Pohl (Ed.), Cognitive illusions: A handbook on fallacies and biases in thinking, judgment and memory (pp. 23–42). Hove, England: Psychology Press. Foley, R. (1987). The theory of epistemic rationality. Cambridge, MA: Harvard University Press. Fong, G. T., Krantz, D. H., & Nisbett, R. E. (1986). The effects of statistical training on thinking about everyday problems. Cognitive Psychology, 18, 253–292. Geary, D. C. (2005). The origin of the mind: Evolution of brain, cognition, and general intelligence. Washington, DC: American Psychological Association. Gilbert, D. T. (1991). How mental systems believe. American Psychologist, 46, 107–119. Gilbert, D. T. (2006). Stumbling on happiness. New York: Alfred A. Knopf. Gilovich, T., Grifﬁn, D., & Kahneman, D. (Eds.) (2002). Heuristics and biases: The psychology of intuitive judgment. New York: Cambridge University Press. Handley, S. J., Capon, A., Beveridge, M., Dennis, I., & Evans, J. S. B. T. (2004). Working memory, inhibitory control and the development of children’s reasoning. Thinking and Reasoning, 10, 175–195. Hardman, D. (1998). Does reasoning occur on the selection task? A comparison of relevancebased theories. Thinking and Reasoning, 4, 353–376. Harman, G. (1995). Rationality. In E. E. Smith & D. N. Osherson (Eds.), Thinking (Vol. 3, pp. 175–211). Cambridge, MA: The MIT Press. Harvey, N. (2007). Use of heuristics: Insights from forecasting research. Thinking and Reasoning, 13, 5–24. Hasher, L., Lustig, C., & Zacks, R. (2007). Inhibitory mechanisms and the control of attention. In A. Conway, C. Jarrold, M. Kane, A. Miyake, & J. Towse (Eds.), Variation in working memory (pp. 227–249). New York: Oxford University Press. Hastie, R., & Dawes, R. M. (2001). Rational choice in an uncertain world. Thousand Oaks, CA: Sage. Hsee, C. K., & Hastie, R. (2006). Decision and experience: Why don’t we choose what makes us happy? Trends in Cognitive Sciences, 10, 31–37. Jacobs, J. E., & Potenza, M. (1991). The use of judgment heuristics to make social and object decisions: A developmental perspective. Child Development, 62, 166–178. Jeffrey, R. C. (1983). The logic of decision (2nd ed.). Chicago: University of Chicago Press. Johnson-Laird, P. N. (1999). Deductive reasoning. Annual Review of Psychology, 50, 109–135.

280

Keith E. Stanovich et al.

Johnson-Laird, P. N. (2005). Mental models and thought. In K. J. Holyoak & R. G. Morrison (Eds.), The Cambridge handbook of thinking and reasoning (pp. 185–208). New York: Cambridge University Press. Johnson-Laird, P. N. (2006). How we reason. Oxford: Oxford University Press. Kahneman, D. (2000). A psychological point of view: Violations of rational rules as a diagnostic of mental processes. Behavioral and Brain Sciences, 23, 681–683. Kahneman, D. (2003). A perspective on judgment and choice: Mapping bounded rationality. American Psychologist, 58, 697–720. Kahneman, D., & Frederick, S. (2002). Representativeness revisited: Attribute substitution in intuitive judgment. In T. Gilovich, D. Grifﬁn, & D. Kahneman (Eds.), Heuristics and biases: The psychology of intuitive judgment (pp. 49–81). New York: Cambridge University Press. Kahneman, D., & Frederick, S. (2005). A model of heuristic judgment. In K. J. Holyoak & R. G. Morrison (Eds.), The Cambridge handbook of thinking and reasoning (pp. 267–293). New York: Cambridge University Press. Kahneman, D., & Tversky, A. (1972). On prediction and judgment. Oregon Research Institute Monograph, 12(4). Kahneman, D., & Tversky, A. (1973). On the psychology of prediction. Psychological Review, 80, 237–251. Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decision under risk. Econometrica, 47, 263–291. Kahneman, D., & Tversky, A. (1982). The simulation heuristic. In D. Kahneman, P. Slovic, & A. Tversky (Eds.), Judgment under uncertainty: Heuristics and biases (pp. 201–208). Cambridge: Cambridge University Press. Kahneman, D., & Tversky, A. (1984). Choices, values, and frames. American Psychologist, 39, 341–350. Kahneman, D. & Tversky, A. (Eds.) (2000). Choices, values, and frames. Cambridge: Cambridge University Press. Kahneman, D., Diener, E., & Schwarz, N. (Eds.) (1999). Well-being: The foundations of hedonic psychology. Thousand Oaks, CA: Sage. Kahneman, D., Krueger, A. B., Schkade, D., Schwarz, N., & Stone, A. (2006). Would you be happier if you were richer? A focusing illusion. Science, 312, 1908–1910. Kail, R. (2000). Speed of information processing: Developmental change and links to intelligence. Journal of School Psychology, 38, 51–61. Kirkpatrick, L., & Epstein, S. (1992). Cognitive-experiential self-theory and subjective probability: Evidence for two conceptual systems. Journal of Personality and Social Psychology, 63, 534–544. Klaczynski, P. A. (1997). Bias in adolescents’ everyday reasoning and its relationship with intellectual ability, personal theories, and self-serving motivation. Developmental Psychology, 33, 273–283. Klaczynski, P. A. (2000). Motivated scientiﬁc reasoning biases, epistemological beliefs, and theory polarization: A two-process approach to adolescent cognition. Child Development, 71, 1347–1366. Klaczynski, P. A. (2001a). Analytic and heuristic processing inﬂuences on adolescent reasoning and decision making. Child Development, 72, 844–861. Klaczynski, P. A. (2001b). Framing effects on adolescent task representations, analytic and heuristic processing, and decision making: Implications for the normative-descriptive gap. Journal of Applied Developmental Psychology, 22, 289–309. Klaczynski, P. A. (2004). A dual-process model of adolescent development: Implications for decision making, reasoning, and identity. Advances in Child Development and Behavior, 32, 73–123.

Development of Rational Thought

281

Klaczynski, P. A., & Gordon, D. H. (1996). Self-serving inﬂuences on adolescents’ evaluations of belief-relevant evidence. Journal of Experimental Child Psychology, 62, 317–339. Klaczynski, P. A., & Lavallee, K. L. (2005). Domain-speciﬁc identity, epistemic regulation, and intellectual ability as predictors of belief-based reasoning: A dual-process perspective. Journal of Experimental Child Psychology, 92, 1–24. Klaczynski, P. A., & Narasimham, G. (1998). Development of scientiﬁc reasoning biases: Cognitive versus ego-protective explanations. Developmental Psychology, 34, 175–187. Klaczynski, P. A., & Robinson, B. (2000). Personal theories, intellectual ability, and epistemological beliefs: Adult age differences in everyday reasoning tasks. Psychology and Aging, 15, 400–416. Klaczynski, P. A., Gordon, D. H., & Fauth, J. (1997). Goal-oriented critical reasoning and individual differences in critical reasoning biases. Journal of Educational Psychology, 89, 470–485. Klauer, K. C., Stahl, C., & Erdfelder, E. (2007). The abstract selection task: New data and an almost comprehensive model. Journal of Experimental Psychology: Learning, Memory, and Cognition, 33, 688–703. Klayman, J., & Ha, Y. (1987). Conﬁrmation, disconﬁrmation, and information in hypothesis testing. Psychological Review, 94, 211–228. Koehler, J. J. (1996). The base rate fallacy reconsidered: Descriptive, normative and methodological challenges. Behavioral and Brain Sciences, 19, 1–53. Kokis, J., Macpherson, R., Toplak, M., West, R. F., & Stanovich, K. E. (2002). Heuristic and analytic processing: Age trends and associations with cognitive ability and cognitive styles. Journal of Experimental Child Psychology, 83, 26–52. Kuhberger, A. (1998). The inﬂuence of framing on risky decisions: A meta-analysis. Organizational Behavior and Human Decision Processes, 75, 23–55. Kuhn, D. (1991). The skills of argument. Cambridge: Cambridge University Press. Larrick, R. P. (2004). Debiasing. In D. J. Koehler & N. Harvey (Eds.), Blackwell handbook of judgment and decision making (pp. 316–337). Malden, MA: Blackwell. Legrenzi, P., Girotto, V., & Johnson-Laird, P. N. (1993). Focussing in reasoning and decision making. Cognition, 49, 37–66. Leslie, A. M. (1987). Pretense and representation: The origins of ‘‘Theory of Mind’’. Psychological Review, 94, 412–426. Levesque, H. J. (1986). Making believers out of computers. Artificial Intelligence, 30, 81–108. Levesque, H. J. (1989). Logic and the complexity of reasoning. In R. H. Thomason (Ed.), Philosophical logic and artificial intelligence (pp. 73–107). Dordrecht, The Netherlands: Kluwer Academic Publishers. Levin, I. P., & Hart, S. S. (2003). Risk preferences in young children: Early evidence of individual differences in reaction to potential gains and losses. Journal of Behavioral Decision Making, 16, 397–413. Levin, I. P., Gaeth, G. J., Schreiber, J., & Lauriola, M. (2002). A new look at framing effects: Distribution of effect sizes, individual differences, and independence of types of effects. Organizational Behavior and Human Decision Processes, 88, 411–429. Levin, I. P., Hart, S. S., Weller, J. A., & Harshman, L. A. (2007). Stability of choices in a risky decision-making task: A 3-year longitudinal study with children and adults. Journal of Behavioral Decision Making, 20, 241–252. Loewenstein, G. F., Read, D., & Baumeister, R. (Eds.) (2003). Time and decision: Economic and psychological perspectives on intertemporal choice. New York: Russell Sage.

282

Keith E. Stanovich et al.

Lubinski, D. (2004). Introduction to the special section on cognitive abilities: 100 years after Spearman’s (1904) ‘‘General Intelligence, Objectively Determined and Measured’’. Journal of Personality and Social Psychology, 86, 96–111. Lucas, E. J., & Ball, L. J. (2005). Think-aloud protocols and the selection task: Evidence for relevance effects and rationalisation processes. Thinking and Reasoning, 11, 35–66. Luce, R. D., & Raiffa, H. (1957). Games and decisions. New York: Wiley. Lynch, A. (1996). Thought contagion. New York: Basic Books. Macpherson, R., & Stanovich, K. E. (2007). Cognitive ability, thinking dispositions, and instructional set as predictors of critical thinking. Learning and Individual Differences, 17, 115–127. Manktelow, K. I. (2004). Reasoning and rationality: The pure and the practical. In K. I. Manktelow & M. C. Chung (Eds.), Psychology of reasoning: Theoretical and historical perspectives (pp. 157–177). Hove, England: Psychology Press. Margolis, H. (1987). Patterns, thinking, and cognition. Chicago: University of Chicago Press. Markovits, H., & Nantel, G. (1989). The belief-bias effect in the production and evaluation of logical conclusions. Memory and Cognition, 17, 11–17. Maule, J., & Villejoubert, G. (2007). What lies beneath: Reframing framing effects. Thinking and Reasoning, 13, 25–44. McFadden, D. (1999). Rationality for economists? Journal of Risk and Uncertainty, 19, 73–105. Metcalfe, J., & Mischel, W. (1999). A hot/cool-system analysis of delay of gratiﬁcation: Dynamics of will power. Psychological Review, 106, 3–19. Mischel, W., Shoda, Y., & Rodriguez, M. L. (1989). Delay of gratiﬁcation in children. Science, 244(May 26), 933–938. Miyake, A., Friedman, N., Emerson, M. J., & Witzki, A. H. (2000). The utility and diversity of executive functions and their contributions to complex ‘‘frontal lobe’’ tasks: A latent variable analysis. Cognitive Psychology, 41, 49–100. Nichols, S., & Stich, S. P. (2003). Mindreading: An integrated account of pretence, selfawareness, and understanding other minds. Oxford: Oxford University Press. Nickerson, R. S. (1998). Conﬁrmation bias: A ubiquitous phenomenon in many guises. Review of General Psychology, 2, 175–220. Nisbett, R. E., & Ross, L. (1980). Human inference: Strategies and shortcomings of social judgment. Englewood Cliffs, NJ: Prentice-Hall. Nisbett, R. E., & Wilson, T. D. (1977). Telling more than we can know: Verbal reports on mental processes. Psychological Review, 84, 231–259. Osman, M., & Laming, D. (2001). Misinterpretation of conditional statements in Wason’s selection task. Psychological Research, 65, 128–144. Over, D. E. (2004). Rationality and the normative/descriptive distinction. In D. J. Koehler & N. Harvey (Eds.), Blackwell handbook of judgment and decision making (pp. 3–18). Malden, MA: Blackwell Publishing. Overton, W. F., Byrnes, J. P., & O’Brien, D. P. (1985). Developmental and individual differences in conditional reasoning: The role of contradiction training and cognitive style. Developmental Psychology, 21, 692–701. Pacini, R., & Epstein, S. (1999). The relation of rational and experiential information processing styles to personality, basic beliefs, and the ratio-bias phenomenon. Journal of Personality and Social Psychology, 76, 972–987. Parker, A. M., & Fischhoff, B. (2005). Decision-making competence: External validation through an individual differences approach. Journal of Behavioral Decision Making, 18, 1–27. Perkins, D. N. (1985). Postprimary education has little impact on informal reasoning. Journal of Educational Psychology, 77, 562–571.

Development of Rational Thought

283

Perkins, D. N. (1995). Outsmarting IQ: The emerging science of learnable intelligence. New York: Free Press. Perkins, D. N. (2002). The engine of folly. In R. J. Sternberg (Ed.), Why smart people can be so stupid (pp. 64–85). New Haven, CT: Yale University Press. Perkins, D. N., Farady, M., & Bushey, B. (1991). Everyday reasoning and the roots of intelligence. In J. Voss, D. Perkins, & J. Segal (Eds.), Informal reasoning and education (pp. 83–105). Hillsdale, NJ: Erlbaum. Perner, J. (1991). Understanding the representational mind. Cambridge, MA: MIT Press. Pronin, E. (2006). Perception and misperception of bias in human judgment. Trends in Cognitive Sciences, 11, 37–43. Rachlin, H. (2000). The science of self-control. Cambridge, MA: Harvard University Press. Reyna, V. F., & Ellis, S. (1994). Fuzzy-trace theory and framing effects in children’s risky decision making. Psychological Science, 5, 275–279. Reyna, V. F., & Farley, F. (2006). Risk and rationality in adolescent decision making. Psychological Science in the Public Interest, 7, 1–44. Reyna, V. F., Lloyd, F. J., & Brainerd, C. J. (2003). Memory, development, and rationality: An integrative theory of judgment and decision making. In S. L. Schneider & J. Shanteau (Eds.), Emerging perspectives on judgment and decision research (pp. 201–245). New York: Cambridge University Press. Roberts, M. J., & Newton, E. J. (2001). Inspection times, the change task, and the rapidresponse selection task. Quarterly Journal of Experimental Psychology, 54A, 1031–1048. Savage, L. J. (1954). The foundations of statistics. New York: Wiley. Shaﬁr, E. (1994). Uncertainty and the difﬁculty of thinking through disjunctions. Cognition, 50, 403–430. Shaﬁr, E., & LeBoeuf, R. A. (2002). Rationality. Annual Review of Psychology, 53, 491–517. Sloman, S. A. (1996). The empirical case for two systems of reasoning. Psychological Bulletin, 119, 3–22. Sloman, S. A. (2002). Two systems of reasoning. In T. Gilovich, D. Grifﬁn, & D. Kahneman (Eds.), Heuristics and biases: The psychology of intuitive judgment (pp. 379–396). New York: Cambridge University Press. Sperber, D., Cara, F., & Girotto, V. (1995). Relevance theory explains the selection task. Cognition, 57, 31–95. Stanovich, K. E. (1999). Who is rational? Studies of individual differences in reasoning. Mahwah, NJ: Erlbaum. Stanovich, K. E. (2002). Rationality, intelligence, and levels of analysis in cognitive science: Is dysrationalia possible? In R. J. Sternberg (Ed.), Why smart people can be so stupid (pp. 124–158). New Haven, CT: Yale University Press. Stanovich, K. E. (2004). The robot’s rebellion: Finding meaning in the age of Darwin. Chicago: University of Chicago Press. Stanovich, K. E. (2008a). Distinguishing the reﬂective, algorithmic, and autonomous minds: Is it time for a tri-process theory? In J. Evans & K. Frankish (Eds.), In two minds: Dual processes and beyond. Oxford: Oxford University Press. Stanovich, K. E. (2008b). Rationality and the reflective mind: Toward a tri-process model of cognition. New York: Oxford University Press. Stanovich, K. E. (2008c). The psychology of rational thought: What intelligence tests miss. New Haven, CT: Yale University Press. Stanovich, K. E., & West, R. F. (1998a). Cognitive ability and variation in selection task performance. Thinking and Reasoning, 4, 193–230. Stanovich, K. E., & West, R. F. (1998b). Individual differences in rational thought. Journal of Experimental Psychology: General, 127, 161–188.

284

Keith E. Stanovich et al.

Stanovich, K. E., & West, R. F. (1998c). Who uses base rates and P(D/BH)? An analysis of individual differences. Memory and Cognition, 26, 161–179. Stanovich, K. E., & West, R. F. (1999). Discrepancies between normative and descriptive models of decision making and the understanding/acceptance principle. Cognitive Psychology, 38, 349–385. Stanovich, K. E., & West, R. F. (2000). Individual differences in reasoning: Implications for the rationality debate? Behavioral and Brain Sciences, 23, 645–726. Stanovich, K. E., & West, R. F. (2007). Natural myside bias is independent of cognitive ability. Thinking and Reasoning, 13, 225–247. Stanovich, K. E., & West, R. F. (2008a). On the failure of intelligence to predict myside bias and one-sided bias. Thinking and Reasoning. Stanovich, K. E., & West, R. F. (2008b). On the relative independence of thinking biases and cognitive ability. Journal of Personality and Social Psychology, 94, 672–695. Sternberg, R. J. (2003). Wisdom, intelligence, and creativity synthesized. Cambridge: Cambridge University Press. Sternberg, R. J. (2006). The Rainbow Project: Enhancing the SAT through assessments of analytical, practical, and creative skills. Intelligence, 34, 321–350. Sa´, W., West, R. F., & Stanovich, K. E. (1999). The domain speciﬁcity and generality of belief bias: Searching for a generalizable critical thinking skill. Journal of Educational Psychology, 91, 497–510. Sa´, W., Kelley, C., Ho, C., & Stanovich, K. E. (2005). Thinking about personal theories: Individual differences in the coordination of theory and evidence. Personality and Individual Differences, 38, 1149–1161. Taylor, S. E. (1981). The interface of cognitive and social psychology. In J. H. Harvey (Ed.), Cognition, social behavior, and the environment (pp. 189–211). Hillsdale, NJ: Erlbaum. Toplak, M. E., & Stanovich, K. E. (2002). The domain speciﬁcity and generality of disjunctive reasoning: Searching for a generalizable critical thinking skill. Journal of Educational Psychology, 94, 197–209. Toplak, M. E., & Stanovich, K. E. (2003). Associations between myside bias on an informal reasoning task and amount of post-secondary education. Applied Cognitive Psychology, 17, 851–860. Toplak, M., Liu, E., Macpherson, R., Toneatto, T., & Stanovich, K. E. (2007). The reasoning skills and thinking dispositions of problem gamblers: A dual-process taxonomy. Journal of Behavioral Decision Making, 20, 103–124. Tversky, A., & Kahneman, D. (1974). Judgment under uncertainty: Heuristics and biases. Science, 185, 1124–1131. Tversky, A., & Kahneman, D. (1981). The framing of decisions and the psychology of choice. Science, 211, 453–458. Tversky, A., & Kahneman, D. (1983). Extensional versus intuitive reasoning: The conjunction fallacy in probability judgment. Psychological Review, 90, 293–315. Tversky, A., & Kahneman, D. (1986). Rational choice and the framing of decisions. Journal of Business, 59, 251–278. Tweney, R. D., & Yachanin, S. (1985). Can scientists rationally assess conditional inferences? Social Studies of Science, 15, 155–173. Valentine, E. R. (1975). Performance on two reasoning tasks in relation to intelligence, divergence and interference proneness: Content and context effects in reasoning. British Journal of Educational Psychology, 45, 198–205. von Neumann, J., & Morgenstern, O. (1944). The theory of games and economic behavior. Princeton, NJ: Princeton University Press.

Development of Rational Thought

285

Wason, P. C. (1966). Reasoning. In B. Foss (Ed.), New horizons in psychology (pp. 135–151). Harmonsworth, England: Penguin. Wason, P. C. (1968). Reasoning about a rule. Quarterly Journal of Experimental Psychology, 20, 273–281. Wilson, T. D., Wheatley, T., Meyers, J. M., Gilbert, D. T., & Axsom, D. (2000). Focalism: A source of durability bias in affective forecasting. Journal of Personality and Social Psychology, 78, 821–836. Wu, G., Zhang, J., & Gonzalez, R. (2004). Decision under risk. In D. J. Koehler & N. Harvey (Eds.), Blackwell handbook of judgment and decision making (pp. 399–423). Malden, MA: Blackwell Publishing. Zelazo, P. D. (2004). The development of conscious control in childhood. Trends in Cognitive Sciences, 8, 12–17.

This page intentionally left blank

LESSONS LEARNED: RECENT ADVANCES IN UNDERSTANDING AND PREVENTING CHILDHOOD AGGRESSION

Nancy G. Guerra and Melinda S. Leidy DEPARTMENT OF PSYCHOLOGY, UNIVERSITY OF CALIFORNIA, RIVERSIDE, CA 92521, USA

I. INTRODUCTION II. THE ADAPTIVE FUNCTIONS OF AGGRESSION A . EVOLUTIONARY PERSPECTIVES B . DEVELOPMENTAL PERSPECTIVES C . MOTIVATION FOR CHANGE III. AGGRESSION AND THE ECOLOGY OF DEVELOPMENT A . AGGRESSION AS A MULTIPLY-DETERMINED BEHAVIOR B . RISK FOR AGGRESSION ACROSS INDIVIDUALS AND CONTEXTS C . THE CUMULATIVE AND INTERACTIVE NATURE OF RISK IV. RISK, CAUSALITY, AND PREVENTION A . PREVENTION AND THE MULTIPLE DETERMINANTS OF AGGRESSIVE BEHAVIOR B . THE METROPOLITAN AREA CHILD STUDY C . RESEARCH-DRIVEN PROGRAMS AND POLICIES V. TRANSLATING RESEARCH TO PRACTICE: BUILDING AN EVIDENCE BASE A . EVIDENCE-BASED PROGRAMS AND PRINCIPLES B . DISSEMINATION AND IMPLEMENTATION C . LINKING PREVENTION WITH POSITIVE YOUTH DEVELOPMENT VI. CONCLUSION REFERENCES

I. Introduction In the present chapter, we summarize and integrate recent developments in the ﬁeld of childhood aggression and prevention science. Our primary goal is to synthesize these developments into a core set of advances or ‘‘lessons learned’’ since the 1970s in understanding and preventing childhood aggression, and to suggest areas where exciting new advances are most likely to occur in the years to come. For purposes of clarity, we use

287 Advances in Child Development and Behavior R.V. Kail : Editor

Copyright r 2008 Elsevier B.V. All rights reserved.

288

Nancy G. Guerra and Melinda S. Leidy

the term childhood aggression to describe behavior aimed at harming or injuring others during childhood and adolescence. In practice, childhood aggression is most commonly used to describe children’s behavior, whereas such behavior in adolescence is called youth violence, even though in some cases violence is conceptualized as a more extreme form of aggression (Guerra & Knox, 2002). We use the term prevention according to guidelines of the Institute of Medicine that includes universal interventions targeting entire populations or groups, selected interventions for individuals or groups at high risk, and indicated interventions for participants who show some signs of aggression but have not met diagnostic criteria for treatment, for example a conduct disorder diagnosis (Mrazek & Haggerty, 1994). Highly publicized events such as school shootings and gang violence have sensitized the public to the urgency of the problem in recent years. Still, concerns about understanding and preventing childhood aggression are not limited to contemporary society. More than two centuries ago, debates about the causes of childhood aggression reﬂected philosophical distinctions about the essence of human nature. On the one hand, children were seen as born unruly only to be made ﬁt for society by training (Hobbes, 1651/1958). On the other, children were seen as born innocent only to be corrupted by social forces (Rousseau, 1762/1979). The nature versus nurture dichotomy continued well into the latter part of the 20th century bolstered by empirical ﬁndings, with general theories of aggression based on biological factors such as instinct (Lorenz, 1966) or drive (Dollard et al., 1939) contrasted with environmental theories derived from operant or social learning processes (Bandura, 1973). Although biological theories of aggression often acknowledged environmental constraints and vice versa, it is since the 1970s that the traditional ‘‘either/or’’ approach to nature versus nurture gradually has been replaced by a more integrated framework (deWaal, 1999). A good example of this can be found in contemporary behavior genetics research on aggression. Since the late 1970s there have been more than 100 quantitative genetic studies on aggression and antisocial behavior highlighting the relative inﬂuences of genes and environment (e.g., Arseneault et al., 2003; O’Connor et al., 1998). Evidence across studies points to moderate genetic and nonshared environmental inﬂuences and small shared environmental inﬂuences on antisocial behavior, particularly for more persistent types of aggression that begin early in development (Mofﬁtt, 2003). Furthermore, rather than acting independently, research suggests that environmental and genetic risks interact, for example, with stronger environmental effects among groups at higher genetic risk (Caspi et al., 2002; Fox et al., 2005; Jaffee et al., 2005).

Advances in Understanding and Preventing Childhood Aggression

289

A different but equally important perspective on the interaction between innate and environmental inﬂuences has emerged from the application of evolutionary psychology to the study of aggression. From this point of view, behaviors such as aggression are seen as products of mechanisms internal to the person activated by inputs that arise from evolution by selection designed to solve adaptive problems in speciﬁc contexts. For instance, Buss and Shackelford (1997) described seven adaptive problems to which aggression may have evolved as a solution: (a) co-opt the resources of others; (b) defend against attack; (c) negotiate power and status hierarchies; (d) inﬂict costs on intrasex rivals; (e) deter rivals from future aggression; (f) deter long-term mates from sexual inﬁdelity; and (g) reduce resources expended on unrelated children. This does not mean that aggression is common to all humans; rather, aggression is context speciﬁc triggered by speciﬁc environmental and social factors where speciﬁc problems are confronted, beneﬁts are likely, and costs are minimized (Hawley, 2003). An individual’s own phenotype (e.g., size, personality) provides information about the feasibility of select strategies under speciﬁc conditions (Tooby & Cosmides, 1990). Both behavior genetics and evolutionary approaches indicate important roles for persons and environments. This work is complemented by a large number of empirical studies of speciﬁc individual and contextual predictors of risk for aggression. Until the 1990s, much of this risk research yielded a rather haphazard collection of diverse risk and protective factors. Only since the start of the 21st century have attempts been made to develop more integrative theories that emphasize multiple inﬂuences on development, multiple levels of inﬂuence, and how they operate together over time during childhood and adolescence. In part, these efforts were informed by ecological models of human development emphasizing nested systems and their mutual interdependence (Bronfenbrenner, 1979), as well as transactional models of behavior grounded in developmental psychopathology (Cicchetti & Carlson, 1989; Shaw, 2003). The integrative orientation of these models is reﬂected in their descriptions; for instance, a developmental-ecological model of antisocial behavior (Tolan, Guerra, & Kendall, 1995) or a biopsychosocial model of the development of chronic conduct problems in adolescence (Dodge & Pettit, 2003). Ecological and integrated theoretical approaches to childhood aggression signaled a growing recognition and understanding of the complexity of risk factors and how they co-occur, interact, and transact (Dodge, Coie, & Lynam, 2006a). At the same time, this has been accompanied by more careful speciﬁcations of distinct subtypes of aggressive behaviors. At this juncture, there is general consensus that aggression is best deﬁned as a heterogeneous set of behaviors that is aimed at harming or injuring another

290

Nancy G. Guerra and Melinda S. Leidy

person or persons (although there has been considerable debate regarding the centrality of intent). A number of different dimensions and classiﬁcation schemes have been proposed involving variations in both the function and form of aggression. One important distinction related to the function of aggression hinges on whether it is proactive in the strategic service of self goals or reactive to provocation or blocked self goals (Little et al., 2003). Proactive aggression can be more calculating and delayed, whereas reactive aggression is more likely driven by characteristics such as impulsivity, frustration, and low social competence (Crick & Dodge, 1994; Dodge & Coie, 1987). This distinction also may shed light on the inconclusive ﬁndings linking testosterone and aggression in humans—high levels of testosterone in males have been shown to predict defensive or reactive aggression related to heightened threat perception rather than proactive or offensive aggression (Olweus, Mattsson, & Low, 1988). Different forms of aggression also have been identiﬁed. Direct aggression is overt, including face-to-face physical and verbal confrontations; in contrast, indirect aggression is covert and less visible, such as spreading rumors and social exclusion. To the extent that indirect aggression is designed to manipulate and harm others within the context of peer relationships, it has been called relational aggression (Crick & Grotpeter, 1996) or social aggression (Galen & Underwood, 1997). This type of aggression has been shown to be more characteristic of girls, while physical aggression is more characteristic of boys. There has also been a surge of interest in bullying as a distinct form of aggression most frequently described as proactive aggression repeated over time in the context of a disproportionate power imbalance (Olweus, 1993). Bullying can only occur in a social context (e.g., schools, workplace) because it requires repeated interactions and often involves bystanders who can intervene to help or instigate such acts (Salmivalli et al., 1996). Just as there may be distinct patterns associated with speciﬁc subtypes of aggressive behavior, it is also important to understand the broader functions of aggression that may coincide more generally with antisocial behavior. Aggression often occurs with other types of antisocial behavior such as delinquency, drug use, academic failure, and risky sexual behavior, particularly during adolescence (Dryfoos, 1990). It has been suggested that these problem behaviors are highly correlated because they share a common set of personality, behavioral, and environmental predictors (Jessor, 1994; Jessor, Donovan, & Costa, 1993). In other words, various antisocial behaviors may develop together and serve similar psychological functions. Advances in understanding the etiology of childhood aggression also have been accompanied by progress in designing and evaluating preventive

Advances in Understanding and Preventing Childhood Aggression

291

intervention programs. These efforts have been enriched further by signiﬁcant progress in the ﬁeld of prevention science. Two of the most signiﬁcant developments since the 1980s of relevance to the prevention of childhood aggression are: (a) an increase in randomized controlled laboratory and ﬁeld trials; and (b) a programmatic and policy shift towards evidence-based practice in health care and psychology. Since the 1980s, public concern about escalating rates of youth violence and crime in the U.S. has resulted in a proliferation of violence prevention programs (Chaiken, 1998). Most of these programs can be considered psycho-educational by virtue of their emphasis on psychological and learning processes. Although the vast majority of these programs remain untested, there has been a marked increase in randomized controlled trials and quasi-experimental ﬁeld studies of anti-violence programs, particularly since the 1980s (Dodge et al., 2006a). For example, in a meta-analysis of school-based anti-violence programs, Derzon, Wilson, and Cunningham (1999) identiﬁed 83 experimental or quasi-experimental program evaluations for inclusion. In a subsequent review of violence prevention programs for individuals, families, and larger social systems, Kerns and Prinz (2002) identiﬁed 40 empirically evaluated studies. Large-scale randomized trials of comprehensive, multi-component programs have also been conducted, including Fast Track (Conduct Problems Prevention Research Group, 2002), the MACS (Metropolitan Area Child Study Research Group, 2002), and the RECAP program (Weiss et al., 2003). In many cases, prevention experiments have assessed hypothesized mediators of change, providing opportunities to examine causal relations between predictors and violence-related outcomes. A parallel trend towards documentation of evidence-based practices has also gained considerable momentum in recent years (APA Presidential Task Force on Evidence-Based Practice, 2006). In the ﬁeld of prevention of childhood aggression and youth violence, this has spawned a number of efforts to identify model programs with experimental evaluations and replication studies (Elliott & Mihalic, 2004). The push towards evidencebased practice has also informed policymaking, with state and federal funding of aggression and violence prevention programs often mandating inclusion of evidence-based programs. In sum, there have been a number of important developments that have greatly enhanced our understanding of the causes, course, and prevention of childhood aggression since the 1970s. In particular, research has highlighted the adaptive functions of speciﬁc types of aggression in speciﬁc contexts, the multiple predictors of aggression and how they interact over time, the intertwined nature of social contexts within a given developmental ecology, and the complex interplay between innate and learned

292

Nancy G. Guerra and Melinda S. Leidy

contributions to aggression. There also has been a signiﬁcant increase in the number of randomized controlled and quasi-experimental ﬁeld trials of preventive interventions, with a focus on the development of evidencebased practice for the prevention of childhood aggression. As we mentioned at the outset of this chapter, our primary goal is to summarize and integrate these recent developments into a core set of advances or ‘‘lessons learned’’ in understanding and preventing childhood aggression, including a discussion of challenges and future directions. We highlight four major advances and organize this review accordingly. In Section II, we provide evidence from evolutionary and developmental studies suggesting that aggression is both adaptive and normative during childhood and adolescence. As we point out, at extreme levels and under speciﬁc contextual conditions, aggression becomes maladaptive. However, as we discuss, it is important for preventive intervention programs to recognize how the adaptive function of aggression might interfere with an individual’s motivation to change, for instance, in the case of the ‘‘popular’’ bully (Vaillancourt, Hymel, & McDougall, 2007). In Section III, we review the literature on individual and environmental risk for aggression, emphasizing the ecology of development and the cumulative and interactive inﬂuences of risk factors. We emphasize the distinction between modiﬁable and non-modiﬁable risk factors. As we point out, non-modiﬁable risk factors (e.g., difﬁcult temperament) can be useful in identifying individuals for focused intervention, with modiﬁable risk factors (e.g., parenting skills) more amenable to change. We highlight the most robust risk factors, differentiating between characteristics of individuals, close interpersonal relationships (e.g., peers, families), proximal contexts (e.g., neighborhoods, schools), and societal conditions that can be viable targets for participant selection, prevention, and intervention (Tolan & Guerra, 1994). In Section IV, we turn to a more focused discussion of the linkages between causal models of childhood aggression and preventive interventions. As we point out, one of the most signiﬁcant challenges in recent years has been the design, implementation, and evaluation of programs that address the multi-component, multi-context, and transactional nature of risk. As an example of such an effort, we discuss ﬁndings from the MACS, a large-scale development and prevention study conducted over the course of eight years (Metropolitan Area Child Study Research Group, 2002). As we note, these large-scale studies also provide an opportunity to assess the speciﬁc mechanisms (i.e., mediators) of inﬂuence, as well as the speciﬁc conditions under which prevention programs are most likely to be effective (i.e., moderators).

Advances in Understanding and Preventing Childhood Aggression

293

In Section V, we address both accomplishments and challenges related to the translation of research to practice and the implementation of evidencebased programs. As we point out, the push to develop a solid evidence base of ‘‘what works’’ in the prevention of childhood aggression has often obscured issues related to ‘‘what works for whom and under what conditions’’ (Guerra, Boxer, & Cook, 2006). For example, it is unclear whether programs that have been effective for boys will be equally effective for girls, or whether programs can be generalized across ethnic and cultural groups (Guerra & Phillips-Smith, 2005). Furthermore, programs designed with optimal funding and under ideal conditions may be less feasible to implement in everyday settings. We also discuss the challenge of aligning and integrating anti-violence programming within larger systems that emphasize prevention of multiple youth problem behaviors and promotion of positive youth development (Guerra & Bradshaw, in press). Finally, in Section VI, we conclude by brieﬂy reviewing these advances and suggesting areas where important new developments are most likely to occur.

II. The Adaptive Functions of Aggression Although progress in understanding childhood aggression since the 1970s highlighted its complexity, the notion that childhood aggression is largely maladaptive still prevailed. Empirical studies painted a picture of the aggressive child as socially inept and generally disliked by peers—low social status and peer rejection were consistently identiﬁed as correlates of aggression (Coie & Kupersmidt, 1983; Dodge et al., 1990). For the most aggressive children, chronic peer disapproval often led to increased individual aggressiveness (Dodge et al., 2003). Yet the notion of aggression as dysfunctional behavior cast aside evolutionary and developmental perspectives on the adaptive functions of aggression for species survival and its normative status from infancy through adolescence. In other words, although excessive levels of aggression may portend suffering and misfortune, the strategic use of aggression under some conditions may serve adaptive functions from birth onward. Only since the late 1990s or so has there been the recognition that aggression can also be adaptive.

A. EVOLUTIONARY PERSPECTIVES As historical and cross-cultural evidence shows, our evolutionary history is laced with examples of violence. Paleontological data reveal a continuous

294

Nancy G. Guerra and Melinda S. Leidy

stream of human aggression dating back thousands of years. Violence is not restricted to early historical periods or particular cultural groups. Ironically, in spite of recent concerns about the escalating rates of violence in the U.S. and elsewhere, evidence suggests that there is actually less violence now than in ancient times (Guerra & Knox, 2002). From an evolutionary perspective, violence may represent a context-speciﬁc solution to particular problems of social living that may ebb and ﬂow in accordance with changing conditions. As we noted earlier, several adaptive functions of violence have been suggested; for instance, Buss and Shackelford (1997) describe seven problems for which violence may have evolved as a solution. From a developmental vantage point, ﬁve of these are particularly relevant for children and youth: (a) co-opting the resources of others; (b) defending against attack; (c) deterring rivals from future aggression; (d) negotiating status and power hierarchies, and (e) inﬂicting costs on same-sex rivals (with the latter two problems more relevant for older children and adolescents). Childhood aggression is often about co-opting the resources of others, whether the speciﬁc focus is the toys of a two-year-old, the lunch money of an eight-year-old, or the designer tennis shoes of a teenager (Campbell, 1993). In many cases, the threat of aggression is sufﬁcient to engender compliance. In contrast, by defending against an aggressive attack, individuals can build a reputation that can deter future aggression and prevent identiﬁcation as a victim and accompanying loss of status. Indeed, innovative research using neuro-imaging techniques suggests that the human brain may be prewired to exact consequences for misdeeds; for instance, de Quervain et al. (2004) found that areas of the brain linked to anticipated satisfaction were activated with actual but not symbolic punishment. Taking this even further, merely cultivating a reputation as an aggressor may function to deter rivals from future aggression. In group settings where aggression is valued because it facilitates access to resources, successful aggressors often achieve positions of status and dominance within the group hierarchy (Hawley, 1999). Status and honor within a group add to one’s reproductive and survival currency. Within groups deﬁned by violence, such as street gangs, the most aggressive individuals often experience the greatest status elevation (Campbell, 1993). However, status elevation only occurs in groups and under conditions where aggression is normative or desirable (Espelage, Holt, & Henkel, 2003; Vaillancourt et al., 2007; Wright, Giammarino, & Parad, 1986). Finally, aggression can regulate access to valuable members of the opposite sex—by inﬂicting costs on same-sex rivals through indirect or direct aggression, they become less desirable to members of the opposite sex (Buss & Shackelford, 1997).

Advances in Understanding and Preventing Childhood Aggression

295

From a developmental standpoint, the form and functions of aggression vary by age from infancy through adolescence. Toddlers may grab things and throw tantrums, but they are unlikely to spread rumors and tell lies about other children, just as adolescents are unlikely to throw public tantrums. Normative aggression must be understood in the context of agegraded standards. However, at any age, aggression may become excessive or chronic. This is particularly troublesome during the early years, when oppositional behaviors that are expected during preschool continue and escalate during the elementary years and beyond. Considerable attention has been paid to this ‘‘early starter’’ group of aggressive children whose behavior seems to persist over time (Mofﬁtt, 2003). We now turn to a discussion of the adaptive functions of aggression during childhood and adolescence, considering the distinction between normative and troublesome behaviors. B. DEVELOPMENTAL PERSPECTIVES As we have discussed previously, aggression is functional and adaptive for human survival. Signs of anger and aggression are evident in infancy, but escalation and regular use of aggression emerges around the end of the ﬁrst year of life. Most 1- and 2-year-olds engage in regular aggression with peers including retaliation (Caplan et al., 1991). For young children, aggression serves primarily to signal discomfort, gain attention, access resources, and defend one’s possessions and territory. Retaliation seems to serve a further purpose in sending a message to playmates that their aggressive acts will not go unpunished—children who are unwilling to retaliate are more likely to be targeted for future aggression. During the preschool years, the onset of language provides a new venue for aggressive behavior. This age period is also associated with an increase in verbal demands for appropriate behavior by adults either at home or in the preschool setting, with aggressive noncompliance (e.g., screaming, hitting, tantrums) increasing dramatically (Klimes-Dougan & Kopp, 1999). Although young children differ in their temperamental and individual propensities to use aggression, such behavior will be used more regularly and become more habitual if it leads to successful outcomes. Consider an irritable boy who wants his playmate’s toy. The playmate does not want to share, so the boy grabs the toy and the playmate starts to cry, relinquishing the toy. Aggression works. In other words, to the extent that aggressive behavior helps meet the child’s needs, it is likely to be sustained or increase, possibly leading to more extreme and maladaptive levels of aggression. However, during the preschool years children also learn to regulate and control their aggression according to the demands of the situation.

296

Nancy G. Guerra and Melinda S. Leidy

For example, Besevegis and Lore (1983) found that preschool children who played together with a teacher in the room were more aggressive than when the teacher left the room. Even at this early age children recognized that the risk of counterattack was higher without a teacher present and adjusted their behavior accordingly. As this illustrates, not only does adaptation depend on the ability to use aggression and to control aggression, but also on the ability to determine whether aggressive or non-aggressive strategies are optimal under speciﬁc environmental conditions. This has been described as ‘‘calibration’’ of response systems, meaning the ability to match responses to the demands of the environment (Malamuth & Heilmann, 1998). As children enter elementary school, their behavior becomes more compliant and aggression gradually declines. They are better able to delay gratiﬁcation and regulate their emotions and behavior according to the dictates of their social worlds. Their increasing cognitive sophistication also renders them better at understanding the nuances of aggression, for instance, whether an action was intentional or accidental. A robust literature has demonstrated that aggressive children are more likely to attribute hostile intent to others under ambiguous circumstances (Crick & Dodge, 1994). Another marked feature of peer relationships from the elementary school years onward is the establishment of social hierarchies. Direct and indirect aggression can serve to elevate an individual’s status in the peer hierarchy. This often begins as ‘‘rough and tumble’’ play during childhood, through which children build afﬁliations and establish dominance patterns (Humphreys & Smith, 1987). As children enter adolescence, the tactics become more subtle, involving gossip, social exclusion, and other forms of indirect aggression, often as part of membership in emerging social cliques. Under some circumstances, this type of aggression (often considered bullying) can lead to high levels of power and inﬂuence within a social group, particularly as children move into adolescence. A literature that emerged in the late 1990s has shown that although some aggressive children are rejected, many aggressive children and adolescents are afforded high levels of status, popularity, and admiration within their peer group (Adler & Adler, 1998; Bukowski, Sippola, & Newcomb, 2000; Farmer & Rodkin, 1996). Thus, as aggression becomes more normative (and often more indirect) during adolescence, it is less likely to engender peer rejection and more likely to elevate one’s social status. In disadvantaged contexts where resources are scarce and danger is high, adolescent aggression may not only result in elevated status but in a wide range of beneﬁts including material goods, protection, deterring rivals from aggression, and power (Fagan & Wilkinson, 1998; Guerra, 1998).

Advances in Understanding and Preventing Childhood Aggression

297

C. MOTIVATION FOR CHANGE The majority of preventive intervention programs for childhood aggression are based on the premise that aggression is maladaptive and dysfunctional and that aggressive children lack the social, emotional, and cognitive skills necessary for positive social interactions. An underlying assumption is that socially incompetent children and/or children with low social status should be highly motivated to change their aggressive behavior in order to ﬁt in better with peers. This may, indeed, be the case for some aggressive children who display social-cognitive biases and deﬁcits, poor social skills, and low peer status (Coie & Dodge, 1998). However, two important caveats should be considered in understanding motivation for change and how it can inﬂuence receptivity to prevention programs. First, as discussed earlier, aggression has an adaptive function that varies with age and across contexts. A marker of adjustment is the ability to ‘‘calibrate’’ one’s aggression according to the demands of the situation. For example, a child who is threatened by a peer may need to display a willingness to retaliate in order to avoid future victimization, just as a child who perceives every glance as hostile and reacts with aggression would need to improve cognitive cue search and interpretation skills. The important point is that extreme non-aggression may be just as maladaptive as excessive aggression. Second, some aggressive children are socially competent, high-status youth. Using cluster-analytic techniques, several studies have provided support for a subgroup of youth who are both popular and aggressive (Luthar & McMahon, 1996; Rodkin et al., 2000). In other words, for some youth in some settings, aggression can lead to high status and dominance within the social group. This does not mean that aggressive children are well-liked. For example, Prinstein and Cillessen (2003) found that aggression was associated with both low and high popularity among adolescents. However, the popular and aggressive youth generally were not well-liked by peers, suggesting a rather complex association between aggression and peer social status. To the extent that aggressive children have power and status, they may resist intervention efforts designed to reduce this behavior. As Vaillancourt et al. (2007) note, ‘‘convincing popular students to reduce bullying behavior will be difﬁcult, if not impossible, when such behavior is viewed as a source of privilege, power, and/or status among peers, and when the status afforded them leads them to view their social interactions as effective and successful’’ (p. 332). Furthermore, when more aggressive children are mixed together for small-group interventions, group effects may elevate the status of aggression so that it becomes more normative and acceptable.

298

Nancy G. Guerra and Melinda S. Leidy

For instance, analyses of data from the MACS project revealed that when highly aggressive youth were together in a small-group program, they socialized each other to become more aggressive over time—while children in groups who were initially comprised of less aggressive youth became less aggressive over time (Boxer et al., 2005). This is also consistent with research on deviant peer contagion, whereby grouping aggressive and delinquent offenders together has been found to result in increased antisocial behavior (Dodge, Dishion, & Lansford, 2006b). Considering the adaptive functions of aggression for social status, popularity, and control and the potential for group dynamics to elevate the status of aggression within the context of preventive intervention programs, it is important to recognize that motivation to change cannot be assumed. Rather, to the extent that aggression and popularity are linked in a given peer context, being ‘‘tough’’ and aggressive might be seen as a desired goal. This may also hinder efforts to encourage bystanders to intervene to stop aggression, particularly if this behavior carries a risk for loss of social status (Salmivalli et al., 1996). This suggests that an important strategy for preventive interventions is to change the adaptive value of aggression in a given setting. This may require moving beyond zero-tolerance policies in order to account for the normative reward structure within the peer group.

III. Aggression and the Ecology of Development A. AGGRESSION AS A MULTIPLY-DETERMINED BEHAVIOR Beyond the adaptive value of aggression in a given social context, still many other factors play a role in the etiology of aggression and help explain variations in this behavior across individuals and groups. There is general consensus that aggression is a multiply-determined behavior, inﬂuenced by individual factors such as personality, temperament, neuropsychological functioning, and biological predispositions, as well as contextual factors such as peer inﬂuences, family socialization, parenting practices, and community disadvantage (Eron, 1987). With aggression and violence increasingly seen as public health problems, much emphasis has been placed on the identiﬁcation of ‘‘risk factors’’ that increase the likelihood of aggression and ‘‘protective factors’’ that moderate the risk–aggression relation or act to promote healthy development when risk is absent (Jessor, 2007, personal communication). A further distinction has been made between non-modiﬁable or static risk factors that can be used to select high-risk youth for intervention (e.g., parental criminality, socioeconomic disadvantage) and modiﬁable or dynamic risk factors (e.g., cognitive

Advances in Understanding and Preventing Childhood Aggression

299

distortions, social skills, parenting practices) that can be targeted for change. However, one limitation of the risk and protective factor approach is that it has led to long lists of factors with very little theoretical integration highlighting the mechanisms or process by which aggression develops. There is clearly redundancy among risk factors suggesting that many risk factors reﬂect a common theme by virtue of their interrelatedness (e.g., socioeconomic disadvantage, parental stress, lack of social support). There are also varying paths to aggressive behavior in childhood and adolescence. An important contribution since the 1990s has been the development of more integrative theories and related empirical studies that emphasize multiple inﬂuences on development, multiple levels of inﬂuence, and mechanisms that explain the risk–behavior relation. For example, in a study of the process by which community violence exposure impacted children’s aggression, Guerra, Huesmann, and Spindler (2003) proposed that observational learning by witnessing violence would lead to an increase in normative beliefs about the acceptability of violence that, in turn, would lead to increased aggression. Indeed, in a large sample of inner-city elementary school children, normative beliefs about aggression were found to mediate the violence exposure–aggression relation. Speciﬁcation of the mechanisms by which risk operates can also have important implications for interventions, particularly when participants are selected based on non-modiﬁable (or difﬁcult to modify in the short run) risk factors. A common approach has been to offer selected preventive interventions for groups of youth at-risk by virtue of their living circumstances, for instance, economically disadvantaged, inner-city youth who are exposed to high levels of community violence. However, in order to counteract the effects of speciﬁc community risk factors, we must understand how they operate vis-a`-vis aggression. If economic disadvantage primarily compromises academic achievement, leading to school dropout and involvement in juvenile gangs, then the best intervention would be to provide academic tutoring and enhanced instruction during the early school years. Similarly, as Guerra et al. (2003) suggest, if violence exposure leads to approval of aggression, then the best intervention would be to counteract normative beliefs about the acceptability of aggression, particularly for children who regularly witness community violence. Of course, a more sustainable (albeit more difﬁcult) strategy would be to minimize factors that cause initial risk, including economic disadvantage and high levels of community violence. An extensive review of risk and protective factors for childhood aggression and violence is beyond the scope of this chapter and can be found elsewhere (see Dodge et al., 2006a for a comprehensive review).

300

Nancy G. Guerra and Melinda S. Leidy

Rather, we provide a brief review of the most robust risk factors for aggression that can be used to select participants for intervention and/or to design focused intervention programs. We group risk factors into characteristics of individuals, close interpersonal relationships (e.g., peers, families), and proximal contexts (e.g., neighborhoods, schools), emphasizing the ecology of development and the cumulative and interactive nature of risk factors (Tolan & Guerra, 1994; Tolan et al., 1995). We recognize the importance of the larger societal context (for instance, cultural norms and ﬁrearm policies) but concede that these inﬂuences are unlikely to be impacted by the short-term, psycho-educational preventive interventions that are the focus of this chapter.

B. RISK FOR AGGRESSION ACROSS INDIVIDUALS AND CONTEXTS 1. Characteristics of Individuals As we have discussed previously, a variety of individual characteristics have been identiﬁed that increase risk for childhood aggression. Some of these individual factors (such as perinatal trauma) begin in utero (Mungas, 1983), whereas others (such as difﬁcult temperament, fearlessness, impulsivity, low verbal ability, and lack of control) begin at birth or shortly after (Bates et al., 1991; Tremblay et al., 1994). Over time, distinct dimensions of personality including low agreeableness and low conscientiousness also crystallize and increase the likelihood of aggression (Miller & Lynam, 2001). In other words, a host of individual predispositions, whether written on a child’s biological birth certiﬁcate or emerging early in the course of development, render certain children more prone to aggression than others from a very early age. Without intervention, children who develop aggressive behavioral patterns early in life are also more likely to graduate to more serious aggression in adolescence and continue such behavior chronically (Mofﬁtt, 2003). For this reason, elevated aggression and its precursors in early childhood are among the best factors for selecting individuals or subgroups for focused prevention and intervention programs (Tolan & Loeber, 1993). However, selecting children based on early aggression does not provide speciﬁc guidance for the content and scope of the intervention itself. Indeed, many individual risk factors linked to temperament, personality, and neuropsychological functioning are difﬁcult to change, although how these unfold in a given context can dictate their course. It is important to bear in mind that children both shape and are shaped by their environments, a point we will return to in our subsequent discussion of contextual risk for

Advances in Understanding and Preventing Childhood Aggression

301

aggression and the cumulative and interactive inﬂuence of risk factors. For instance, difﬁcult temperament is more likely to result in ineffective parenting and ineffective parenting is more likely to exacerbate the relation between difﬁcult temperament and later aggression (Bates et al., 1991, 1998). It is also the case that children actively navigate and interpret their social worlds. How they come to understand both their own behavior and the behavior of others has important implications for action. Over time, children learn speciﬁc patterns of cognition that make aggression more or less likely. For example, one of the most robust ﬁndings in the socialcognitive literature on children’s aggression is the tendency of more aggressive children to attribute hostile intent to others under ambiguous circumstances, known as hostile attributional bias (Crick & Dodge, 1994; Dodge, 1986; Graham & Hudley, 1994; Guerra & Slaby, 1990). This means that a child who interprets another’s glance as hostile is more likely to respond with aggression than a child who believes the same glance is neutral or benign. Beginning in the 1960s, there has been an increasing recognition of the cognitive underpinnings of aggression (Anderson & Bushman, 2002; Bandura, 1986; Crick & Dodge, 1994; Huesmann, 1998). Most socialcognitive models of childhood aggression draw heavily from cognitive information-processing theory, emphasizing both discrete social informationprocessing skills as well as speciﬁc types of social knowledge stored in memory (the ‘‘data base’’ that individuals develop over time). Furthermore, because the child’s cognitive system develops over time, it is amenable to early preventive efforts while cognitions are most malleable (Huesmann & Guerra, 1997) as well as later efforts to modify maladaptive patterns of thought (Guerra & Slaby, 1990). Indeed, cognitive-behavioral prevention and intervention programs consistently have been shown to be effective for aggression, violence, and delinquency (Lipsey & Wilson, 1998). This leads us to ask what speciﬁc social information-processing skills and/or speciﬁc types of social knowledge are the most robust risk factors for childhood aggression and are the most viable targets for prevention and intervention? Much of the work in this area has emphasized discrete and sequential social information-processing skills that involve encoding and interpretation of cues, response search, evaluation, decision, and action (Crick & Dodge, 1994; Guerra & Huesmann, 2004). In short, the cognitive system is seen as processing inputs of social stimuli (what happened and why?), searching memory for relevant information (what does this mean?), and generating outputs accordingly (what should I do and what are the consequences?). In addition to hostile attributional bias, aggression is associated with increased attention to aggressive cues (Baumeister, Smart, & Boden, 1996; Dill et al., 1997), generation of more aggressive

302

Nancy G. Guerra and Melinda S. Leidy

solutions, and anticipation of positive outcomes such as tangible rewards for aggression (Guerra & Slaby, 1990; Perry, Perry, & Rasmussen, 1986). However, a child’s choice of an appropriate response also hinges on what is encoded in memory as acceptable behavior. We have referred to these internalized standards as normative beliefs about the appropriateness of aggression. These beliefs develop from observation of one’s own behavior and the behavior of inﬂuential models as well as from direct instruction across contexts. As children get older, normative beliefs about aggression become increasingly predictive of their own aggressive behavior (Huesmann & Guerra, 1997). A normative context that supports or sanctions aggression can also inﬂuence individual children’s behavior within that context and is thus an important focus for prevention and intervention programs (e.g., Olweus, 1993). Information-processing shortcuts and memory structures help decrease the cognitive workload. Over time, many of these biases and beliefs are invoked automatically without deliberate attention. Furthermore, expected events and actions often are linked together in scripts or event schemas that serve as guides for behavior in everyday situations. Because scripts also simplify cognitive processing, in many cases a particular scripted response becomes dominant or automatic. More aggressive children presumably have more well-connected and dominant aggressive scripts encoded in memory (Huesmann, 1998). This highlights the need to consider the importance of automatic as well as controlled processing for socialcognitive interventions.

2. Close Interpersonal Relationships As we have discussed previously, individual risk for aggression is molded and shaped by contextual inﬂuences. Even highly heritable characteristics such as temperament have been shown to interact with contextual factors such as parenting styles to exacerbate risk (Bates et al., 1991, 1998). Furthermore, individual factors that are primarily learned are highly inﬂuenced by models and reward structures across settings (Bandura, 1986). From birth, children are embedded in a series of close interpersonal relationships with parents, relatives, caring adults, siblings, and peers that shape their development rather than rubber stamp their genetic destiny. There is now a substantial literature documenting the effects of these relationships on aggressive behavior, with particular emphasis on the inﬂuence of parents and peers. Several aspects of the parent–child relationship have been shown to inﬂuence the development of aggression, including the quality of the parent–child relationship, parenting practices, and parental monitoring.

Advances in Understanding and Preventing Childhood Aggression

303

A consistent ﬁnding in the research literature is that certain parenting practices and parent–child relationships can increase the likelihood of child aggression, and that the inﬂuence of these factors is particularly salient for younger children. Children who experience rejection, neglect, or indifference from parents are more likely to display aggressive behavior (Bousha, & Twentyman, 1984; Dahlberg, 1998; Loeber & Stouthamer-Loeber, 1986). Parents who are neglectful or disengaged are often unresponsive to the needs of their children and demand little of them. These children may engage in aggressive behaviors to gain attention from their parents. In contrast, parents who are warm, supportive, and responsive have children who are less aggressive and exhibit less behavioral problems (Bates & Bayles, 1988). The quality of the parent–child relationship also inﬂuences child aggression. Although consistent discipline practices have been linked to lower levels of aggression, problematic discipline practices and erratic expressions of anger promote aggression in children (Patterson, 1982, 2002). Children become less inhibited from displaying aggression when discipline is inconsistent and parenting practices are inept. This often leads to cycles of mutually coercive behavior. Parents who use inconsistent discipline tactics have been found to punish children not only for deviant behaviors but for prosocial behaviors as well (Patterson, 1982). However, children are also part of this coercive cycle. Children will purposely use aversive behaviors, such as whining or tantrums, to coerce their parents into giving them what they want. The children are then rewarded for this behavior, because the parents give in, which reinforces the aggressive or aversive behavior. The use of corporal punishment also has been associated with increased aggression in children. There are several reasons for this. First, when parents resort to physical means of controlling and punishing their children they send a message that aggression is a normative, acceptable, and effective way to gain compliance (Bandura, 1973, 1986). When corporal punishment is used in response to children’s aggression, in essence, parents are punishing children with the very behavior they are trying to eliminate. This, in turn, communicates to the child that it is acceptable to hit others when they behave in ways they do not like. Second, the use of this disciplinary tactic leads to avoidance of the disciplinary ﬁgure, reducing parental opportunities to direct and inﬂuence their child. Third, corporal punishment also promotes hostile attributions, which in turn, predicts aggressive behavior. Experience with harsh treatment from parents results in children who are hypervigilant to hostile cues, who attribute hostile intent to others, access more aggression potential responses, and view aggression as a way to attain social beneﬁts (Dodge et al., 1986). Taken to

304

Nancy G. Guerra and Melinda S. Leidy

the extreme case of physical abuse, the evidence is compelling, with physical abuse linked to early aggression as well as violent and delinquent behavior during adolescence (Luntz & Widom, 1994). One of the goals of parenting is to teach children to behave independently in morally and socially acceptable ways. Attributing compliance to internal rather than external sources is an integral part of this process, and corporal punishment also has been found to interfere with this process by promoting external attributions (Gershoff, 2002; Hoffman, 1983). Physical force by the parent provides external controls to which children can attribute their compliance, and therefore, can propel children to avoid misbehaviors in order to avoid future punishment but does not teach children the responsibility to behave independently in morally and socially acceptable ways (Hoffman, 1983). Thus, the child may never learn socially acceptable ways of handling situations and instead views aggression and violence as a reasonable option for solving social conﬂicts. As children grow and become adolescents, lack of parental monitoring is associated with higher levels of aggression, violence, delinquency, as well as poorer relations with peers and teachers (Pettit et al., 2001). Monitoring refers to parents knowing where their children are, whom they are with, and what they are doing. Good supervision allows parents to respond appropriately to antisocial and delinquent behaviors, as well as minimizes the adolescent’s contact with risky circumstances. In addition to parental inﬂuences, characteristics of a child’s peer group can increase risk of aggression, although the speciﬁc mechanisms seem to vary by age. For younger children, aggression can lead to peer rejection (which then leads to increased aggression), particularly when this behavior is ineffective and/or excessive. Indeed, by the time children are in second and third grades, children demand more social competence from their friends where problem solving with less physical coercion is expected (Dodge et al., 1990; Kupersmidt & Patterson, 1991). Aggressive children who are quick to ﬁght and slow to employ negotiation, bargaining, and other forms of problem solving are more likely to be rejected by peers (Bierman, Smoot, & Aumiller, 1993; Fraser, 1996). However, as we mentioned earlier, aggression does not always lead to peer rejection. When children are viewed as defending themselves, they are usually viewed positively by their peers (Fraser, 1996; Lancelotta & Vaughn, 1989). In some settings and particularly as children get older, aggression and bullying can lead to increased popularity and social status (Luthar & McMahon, 1996; Rodkin et al., 2000; Vaillancourt et al., 2007). To the extent that aggression becomes more normative for certain youth during adolescence, it is less likely to engender peer rejection and more likely to elevate one’s social status.

Advances in Understanding and Preventing Childhood Aggression

305

During adolescence the inﬂuence of the peer social clique or network also increases, with peer groups providing further validation and support for the standards of behavior they are deﬁned by. Aggressive, antisocial, or delinquent peer groups tend to attract like-minded youth (a phenomenon known as homophily), and being in a deviant peer group tends to increase antisocial behavior, particularly for the moderately deviant youth who may still be experimenting with different behavioral styles (Tremblay et al., 1995a). The peer group can provide an organizational context for more sophisticated displays of aggressive and antisocial behavior, attracting more aggressive youth and also legitimizing their behavior as normative. In more extreme cases, such as high violence juvenile gangs, this context becomes highly structured and proscriptive with clear mandates for aggressive and delinquent behavior.

3. Proximal Contexts Two of the most important proximal developmental contexts for children are neighborhoods and schools. These contexts exert independent inﬂuences on children’s development and behavior, but also inﬂuence the quality and capacity of caregivers and others. Consistent with ecological principles, contexts are nested and interdependent (Bronfenbrenner, 1979). Consider the effects of community economic disadvantage. Family poverty increases the probability of peer-directed aggressive behavior by children, adolescents, and adults (Bradley & Corwyn, 2002; Sampson & Laub, 1994; Spencer, Dobbs, & Phillips, 1988). One potential mechanism of inﬂuence involves the effect of poverty on parents’ ability to raise their children. Faced with limited resources and support, multiple stressors, and unemployment (or multiple jobs), parents may have little time and energy left to actively participate in childrearing (McLoyd, 1990). For instance, Sampson and Laub (1994) found that family poverty was associated with harsh discipline, low supervision, and poor parent–child attachment, which was in turn related to delinquency. Neighborhood inﬂuences can also operate independent of their effect on families or other relationships. Consider the effect of exposure to community violence. Children (particularly boys) who are exposed to higher levels of community violence are more likely to be aggressive (Attar, Guerra, & Tolan, 1994; Morales & Guerra, 2006). Children who witness violence more regularly come to see it as acceptable behavior and internalize normative beliefs supporting aggression (Guerra et al., 2003). It may also be that high levels of community violence create a climate of fear where children are more attentive to aggressive cues and more willing to interpret ambiguous cues as threats (for their own safety).

306

Nancy G. Guerra and Melinda S. Leidy

Other neighborhood factors can decrease the risk of violence, even within disadvantaged and more violent communities. Sampson and colleagues (e.g., Sampson, Morenoff, & Gannon-Rowley, 2002; Sampson, Raudenbush, & Earls, 1997) coined the term collective efficacy to refer to the willingness of residents to intervene for the common good based on mutual trust and solidarity. Juvenile crime rates are lower in neighborhoods where residents monitor children’s play groups, intervene to prevent deviant behaviors such as truancy, confront people who are disturbing public space, and organize to maximize community resources (Sampson et al., 1997). In essence, the community assumes a parenting role in monitoring children’s behavior and garnering resources beyond what is done by individual families in their own homes. Characteristics of schools can also increase the likelihood of childhood aggression. Some of these characteristics are directly related to the communities they serve. Schools in more disadvantaged neighborhoods typically have fewer resources, higher student–teacher ratios, and higher turnover rates (McLoyd, 1990). These schools may simply be less able to educate children effectively. Not only do academic difﬁculties portend heightened aggression, but children who are struggling with school are less likely to feel connected to their school and more likely to dropout or engage in risky behaviors (Hawkins et al., 1999). In addition, speciﬁc school practices such as ability tracking (Dahlberg, 1998; Kerckhoff, 1988), assignment to classrooms with deviant peers for special education (Dodge et al., 2006a; Peetsma et al., 2001), and temporary suspension programs (Arum & Beattie, 1999) can foster negative peer group interactions and antisocial behavior. Even at the classroom level, the proportion of classmates who are aggressive and endorse aggressive normative beliefs has an inﬂuence on individual levels of aggression (Henry et al., 2000).

C. THE CUMULATIVE AND INTERACTIVE NATURE OF RISK Speciﬁc characteristics of individuals, close interpersonal relationships, and proximal social contexts increase risk for childhood aggression. Still, no single factor explains more than a modest proportion of variance. As suggested by an ecological framework, individuals are nested within a social system comprised of relationships, settings, and larger societal inﬂuences, all of which reciprocally inﬂuence each other as well. The effects of risk on aggression can accumulate over time and/or across settings, but the effects of risk also can be triggered only when other risk factors are present (or in direct proportion to the amount of other risk factors present). Cumulative models emphasize the additive nature of risk such that the number of risk

Advances in Understanding and Preventing Childhood Aggression

307

factors rather than any particular factor augments risk. An emphasis on how many risk factors are present suggests that children with the greatest number of risk factors should be identiﬁed for focused prevention and intervention, and that interventions should be multi-component and multicontext (Tolan et al., 1995). Interactive models emphasize the fact that many risk factors exert their inﬂuence contingently. For example, children with an early propensity to behave aggressively appear to be more sensitive to the effects of peer rejection than their less aggressive peers (Dodge & Pettit, 2003). Interactive models suggest that interventions should identify those at-risk children most likely to be impacted by a speciﬁc malleable risk factor and target services accordingly; however, in both cases greater risk exposure is linked to more negative outcomes. It is also important to consider how early risk factors, if left unchecked, can set in motion a downward spiral of events that increase the likelihood of aggression and other negative outcomes. A boy with a difﬁcult temperament who is spanked and harshly disciplined may come to see hitting as an effective strategy (and learn few other social skills), triggering rejection by peers at school and withdrawal from his parents, leading to even harsher discipline and more aggression. As his social-cognitive understanding crystallizes, he may develop a hostile attributional bias and aggressive scripts leading to more aggressive social interactions. The implication is that risk is transactional. In other words, risk factors for aggression exert a reciprocal inﬂuence on each other across time. This is consistent with a robust literature showing that children whose aggression becomes more marked early in development are more likely to develop chronic and persistent patterns of antisocial behavior later in life (Farrington, 1991; Mofﬁtt, 2003). In sum, advances in understanding the complex nature of individual and contextual risk and how it unfolds over time have signiﬁcant implications for the prevention of childhood aggression. We now turn to a discussion of the speciﬁc ‘‘lessons learned’’ from ecological models of risk, as illustrated by a large-scale prevention research trial, the Metropolitan Area Child Study.

IV. Risk, Causality, and Prevention A. PREVENTION AND THE MULTIPLE DETERMINANTS OF AGGRESSIVE BEHAVIOR Given the multiple determinants of aggressive behavior, the multiple processes by which risk can be exacerbated or reduced, variations in these

308

Nancy G. Guerra and Melinda S. Leidy

processes by age, and a ﬁnite amount of resources dedicated to prevention of childhood aggression, a critical issue involves how best to prioritize and direct our efforts. A starting point is to consider key questions that have yielded signiﬁcant advances and that have clear program and policy implications. We propose the following questions: (1) (2)

(3)

(4)

At what age should systematic prevention programming begin? Should the prevention net be cast widely for all youth or should we identify children (or populations) most at-risk for targeted preventive interventions? Which speciﬁc individual or contextual risk factors are the best candidates for preventive efforts and what types of programs have been proven most effective? Should we integrate programming across multiple contexts of development so that several risk factors are addressed simultaneously and anti-violence socialization mechanisms are consistent over time and across settings?

We draw on the extant literature to address each of these questions. To provide a speciﬁc example of how these issues have been addressed in prevention research trials, we discuss ﬁndings from the MACS, a large-scale development and prevention study conducted over the course of eight years (Metropolitan Area Child Study Research Group, 2002). We then summarize strategies for incorporating sound, empirically derived theories into program design.

B. THE METROPOLITAN AREA CHILD STUDY The Metropolitan Area Child Study (MACS) is a longitudinal schoolbased development and prevention study conducted during the 1990s with elementary school children from inner-city and urban communities (Guerra et al., 1993). It was funded under a request for applications issued by the National Institute of Mental Health in 1990 with the primary purpose being to develop, implement, and evaluate multi-component, multi-context antiviolence programs for at-risk children and youth. The study was grounded in a cognitive-ecological model of the development of aggression that stressed the social-cognitive and contextual factors empirically linked to the learning of aggression in childhood (Guerra et al., 1997). The speciﬁc socialcognitive areas targeted were self-understanding/self-efﬁcacy, social perspective taking, normative beliefs about aggression, social problem-solving

Advances in Understanding and Preventing Childhood Aggression

309

skills, and cognitive scripts. The contexts targeted for intervention were the classroom, peer group, and family. Although the cognitive-ecological model driving the intervention suggested it was important to modify multiple cognitions across these multiple contexts as they developed during the elementary years, from a practical standpoint it is unlikely that schools or community agencies would be able provide interventions for all children in their classrooms, peer groups, and families across all school years. To address the potential of this research to be translated into everyday settings, the study was designed to answer the question of how much intervention, at what age, and in which contexts is necessary to prevent aggression among the most aggressive children. Three intervention conditions were evaluated at two grade levels (in addition to a no-treatment control condition). The three intervention conditions (labeled Levels A, B, and C) represented increase in the number of contexts involved and the dose received through this cognitive-ecological model. The two grade levels were early elementary (Grades 2–3) and late elementary (Grades 5–6). The Level A intervention was seen as the most cost-effective and least intrusive method of intervention delivered for all children at the classroom level. This general enhancement classroom intervention provided a 2-year program that included teacher consultation on classroom management and a 40-lesson social-cognitive curriculum (Yes I Can) delivered by teachers in the classroom during the regular school day. The curriculum covered the ﬁve areas of social cognition described above. The Level B intervention provided the general enhancement classroom component plus a 2-year, small-group training for the most aggressive children. This general enhancement plus small-group peer-skills training intervention was designed to change cognitions and behavior among the most aggressive children as well as to minimize peer reinforcement of aggression by changing peer group norms about the acceptability of aggression (Eargle, Guerra, & Tolan, 1994). The Level C intervention provided the most costly and comprehensive intervention by adding a 1-year family intervention to the classroom enhancement and small-group program. The family intervention was designed primarily to help parents recognize and reinforce prosocial behavior, improve parenting skills, enhance family communication, and provide an opportunity for family support (Tolan & McKay, 1996). Finally, an important consideration was the extent to which community context and school resources moderated intervention effects. Although the need for preventive interventions may be greatest in the most distressed, inner-city contexts, the effects of psycho-educational interventions may simply be overwhelmed by the economic and social strain present in these

310

Nancy G. Guerra and Melinda S. Leidy

settings. To examine how efﬁcacy varied as a function of these school and community characteristics, each of the intervention conditions included schools from both low-income inner-city communities (average poverty rates of 40.25%) and moderate-income urban communities (average poverty rates of 25%). As has been reported previously (Metropolitan Area Child Study, 2002) there were signiﬁcant effects on aggressive behavior for high-risk children only for the most comprehensive intervention (Level C) when delivered early (Grades 2–3) and in the moderate resource communities. Furthermore, when the early intervention was followed by an additional 2-year intervention delivered later (Grades 5–6), the magnitude of the effect doubled. It is important to note that none of the intervention conditions yielded signiﬁcant positive effects for the older elementary school children; in fact, the Level B intervention that included small-group peer-skills training resulted in modest iatrogenic effects, consistent with research on deviant peer contagion (Dodge et al., 2006b). Taken in conjunction with the expanding research base in the ﬁeld of prevention science, these ﬁndings can provide direction for research-driven programs and policies.

C. RESEARCH-DRIVEN PROGRAMS AND POLICIES A noteworthy development in anti-violence prevention and intervention programming has been an increasing recognition of the importance of theory-driven versus problem-driven programs (Kerns & Prinz, 2002). Many problem-driven programs evolved in response to a particular community problem but with little rationale for the particular approach employed or emphasis on systematic evaluation. In contrast, theory-driven programs are based on sound, empirically derived theories, with evaluations emphasizing short- and long-term outcomes as well as mediators and moderators of change. A focus on theory-driven programs also has led to an expanding number of scientiﬁcally evaluated aggression prevention programs. This growing evidence-base provides direction on several important issues that bear directly on programs and policies. 1. At What Age Should Systematic Prevention Programming Begin? An important advance in the prevention and intervention of childhood aggression has been the convergence of evidence supporting the ‘‘earlier is better’’ dictum. It is clear that by the elementary school years, childhood aggression is predictive of later aggressive and antisocial behavior across cultures and contexts (Farrington, 1991; Mofﬁtt, 2003). This early

Advances in Understanding and Preventing Childhood Aggression

311

aggression does not appear spontaneously upon school entry, but is related to a myriad of individual, family, and community risk factors that exert their inﬂuence from birth onward. The importance of preventive efforts during the preschool period also is supported by several comprehensive reviews showing that programs for very young children can have short-term impacts on behavior as well as long-term impacts on the prevention of delinquency (Yoshikawa, 1994; Zigler, Taussig, & Black, 1992). Even within the elementary school years up until approximately age 12, the research evidence suggests that earlier intervention is better. An aggressive child who is disruptive at school entry is likely to alienate peers and teachers. In turn, this can lead to social rejection and academic failure that further escalate risk of aggression. Even a child who is not aggressive at school entry may experience new social or academic challenges leading to aggressive behavior. Cognitions and characteristic behavioral styles also appear to crystallize during the later elementary years, suggesting they are more malleable with younger children (Huesmann & Guerra, 1997). Findings from MACS clearly illustrated the importance of early intervention for this age group. This is consistent with a number of other early intervention studies that have found preventive beneﬁts for programs beginning in kindergarten or shortly after (Conduct Problems Prevention Research Group, 2002; Kellam et al., 1994; Tremblay et al., 1995b). Although some programs have proven effective with older elementary school children (e.g., Graham & Hudley, 1993), the ﬁfth and sixth grade children who participated in the MACS intervention did not display reductions in aggression, even with the most comprehensive program. However, the effects of the early intervention were signiﬁcantly enhanced when followed by a late intervention, suggesting that later interventions are most effective for children whose aggression has not yet become habitual. In sum, the evidence from prediction and prevention studies suggests that ‘‘earlier is better’’ when cognition and behavior are most malleable. As Dodge and Pettit (2003) note, ‘‘prevention during the early stages of the evolution of chronic conduct problems is more likely to be successful than intervention in adolescence, after antisocial outcomes have become inevitably overdetermined’’ (p. 363). Of course this does not mean that prevention programs should not try to reach older youth—given that adolescence is a time of heightened violence and victimization, it is also important to develop effective prevention programs for this age group. In practice, most programs for adolescents would be considered treatment rather than prevention because they involve identiﬁed (often adjudicated delinquent) youth. Unfortunately, there is scant evidence for effective prevention programs for adolescents, and only a limited number of effective

312

Nancy G. Guerra and Melinda S. Leidy

treatment programs for seriously antisocial and violent youth (Guerra, Kim, & Boxer, 2008).

2. Should the Prevention Net be Cast Widely for All Youth or Should We Identify Children (or Populations) Most At-Risk for Targeted Interventions? Universal interventions cast a wide net to include all individuals in a given setting. A common strategy is to provide programs for children in a classroom or school without identiﬁcation of those most at-risk for aggression. School-based universal programs often emphasize social skills deemed important regardless of risk status and/or normative standards (such as ‘‘no bullying’’) that apply to all students. Universal programs can provide a foundation for more focused programs by promoting antiviolence messages and skills. They may be successful in making aggression somewhat less normative and/or adaptive. However, it is less likely that they can provide the individualized attention and intensive intervention needed by the most at-risk youth. For this reason, a number of preventive interventions have included or been limited to a selected group of high-risk children. Given the constellation of risk factors, many of which are present from an early age, the question then becomes how best to determine risk status in order to select participants. The most common strategy has been to identify the most aggressive children (typically beginning in elementary school) from populations most at risk such as economically disadvantaged children living in violent urban neighborhoods. The success of this strategy hinges on the predictive accuracy of early aggression as a marker for later aggression, typically around 50%. In other words, approximately half of the children identiﬁed as aggressive in early childhood continue at elevated aggression levels or escalate to serious antisocial behavior. This raises a concern that many children will be unnecessarily involved in prevention programs, using valuable resources, potentially being labeled as aggressive or at-risk, and possibly being exposed unnecessarily to more aggressive youth. Still, the fact that children are being identiﬁed based on a behavior which is disruptive at the moment (regardless of future predictive accuracy) supports the inclusion of aggressive children in focused interventions. A primary caution is that programs are framed so as to reduce possible stigma and iatrogenic effects. In the MACS intervention, the peer intervention for the more aggressive children was described as a ‘‘leadership training program’’ to reinforce skills and beliefs of children who were likely to be inﬂuential in their classrooms. Being cognizant of the potential for peer contagion, the groups included children above the school median for

Advances in Understanding and Preventing Childhood Aggression

313

aggression, resulting in a mix of moderate and high aggressive youth. However, as we learned, this still resulted in iatrogenic effects—when highly aggressive youth were together in the small-group program, they socialized each other to become more aggressive over time, while children in groups who were initially comprised of less aggressive youth became less aggressive over time (Boxer et al., 2005). This suggests that interventions based on identiﬁcation of aggressive children should not group them together in small-group programs.

3. Which Specific Individual or Contextual Risk Factors are the Best Candidates for Preventive Efforts and What Types of Programs have been Proven Most Effective? Consistent with the organization of our earlier review of risk factors for childhood aggression, most psycho-educational preventive interventions can be grouped into one of three categories: individual-level interventions; close interpersonal relations interventions; and proximal social contexts interventions. A fourth category of multi-dimensional, multi-context programming represents different combinations of the above approaches and is discussed in the next section. In an earlier review of adolescent violence prevention programs, Tolan and Guerra (1994) noted that approximately half of the preventive interventions as of the mid-1990s would be considered individual-level interventions. These included a range of approaches including psychotherapy, behavior modiﬁcation, cognitive-behavioral programs, and social skills training. There has been a shift in the last two decades towards an increasing focus on social-cognitive and social-skills development programs. This is based on increasing evidence supporting the inﬂuence of social-cognitive factors on aggression from an early age and related support for the effectiveness of cognitive-behavioral programs (and general lack of empirical support for non-cognitive-behavioral counseling, social work, and other therapeutic preventive interventions). Furthermore, socialcognitive skills and beliefs are amenable to change through structured interventions in classrooms, youth agencies, and other intervention settings so that they are appropriate for universal, selected, and indicated programming. Individual-level interventions targeting social cognition and social skills now form the majority of prevention programs, particularly for school children (Wilson et al., 2001). A number of different social-cognitive skill programs have been developed, primarily for elementary school children, although some programs have been evaluated with preschool children and adolescents. Programs developed during the 1970s and 1980s generally

314

Nancy G. Guerra and Melinda S. Leidy

focused on a speciﬁc skill or area of social cognition, demonstrating improvements in areas such as moral reasoning (e.g., Arbuthnot & Gordon, 1986) and anger coping (e.g., Lochman et al., 1984). Although some single component programs were successful, in many cases social-cognitive gains did not translate into signiﬁcant behavioral improvements or long-term effects, suggesting that single component programs may be necessary but not sufﬁcient to change behavior. Subsequent programs were more likely to provide integrated programs that addressed multiple aspects of social cognition and skills related to individual risk for aggression. One of the most widely used multi-dimensional social-cognitive/social skills interventions aimed directly at reducing aggression and violence is Aggression Replacement Training. This 30-hour, multi-modal program for identiﬁed aggressive children and youth emphasizes skill acquisition, impulse and anger control, and moral reasoning development. Outcome evaluations have revealed some positive effects, particularly with older youth (Goldstein, 2004). Another popular program for elementary, middle, and high-school youth with demonstrated effectiveness is Life Skills Training (Botvin, Mihalic, & Grotpeter, 1998). Life Skills Training is a classroom-based universal program emphasizing decision-making, anger control, social competence, and peer resistance skills. Although originally developed as a drug abuse prevention program, it has also been shown to be effective in preventing aggressive and antisocial behavior (Botvin, Grifﬁn, & Nichols, 2006). Other universal interventions targeting multiple aspects of social cognition have been shown to be effective in changing cognition and behavior at the elementary school level, including the Providing Alternative Thinking Strategies Program (Greenberg, Kusche, & Mihalic, 1998), the Resolving Conﬂict Creatively Program (Aber, Brown, & Henrich, 1999), and the Second Step Violence Prevention Program (Grossman et al., 1997; McMahon & Washburn, 2003). Most interventions targeting close interpersonal relationships focus on families. Families are the primary socialization context children and their inﬂuence endures throughout childhood (although the salience of the peer group increases during adolescence). As we mentioned earlier in this chapter, a consistent ﬁnding in the research literature is that certain parenting practices and parent–child relationships can increase the likelihood of child aggression making them viable targets for preventive interventions. Furthermore, family risk begins early in development, although the nature of risk changes over time. Accordingly, family interventions have been developed for parents of infants, preschoolers, children, and teenagers, and the speciﬁc focus of the intervention has been connected to the nature of risk during these different developmental periods. Furthermore, family interventions are by and large selected or indicated due to the need to

Advances in Understanding and Preventing Childhood Aggression

315

identify participants most likely to beneﬁt from interventions. In some cases, services are offered to high-risk populations, such as low-income, ﬁrst-time mothers (regardless of their individual risk status). This practice is more typical for programs targeting infants and young children. One of the most widely cited early family intervention programs with demonstrated effectiveness for preventing childhood aggression is the Nurse Home Visitation Program (Olds et al., 1998). This program was designed to help women experiencing the transition to motherhood for the ﬁrst time, providing them with skills needed to gain conﬁdence and increase their self-efﬁcacy as a parent. The program also draws on attachment theory to highlight the signiﬁcance of trusting and warm relationships. Programs for parents of preschoolers tend to be more directly focused on effective parenting practices for high-risk families that encourage prosocial behavior and reduce aggression, such as the Incredible Years Training for Parents Program (Webster-Stratton et al., 2001). Family interventions for children and adolescents reﬂect a broad range of theoretical underpinnings and techniques. Some interventions emphasize behavioral parent training. One of the most well-known of these approaches is Parent Management Training developed by Patterson and colleagues for antisocial boys and their families (Patterson, 1982, 2002; Patterson, Reid, & Dishion, 1992). The emphasis of Parent Management Training is on changing interaction patterns of parents and children in order to decrease the ‘‘coercive’’ style of interacting that promotes child aggression and later delinquency. Other approaches to family intervention that have been found effective in preventing aggression and antisocial behavior include parent training but also address issues related to overall family functioning. For example, Functional Family Therapy is a family behavioral intervention designed several decades ago to work with less serious and generally younger, aggressive, and delinquent youth. It is a structured intervention that combines family systems concepts, social learning theory, behavior management, and most recently cognitive processes (Sexton & Alexander, 2000). A main focus of the program is to improve family functioning through increased family problem-solving, enhanced emotional bonds among family members, and improved ability of parents to provide structure and guidance to their children. Still other approaches that build on behavioral parent training also provide direct instruction for children in social and life skills as well as family practice sessions based on therapeutic play or parent–child interactive therapy (Herschell et al., 2002). For example, Families and Schools Together is a multi-family group intervention that uses a systems-based family strategy to build skills in children and empower

316

Nancy G. Guerra and Melinda S. Leidy

parents to be primary prevention agents for their children. The program includes play therapy, family therapy, and behavioral skills training for parents of elementary-school age children (although the program has also been used with younger and older children). Effectiveness has been demonstrated on a variety of positive behavioral and prevention outcomes during the elementary school years across diverse ethnic and economic groups (McDonald & Frey, 1999). For older youth who are most at-risk of violence and delinquency, comprehensive family interventions that incorporate parent training, family functioning, and management of external demands have been found to be effective. Perhaps the most widely cited and well-evaluated program of this type is Multisystemic Family Therapy. This family-based intervention targets family risk factors for adolescent antisocial behavior including low levels of parent monitoring, poor discipline practices, association with antisocial peers, and poor school performance. In addition to improving parents’ abilities to address these risk factors, Multisystemic Family Therapy also addresses barriers to family empowerment and effective functioning within the family ecology (Henggeler et al., 1998). Proximal social context interventions focus on changing the system-level or organizational inﬂuences on behavior rather than changing individuals or close interpersonal relationships directly. For example, although individual-level social-cognitive programs often are implemented in school or community settings, the primary focus is on changing the individual. The social settings most frequently targeted for change in programs to prevent childhood aggression are the classroom and the school. The most common types of classroom and school interventions are efforts to establish appropriate norms and expectations for classroom behavior and classroom or instructional management programs emphasizing effective teacher practices (Wilson et al., 2001). School-wide interventions emphasize different strategies including coordinated school-level planning and development (Cook, Murphy, & Hunt, 2000), creation of caring communities and enhancing school climate (Battistich et al., 1996), and strengthening teacher instructional practices (Metropolitan Area Child Study Research Group, 2002). An example of a school-wide approach that has gained considerable popularity in recent years is the Olweus Bullying Prevention Program (Olweus, 1993; Olweus, Limber, & Mihalic, 1999). This is a universal intervention engaging all adults and students to create a normative climate and set standards against bullying. The intervention does not include a curriculum or speciﬁc lessons but rather emphasizes the formation of a bullying prevention coordinating committee, increased supervision of locations where bullying is most likely to occur, class meetings and the

Advances in Understanding and Preventing Childhood Aggression

317

enforcement of class rules, and targeted strategies for working with bullies and victims. The overarching goal of this program is to change the normative context in order to discourage bullying and victimization. Although this program has shown effectiveness internationally (Olweus, 1993; Olweus et al., 1999), there is limited support for its success across schools in the U.S. It has been suggested that school-based prevention focused on changing the social context may need to carefully consider the combinations or sequences of programs and strategies that work best in order to design comprehensive packages of prevention strategies (Wilson et al., 2001). In other words, the complexity of risk suggests that a more comprehensive approach crossing multiple domains may have more powerful effects.

4. Should We Integrate Programming Across Multiple Contexts of Development So That Several Risk Factors are Addressed Simultaneously and Anti-Violence Socialization Mechanisms are Consistent Over Time and Across Settings? Throughout this chapter we have highlighted the multi-component, multi-context nature of risk for childhood aggression. The convergence of risk factors across domains and contexts points to the need for multicomponent, multi-context interventions. A signiﬁcant advance since the early 1990s has been the development and evaluation of various combined approaches, often assessed by testing the relative efﬁcacy of components alone or in various combinations. For example, the MACS intervention compared a classroom enhancement program for all students (Level A) with the classroom program plus a small-group peer-skills program (Level B), with the classroom plus small group plus family intervention (Level C). The research question was whether the extension in context (including peers and families) resulted in greater preventive effects that, in turn, warranted the additional costs involved. As discussed previously, for the moderately disadvantaged urban children participating in the intervention, only the combined condition (Level C) was effective in preventing aggression and only for the younger elementary school children. However, in the most distressed inner-city communities, even the most comprehensive and multi-context program was not found effective in preventing aggression. It may be that the community context of scarce resources, residential mobility, and high levels of violence simply overwhelmed any potential effects from the intervention (given that the intervention did not address these community factors). A number of other studies have added components to expand contexts impacted beyond individuals. For example, the Coping Power Program developed by Lochman and colleagues targeted an array of social-cognitive

318

Nancy G. Guerra and Melinda S. Leidy

problem solving skills in aggressive 4th and 5th grade boys (Lochman & Wells, 2004). An enhanced version of the program that also included 16 parent group sessions providing behavioral parent management training was more effective than the child program alone or a control (Lochman & Wells, 2004). Similar beneﬁts from adding a parent component to socialcognitive/social skills program have been found in other studies (Kazdin, 2003; Tremblay et al., 1995b). However, in some cases, multi-context interventions can have paradoxical effects, particularly when the peer context is addressed. For instance, Dishion and Andrews (1995) found that adding a peer intervention to a family program actually undermined the effects of the family intervention. This is consistent with the MACS ﬁndings wherein students in the classroom plus small-group intervention evidenced iatrogenic effects. In sum, the evidence suggests that the most effective combination is individual social-cognitive/social skills and family interventions. However, there is clearly a need for continued evaluation of multi-component, multi-context prevention programs. As Weissberg, Kumpfer, and Seligman (2003) note, ‘‘One of the ﬁeld’s highest priorities and payoffs will come from systematically evaluating multiyear, multi component programs that target multiple social and health outcomes’’ (p. 430).

V. Translating Research to Practice: Building an Evidence Base A. EVIDENCE-BASED PROGRAMS AND PRINCIPLES As we mentioned at the outset of this chapter, one of the most signiﬁcant developments in the ﬁeld of prevention science since the 1990s has been the recognition of the value of research-based preventive interventions and the importance of identifying and disseminating such empirically supported programs. In order to accomplish this goal there has been considerable dialogue about the appropriate consensus standards for identifying programs worthy of adoption, a push to create readily accessible registries of efforts, and the development of a new infrastructure of organizations speciﬁcally tasked with translating research into practice (Biglan et al., 2003). Most of the translational work in the ﬁeld of childhood aggression and youth violence has emphasized the identiﬁcation of empirically supported interventions based on standards set by panels or study groups. One of the most well-known efforts is the Blueprints project at the Center for the Study

Advances in Understanding and Preventing Childhood Aggression

319

and Prevention of Violence at the University of Colorado (Elliott & Mihalic, 2004). Using a rigorous standard of evidence (e.g., strong research design, sustained effects, multi-site replication), the center has reviewed over 600 aggression and violence prevention programs, identifying 10 original ‘‘Blueprint’’ programs, although one was subsequently dropped and two were added (for a current total of 11 programs). These programs include several interventions discussed in this chapter—the Olweus Bullying Prevention Program, Multisystemic Family Therapy, Functional Family Therapy, Nurse-Home Visitation, Life Skills Training, The Incredible Years, and Promoting Alternative Thinking Strategies. The center has also identiﬁed 18 promising programs with good scientiﬁc evidence. The central premise of this evidence-based approach is that programs must be implemented with strict ﬁdelity to the model as designed and evaluated. However, in practice there are many barriers to broad or consistent implementation of evidence-based programs, a point we will return to shortly. A parallel development in translational research has been a ‘‘common factor’’ approach based on evidence-based principles (Tashiro & Mortensen, 2006). Common factors approaches emphasize the speciﬁc elements or components of interventions that are most effective. For example, Nation et al. (2003) identiﬁed nine characteristics or principles of effective prevention for youth problem behaviors including opportunities for positive relationships, sociocultural relevance, and well-trained staff. These characteristics address both the program emphasis (for instance, on positive relationships) as well as important implementation characteristics (for instance, well-trained staff). Although proponents of the evidencebased model programs and common factors approach do not always agree, there is agreement that dissemination and implementation concerns merit focused attention (Biglan et al., 2003).

B. DISSEMINATION AND IMPLEMENTATION The science behind evidence-based programs does not automatically guarantee that they will be adopted in different settings and improve outcomes under all conditions. In practice, although policy shifts have favored evidence-based programs (and often state this as a requirement for funding), selection and careful implementation of evidence-based programs is the exception rather than the rule (Backer, 2000). In part, this is due to limitations in dissemination—many programs are not manualized or packaged to allow for easy distribution and adoption (although the intent of projects such as Blueprints is to provide careful guidelines for implementation). In some cases, widespread dissemination becomes a business enterprise,

320

Nancy G. Guerra and Melinda S. Leidy

with program costs exceeding many agency and school budgets. Very little research has been conducted to understand the process of program adoption, that is, how educators and other service providers make decisions to select, use, adapt, or combine speciﬁc evidence-based programs (Greenberg et al., 2003). Difﬁculties with dissemination also extend to efforts to infuse evidence-based strategies in agencies and systems. It is often the case that certain strategies become popularized at the expense of others and that a culture emerges supporting these strategies (which are sometimes misunderstood). The research on essential core components or effective strategies has lagged behind the documentation of evidence-based programs. It is also the case that evidence-based programs and strategies will not improve outcomes unless they are implemented properly. Although a clear premise of an evidence-based approach is to implement a program with ﬁdelity (including ﬁdelity in adhering to program principles), in practice, this has proven difﬁcult for several reasons. Programs that have proven efﬁcacious tend to be costly and demanding of both staff and participant engagement; however, few studies have examined the minimum intensity needed to produce meaningful change, assuming that a program will be implemented with ﬁdelity in its entirety. Although ﬁdelity is a worthwhile goal, in practice it is also likely that some modiﬁcations will be needed in order to adapt a program to local cultural conditions, resources, and needs. If practitioners do not see the relevance of a given intervention to a particular setting, they are unlikely to implement the program as planned. Furthermore, only recently are evaluations emphasizing program costs and providing cost-beneﬁt analyses of program impact that can impact both adoption and sustainability. It is typically the case that aggression prevention programs are designed and evaluated without consideration of whether they are low cost and sustainable within youth-serving systems, suggesting a need for greater dialogue and coordination among researchers and practitioners.

C. LINKING PREVENTION WITH POSITIVE YOUTH DEVELOPMENT Another issue related to dissemination and implementation is how to coordinate prevention efforts within a given setting and connect them with positive youth development activities. Although a focus on risk factors has dominated the ﬁeld of prevention of aggression and other problem behaviors since the 1980s, there has been a subsequent backlash generated by the negative connotations of risk models and their role in encouraging

Advances in Understanding and Preventing Childhood Aggression

321

a deﬁcit-oriented, problem-centered vision of youth (Damon, 2004). This has resulted in support for a positive youth development approach that focuses on building strengths and assets for all youth rather than correcting deﬁciencies in identiﬁed youth (Eccles & Gootman, 2002). Rather than pitting a risk strategy against a positive youth development strategy, a number of prevention researchers have called for a synthesis of prevention and promotion approaches (Guerra & Bradshaw, in press; Weissberg et al., 2003). Many of the risk factors for childhood aggression as well as the protective factors that prevent aggression can be recast as core competencies and supports for success that, when absent, lead to problem behaviors. In other words, positive outcomes can protect youth from adversity and support healthy development and success (Cicchetti, et al., 2000).

VI. Conclusion As we have illustrated throughout this chapter, a number of important advances in the understanding and prevention of childhood aggression have emerged since the 1980s. Several major shifts are worth highlighting. First, our understanding of the causes of aggression has shifted from general theories of aggression that emphasized nature versus nurture to integrated theories of development that emphasize the multiple predictors of aggression and how they interact across contexts and over time from conception onward. Rather than contrast nature versus nurture, the focus has shifted to the complex interplay between innate and learned contributions to aggression. From a developmental perspective, the child is seen as possessing certain individual propensities and temperamental risk that can escalate or decrease over time as a function of contextual inﬂuences and how they unfold. This individual risk is evident from an early age and certainly by elementary school when characteristic patterns of aggression emerge. Not only can contextual risk exacerbate the effects of individual risk, for instance the interaction of difﬁcult child temperament and ineffective parenting, but environmental contingencies also determine the adaptive value of aggression in a given setting. An important conclusion is that prevention should begin early in development when behavior is more malleable. Second, there has been an increasing emphasis on the child’s emerging pattern of social cognition. As we have seen, how a child interprets and understands his or her social world can impact patterns of responding, including aggression. Aggressive children are more likely than their less aggressive peers to overattribute hostile intent to others under ambiguous

322

Nancy G. Guerra and Melinda S. Leidy

circumstances, generate aggressive solutions, perceive the consequences of aggression to be more positive and less negative, endorse the legitimacy of aggression as a response, and develop aggressive scripts that render this behavior more automatic under commonplace circumstances. Fortunately, these social-cognitive patterns are quite amenable to modiﬁcation through intervention. Indeed, cognitive-behavioral programs that emphasize the link between social cognition and aggression have proven to be among the most effective preventive interventions (Guerra & Huesmann, 2004). Third, developments in the ﬁeld of prevention science have highlighted the importance of randomized controlled trials as opportunities to test developmental theories and to develop an evidence-base of effective programs. These trials have increasingly been used to test multi-component, multi-context interventions that address the complex nature of risk. As we have pointed out, many of the challenges and areas where new developments in understanding and preventing aggression are likely to occur will most likely be informed by advances in prevention science. Of particular importance is the need to specify more carefully the speciﬁc mediators of prevention outcomes in order to best identify critical program components or principles. Another concern is the conditions under which prevention is most effective. As we illustrated in our discussion of the MACS intervention study, important moderators such as community resources can also render programs more or less effective under different conditions. Finally, the challenge remains to understand the implementation conditions that must be met for programs to improve outcomes, including the need to align prevention programming with school and community-wide efforts to enhance positive youth development.

REFERENCES Aber, J. L., Brown, J. L., & Henrich, C. C. (1999). Teaching conflict resolution: An effective school-based approach to violence prevention. New York: National Center for Children in Poverty. Adler, P. A., & Adler, P. (1998). Peer power: Preadolescent culture and identity. New York: Rutgers University Press. Anderson, C. A., & Bushman, B. J. (2002). Human aggression. Annual Review of Psychology, 53, 27–51. APA Presidential Task Force on Evidence-Based Practice. (2006). Evidence-based practice in psychology. American Psychologist, 61, 271–285. Arbuthnot, J., & Gordon, D. A. (1986). Behavioral and cognitive effects of a moral reasoning development intervention for high-risk behavior-disordered adolescents. Journal of Consulting and Clinical Psychology, 54, 208–216. Arseneault, L., Mofﬁtt, T. E., Caspi, A., Taylor, A., Rijsdijk, F. V., Jaffee, S. R., Ablow, J. C., & Measelle, J. R. (2003). Strong genetic effects on cross-situational antisocial behavior

Advances in Understanding and Preventing Childhood Aggression

323

among 5-year-old children according to mothers, teachers, examiner-observers, and twins’ self reports. Journal of Child Psychology and Psychiatry, 44, 832–848. Arum, R., & Beattie, I. (1999). High school experience and the risk of adult incarceration. Criminology, 37, 515–537. Attar, B., Guerra, N. G., & Tolan, P. (1994). Neighborhood disadvantage, stressful life events, and adjustment in elementary school children. Journal of Clinical Child Psychology, 23, 394–400. Backer, T. E. (2000). The failure of success: Challenges of disseminating effective substance abuse prevention programs. Journal of Community Psychology, 28, 363–373. Bandura, A. (1973). Aggression: A social learning analysis. Englewood Cliffs, NJ: PrenticeHall. Bandura, A. (1986). Social foundations of thought and action: A social-cognitive theory. Englewood Cliffs, NJ: Prentice-Hall. Bates, J. E., & Bayles, K. (1988). Attachment and the development of behavior problems. In J. Belsky & T. Nezworski (Eds.), Clinical implications of attachment (pp. 253–299). Hillsdale, NJ: Erlbaum. Bates, J. E., Bayles, K., Bennett, D. S., Ridge, B., & Brown, M. M. (1991). Origins of externalizing behavior problems at eight years of age. In D. Pepler & K. Rubin (Eds.), Development and treatment of childhood aggression (pp. 93–120). Hillsdale, NJ: Erlbaum. Bates, J. E., Pettit, G. S., Dodge, K. A., & Ridge, B. (1998). Interaction of temperamental resistance to control and restrictive parenting in the development of externalizing behavior. Developmental Psychology, 34, 982–995. Battistich, V., Schaps, E., Watson, M., & Solomon, D. (1996). Prevention effects of the Child Development Project: Early ﬁndings from an ongoing multi-site demonstration trial. Journal of Adolescent Research, 11, 12–25. Baumeister, R. F., Smart, L., & Boden, J. M. (1996). Relation of threatened egotism to violence and aggression: The dark side of self esteem. Psychological Review, 103, 5–33. Besevegis, E., & Lore, R. (1983). Effects of an adult’s presence on the social behavior of preschool children. Aggressive Behavior, 9, 243–252. Bierman, K. L., Smoot, D. L., & Aumiller, K. (1993). Characteristics of aggressive-rejected, aggressive-nonrejected, and rejected-nonaggressive boys. Child Development, 64, 139–151. Biglan, A., Mrazek, P. J., Carnine, D., & Flay, B. (2003). The integration of research and practice in the prevention of youth problem behaviors. American Psychologist, 58, 433–440. Botvin, G. J., Grifﬁn, K. W., & Nichols, T. R. (2006). Preventing youth violence and delinquency through a universal school-based prevention approach. Prevention Science, 7, 403–408. Botvin, G. J., Mihalic, S. F., & Grotpeter, J. K. (1998). Blueprints for violence prevention, Book five: Life skills training. Boulder, CO: Center for the Study and Prevention of Violence. Bousha, D. M., & Twentyman, C. T. (1984). Mother–child interactional style in abuse, neglect, and control groups: Naturalistic observations in the home. Journal of Abnormal Psychology, 93, 106–114. Boxer, P., Guerra, N. G., Huesmann, L. R., & Morales, J. (2005). Proximal effects of a smallgroup selected prevention program on aggression in elementary school children: An investigation of the peer contagion hypothesis. Journal of Abnormal Child Psychology, 33, 325–338. Bradley, R. H., & Corwyn, R. F. (2002). Socioeconomic status and child development. Annual Review of Psychology, 53, 371–399. Bronfenbrenner, U. (1979). The ecology of human development: Experiments by nature and design. Cambridge, MA: Harvard University Press.

324

Nancy G. Guerra and Melinda S. Leidy

Bukowski, W. M., Sippola, L. K., & Newcomb, A. F. (2000). Variations in patterns of attraction to same- and other-sex peers during adolescence. Developmental Psychology, 36, 147–154. Buss, D. M., & Shackelford, T. K. (1997). Human aggression in evolutionary psychological perspective. Clinical Psychology Review, 17, 605–619. Campbell, A. (1993). Men, women, and aggression. New York: Basic Books. Caplan, M., Vespo, J. E., Pedersen, J., & Hay, D. F. (1991). Conﬂict over resources in small groups of 1- and 2-year-olds. Child Development, 62, 1513–1524. Caspi, A., McClay, J., Mofﬁtt, T. E., Mill, J., Martin, J., Craig, I. W., Taylor, A., & Poulton, R. (2002). Role of genotype in the cycle of violence in maltreated children. Science, 297, 851–854. Chaiken, M. R. (1998). Tailoring established after-school programs to meet urban realities. In D. S. Elliott, B. A. Hamburg, & K. R. Williams (Eds.), Violence in American schools: A new perspective (pp. 348–375). New York: Cambridge University Press. Cicchetti, D. & Carlson, V. (Eds.) (1989). Child maltreatment: Theory and research on the causes and consequences of child abuse and neglect. New York: Cambridge University Press. Cicchetti, D., Rappaport, J., Sandler, I. N., & Weissberg, R. P. (Eds.) (2000). The promotion of wellness in children and adolescents. Washington, DC: Child Welfare League of America Press. Coie, J. D., & Dodge, K. A. (1998). Aggression and antisocial behavior. In N. Eisenberg (Ed.), Handbook of child psychology (pp. 779–862). New York: Wiley. Coie, J. D., & Kupersmidt, J. (1983). A behavioral analysis of emerging social status in boys’ groups. Child Development, 54, 1400–1416. Conduct Problems Prevention Research Group. (2002). Evaluation of the ﬁrst three years of the Fast Track Prevention Trial with children at high risk for adolescent conduct problems. Journal of Abnormal Child Psychology, 30, 19–35. Cook, T. D., Murphy, R. F., & Hunt, H. D. (2000). Comer’s School Development Program in Chicago: A theory-based evaluation. American Educational Research, 37, 535–597. Crick, N. R., & Dodge, K. A. (1994). A review and reformulation of social informationprocessing mechanisms in children’s social adjustment. Psychological Bulletin, 115, 74–101. Crick, N. R., & Grotpeter, J. K. (1996). Children’s treatment by peers: Victims of relational and overt aggression. Development and Psychopathology, 8, 367–380. Dahlberg, L. L. (1998). Youth violence in the United States: Major trends, risk factors, and prevention approaches. American Journal of Preventive Medicine, 14, 259–272. Damon, W. (2004). What is positive youth development? Annals, AAPSS, 591, 13–24. Derzon, J. H., Wilson, S. J., & Cunningham, C. A. (1999). The effectiveness of school-based interventions for preventing and reducing violence. Final Report, Center for Evaluation Research and Methodology, Vanderbilt University. deWaal, F. (1999). The end of nature versus nurture. Scientific American, 281(6), 94–99. Dill, K. E., Anderson, C. A., Anderson, K. B., & Deuser, W. E. (1997). Effects of aggressive personality on social expectations and social perceptions. Journal of Research in Personality, 31, 272–292. Dishion, T. J, & Andrews, D. W. (1995). Parenting escalation in problem behaviors with highrisk young adolescents: Immediate and 1-year outcomes. Journal of Consulting and Clinical Psychology, 63, 538–548. Dodge, K. A. (1986). A social information processing model of social competence in children. In M. Perlmutter (Ed.), The Minnesota symposium on child psychology (pp. 77–1125). Hillsdale, NJ: Erlbaum. Dodge, K. A., & Coie, J. D. (1987). Social information-processing factors in reactive and proactive aggression in children’s peer groups. Journal of Personality and Social Psychology, 53, 1146–1158.

Advances in Understanding and Preventing Childhood Aggression

325

Dodge, K. A., & Pettit, G. S. (2003). A biopsychosocial model of the development of chronic conduct problems in adolescence. Developmental Psychology, 39, 349–371. Dodge, K. A., Pettit, G. S., McClaskey, C. L., & Brown, M. M. (1986). Social competence in children. Monographs of the Society for Research in Child Development, 51, (2, Serial No. 213). Dodge, K. A., Coie, J. D., Pettit, G. S., & Price, J. M. (1990). Peer status and aggression in boys’ groups: Developmental and contextual analysis. Child Development, 61, 1289–1309. Dodge, K. A., Lansford, J. E., Burks, V. S., Bates, J. E., Pettit, G. S., Fontaine, R., & Price, J. M. (2003). Peer rejection and social information-processing factors in the development of aggressive behavior problems in children. Child Development, 74, 374–393. Dodge, K. A., Coie, J. D., & Lynam, D. (2006a). Aggression and antisocial behavior in youth. In W. Damon (Ed.) & N. Eisenberg (Vol. Ed.), Handbook of child psychology: Vol. 3, Social, emotional, and personality development (6th ed., pp. 719–788). Hoboken, NJ: Wiley. Dodge, K. A., Dishion, T. J., & Lansford, J. E. (Eds.) (2006b). Deviant peer influences in programs for youth. New York: Guilford. Dollard, J., Doob, L. W., Miller, N. E., Mowrer, O. H., & Sears, R. R. (1939). Frustration and aggression. New Haven: Yale University Press. Dryfoos, J. (1990). Adolescents at risk: Prevalence and prevention. New York: Oxford University Press. Eargle, A., Guerra, N. G., & Tolan, P. H. (1994). Small-group treatment for children living in high violence neighborhoods. Journal of Child and Adolescent Group Therapy, 4, 229–243. Eccles, J. & Gootman, J. (Eds.) (2002). Community programs to promote youth development. Washington, DC: National Academy Press. Elliott, D. S., & Mihalic, S. (2004). Issues in disseminating and replicating effective prevention programs. Prevention Science, 5, 47–52. Eron, L. D. (1987). The development of aggressive behavior from the perspective of a developing behaviorism. American Psychologist, 42, 435–442. Espelage, D. L., Holt, M. K., & Henkel, R. R. (2003). Examination of peer-group contextual effects on aggression during early adolescence. Child Development, 74, 205–220. Fagan, J., & Wilkinson, D. L. (1998). The social contexts and functions of adolescent violence. In D. S. Elliott, B. A. Hamburg, & K. R. Williams (Eds.), Violence in American schools (pp. 55–93). New York: Cambridge University Press. Farmer, T. W., & Rodkin, P. C. (1996). Antisocial and prosocial correlates of social positions: The social network centrality perspective. Social Development, 5, 174–188. Farrington, D. P. (1991). Childhood aggression and adult violence: Early precursors and later life outcomes. In D. Pepler & K. Rubin (Eds.), The development and treatment of childhood aggression (pp. 5–29). Hillsdale, NJ: Erlbaum. Fox, N. A., Nichols, K. E., Henderson, H. A., Rubin, K., Schmidt, L., Hamer, D., Ernst, M., & Pine, D. S. (2005). Evidence for a gene–environment interaction in predicting behavioral inhibition in middle childhood. Psychological Science, 16, 921–926. Fraser, M. W. (1996). Aggressive behavior in childhood and early adolescence: An ecologicaldevelopmental perspective on youth violence. Social Work, 41, 347–361. Galen, B. R., & Underwood, M. K. (1997). A developmental investigation of social aggression among children. Developmental Psychology, 33, 589–600. Gershoff, E. T. (2002). Corporal punishment by parents and associated child behaviors and experiences: A meta-analytic and theoretical review. Psychological Bulletin, 128, 539–579. Goldstein, A. P. (2004). Evaluations of effectiveness. In A. P. Goldstein, R. Nensen, B. Daleﬂod, & M. Kalt (Eds.), New perspectives on aggression replacement training (pp. 230–244). Chichester: Wiley.

326

Nancy G. Guerra and Melinda S. Leidy

Graham, S., & Hudley, C. (1993). An attributional intervention to reduce peer-directed aggression among African-American boys. Child Development, 64, 124–138. Graham, S., & Hudley, C. (1994). Attributions of aggressive and nonaggressive AfricanAmerican male early adolescents: A study of construct accessibility. Developmental Psychology, 30, 365–373. Greenberg, M. T., Kusche, C., & Mihalic, S. F. (1998). Blueprints for violence prevention, Book ten: Promoting alternative thinking strategies (PATHS). Boulder, CO: Center for the Study and Prevention of Violence. Greenberg, M., Weissberg, R., O’Brien, M., Zins, J. E., Fredericks, L., Resnick, H., & Elias, M. J. (2003). Enhancing school-based prevention and youth development through coordinated social, emotional, and academic learning. American Psychologist, 58, 466–474. Grossman, D. C., Neckerman, H. J., Koepsell, T. D., Liu, P. Y., Asher, K. N., Beland, K., & Rivara, F. (1997). Effectiveness of a violence prevention curriculum among children in elementary school: A randomized controlled trial. Journal of the American Medical Association, 277, 1605–1611. Guerra, N. G. (1998). Intervening to prevent childhood aggression in the inner city. In J. McCord (Ed.), Children at risk: Growing up in the inner city (pp. 256–312). New York: Cambridge University Press. Guerra, N. G., & Huesmann, L. R. (2004). A cognitive-ecological model of aggression. Revue Internationale de Psychologie Sociale, 17, 177–203. Guerra, N. G., & Knox, L. (2002). Violence. Encyclopedia of crime and justice (pp. 1649–1655). New York: Macmillan. Guerra, N. G. & Phillips-Smith, E. (Eds.) (2005). Preventing youth violence in a multicultural society. Washington, DC: American Psychological Association Press. Guerra, N. G., & Slaby, R. G. (1990). Cognitive mediators of aggression in adolescent offenders: II. Intervention. Developmental Psychology, 26, 269–277. Guerra, N. G., Tolan, P. H., Huesmann, L. R., VanAcker, R., Eron, L. D., & Henry, D. (1993). Prevention of antisocial behavior in inner city and urban youth. National Institute of Mental Health Research Grant No. MH-48034. Guerra, N. G., Eron, L. D., Huesmann, L. R., Tolan, P. H., & VanAcker, R. (1997). A cognitive-ecological approach to the prevention and mitigation of violence and aggression in inner-city youth. In P. Fry & K. Bjorkqvist (Eds.), Cultural variation in conflict resolution: Alternatives for reducing violence (pp. 199–213). New York: Polonium. Guerra, N. G., Huesmann, L. R., & Spindler, A. (2003). Community violence exposure, social cognition, and aggression among urban elementary-school children. Child Development, 74, 1507–1522. Guerra, N. G., Boxer, P., & Cook, C. (2006). What works (and what does not) in preventing youth violence. Special issue. New Directions in Evaluation Research, 110, 59–72 Guerra, N. G., & Bradshaw, C. (Eds.) (in press). Core competencies to prevent problem behaviors and promote positive youth development. New Directions in Child and Adolescent Development. Guerra, N. G., Kim, T., & Boxer, P (2008). What works: Best practices with juvenile offenders. In R. Hoge, N. G. Guerra, & P. Boxer (Eds.), Treating the juvenile offender (pp. 79–102). New York: Guilford Press. Hawkins, J. D., Catalano, R. F., Kosterman, R., Abbott, R., & Hill, K. G. (1999). Preventing adolescent health-risk behaviors by strengthening protection during childhood. Archives of Pediatrics and Adolescent Medicine, 153, 226–324. Hawley, P. H. (1999). The ontogenesis of social dominance: A strategy-based evolutionary perspective. Developmental Review, 19, 97–132. Hawley, P. H. (2003). Strategies of control, aggression, and morality in preschoolers: An evolutionary perspective. Journal of Experimental Child Psychology, 85, 213–235.

Advances in Understanding and Preventing Childhood Aggression

327

Henggeler, S. W., Schoenwald, S. K., Borduin, C. M., Rowland, M. D., & Cunningham, P. B. (1998). Multisystemic treatment of antisocial behavior in children and adolescents. New York: Guilford Press. Henry, D., Guerra, N. G., Huesmann, L. R., Tolan, P., VanAcker, R., & Eron, L. D. (2000). Normative inﬂuences on aggression in urban elementary school classrooms. American Journal of Community Psychology, 28, 59–81. Herschell, A., Calzada, E., Eyberg, S., & McNeil, C. B. (2002). Parent–child interaction therapy: New directions in research. Cognitive and Behavioral Practice, 9, 9–15. Hobbes, T. (1651/1958). Leviathan. Indianapolis, IN: Liberal Arts Press. Hoffman, M. L. (1983). Affective and cognitive processes in moral internalization. In E. T. Higgins, D. N. Ruble, & W. W. Hartup (Eds.), Social cognition and social development (pp. 236–274). New York: Cambridge University Press. Huesmann, L. R. (1998). The role of social information processing and cognitive schema in the acquisition and maintenance of habitual aggressive behavior. In R. Geen & E. Donnerstein (Eds.), Human aggression (pp. 73–109). San Diego, CA: Academic Press. Huesmann, L. R., & Guerra, N. G. (1997). Social norms and children’s aggressive behavior. Journal of Personality and Social Psychology, 72, 408–419. Humphreys, A. P., & Smith, P. K. (1987). Rough and tumble, friendship, and dominance in school children: Evidence for continuity and change with age. Child Development, 38, 201–212. Jaffee, S. R., Caspi, A., Mofﬁtt, T. E., Dodge, K. A., Rutter, M., & Taylor, A. (2005). Nature x nature: Genetic vulnerabilities interact with physical maltreatment to promote conduct problems. Development and Psychopathology, 17, 67–84. Jessor, R. (1994). Problem-behavior theory and the life course in adolescence: Epistemology in action. Mind and Medicine, 8, 57–68. Jessor, R. (2007). Personal communication. Jessor, R., Donovan, J. E., & Costa, F. M. (1993). Problem behavior theory and young adulthood: Beyond methodological orthodoxy. Contemporary Psychology, 38, 898–900. Kazdin, A. (2003). Psychotherapy for children and adolescents. Annual Review of Psychology, 54, 253–276. Kellam, S. G., Rebok, G. W., Ialongo, N., & Mayer, L. S. (1994). The course and malleability of aggressive behavior from early ﬁrst grade into middle school: Results of a developmental epidemiologically-based preventive trial. Journal of Child Psychology and Psychiatry and Allied Disciplines, 35, 259–281. Kerckhoff, A. C. (1988). Effects of ability and grouping in British secondary schools. American Sociological Review, 51, 842–858. Kerns, S. E. U., & Prinz, R. J. (2002). Critical issues in the prevention of violence-related behavior in youth. Clinical Child and Family Psychology Review, 5, 133–160. Klimes-Dougan, B., & Kopp, C. B. (1999). Children’s conﬂict tactics with mothers: A longitudinal investigation of the toddler and preschool years. Merrill-Palmer Quarterly, 45, 226–241. Kupersmidt, J. B., & Patterson, C. J. (1991). Childhood peer rejection, aggression, withdrawal, and perceived competence as predictors of self-reported behavior problems in preadolescence. Journal of Abnormal Child Psychology, 19, 427–449. Lancelotta, G. X., & Vaughn, S. (1989). Relation between types of aggression and sociometric status: Peer and teacher perceptions. Journal of Educational Psychology, 81, 86–90. Lipsey, M. W., & Wilson, D. (1998). Effective intervention for serious offenders: A synthesis of research. In R. Loeber & D. Farrington (Eds.), Serious and violent juvenile offenders (pp. 313–345). Thousand Oaks, CA: Sage. Little, T., Brauner, J., Jones, S., Nock, M., & Hawley, P. (2003). Rethinking aggression: A typological examination of the functions of aggression. Merrill-Palmer Quarterly, 49, 343–369.

328

Nancy G. Guerra and Melinda S. Leidy

Lochman, J. E., Burch, P. R., Curry, J. F., & Lampron, L. B. (1984). Treatment and generalization effects of cognitive-behavioral and goal-setting interventions with aggressive boys. Journal of Consulting and Clinical Psychology, 52, 915–916. Lochman, J. E., & Wells, K. C. (2004). The coping power program for preadolescent aggressive boys and their parents: Outcome effects at the 1-year follow-up. Journal of Consulting and Clinical Psychology, 72, 571–578. Loeber, R., & Stouthamer-Loeber, M. (1986). Family factors as correlates and predictors of juvenile conduct problems and delinquency. In T. M. Morris (Ed.), Crime and justice: An annual review of research (Vol. 7, pp. 29–149). Chicago: University of Chicago Press. Lorenz, K. (1966). On aggression. New York: Harcourt. Luntz, B. K., & Widom, C. S. (1994). Antisocial personality disorders in abused and neglected children grown up. American Journal of Psychiatry, 151, 670–674. Luthar, S. S., & McMahon, T. J. (1996). Peer reputation among inner-city adolescents: Structure and correlates. Journal of Research on Adolescence, 6, 581–603. Malamuth, N. M., & Heilmann, M. F. (1998). Evolutionary psychology and sexual aggression. In C. H. Crawford & D. L. Krebs (Eds.), Handbook of evolutionary psychology (pp. 515–542). Hillsdale, NJ: Erlbaum. McDonald, L., & Frey, H. E. (1999). Families and schools together: Building relationships. OJJDP Bulletin. Washington, DC: U.S. Department of Justice, Ofﬁce of Justice Programs, Ofﬁce of Juvenile Justice and Delinquency Prevention. McLoyd, V. (1990). The impact of economic hardship on black families and children: Psychological distress, parenting, and socioemotional development. Child Development, 61, 311–346. McMahon, S. D., & Washburn, J. (2003). Violence prevention: An evaluation of program effects with urban African-American students. The Journal of Primary Prevention, 24, 43–62. Metropolitan Area Child Study Research Group. (2002). A cognitive-ecological approach to preventing aggression in urban settings: Initial outcomes for high-risk children. Journal of Consulting and Clinical Psychology, 70, 179–194. Miller, J. D., & Lynam, D. R. (2001). Structural models of personality and their relation to antisocial behavior: A meta-analytic review. Criminology, 39, 765–792. Mofﬁtt, T. E. (2003). Life-course persistent and adolescence-limited antisocial behavior: A 10-year research review and a research agenda. In B. Lahey, T. E. Mofﬁtt, & A. Caspi (Eds.), The causes of conduct disorder and serious juvenile delinquency (pp. 49–75). New York: Guilford. Morales, J., & Guerra, N. G. (2006). Effects of multiple context and cumulative stress on urban children’s adjustment in elementary school. Child Development, 77, 907–923. Mrazek, P. J. & Haggerty, R. J. (Eds.) (1994). Reducing risks for mental disorders: Frontiers for preventive intervention research. Washington, DC: National Academy Press. Mungas, D. (1983). An empirical analysis of speciﬁc syndromes of violent behavior. Journal of Nervous and Mental Diseases, 171, 354–361. Nation, M., Crusto, C., Wandersman, A., Kumpfer, K. L., Seybolt, D., Morrissey-Kane, E., & Davino, K. (2003). What works in prevention: Principles of effective prevention. American Psychologist, 58, 449–456. O’Connor, T. G., McGuire, S., Reiss, D., Hetherington, E., & Plomin, R. (1998). Co-occurrence of depressive symptoms and antisocial behavior in adolescence: A common genetic liability. Journal of Abnormal Psychology, 107, 27–37. Olds, D. L., Hill, P. L., Mihalic, S. F., & O’Brien, R. A. (1998). Blueprints for violence prevention, Book seven: Prenatal and infancy home visitation by nurses. Boulder, CO: Center for the Study and Prevention of Violence. Olweus, D. (1993). Bullying at school: What we know and what we can do. Oxford, UK: Blackwell Publishers.

Advances in Understanding and Preventing Childhood Aggression

329

Olweus, D., Mattson, A., & Low, H. (1988). Circulating testosterone levels and aggression in adolescent males: A causal analysis. Psychosomatic Medicine, 50, 261–272. Olweus, D., Limber, S., & Mihalic, S. F. (1999). Blueprints for violence prevention, Book nine: Bullying Prevention Program. Boulder, CO: Center for the Study and Prevention of Violence. Patterson, G. R. (1982). Coercive family process. Eugene, OR: Castalia Press. Patterson, G. R. (2002). The early development of coercive family processes. In J. B. Reid, G. R. Patterson, & J. Snyder (Eds.), Antisocial behavior in children and adolescents (pp. 25–44). Washington, DC: American Psychological Association. Patterson, G. R., Reid, J. B., & Dishion, T. J. (1992). Antisocial boys: A social interactional approach. Vol. 4. Eugene, OR: Castalia. Peetsma, T., Vergeer, M., Roeleveld, J., & Karsten, S. (2001). Inclusion in education: Comparing pupils’ development in special and regular education. Educational Review, 53, 125–135. Perry, D. G., Perry, L. C., & Rasmussen, P. (1986). Cognitive social learning mediators of aggression. Child Development, 57, 700–711. Pettit, G. S., Laird, R. D., Dodge, K. A., Bates, J. E., & Criss, M. M. (2001). Antecedents and behavior-problem outcomes of parental monitoring and psychological control in early adolescence. Child Development, 72, 583–598. Prinstein, M. J., & Cillessen, A. H. N. (2003). Forms and functions of adolescent peer aggression associated with high levels of peer status. Merrill-Palmer Quarterly, 49, 310–342. de Quervain, D. J. F., Fischbacher, U., Treyer, V., Schellhammer, M., Schnyder, U., Buck, A., & Fehr, E. (2004). The neural basis of altruistic punishment. Science, 305, 1254–1258. Rodkin, P. C., Farmer, T. W., Pearl, R., & VanAcker, R. (2000). Heterogeneity of popular boys: Antisocial and prosocial conﬁgurations. Developmental Psychology, 36, 14–24. Rousseau, J. J. (1762/1979). E´mile. New York: Basic Books. Salmivalli, C., Lagerspetz, K., Bjorkqvist, K., Osterman, K., & Kaukiainen, A. (1996). Bullying as a group process: Participant roles and their relations to social status within the group. Aggressive Behavior, 22, 1–15. Sampson, R. J., & Laub, J. H. (1994). Urban poverty and the family context of delinquency: A new look at structure and process in a classic study. Child Development, 65, 523–540. Sampson, R. J., Raudenbush, S. W., & Earls, F. (1997). Neighborhoods and violent crime: A multilevel study of collective efﬁcacy. Science, 277, 918–924. Sampson, R. J., Morenoff, J. D., & Gannon-Rowley, T. (2002). Assessing ‘‘neighborhood effects’’: Social processes and new directions in research. Annual Review of Sociology, 28, 443–478. Sexton, T. L., & Alexander, J. F. (2000). Functional family therapy (Bulletin). Washington, DC: Department of Justice, Ofﬁce of Justice Programs, Ofﬁce of Juvenile Justice and Delinquency Prevention. Shaw, D. S. (2003). Innovative approaches and methods to the study of children’s conduct problems. Social Development, 12, 322–326. Spencer, M. B., Dobbs, B., & Phillips, D. (1988). African-American adolescents: Adaptational processes and socioeconomic diversity in behavioral outcomes. Journal of Adolescence, 11, 117–137. Tashiro, T., & Mortensen, L. (2006). Translational research: How social psychology can improve psychotherapy. American Psychologist, 61, 959–966. Tolan, P. H., & Guerra, N. G. (1994). Prevention of delinquency: Current status and issues. Applied and Preventative Psychology, 3, 251–273. Tolan, P. H., & Loeber, R. (1993). Antisocial behavior. In P. H. Tolan & B. J. Cohler (Eds.), Handbook of clinical research and practice with adolescents (pp. 307–331). New York: Wiley.

330

Nancy G. Guerra and Melinda S. Leidy

Tolan, P. H., & McKay, M. (1996). Preventing serious antisocial behavior in inner-city children: An empirically based family intervention program. Family Relations: Journal of Applied Family and Child Studies, 45, 145–155. Tolan, P. H., Guerra, N. G., & Kendall, P. C. (1995). A developmental-ecological perspective on antisocial behavior in children and adolescents: Toward a uniﬁed risk and intervention framework. Journal of Consulting and Clinical Psychology, 63, 579–584. Tooby, J., & Cosmides, L. (1990). On the universality of human nature and the uniqueness of the individual: The role of genetics and adaptation. Journal of Personality, 58, 17–67. Tremblay, R. E., Pihl, R. O., Vitaro, R., & Dobkin, P. L. (1994). Predicting early onset of male antisocial behavior from preschool behavior. Archives of General Psychiatry, 51, 732–739. Tremblay, R. E., Maˆsse, L. C., Vitaro, F., & Dobkin, P. L. (1995a). The impact of friend’s deviant behavior on early onset of delinquency: Longitudinal data from 6 to 13 years of age. Development and Psychopathology, 7, 649–667. Tremblay, R. E., Kurtz, L., Maˆsse, L. C., Vitaro, F., & Phil, R. O. (1995b). A bimodal preventive intervention for disruptive kindergarten boys: Its impact through adolescence. Journal of Consulting and Clinical Psychology, 63, 560–568. Vaillancourt, T., Hymel, S., & McDougall, P. (2007). Bullying is power: Implications for school-based intervention strategies. In J. E. Zins, M. J. Elias, & C. A. Maher (Eds.), Bullying, victimization, and peer harassment (pp. 317–337). New York: Haworth Press. Webster-Stratton, C., Mihalic, S., Fagan, A., Arnold, D., Taylor, T., & Tingley, C. (2001). Blueprints for violence prevention, Book eleven: The incredible years: Parent, teacher, and child training series. Boulder, CO: Center for the Study and Prevention of Violence. Weiss, B., Harris, V., Catron, T., & Han, S. H. (2003). Efﬁcacy of the RECAP intervention program for children with concurrent internalizing and externalizing problems. Journal of Consulting and Clinical Psychology, 71, 364–374. Weissberg, R. P., Kumpfer, K. L., & Seligman, M. E. P. (2003). Prevention that works for children and youth: An introduction. American Psychologist, 58, 425–432. Wilson, D. B., Gottfredson, D. C., & Najaka, S. S. (2001). School-based prevention of problem behaviors: A meta-analysis. Journal of Quantitative Criminology, 17, 247–272. Wright, J. C., Giammarino, M., & Parad, H. W. (1986). Social status in small groups: Individual-group similarity and the social ‘‘misﬁt’’. Journal of Personality and Social Psychology, 50, 523–536. Yoshikawa, H. (1994). Prevention as cumulative protection: Effects of early family support and education on chronic delinquency and its risks. Psychological Bulletin, 115, 28–54. Zigler, E., Taussig, C., & Black, K. (1992). Early childhood intervention: A promising preventive for juvenile delinquency. American Psychologist, 47, 997–1006.

THE SOCIAL COGNITIVE NEUROSCIENCE OF INFANCY: ILLUMINATING THE EARLY DEVELOPMENT OF SOCIAL BRAIN FUNCTIONS

Mark H. Johnsona, Tobias Grossmanna, and Teresa Farronia,b a

CENTRE FOR BRAIN AND COGNITIVE DEVELOPMENT, SCHOOL OF PSYCHOLOGY, BIRKBECK, LONDON, UK b UNIVERSITY OF PADUA, ITALY

I. INTRODUCTION II. FACE PROCESSING A. THE SENSORY HYPOTHESIS B. NON-FACE STRUCTURAL PREFERENCES C. NEWBORNS HAVE COMPLEX FACE REPRESENTATIONS D. DEVELOPING CORTICAL AREAS FOR FACE PROCESSING III. EYE GAZE PROCESSING A. NEWBORNS DETECT DIRECT GAZE B. GAZE FOLLOWING C. NEURAL PROCESSING OF DIRECT GAZE IN INFANTS D. UNDERSTANDING EYE–OBJECT RELATIONS IV. PERCEPTION OF EMOTIONS A. PROCESSING EMOTION IN THE FACE B. PROCESSING EMOTION IN THE VOICE C. INTEGRATION OF EMOTIONAL INFORMATION FROM FACE AND VOICE V. INTERACTIONS BETWEEN FACE IDENTITY, EYE GAZE, AND EMOTION A. DISSOCIABLE NEURAL PATHWAYS FOR PROCESSING FACES B. INTERACTIONS BETWEEN FACE PROCESSING PATHWAYS IN ADULTS C. INTERACTING FACE PATHWAYS IN INFANTS VI. CONCLUSIONS REFERENCES

I. Introduction A distinct network of cortical and sub-cortical structures is selectively activated when we are presented with, or need to process, information about fellow humans. This network has been termed the ‘‘social brain’’

331 Advances in Child Development and Behavior R.V. Kail : Editor

Copyright r 2008 Elsevier B.V. All rights reserved.

332

Mark H. Johnson et al.

(Adolphs, 1999, 2003; Brothers, 1990), and the speciﬁc ﬁeld of research around it has become known as ‘‘social (cognitive) neuroscience.’’ Although hundreds of studies have focused on this network in adults, its origins in infancy remain unclear. Studying the social brain in infancy is important not only for the developmental psychologists, but also for those interested in fundamental questions about the adult system. For example, there is a longstanding debate in the adult literature about the degree to which a cortical area known to be activated by faces, the ‘‘fusiform face area (FFA)’’ (Kanwisher, McDermott, & Chun, 1997), is speciﬁcally dedicated for processing this type of stimulus (to the exclusion of others). While some argue that the region undertakes (face) category-speciﬁc processing (Kanwisher, 2000), others have presented evidence that the region can be recruited by any class of stimulus to which the participant acquires perceptual expertise (Gauthier et al., 1999). Evidence relevant to this debate comes from developmental studies in which the effects of experience can be more easily assessed (Cohen-Kadosh & Johnson, 2007). More importantly, accounts of the emergence of specialized functions in the brain that are inspired by developmental considerations can reconcile apparently conﬂicting views based only on data from adults (Johnson, 2005a). The earliest stage of postnatal development, infancy (0–2 years), is the time of life during which the most rapid changes take place. The dependent newborn seems almost a different species from the active, inquisitive 2-yearold toddler. During this period, human infants are faced with at least two vital and daunting challenges. The ﬁrst is to understand their physical world and its properties, such as that objects still exist even when they are out of sight, are solid, and tend to fall downwards from tables. The second major challenge is for the infant to gain understanding of the social world, ﬁlled with other people, including parents, siblings, and other family members. Learning how to act in the physical world is difﬁcult, but it is generally predictable in the sense that most events are contingent and highly replicable. In contrast, learning about the social world means not only developing the ability to predict the general behavior and responses of fellow human beings, but also going beyond the sensory input to understand the intentions, beliefs, and desires of others. Although the social world has some predictability, some cues are less reliable than others, and events are rarely exactly the same from one time to the next. Furthermore, relating socially to others has not only profound effects on what infants feel, think, and do, it is also essential for healthy development and for optimal functioning throughout life. Indeed, the ability to learn from other humans is perhaps one of the most important adaptations of our species (Csibra & Gergely, 2006). Therefore, developing an understanding of other people is arguably the most fundamental task that infants face in learning about their world.

The Social Cognitive Neuroscience of Infancy

333

The development of the brain circuitries involved in all kinds of cognitive processes depends upon the interaction of two broad factors: nature (our inheritance or genetic factors), and nurture (environmental inﬂuences). If we aim to understand how these factors interact to ‘‘build’’ the mature social brain network, it seems of particular importance to look at how the human brain deals with social information during infancy. In this chapter we review several aspects of the emerging social brain in infancy from the perspective of two related theoretical viewpoints. The ﬁrst of these viewpoints derives from Johnson’s (2005a) updated version of Johnson and Morton’s (1991; Morton & Johnson, 1991) two-process model. The original two-process theory sought to reconcile apparently conﬂicting lines of evidence about the development of face processing by postulating the existence of two systems: a tendency for newborns to orient to faces (Conspec), and an acquired specialization of cortical circuits for other aspects of face processing (Conlern). We hypothesized that Conspec served to bias the input to developing cortical circuitry over the ﬁrst weeks and months of life thus ensuring that appropriate cortical specialization occurred in response to the social and survival-relevant stimulus of faces. The Conspec notion occupies the middle ground between those who argue that face preferences in newborns are due to low-level psychophysical biases, and others who propose that infants’ representations of faces are, in fact, richer and more complex than we supposed (which we describe in detail in the next section). Johnson and Morton (1991) speculated that Conspec was mediated largely, but not necessarily exclusively, by subcortical visuo-motor pathways. Brieﬂy, this proposal was made for several reasons: (a) the newborn behavioral preference declined at the same age as other newborn reﬂexes assumed to be under sub-cortical control, (b) evidence from the maturation of the visual system indicating later development of cortical visual pathways (e.g. Johnson, 1990), and (c) evidence from another species (the domestic chick) (see Horn, 2004). Subsequently, cognitive neuroscience, electrophysiological, and neuropsychological studies with adults provided evidence for a rapid, low-spatial frequency, sub-cortical face detection system in adults that involves the superior colliculus, pulvinar, and amygdala (Johnson, 2005a). For example, evidence from patients with hemi-spatial neglect indicates that they display visual extinction to stimuli in the neglected ﬁeld unless arranged in the pattern of a face (Vuilleumier, 2000; Vuilleumier & Sagiv, 2001). Vuilleumier uses this and other evidence to propose that the human amygdala is part of a low-spatial frequency sub-cortical face processing route that boosts cortical processing of faces when these stimuli are most relevant—a system for ‘‘emotional attention.’’ These and other lines of evidence strongly support the view that there is a ‘‘quick and dirty’’ neural

334

Mark H. Johnson et al.

route for face processing in adults, and that this same system operates in young infants, including newborns (Johnson, 2005a). The second theoretical viewpoint that motivates this review of the social brain in infancy concerns a model of functional organization in the cerebral cortex called ‘‘Interactive Specialization.’’ Johnson (2001) presented three distinct, but not necessarily mutually exclusive, frameworks for understanding postnatal development in humans. According to a ‘‘Maturational’’ perspective, functional brain development involves the sequential coming ‘‘on line’’ of a number of modular cortical regions. The maturation of a given region is thought to allow or enable advances in the perceptual, cognitive, or motor abilities of the child. As applied to the neurodevelopment of social cognition, this implies that more complex aspects (such as ‘‘theory of mind’’ computations) will depend on the maturation of associated cortical regions (possibly within the prefrontal cortex). The second perspective, ‘‘Skill learning,’’ argues for parallels between the dynamic brain events associated with the acquisition of complex perceptual and motor skills in adults, and the changes in brain function seen in infants and children come to successfully perform simpler tasks. From this perspective, it has been argued that some cortical regions become recruited for processing social information because typical humans become perceptual experts in this domain (Gauthier & Nelson, 2001). A third perspective, ‘‘Interactive specialization,’’ posits that functional brain development, at least in cerebral cortex, involves a process of specialization in which regions go from initially having very broadly tuned functions, to having increasingly ﬁnely tuned (more specialized) functions (Johnson, 2001). A consequence of increased specialization of cortical regions is the increasing focal patterns of cortical activation resulting from a given task demand or stimulus. By this view, some regions of cortex gradually become increasingly specialized for processing social stimuli and thus become recruited into the social brain network. In this chapter, we use the sub-cortical route hypothesis and the interactive specialization model to review and interpret studies on the development of face perception and identity, eye gaze perception, and the perception of emotional expressions. We then discuss how these different facets of social brain function interact, and speculate on useful directions for future research.

II. Face Processing The face provides a wealth of socially relevant information. To detect and recognize faces is therefore commonly considered to be an important

The Social Cognitive Neuroscience of Infancy

335

adaptation of social animals. From birth, human infants preferentially attend to some face-like patterns (Morton & Johnson, 1991; Johnson, 2005a) suggesting a powerful mechanism to bias the input that is processed by the newborn’s brain. Perhaps the most controversial aspect of the twoprocess theory was the proposal that ‘‘infants possess some information about the structural characteristics of faces from birth’’ (Morton & Johnson, 1991, p. 164). Although this committed us to the view that the general conﬁguration that composes a face was important, we did not commit to a speciﬁc representation. Nevertheless, our empirical observation from the early experiments with newborns, and evidence from other species, indicated that a stimulus with three high-contrast blobs corresponding to the approximate location of the eyes and mouth might be sufﬁcient. A stimulus that appeared minimally sufﬁcient to activate this representation (Figure 1) was termed ‘‘Conﬁg.’’ We did not claim that this stimulus was identical to the representation supporting the hypothesized Conspec system, and we remained open to the view that more ideal stimuli for activating Conspec could potentially be created. At the time of publication the idea that infants were born with face-speciﬁc information had been rejected by most in the ﬁeld, largely on the basis of experiments with 1- and 2-month-old infants that failed to show face preferences (see Johnson & Morton, 1991, for review). Because infants beyond the newborn stage did not prefer schematic faces over scrambled faces, it was concluded that newborns could not discriminate faces from other stimuli. At present, the Conspec notion occupies the middle ground between those who argue that face preferences in newborns are due to low-level psychophysical biases, and others who propose that infants’ representations of faces are, in fact, richer and more complex than we supposed. We defend the view that conﬁgurational information about faces is important, and argue that this conﬁgurational information is sufﬁcient to account for most of the currently available evidence pertaining to schematic and naturalistic face-like images. Between 1991 and 2007, there were at least 16 studies conducted on face preferences in newborns (see Johnson, 2005a, for review). All of these studies found some evidence of discrimination of face-like patterns. However, one study failed to demonstrate a preference for face conﬁguration (Easterbrook et al., 1999). In this study the authors failed to ﬁnd evidence of preferential

Fig. 1. Examples of schematic face stimuli used in the newborn experiments.

336

Mark H. Johnson et al.

tracking of face-like patterns, but did ﬁnd discrimination between face and non-face-like patterns in a habituation–dishabituation procedure. In considering the reasons for this apparent failure to replicate it is useful to refer to the reasons why Johnson et al. (1991) changed their procedure for testing infants after the newborn period. Speciﬁcally, they found that the procedure of moving the test stimulus around the newborn’s head, and measuring the extent of eye and head turning to keep the stimulus in view, was not effective after the immediate newborn period due to maximal (ceiling) tracking to all patterned stimuli. They therefore used a different testing procedure in which the infant was oriented relative to the stimulus. In the experiments conducted by Easterbrook and colleagues, all of the patterned stimuli used elicited greater degrees of orienting than in prior studies. Although the precise reasons for this greater extent of tracking in Easterbrook’s study are not known, differential responding is unlikely when near-ceiling performance is achieved. This methodological point aside, several types of explanations have been advanced for newborn face preferences that have been observed in at least ﬁve independent laboratories and the vast majority of studies conducted.

A. THE SENSORY HYPOTHESIS A number of authors have advanced the view that newborn visual preferences, including those for face-related stimuli, can be accounted for simply in terms of the relative visibility of the stimuli. For example, Kleiner and Banks (1987; Kleiner, 1993) originally argued that the ‘‘linear systems model’’ of infant visual preferences could be used to account for newborn face preferences. Speciﬁcally, by this account newborn face preferences could be accounted for in terms of a simple psychophysical measure, the energy in the amplitude spectrum of the stimulus (following Fourier decomposition). Morton, Johnson, and Maurer (1990) reviewed critical experiments on this topic, and concluded that even when amplitude is held constant, phase information (roughly speaking, conﬁguration) still inﬂuences preference toward faces. In addition to these arguments, Johnson and Morton (1991) conducted spatial frequency analyses on the range of schematic face stimuli discussed in their book. This analysis revealed that although spatial frequency is an important predictor of newborn visual preferences, stimuli with the phase appropriate for a face were always more preferred than predicted by spatial frequency alone. In a later revival of the sensory hypothesis, Acerra, Burnod, and de Schonen (2002) constructed a neural network model of visual cortical processing in infants. This model, based on four properties of the infant

The Social Cognitive Neuroscience of Infancy

337

visual system (the spatial frequencies related to an immature retina; the tuning properties of neurons in area V1; the activity-dependent learning properties of both feedforward and lateral connections in cortical areas; and the absence of interhemispheric connections) successfully simulated the results of some of the experiments on newborns preferences for schematic face stimuli. In one critical comparison, that between the upright and inverted ‘‘Conﬁg’’ stimulus (see Figure 1), differences in the spacing of high-contrast elements relative to the outline frame yields slight differences in spatial frequency that can be detected and ampliﬁed within the model. To produce the correct predictions for the behavioral data simulated, however, depends precisely on the exact details of the stimuli used in one experiment (Valenza et al., 1996). Speciﬁcally, Valenza et al. (1996) inverted the internal facial features in their stimuli in a way that resulted in the inverted stimulus having less regular spacing between the white and black areas than the upright case. The model exploited and ampliﬁed this small difference in spatial frequency proﬁle, and was not tested with the equivalent patterns from other newborn experiments (such as Johnson et al., 1991). Furthermore, Bednar and Miikkulainen (2003) doubt that the Acerra et al. model will be able to simulate results from experiments involving real faces, and especially the complex visual scenes that newborns are exposed to in real life. This is because very small irrelevant differences between stimuli (such as where the hairline falls on a face) would have ampliﬁed effects on visual preferences.

B. NON-FACE STRUCTURAL PREFERENCES Debates about the mechanisms underlying face preferences in infants have often revolved around a contrast between structural and sensory preferences. For much of this debate, structural preferences referred to information about the conﬁguration of features that compose a face. An alternative proposal is that infants’ preference behavior to face and nonface stimuli can be explained by ‘‘domain-general’’ structural preferences, such as those based on known adult ‘‘Gestalt’’ principles (see Simion et al., 2003). Speciﬁcally, Simion and colleagues argue that ‘‘newborns’ preference for faces may simply result from a number of non-speciﬁc attentional biases that cause the human face to be a frequent focus of newborns’ visual attention’’ (Simion et al., 2003, p. 15). These authors still believe that the purpose of these biases is to direct attention to faces in the natural environment of the newborn, so that they are debating the nature of the representation(s) that underlie this bias, not the existence of the mechanism itself.

338

Mark H. Johnson et al.

The behavioral data discussed by Simion and colleagues is consistent with a preference for up–down asymmetrical patterns with more elements or features being in the upper half of a bounded object or area. Such preferences are sometimes said to be due to an upper visual ﬁeld bias (Simion et al., 2002). However, all the experiments to date indicate a crucial interdependency between the borders of the object/area and the elements within it. Indeed, some existing experiments from this group suggest that the shape of boundary needs to correspond with the number of elements enclosed within it (Macchi Cassia et al., 2007). Experiments that independently manipulate upper visual ﬁeld elements and bounded areas, and experiments that measure eye movements sufﬁciently to control upper/ lower visual ﬁeld presentation have not yet been done. Therefore the attentional bias being discussed by these authors—a bounded object or area with a greater number of elements or features in the upper half—is likely the product of one system or representation rather than being independent non-speciﬁc biases. Skepticism about the ‘‘top-heavy’’ account of newborn face preferences seems warranted for a number of additional reasons. First, an assumption behind the ‘‘top-heavy’’ view is that a series of non-speciﬁc biases is a computationally simpler and more parsimonious account of newborn preference than Conspec. However, as discussed previously, the extent literature suggests that the hypothesized underlying mechanism has, at a minimum, to integrate a boundary with the inner elements/features. More likely, object segregation followed by numerical assessment of the number of elements present will be required. In contrast, Conspec has been simulated by computational neural network model in which a basic Conﬁg representation arises within the artiﬁcial cortex as a result of spontaneous activity (Bednar & Miikkulainen, 2003). Another reason for doubting that an upper visual ﬁeld bias alone is sufﬁcient to account for newborn preferences comes from studies on the effects of phase contrast (Farroni et al., 2005). In these experiments newborn preferences for upright compared to inverted Conﬁg patterns and photographic face images were assessed with both black elements on white (as in previous studies) and the converse (Figure 2). If the newborns are seeking elements or features then phase contrast should make either no difference, or possibly cause them to prefer white elements on a black background (because lighter elements are typically closer to the viewer in natural scenes). In contrast, if the purpose of the representation is to detect faces then black elements on white should be more effective, because the eyes and mouth region are recessed into the face, and appear in shadow under natural (top–down) lighting conditions. Consistent with the latter view, we (Farroni et al., 2005) found the preference for an upright face only

The Social Cognitive Neuroscience of Infancy

339

Fig. 2. Newborns’ looking time measured in response to schematic face stimuli which differed in their contrast polarity (Figure adapted from Farroni et al., 2005).

under conditions of black on white or when smaller dark blobs are placed inside the light elements. Infants preferred face-like conﬁgurations only when some contrast relations within the image resembled the natural difference between the darker iris and the lighter sclera of human eyes. This was also shown to be the case with photographic images. Newborns seem to seek face-like stimuli that provide them with gaze information as well. Another area that casts doubt on the ‘‘top-heavy’’ hypothesis is evidence taken to support the existence of complex face processing abilities in newborns. This evidence is discussed next.

C. NEWBORNS HAVE COMPLEX FACE REPRESENTATIONS Some empirical results have lead to the hypothesis that newborns already have complex processing of faces (for review see Quinn & Slater, 2003). These ﬁndings, usually obtained with naturalistic face images, include a

340

Mark H. Johnson et al.

preference for attractive faces (Slater et al., 1998, 2000), data indicating that newborns are sensitive to the presence of eyes in a face (Batki et al., 2000), and that they prefer to look at faces with direct gaze that engage them in eye contact (Farroni et al., 2002). Such ﬁndings have led some authors to conclude that ‘‘y face recognition abilities are well developed at birth’’ (Quinn & Slater, 2003, p. 4). The usual interpretation of attractiveness preferences in older infants is that these stimuli are seen as more face-like because they best match an average or prototypical face (e.g. Langlois & Roggman, 1990). We note in passing that the newborn effect is not just due to increased symmetry because no preference is found with inverted stimuli. However, if we assume that Conspec is activated most strongly by the most typical spatial arrangement of features (rather than equidistant blobs), then a preference for more average of typical arrangements of facial features would be expected. Furthermore, inspection of realistic face images through the appropriate spatial frequency ﬁlters for newborns reveals that a mechanism sensitive to the arrangement of high-contrast, low-spatial frequency components of a face would be preferentially activated by (a) the typical conﬁgural arrangement of eye and mouth regions, (b) the presence (or absence) of open eyes, and (c) direct versus averted gaze (see Farroni et al., 2002). Clearly, the extent to which Conspec can explain newborn responses to a variety of naturalistic faces requires further investigation. However, it is very difﬁcult to see how the ‘‘top-heavy’’ hypothesis can explain this body of data, suggesting that this theory can only account for a narrow range of preference phenomena in newborns. Regardless of what the exact basis of the newborn preference is, most investigators agree that it is sufﬁcient to cause newborns to orient to faces in their early visual environment. Besides the detection of faces, another important aspect is to recognize familiar faces. Newborns ﬁrst recognize their mother’s face on the basis of information from the outer contour of the head, hairline, and the internal conﬁguration of eyes, nose, and mouth (Bushnell, Sai, & Mullin, 1989). But not until 6 weeks of life infants are able to do this recognition solely on the face’s internal conﬁguration (de Schonen & Mathivet, 1990). The face preference observed in newborns is thought to be guided largely by sub-cortical brain structures, but the maturation of visual cortical areas is necessary for the emergence of the more sophisticated competencies underlying identity recognition from faces (for a discussion, see Johnson, 2005a).

D. DEVELOPING CORTICAL AREAS FOR FACE PROCESSING Investigations of brain areas involved in face processing in infants have been limited by ethical and technical issues (de Haan & Thomas, 2002). One

The Social Cognitive Neuroscience of Infancy

341

exception has been a study by Tzourio-Mazoyer and colleagues, who took advantage of the opportunity to perform positron emission tomography (PET) on infants in an intensive care unit as part of a clinical follow-up (Tzourio-Mazoyer et al., 2002). In this small-scale study, a group of 6 2-month-olds were imaged while they watched a face or an array of colored diodes used as a control stimulus. A subtraction analysis revealed that faces activated a network of areas in 2-month-old infants’ brains similar to that described as the core system for face processing in adults (Haxby, Hoffman, & Gobbini, 2000). More speciﬁcally, activation was reported in an area in infants’ right inferior temporal gyrus, which is thought to be the homologue of the adult FFA (Gauthier et al., 1999; Kanwisher, 2000). It is interesting to note that a cortical region at the age of 2 months is neuroanatomically immature (Huttenlocher, 2002; Huttenlocher & Dabholkar, 1997) and has only a low level of metabolic activity (Chugani & Phelps, 1986; Chugani, Phelps, & Mazziotta, 1987) can exhibit functional activity. Furthermore, as we will see later, activation of an area in response to faces does not mean that the area is speciﬁcally tuned for this function. Face perception also activated bilateral inferior occipital and right inferior parietal areas in infants. The former has been implicated in early perceptual analysis of facial features, whereas the latter is thought to support spatially directed attention in adults (Haxby et al., 2000). Interestingly, and contrary to what is known from adults, face processing in 2-month-olds also recruited areas in the inferior frontal and superior temporal gyrus, which have been identiﬁed as a part of the adult language network. One possible interpretation is that the coactivation of face and future language networks has a facilitative effect on social interactions guiding language learning by looking at the speaker’s face (TzourioMazoyer et al., 2002). While bearing in mind the small sample size and non-optimal control stimulus, the study by Tzourio-Mazoyer and colleagues provided important information on a structural level by identifying the neural substrates of the infant face processing network. However, as mentioned earlier, due to ethical and technical concerns with the use of neuroimaging methods like PET and functional magnetic resonance imaging (fMRI) in infants, evidence coming from these kinds of studies is rare. Therefore, most studies that have examined infant brain function rely on the more readily applicable recording of EEG measures, which provide excellent temporal resolution but relatively poor spatial resolution. We now discuss brieﬂy the empirical evidence available on infants’ face processing using event-related brain potentials (ERPs; for a more detailed review, see de Haan, Johnson, & Halit, 2003). In adults, human faces elicit an N170 response, which is most prominent over posterior temporal sites and is larger in amplitude and longer in latency to inverted than to upright

342

Mark H. Johnson et al.

faces (Bentin et al., 1996; de Haan, Pascalis, & Johnson, 2002). This component is not modulated by the inversion of monkey faces (de Haan et al., 2002), nor when upright objects are compared to inverted objects (Bentin et al., 1996). This selective inversion effect for human faces has been taken as evidence for a face-related processing mechanism generating the N170. On the basis of waveform morphology and some of its response properties, it has been suggested that the infant N290 is a precursor to the adult N170. In these studies, infants’ and adults’ ERPs were measured in response to upright and inverted human and monkey faces (de Haan et al., 2002; Halit, de Haan, & Johnson, 2003). The infant N290 is a negativegoing deﬂection observed over posterior electrodes whose peak latency decreases from 350 ms at 3 months to 290 ms at 12 months of age (Halit et al., 2003). This is consistent with the latency of many prominent ERP components reducing with increasing age during childhood. The results of these studies indicate that at 12 months of age the amplitude of the infant N290, like the adult N170, increases to inverted human but not inverted monkey faces when compared to the upright faces. However, the amplitude of the N290 was not affected by stimulus inversion at an earlier age (3 and 6 months). By applying new source separation and localization methods (Richards, 2004) to the infant ERP data, Johnson et al. (2005) identiﬁed candidate cortical sources of the face inversion effect. This analysis suggested that generators in the left and right lateral occipital area, the right FFA, and the right superior temporal sulcus (STS) discriminated between upright and inverted faces in 3- and 12-month-olds. All three cortical areas have been implicated in the core face processing system in adults (Haxby et al., 2000), and they also generally overlap with the areas identiﬁed in the previously discussed PET study with 2-month-olds (Tzourio-Mazoyer et al., 2002). The development of the brain processes reﬂected in the N170/N290 continues well beyond infancy (for a review, see Taylor, Batty, & Itier, 2004). Although latency of the adult N170 is delayed by face inversion, no such effect is observed for the latency of the infant N290 at any age (de Haan et al., 2002; Halit et al., 2003) and, in fact, this latency effect may not appear until 8–11 years (Taylor et al., 2004). Another developmental change is that although the amplitude of the adult N170 is larger to the monkey faces, infants’ N290 shows the opposite pattern. A completely adult-like modulation of the amplitude of the N170 has not been reported until 13–14 years (Taylor et al., 2004). The reason for these changes in the N170 response proﬁle during mid-to-late childhood remains unclear. To date, face processing, besides speech perception (Dehaene-Lambertz, Dehaene, & Hertz-Pannier, 2002; Kuhl, 2004), represents the best-studied

The Social Cognitive Neuroscience of Infancy

343

aspect of the infant social brain. The evidence available from PET and EEG/ERP studies suggests that most of the brain areas and mechanisms implicated in adult face processing can be activated relatively early in postnatal life. However, there are some additional effects, such as the activation of the inferior frontal and superior temporal gyrus in 2-montholds (Tzourio-Mazoyer et al., 2002), and the STS generator of the inversion effect found in 3- and 12-month-olds (Johnson et al., 2005), that do not directly map onto the mature face processing system. In addition to the extra regions involved while infants perceive faces, another important observation in the infant ERP work is that the infant face processing system possesses much broader response properties which are not yet ﬁnely tuned to upright human faces. This suggests that despite the gradual cortical specialization seen throughout the ﬁrst year of life, the system continues to specialize well beyond infancy. These changes in the degree of specialization of processing, and the spatial extent of cortical activation, are in accordance with the interactive specialization perspective alluded to earlier (Johnson, 2001).

III. Eye Gaze Processing An important social signal encoded in faces is eye gaze. The detection and monitoring of eye gaze direction is essential for effective social learning and communication among humans (Bloom, 2000; Csibra & Gergely, 2006). Eye gaze provides information about the target of another person’s attention and expression and also conveys information about communicative intentions and future behavior (Baron-Cohen, 1995). Eye contact is considered to be the most powerful mode of establishing a communicative link between humans (Kampe, Frith, & Frith, 2003). From birth, human infants are sensitive to another person’s gaze as reﬂected in their preference to look at faces that have their eyes open rather than closed (Batki et al., 2000).

A. NEWBORNS DETECT DIRECT GAZE We (Farroni et al., 2002) tested newborn infants by presenting them with a pair of stimuli, one a face with eye gaze directed straight at the newborns, and the other with averted gaze (Figure 3). Fixation times were signiﬁcantly longer for the face with the direct gaze. Furthermore, the number of orientations was greater with the straight gaze than with the averted gaze. These results demonstrate preferential orienting to direct eye gaze from birth.

Mark H. Johnson et al.

344

b Number of Orientations

Looking Time (sec.)

a

100

50

20

10

c Looking Time −1.0

− 0.5

0.5

1.0

Orientations Fig. 3. Results of the Farroni et al. (2002) preferential looking study with newborns. (a) Mean looking times spent at two stimulus types. (b) Mean number of orientations toward each type of stimulus. (c) Filled triangles represent reference score for the direct gaze over the averted for each individual newborn and open triangles represent average preference scores.

The preference is probably a result of a fast and approximate analysis of the visual input, dedicated to ﬁnding socially relevant stimuli for further processing. To examine the speciﬁcity of this newborn preference, we conducted two further experiments. Our goal was to see whether inverting faces has any effect on gaze perception in newborns. If the preference for direct gaze is not found under conditions of inversion, then we may conclude that lowlevel aspects of the faces, such as symmetry or local spatial frequency, are not the basis for the newborn preference for direct gaze in upright faces previously observed. Furthermore, an absence of the effect with inversion would indicate that the factors responsible for the preference require the eyes to be situated within the conﬁguration of an upright face. Newborns did not show signiﬁcant looking time or orientation differences between direct and averted gaze conditions when the faces were presented upside down. These results rule out symmetry and local spatial frequency as possible explanations of the newborn effect. Two underlying mechanisms could account for the observed gaze preferences in newborns. By one account, even newborns have sophisticated face-processing abilities sufﬁcient to extract gaze direction when presented in the context of a face. By an alternative account, the preferences of

The Social Cognitive Neuroscience of Infancy

345

newborns are based on a primitive ‘‘Conspec’’ detector that responds to an optimal conﬁguration of high-contrast elements. Straight-on faces with direct gaze better ﬁt this simple template than do faces with averted gaze (see Farroni et al., 2002, 2003). To test these hypotheses, we conducted a second experiment that involved similar face stimuli, but with averted head angles. We reasoned that a sophisticated face processing system would be able to extract gaze direction even when head angle varied. In contrast, a simple Conspec mechanism may only produce a preference for direct gaze under conditions where the spacing between eyes and mouth is optimal. Changing head angle alters the relative spacing of the two eyes and mouth, and thus may disrupt the preference seen with a straight head. In this experiment newborns looked equally at the direct gaze and at the averted gaze, and they oriented equally to both. The results of these experiments show that the strong preference for faces with direct gaze depends on the eyes being situated within the context of an upright straight-ahead face. This ﬁnding simultaneously rules out many low-level explanations of the original result along with the suggestion that newborns may have sophisticated eye gaze perception abilities. Rather, they are consistent with the view that newborns orient to direct gaze due to a primitive conﬁguration detection system (such as ‘‘Conspec’’).

B. GAZE FOLLOWING Eye gaze has also been shown to effectively cue human infants’ attention to spatial locations. Hood, Willen, and Driver (1998) showed that the perception of an adult’s deviated gaze induces shifts of attention in the corresponding direction in 10- to 28-week-olds. In their experiments they modiﬁed the standard Posner cueing paradigm (Posner, 1980) using as a central cue the direction of gaze of a woman’s face, thus creating a computer generated eye gaze shift (Figure 4 and further for details). Infants’ eye movements toward a peripheral target were faster when the location of the target was congruent with the direction of gaze of a centrally presented face. In subsequent studies, we examined the visual properties of the eyes that enable infants to follow the direction of the gaze (Farroni et al., 2000). We tested 4-month-olds using a cueing paradigm adapted from Hood et al. (1998). Each trial began with the stimulus face eyes blinking (to attract attention), then the pupils shifted to either the right or the left (see Figure 4). A target stimulus was then presented either in the same position where the eyes were looking (congruent position) or in a location incongruent with the direction of gaze. We found that infants looked faster at the location congruent with the direction of gaze of the face.

346

Mark H. Johnson et al.

Fig. 4. Example of the edited video image illustrating the stimulus for Experiment 1 in Farroni et al. (2000). In this trial the stimulus target (the duck) appears on the side which is incongruent with the direction of gaze.

In a second experiment, we manipulated the stimulus face so that the whole face was shifted to one side (right or left) while the pupils remained ﬁxed (see Figure 4). In this case the infants were faster to look in the direction in which the whole face was shifted, and not the direction where the pupils were directed. Therefore, the infants actually followed the biggest object with lateral motion (i.e., the face) and not the eyes. In a third experiment, we used the same paradigm as in the ﬁrst experiment, but this time when the eyes were opened the pupils were already oriented to the left or right, and the infants were not able to perceive the movement of the pupils. In this case the cueing effect disappeared. Up to this point, the results suggested that the critical feature for eye gaze cue in infants is the movement of the pupils, and not the ﬁnal direction of the pupils. To try to understand this cueing effect better, we did three further variants of the same procedure (Farroni et al., 2002). In the ﬁrst experiment in this series we examined the effect of inverting the face on cueing. If infants are merely cued by motion, then an inverted face should produce the same cueing as an upright one. To our surprise, cueing had no effect,

The Social Cognitive Neuroscience of Infancy

347

suggesting that the context of an upright face may be important. In the next experiment infants saw a face with an averted gaze that then shifted to the center. If infants are responding just to the motion of elements they should be cued in the direction opposite to the initial gaze. Again, no cueing effect was observed. These results did not support the hypothesis that directed motion of elements is the only determining factor for the cueing effects. In the last experiment, a more complex gaze shift sequence allowed us to analyze the importance of beginning with a period of mutual gaze: the eyes shifted from center to averted, and then back to center. Here we observed a signiﬁcant cueing effect. Taken together, these results suggest that cueing effects are observed only following a period of mutual gaze with an upright face. In other words, mutual gaze with an upright face may engage mechanisms of attention such that the viewer is more likely to be cued by subsequent motion. In summary, the critical features for eye gaze cueing in infants are (1) lateral motion and (2) at least a brief preceding period of eye contact with an upright face. Finally, eye gaze has also been shown to effectively cue newborn infants’ attention to spatial locations, suggesting a rudimentary form of gaze following (Farroni et al., 2004a).

C. NEURAL PROCESSING OF DIRECT GAZE IN INFANTS The question that arose next concerned how the behaviorally expressed preference for mutual gaze and the capacity to follow gaze are implemented in the infant brain. To test infants’ sensitivity to gaze direction, we recorded ERPs as 4-month-old infants viewed photographs of faces, some with direct gaze and some with the eyes averted randomly to the right or left (Farroni et al., 2002). ERPs have been shown to be sensitive to small differences in transient brain activation due to processing of visual stimuli in infants (Csibra et al., 2000) and they provide the most feasible neuroimaging method to study the brain development of healthy babies. We hypothesized that an early preference for eye contact would facilitate the processing of faces with direct gaze. In our analyses, we focused on the N290 component described earlier (the precursor to the adult N170). As in previous studies with infants, the N290 component peaked around 240–290 ms poststimulus. Its amplitude was greater in response to direct gaze than to averted gaze (Figure 5) with the N290 amplitude being more negative over the medial than over the lateral leads. No differences between left and right hemisphere were observed. By 4 months, the infant brain manifests enhanced processing of faces with direct gaze as indexed by an increased amplitude of the infant N170 when compared to averted gaze (Farroni et al., 2002). This ﬁnding is

348

Mark H. Johnson et al.

Fig. 5. Four-month-old infants’ ERPs recorded to faces with direct and averted gaze in Farroni et al. (2002). An enhanced N170 component (peaking around 240 ms after stimulus onset) was observed over occipital cortex in response to faces with direct gaze when compared to faces with averted gaze.

obtained even when the head is averted but direct mutual gaze is maintained (Farroni, Johnson, & Csibra, 2004b). However, enhanced neural processing to faces with direct gaze is only found when eyes are presented in the context of an upright face. Our results indicated that at this age cortical processing of faces is enhanced with direct gaze under similar conditions to those observed in adults. Overall, these results further the view that relatively simple perceptual biases in the newborn are replaced by adult-like perceptual processing of gaze within a few months of birth. Furthermore, source separation and localization methods were used to identify the cortical sources of the scalp-recorded ERP effects sensitive to eye gaze (Johnson et al., 2005). Contrary to adults, who show speciﬁc activations associated with eye gaze perception in the STS (Allison, Puce, & McCarthy, 2000), cortical generators localized in the fusiform gyrus discriminated gaze direction best in infants. Although the amplitude of the N170 in infants is modulated by eye gaze (Farroni et al., 2002, 2004b) and face orientation (de Haan et al., 2002; Halit et al., 2003), the amplitude of the adult N170 is only affected by face inversion but not by direction of gaze (Grice et al., 2005; Taylor et al., 2001). Taken together, the difference in the response properties of the infant N170 (N290) versus the adult N170 suggest that face and eye gaze share common patterns of cortical activation early in ontogeny which later partially dissociate and become more specialized, a developmental change that is consistent with the interactive specialization view introduced earlier (Johnson, 2001).

The Social Cognitive Neuroscience of Infancy

349

Another technique that can reveal brain activation missed by averaging methods is the analysis of high-frequency oscillations in the gamma band (20–100 Hz). It is thought that brain oscillations in the high-frequency range reﬂect neural mechanisms by which activity of small neuronal networks is synchronized, whereas very large networks are recruited during slow oscillations (Buzsaki & Draguhn, 2004). Synchronous activity of oscillating networks is a prominent feature of neural activity throughout the animal kingdom (Sejnowski & Paulsen, 2006), and it is viewed as the critical middle ground linking single-neuron activity to behavior (Engel, Fries, & Singer, 2001; Hermann, Munk, & Engel, 2004). Such oscillations are either time-locked to eliciting stimuli (evoked gamma activity) or can be detected as induced gamma activity consisting of oscillatory bursts whose latency jitters from trial to trial and its temporal relationship with the stimulus onset is fairly loose. Hence, induced gamma activity is not revealed by classical averaging techniques and speciﬁc methods based on timevarying spectral analysis of single trials are required to detect it (TallonBaudry & Bertrand, 1999). As a theoretical framework, which attempts to assign functional signiﬁcance to early-evoked and late-induced gamma-band responses, Hermann and colleagues (2004) have put forward a match-utilizationmodel. According to this model, the early gamma-band response reﬂects the matching of stimulus-related information with memory contents in primary sensory cortex. After a stimulus has been identiﬁed through this matching stage, this information can be used in all kinds of more complex cognitive operations involving other brain areas and the late-induced gamma response might be a signature of such a utilization process. Gamma oscillations are also of special interest because they have been found to correlate with the BOLD response used in fMRI as shown in invasive work with animals (Niessing et al., 2005) and non-invasive studies combining EEG and fMRI in humans (Fiebach, Gruber, & Supp, 2005; Foucher, Otzenberger, & Gounot, 2003). In this work it also has been shown that whereas fMRI BOLD correlated with gamma-band activity such a correlation was not found for ERP measures (e.g. Foucher et al., 2003). These ﬁndings are consistent with a biophysical model which suggests that increases in hemodynamic signals as measured by fMRI are associated with a shift in the spectral mass from low to high frequencies as measured with EEG (Kilner et al., 2005). Grossmann et al. (2007) examined gamma oscillations and its relation to eye gaze perception in 4-month-old infants. This study analyzed previously published EEG data sets taken from Farroni et al.’s (2002) and (2004b) ERP studies in which infants were presented with upright images of female faces directing their gaze toward them or to the side. We predicted a burst

350

Mark H. Johnson et al.

of gamma oscillation over prefrontal sites to direct gaze if gamma oscillations are indeed related to detecting eye contact/communicative intent as suggested by adult fMRI work (Kampe et al., 2003; Schilbach et al., 2006). Because the right intraparietal sulcus (IPS) and right STS are sensitive to averted gaze in the adult brain (Hoffman & Haxby, 2000), we hypothesized that some activity over right temporo-parietal regions would be associated with the perception of averted gaze. In addition, another group of 4-month-old infants were presented with the same face stimuli upside down, which is thought to disrupt conﬁgural face processing (Rodriguez et al., 1999; Turati et al., 2004) and infants’ preference for mutual gaze (Farroni, Menon, & Johnson, 2006). Thus, we predicted that inverted faces would not induce activity in the gamma band that differs as a function of eye gaze. The data revealed that gamma oscillations varied as a function of gaze direction in the context of an upright face, which extends previous ERP and source localization results (Farroni et al., 2002, 2004b; Johnson et al., 2005). In support of our hypotheses, speciﬁc effects with distinct spatial and temporal characteristics were observed depending upon whether gaze was directed at or directed away from the infant. Direct gaze compared to averted gaze evoked early (100 ms) increased gamma activity (20–40 Hz) at occipital channels. Short-latency phase-locked oscillatory evoked gamma responses have been described in the visual modality in response to brief static stimuli in infant and adult EEG (Csibra et al., 2000; Tallon-Baudry & Bertrand, 1999). In adults, evoked gamma activity is signiﬁcantly larger for items that match memory representations (Hermann et al., 2003, 2004). For infants a face with direct gaze may represent a more familiar and prototypical face (Farroni et al., 2007), which is closer to what is represented in memory than a face with averted gaze, and therefore elicits an enhanced evoked oscillatory response. This interpretation is supported by, and might be linked to, ﬁndings showing an enhanced neural encoding (Farroni et al., 2002) and better recognition of upright faces with direct gaze in infants (Farroni et al., 2007). The modulation of the evoked gamma response in this study was observed much earlier (100 ms) than the effect in the previous ERP study (290 ms, Farroni et al., 2002, 2004). This indicates that gamma activity is a more sensitive measure of some aspects of very early brain processes related to the discrimination of gaze direction. As predicted, direct gaze also elicited a late (300 ms) induced gamma burst over right prefrontal channels. Directing eye gaze at someone (i.e., making eye contact) serves as an important ostensive signal in face-to-face interactions that helps establishing a communicative link between two people. Successful communication between two people may well depend crucially on the ability to detect the intention to communicate, conveyed by

The Social Cognitive Neuroscience of Infancy

351

signals directed at the self such as making eye contact (Kampe et al., 2003). On a neural level, the medial prefrontal cortex (MPFC) is consistently activated when gaze is directed at, but not when gaze is averted away from, the self (Kampe et al., 2003; Schilbach et al., 2006). Because gamma oscillations measured with EEG are correlated with the BOLD response used in fMRI (Fiebach et al., 2005; Foucher et al., 2003), eye contact detection in 4-month-old infants may well recruit very similar brain mechanisms as in adults. The gamma burst distributed over right frontal cortex in infants might reﬂect less-localized functional activity than in adults, which is in accordance with current views of functional brain development (Johnson, 2001), suggesting a more diffuse to a more focal pattern cortical activity with age. Averted gaze also serves an important function during social communication by directing the perceiver’s attention to certain locations or objects, and behavioral evidence indicates that 4-month-olds are sensitive to this aspect of eye gaze (Farroni et al., 2003; Hood et al., 1998). The right IPS and right STS have been identiﬁed as sensitive to averted gaze in the adult human brain (Haxby et al., 2000; Hofman & Haxby, 1999). Activity in the IPS may be speciﬁcally recruited when perceived eye gaze direction elicits a shift in spatial attention, whereas STS is more generally associated with eye and mouth movements (Haxby et al., 2000). Our ﬁnding of a late (300 ms) induced gamma burst in response to averted gaze over right occipito-temporal-parietal regions might reﬂect similar but perhaps more diffuse brain activations in infants. In another study (Grossmann & Farroni, in press), we further examined how head orientation would inﬂuence the infants’ brain responses to eye gaze direction cues observed in the gamma band. This study was based on a time–frequency analysis performed on previously published EEG data (Farroni et al., 2004b). Infants were presented with upright images of female faces orienting their head away from the infant but either directing their gaze towards them or away from them. Corresponding with the ﬁndings reported for straight head angle faces, direct gaze compared to averted gaze elicited early (100 ms) increased gamma activity (20–40 Hz) at occipital channels. However, contrary to the previous ﬁndings (Grossmann, Striano, & Friederici, 2007), no induced gamma burst was observed over prefrontal channels to direct gaze when the head was averted. This suggests that although infants at the age of 4 months can discriminate between direct and averted gaze in the context of averted heads as indicated by the increased evoked occipital gamma activity to direct gaze, in this context, they do not yet recruit brain processes associated with detecting eye contact as a communicative signal. Hence, it follows that by 4 months a frontal face is required to elicit activity in

352

Mark H. Johnson et al.

prefrontal brain structures involved in social communication. This ﬁnding stands in contrast to fMRI work showing that adults show speciﬁc activity in the prefrontal cortex in response to direct gaze even when the head is averted from the perceiver (Kampe et al., 2003) and therefore suggests that development must occur after 4 months which enables the human brain to detect mutual gaze from static face representations regardless of head orientation. In a follow-up study (Grossmann et al., submitted), we were able to show that when infants at this age are presented with dynamic eye gaze cues directed at them in the context of an averted head a burst of prefrontal gamma activity was observed. This suggests that biological motion cues (i.e., gaze shifts) help infants to detect communicative signals and recruit frontal brain regions, which play a fundamental role for the development of social perception and cognition (Anderson et al., 1999). Furthermore, as in the previous study using faces with a frontal orientation (Grossmann et al., 2007), we observed a late (300 ms) induced gamma burst over right temporo-parietal regions (Grossman et al., submitted), but whereas when frontal faces were used this burst was evoked by averted gaze, for oriented faces this activity occurred in response to direct gaze. These seemingly contradictory ﬁndings need explanation. Our suggestion is that the gamma activity observed over right temporoparietal channels is associated with neural computations integrating eye gaze direction information in relation to head angle. When the results are revisited from this perspective, it appears that this gamma burst is observed when eye direction is different/incongruent from the head direction (i.e., averted gaze in a frontal face and direct gaze in an oriented face). This view is in accordance with adult fMRI data showing increased activity in the right STS to averted gaze in frontal face (Hoffman & Haxby, 1999) and to direct gaze in averted face (Pelphrey, Viola, & McCarthy, 2004). The ﬁnding that inverted faces did not elicit gamma-band responses that differed between direct and averted gaze is in line with, and adds further developmental evidence to, the notion that face inversion disrupts face processing (Rodriguez et al., 1999; Turati et al., 2004). This indicates that relatively early in development cortical structures involved in face processing are already somewhat specialized to extract information about gaze direction from upright faces. It further shows that the gamma-band effects observed in response to direct and averted gaze are not simply driven by ‘‘lower level’’ perceptual parameters (e.g., symmetry (direct gaze) and asymmetry (averted gaze)) because then they should have occurred in the inverted condition as well. These gamma-band ﬁndings in infants show a high degree of correspondence in terms of timing and frequency content with previous ﬁndings in adults (Tallon-Baudry & Bertrand, 1999). This suggests continuity throughout development, and further underlines

The Social Cognitive Neuroscience of Infancy

353

the functional importance of gamma-band oscillations also for social perception.

D. UNDERSTANDING EYE–OBJECT RELATIONS Another communicative function of eye gaze is to direct attention to certain locations, events, and objects. Understanding the relations between eye gaze and target objects is particularly important for aspects of development such as word learning. Comprehending that another’s gaze direction refers to a speciﬁc object allows the child to associate the object with a name or emotional expression (Baldwin & Moses, 1996). Adults’ gaze has been found to facilitate object processing at a neural level in infants as young as 4 months (Reid et al., 2004). In this ERP study, objects that were previously cued by eye gaze elicited a diminished positive slow wave observed between 700 and 1000 ms over right fronto-temporal channels. A diminished positive slow wave is thought to indicate deeper memory encoding (Nelson & Collins, 1991). This suggests that eye gaze as a social cue facilitates brain processes involved in memory encoding that might assist infants’ learning. In another ERP study, 9-month-old infants and adults watched a face whose gaze shifted either towards (object-congruent) or away from (objectincongruent) the location of a previously presented object (Senju, Johnson, & Csibra, 2006). This paradigm was based on that used in an earlier fMRI study (Pelphrey et al., 2003) and designed to reveal the neural basis of ‘‘referential’’ gaze perception. When the ERPs elicited by object-incongruent gaze shifts were compared to the object-congruent gaze shifts, an enhanced negativity around 300 ms over occipito-temporal electrodes was observed in both infants and adults. This suggests that infants encode referential information of gaze using similar neural mechanisms to those engaged in adults. However, only infants showed a fronto-central negative component (Nc) that was larger in amplitude for object-congruent gaze shifts. Thus, in the less specialized infant brain, the referential information of gaze may be encoded in broader cortical circuits than in the more specialized adult brain. In summary, infants show speciﬁc modulations of cortical responses associated with eye gaze perception, which differ from what is seen in adults. Nevertheless the ﬁndings reviewed also suggest that they successfully discriminate between direct and averted gaze, and can make use of gaze cues when encoding objects in memory. This indicates that, although the cortical structures involved in the perception of gaze direction may not be fully differentiated in infancy, they are at least partially functional. This may

354

Mark H. Johnson et al.

be critical for social communication and learning from others (see Section VI Conclusions).

IV. Perception of Emotions Social behavior is coupled with emotion. During social interactions, emotions are overtly expressed by the infant, whereas displays of emotion are normally more regulated in adults. Nevertheless, most brain structures that are important in processing emotions in adults are also essential for social behavior (for a review, see Adolphs, 2003). Detecting emotion in others through their facial expressions is an important way to communicate emotions in social interactions (Izard, 1991). Recognizing facial expressions permits us to detect another person’s emotional state and provides cues on how to respond in these social interactions.

A. PROCESSING EMOTION IN THE FACE In an ERP study in which 7-month-olds watched happy versus fearful faces (Nelson & de Haan, 1996), fearful faces elicited an enhanced Nc peaking around 500 ms. The Nc has its maximum at frontal and central sites and has been thought of as an obligatory attentional response sensitive to stimulus familiarity (Courchesne, Ganz, & Norcia, 1981; Snyder, Webb, & Nelson, 2002). Dipole modeling has revealed that the cortical sources of the Nc can be localized in the anterior cingulate and other prefrontal regions (Reynolds & Richards, 2005). This suggests that 7-month-old infants in Nelson and de Haan’s study allocated more attentional processing resources to the unfamiliar fearful than to the familiar happy expression as indicated by an enhanced Nc. Another important question is whether infants can discriminate between different negative emotional facial expressions. De Haan and Nelson (1997) found no difference in 7-month-olds’ ERP responses to fearful and angry faces and suggested that infants did not display different brain responses to the two expressions because they perceived the signal value of both expressions as ‘‘negative’’ or because they perceived both expressions as equally unfamiliar. In a subsequent study (Kobiella et al., 2008) using another set of facial stimuli and measuring from more electrode positions than de Haan and Nelson (1997), infants’ N290, P400, and Nc differed between emotions (for a discussion of how the neural processing of angry in contrast to fearful facial expressions develops during infancy, see Grossmann et al., 2007). This suggests that infants discriminate between

The Social Cognitive Neuroscience of Infancy

355

emotions of negative valence, which conﬁrms ﬁndings that infants at this age discriminate between the angry and fearful expressions behaviorally (Serrano, Iglesias, & Loeches, 1995). It further suggests that the infant N290—which is viewed as the developmental precursor to the adult N170— is sensitive to emotional information, whereas the adult N170 does not show an emotion-dependent modulation (see also Leppa¨nen et al., 2007). This suggests that early in life the N170 is sensitive to other aspects of face processing such as eye gaze (Farroni et al., 2002) and emotional expression (Kobiella et al., 2008; Leppa¨nen et al., 2007), a notion consistent with the interactive specialization view of social brain development (Johnson, 2001, 2005a, 2005b). Processing of facial expressions has been studied as a function of particular experiences such as early institutional rearing (Parker et al., 2005) and maternal personality (de Haan et al., 2004). In general, these studies suggest that experiential factors inﬂuence the ways that infants process facial expressions. It is important to note that the reverse may also be true—that is, neural development occurring at the end of the ﬁrst year (Diamond, 1991, 2000; Johnson, 2001, 2005a, 2005b) may impact infant behavior and subsequently infants’ experiences with others. Although the control of infants’ orienting is partly in the hands of the caregivers’ presentation of relevant information, the infant is clearly involved in soliciting attention and information from adults (Baldwin & Moses, 1996; Stern, 1985). With the development of the anterior attention system during the ﬁrst year of postnatal life, more direct control of attention passes from the caregiver to the infant (Bush, Luu, & Posner, 2000; Posner & Rothbart, 2000), which might inﬂuence and regulate infants’ emotional exchanges and experiences.

B. PROCESSING EMOTION IN THE VOICE Emotion produces not only characteristic changes in facial patterning but also in respiration, phonation, and articulation, which in turn determine the acoustic signal (Scherer, 1989). Thus speech melody, also referred to as emotional prosody, serves as a signal for various emotions. Adult listeners across cultures can reliably and readily recognize different emotions on the basis of vocal cues (Banse & Scherer, 1996; Scherer, Banse, & Wallbott, 2001) and do so using some of the same brain structures used to recognize facial expressions (Adolphs, Tranel, & Damasio, 2002). How the processing of emotional prosody develops over the course of ontogeny is only poorly understood. Unlike the extensive behavioral work on processing facial expressions in infancy, studies on infants’ perception of vocal expressions of emotion are relatively scarce (Walker-Andrews, 1997).

356

Mark H. Johnson et al.

In one of the few studies of infants’ processing of emotional prosody (Grossmann, Striano, & Friederici, 2005), infants heard semantically neutral words spoken with either a happy, angry, or neutral voice. Angry prosody elicited an Nc peaking around 450 ms over fronto-central sites in infants’ ERPs, suggesting greater allocation of attention to the angry prosody. This is consistent with the notion of an evolutionarily driven propensity to react more strongly to negative than to positive or neutral stimuli (Cacioppo & Berntson, 1999). A positive slow wave (700–1000 ms) was elicited by angry and happy prosody over temporal electrode sites, whereas words with neutral prosody returned to baseline. This indicates an enhanced processing in auditory (temporal) brain structures of the emotionally loaded stimuli (happy and angry), which is in line with recent fMRI work on adults’ processing of emotional speech (Grandjean et al., 2005; Mitchell et al., 2003). These ﬁndings suggest that very early in development, the human brain detects emotionally loaded words and shows differential attentional responses depending on their emotional valence. The three emotional speech categories used in the Grossmann et al. (2005) study signiﬁcantly differed in duration and fundamental frequency but did not differ with respect to their mean intensity. Effects due to these perceptual differences, however, are likely to be limited to early (exogenous) ERP components (earlier than 200 ms). Therefore, acoustic differences alone cannot account for the late ERP effects observed in response to the various emotions.

C. INTEGRATION OF EMOTIONAL INFORMATION FROM FACE AND VOICE The way that emotions are perceived when communicated either by the face or the voice has been studied in each modality separately. However, in most social interactions, emotional information is communicated simultaneously by different modalities such as the face and voice. Thus the question arises: how is emotional information from the face and voice integrated? To address this question, processing of emotionally congruent and incongruent face–voice pairs has been studied using ERP measures (Grossmann, Striano, & Friederici, 2006). In this study, infants watched facial expressions (happy or angry) and, after a delay of 400 ms, heard a word spoken with a prosody that was either emotionally congruent or incongruent with the facial expression being presented. Emotionally congruent face–voice pairs elicited similar ERP effects as recognized items in previous memory studies with infants, children, and adults in which a picture (visually presented)–name (acoustically presented) matching

The Social Cognitive Neuroscience of Infancy

357

paradigm was used (Friedman, 1991; Friedman et al., 1991, 1992; Nelson et al., 1998; Rugg & Coles, 1995). Thus, the ERP effects in Grossmann et al.’s (2006) study provides electrophysiological evidence that 7-month-olds integrate emotional information across modalities and recognize common affect in the face and voice, a ﬁnding that is in line with previous behavioral studies (Soken & Pick, 1992; Walker, 1982; Walker-Andrews, 1986). Because the face–voice pairs presented to the infants were novel to them, these ﬁndings not only indicate that infants recognized common affect but, moreover, that they applied their knowledge about emotions in face and voice to draw inferences about what might be appropriate emotional face– voice pairings. Extending previous behavioral ﬁndings, the ERP data revealed insights into the time course and characteristics of the processes underlying the integration of emotional information from face and voice in the infant brain. Emotionally incongruent prosody elicited a larger frontocentral Nc (400–600 ms) in infants’ ERPs than did an emotionally congruent prosody, indicating increased allocation of attention. Conversely, the amplitude of infants’ positive component (600–1000 ms), with a maximum over parietal channels, was larger to emotionally congruent than to incongruent prosody, indexing recognition of the congruity between face and voice (Grossmann et al., 2006). Interestingly, the body of evidence presented here shows that effects of emotion, in contrast to the ‘‘early’’ effects elicited by face and gaze manipulations alone, are reﬂected in speciﬁc modulations of components that occur relatively ‘‘late’’ in the infants’ ERP (W300 ms; in contrast, some emotional expressions can have effects at very short-latency ERP components in adults, Eimer & Holmes, 2002). This might be related to the fact that learning to identify an emotion requires more complex cognitive and brain operations, which integrate various perceptual elements provided by face and/or voice, and retrieve speciﬁc information about the emotion stored in memory. The studies that have been conducted so far have only looked at a limited set of emotions (happiness, fear, and anger). This work should be extended to other basic (disgust, sadness, and surprise) and complex (social) emotions in order to understand the emotionspeciﬁcity of the effects summarized here.

V. Interactions Between Face Identity, Eye Gaze, and Emotion Following from the two theoretical frameworks that have motivated our review so far, several speciﬁc hypotheses emerge with regard to the

358

Mark H. Johnson et al.

interaction, or lack of it, between the different aspects of social brain function reviewed above. In this section we begin by reviewing evidence in adults of partially independent processing streams for processing different kinds of information from the face (identity, eye gaze, and emotion). Interaction between these streams appears to occur in a restricted and taskdependent way. We then review evidence from infants in support of the prediction from the interactive specialization view that as the social brain network develops processing streams will fractionate and become increasingly specialized and dissociable. A speciﬁc prediction is that infants may have a common processing stream for all aspects of face processing.

A. DISSOCIABLE NEURAL PATHWAYS FOR PROCESSING FACES In adults, emotion, eye gaze, and identity are processed by partially independent neural routes, and particular structures become specialized for these different computations. One driver for these differential patterns of specialization may be that different, and even conﬂicting, invariances need to be extracted from the same basic visual input. For example, recognizing the identity of a face requires the invariant aspects of facial structure conﬁguration to be extracted from the dynamic changes associated with eye gaze shifts and emotional expressions. An example of the neural dissociation of these pathways comes from neuropsychological studies that provide evidence that speciﬁc brain damage can impair the perception of facial expressions without impairment in detecting face identity (Kurucz & Feldmar, 1979) or the face inversion perception (Bruyer et al.,1983). Furthermore, neuroimaging studies provide evidence of a spatial and temporal dissociation between face identity and emotion perception (Munte et al., 1998; Sergent et al., 1994). Finally, different facial expressions may be processed by different neural networks that involve both cortical regions (prefrontal, frontal and orbital frontal cortex, occipital temporal junction, cingulated cortex, and sensorial motor cortex) and sub-cortical structures (amygdala, insula, and basal ganglia) (Damasio et al., 2000; Gorno-Tempini et al., 2001; Kesler-West et al., 2001; Lane et al., 1997; Streit et al., 1999).

B. INTERACTIONS BETWEEN FACE PROCESSING PATHWAYS IN ADULTS Although there are dissociable routes for different aspects of face processing in adults, there are also speciﬁc interactions between these

The Social Cognitive Neuroscience of Infancy

359

streams of processing. These interactions can be speciﬁcally restricted to particular stages of processing, and can be speciﬁc to particular emotions, directions of eye gaze, or identities. We begin by considering the effects of gaze direction on other aspects of face processing. George, Driver, and Dolan, (2001) investigated how gaze direction (direct or averted) inﬂuences face processing in adults using a gender recognition task. They presented a face with direct or averted gaze, and the face was either a frontal view or tilted at 45 degrees. Speciﬁc regions of the fusiform gyrus yielded stronger responses to faces when these looked directly at the participant (regardless the orientation of the head). Because the fusiform response can be sensitive to recognition of individuals (Gauthier et al., 2000, 1999; George et al., 1999; Sergent et al., 1994, 1992), and is stronger for attended faces (Wojciulik, Kanwisher, & Driver, 1998), the authors concluded that the stronger activity found for faces with direct gaze could be interpreted in terms of enhanced attention and deeper encoding for faces with direct gaze (George et al., 2001). Furthermore, it has been previously argued that direct gaze is likely involved in the communication of increased emotional intensity, regardless of the particular emotion being expressed (Kimble & Olszewski, 1980; Kleinke, 1986). Despite the intuitive appeal of the long-standing notion that direct gaze may facilitate emotional expression processing, brain and behavioral research with adults has shown that the way in which gaze direction inﬂuences emotion perception depends on the emotion in question (Adams & Kleck, 2003, 2005; Adams et al., 2003). There are some reasons for supporting an interaction between eye gaze and emotional expression processing. First, analysis of facial expression and gaze direction may be partly processed by the same anatomical structures. Areas in the STS are critical in the processing of changeable features of faces, such as expression and gaze direction (Allison et al., 2000; Hoffman & Haxby, 2000; Perrett et al., 1992). Neuroimaging studies have also shown that the amygdala plays a central role in processing emotional facial expressions, particularly fear expression and fearful eyes (Adolphs et al., 1995; Breiter et al., 1996; Whalen et al., 1998, 2001) and monitoring gaze direction (George et al., 2001; Wicker et al., 2003). Adams et al. (2003) also highlighted amygdala’s involvement in the combined processing of facial expression and gaze direction. Namely, they demonstrated that the amygdala is differentially responsive to anger and fear expressions as a function of gaze direction. Both eye gaze and emotional expression are reported to elicit rapid and automatic spatial orienting (Driver et al., 1999; O¨hman, Flykt, & Esteves, 2001). As mentioned previously, eye gaze can trigger orienting in automatic fashion, in that individuals shift their attention to the location of a lookedat target in a fast and efﬁcient manner, even when eye gaze direction is

360

Mark H. Johnson et al.

uninformative and even counterpredictive of the target location (Driver et al., 1999; Friesen & Kingstone, 1998, 2003). As regards emotion, many behavioral observations indicate that people more readily pay attention to emotional than neutral stimuli. These effects also arise in a reﬂexive or involuntary manner. In visual search tasks where a unique target must be found among distracters, detection times are faster when the target has some emotional values, such as an angry or happy face among neutral faces (Eastwood, Smilek, & Merikle, 2001; Fox, Russo, & Dutton, 2002), or a snake or spider among ﬂowers (O¨hman et al., 2001). Furthermore, fearful faces presented peripherally trigger orienting of attention toward their location, thus facilitating target discrimination at this location, more efﬁciently so than neutral or happy faces do (Pourtois et al., 2004). Finally, another reason for supporting that emotional faces may promote attention to gaze direction is that one could expect that an expressive face (e.g., happy or fearful) gazing in a certain direction would be a more powerful attention-orienting stimulus than a comparably gazing neutral face. Such a stimulus would provide information about another individual’s direction of attention and, in addition, about the emotional signiﬁcance of the object or person he or she is attending to. A smiling face with an averted gaze would inform the observer that the other person is looking at something nice or funny, whereas a comparably gazing fearful face would suggest the presence of something threatening. So, logic suggests that the gaze of an emotional face should be a more effective cue to attention. This should be particularly true for fearful faces. Because another’s fearful gaze often signals a source of danger, learning, and/or natural selection may have particularly favored attention to the same location. By comparison, averted gaze in happy faces may indicate sources of reward important for interpersonal communication, but it has less immediate relevance to personal safety or survival. We have seen that in adults there are complex and restricted interactions between the different streams of face processing. Interactions can either be early in processing, such as modulation by a fearful expression, or after the independent parallel processing of identity, gaze, and expression. These interactions are selective with, for example, speciﬁc emotions interacting with speciﬁc directions of eye gaze.

C. INTERACTING FACE PATHWAYS IN INFANTS A speciﬁc prediction of the interacting systems model is that dissociable pathways involving specialized cortical structures in adults may start life as un-differentiated highways of common processing in infants

The Social Cognitive Neuroscience of Infancy

361

(Johnson, 2001, 2005a, 2005b; Johnson & Munakata, 2005). If this is the case we should expect neural and perceptual interactions between identity, eye gaze, and emotion. However, in contrast to adults, this interaction may be general and non-speciﬁc. In Farroni et al. (2006), we examined how facial identity and gaze direction may interact early in infancy. The question was whether direct gaze evokes a deeper analysis of other aspects of the person’s face such as its identity. Speciﬁcally, we predicted that infants who had previously seen a face accompanied by direct gaze would subsequently demonstrate recognition of that face by displaying a novelty preference for an unfamiliar face. In contrast, we predicted no preference when infants initially see a face with averted gaze. These predictions were conﬁrmed. In addition, post-hoc analyses revealed that the order in which the two conditions were presented made a difference. If the experiment began with the direct gaze condition then the averted gaze condition also elicited evidence of subsequent individual recognition. In contrast, if the experiment began with the averted gaze condition, then only the later occurring direct gaze condition resulted in recognition. These effects were not due to differences in looking time between the conditions during the habituation phase. These results conﬁrm the hypothesis that direct gaze is an important modulator of face and social information processing early in life. We concluded that from at least 4 months of age infants show facilitation of cueing of spatial attention, improved individual recognition, and enhancement of neural correlates of face processing, when accompanied, or preceded by, direct gaze. The question of how eye gaze interacts with emotional expression has also been examined in young infants using ERP measures (Striano et al., 2006). In this study, 4-month-old infants were presented with neutral, happy, and angry facial expressions when accompanied with direct and averted gaze. Neural processing of angry faces was enhanced in the context of direct gaze as indexed by an increased amplitude of a late positive frontocentral ERP component, whereas there was no effect of gaze direction on the processing of the other expressions. This neural sensitivity to the direction of angry gaze can be explained in two different ways. First, infants at this age who have very rarely seen angry faces (Campos et al., 2000) might be primed to detect novel/unfamiliar emotions that are directed toward them. Second, infants may detect the potential threat conveyed by an angry face with gaze directed at them, which is reﬂected in an enhanced neural response. To ﬁnd out which explanation (‘‘novelty’’ versus ‘‘threat’’) can account for these ERP ﬁndings, it would be necessary to examine the processing of a facial expression that is equally unfamiliar to the infant but whose perception is enhanced in the context of averted gaze; one such emotion

362

Mark H. Johnson et al.

is fear (Adams & Kleck, 2005, Adams et al., 2003). Nevertheless it is important to note that by using ERP measures it was possible to show that the neural processing of facial expression is inﬂuenced by the direction of eye gaze. In summary, converging lines of evidence suggest more generalized interactions between different aspects of face perception (identity, eye gaze, and emotion) in infants than in adults. These generalized interactions may be indicative of less differentiated processing streams for these functions in infants than in adults. Greater common processing of the different computations applied to faces in early development is a prediction of the interactive specialization approach.

VI. Conclusions In this chapter we have illustrated how adopting a developmental cognitive neuroscience approach sheds light on how the social brain network emerges during infancy. We believe that the combination of neuroscience, cognitive, and behavioral studies has signiﬁcantly advanced our understanding in this foundational area of neuroscience. Our work, and that of other labs, has been presented within two frameworks; the twoprocess model of the development of face processing originally presented by Johnson and Morton (1991; Morton & Johnson, 1991) and the interactive specialization model of functional brain development (Johnson 2001, 2005b). With regard to the two-process model we reviewed a number of studies of newborn face-related preferences, most of which supported the view that newborns have a bias to orient toward faces in their natural visual environment. Although the exact mechanisms that underlie this bias remain the topic of some debate, the proposal that best accounts for the majority of the data currently available is that there is a ‘‘quick and dirty’’ sub-cortical route for face detection that is activated by a face (or eye)-like phase contrast pattern within a bounded surface or object (Johnson, 2005a). Interestingly, some evidence suggests a broader role for Conspec than originally envisaged because strong evidence from adults reveals that the ‘‘quick and dirty’’ sub-cortical route modulates processing in cortical regions within the social brain network. This indicates that the mechanisms that underlie the orienting toward, and foveating of, faces with direct gaze in young infants also facilitate the activation of relevant cortical regions, providing an important foundation for the emerging social brain. Turning to the development of social perception over the early months and years of life, we have reviewed evidence broadly consistent with

The Social Cognitive Neuroscience of Infancy

363

predictions from the Interactive Specialization approach. According to this view, most parts of the social brain network can be activated in infants, though activation may also extend to other regions not activated under these circumstances in adults. Furthermore, the social brain regions activated may have broader functions (be less ﬁnely tuned) than in adults. Some of the evidence consistent with this ‘‘Interactive Specialization’’ view is that, compared to adults, infants activate regions in addition to, and surrounding, the core face processing network. Furthermore, face and eye gaze perception have been shown to share common patterns of cortical activation early in ontogeny, which later partially dissociate and become more specialized (Farroni et al., 2002, 2004a, 2004b; Grice et al., 2005; Johnson et al., 2005; Taylor et al., 2001). These ﬁndings support the view that structures in the social brain network initially have more homogeneous response properties, with common processing of many aspects of faces, bodies, and actions. With experience, these structures may become more differentiated and specialized in their response properties, ﬁnally resulting in the specialized patterns of activation typically observed in adults. This view has implications for atypical development in that some developmental disorders that involve disruption to the social brain network, such as autism, may be characterized in terms of failures or delays of the specialization of structures on the cortical social brain network (see Johnson et al., 2005, for further discussion). For example, Grice et al. (2005) found ERP evidence consistent with common processing of eye gaze and other aspects of face perception in young children with autism, at an age at which there is evidence for different streams of processing having emerged in typically developing children. We have described several neuroimaging methods that have been used to study different aspects of social information processing in infants. The available evidence on the neural processes related to face, gaze, emotion, biological motion, action, and joint attention discussed revealed how the infant brain processes information about the social world. However, many areas of infant social cognition, such as imitation, social (complex) emotions, and ‘‘theory of mind’’ remain unexplored (for recent behavioral studies on infant theory of mind, see Onishi & Baillargeon, 2005; Southgate, Senju, & Csibra, 2007; Surian, Caldi, & Sperber, 2007). Future research is needed to examine the neural correlates of these more complex aspects of social cognitive development, and it is very likely that this kind of work will reveal greater differences between adult and infant/child brain function. Another important issue is that although the cross-talk between developmental psychologists and social cognitive neuroscientists has begun on a theoretical level (Decety & Sommerville, 2003; Meltzoff & Decety, 2003),

364

Mark H. Johnson et al.

there is very little infant brain research that is more directly informed and motivated by already existing theories of infant social cognitive development (Csibra & Gergely, 2006, Meltzoff, 2002, 2005; Tomasello et al., 2005). The available theoretical frameworks explaining the developmental trajectories of social cognitive capacities provide a rich source of hypotheses that are testable using the neuroimaging tools. In this context it will be of particular importance to identify the neural processes that underlie known social behavioral and social cognitive transitions. Thus, further progress in the developmental cognitive neuroscience of the social brain network crucially depends upon a closer integration of human functional brain development with theories of social cognitive development.

REFERENCES Acerra, F., Burnod, Y., & de Schonen, S. (2002). Modelling aspects of face processing in early infancy. Developmental Science, 5, 98–117. Adams, R. B., & Kleck, R. E. (2003). Perceived gaze direction and the processing of facial displays of emotion. Psychological Science, 14, 644–647. Adams, R. B., & Kleck, R. E. (2005). The effects of direct and averted gaze on the perception of facially communicated emotion. Emotion, 5, 3–11. Adams, R. B., Gordon, H. L., Baird, A. A., Ambady, N., & Kleck, R. E. (2003). Effects of gaze on amygdala sensitivity to anger and fear faces. Science, 300, 1536. Adolphs, R. (1999). Social cognition and the human brain. Trends in Cognitive Science, 3, 469–479. Adolphs, R. (2003). Cognitive neuroscience of human social behaviour. Nature Reviews Neuroscience, 4, 165–178. Adolphs, R., Tranel, D., Damasio, H., & Damasio, A. R. (1995). Fear and the human amygdala. Journal of Neuroscience, 15, 5879–5891. Adolphs, R., Tranel, D., & Damasio, H. (2002). Neural systems for recognizing emotion from prosody. Emotion, 2, 23–51. Allison, T., Puce, A., & McCarthy, G. (2000). Social perception: Role of the STS region. Trends in Cognitive Science, 4, 267–278. Anderson, S. W., Bechara, A., Damasio, H., Tranel, D., & Damasio, A. R. (1999). Impairment of social and moral behavior related to early damage in human prefrontal cortex. Nature Neuroscience, 2, 469–479. Baldwin, D. A., & Moses, L. J. (1996). The ontogeny of social information gathering. Child Development, 67, 1915–1939. Banse, R., & Scherer, K. (1996). Acoustic proﬁles in vocal emotion expression. Journal of Personality and Social Psychology, 70, 614–636. Baron-Cohen, S. (1995). Mindblindness: An essay on autism and theory of mind. Cambridge: MIT Press. Batki, A., Baron-Cohen, S., Wheelwright, S., Connellan, J., & Ahluwalia, J. (2000). Is there an innate gaze module? Evidence from human neonates. Infant Behavior and Development, 23, 223–229. Bednar, J. A., & Miikkulainen, R. (2003). Learning innate face preferences. Neural Computation, 15, 1525–1557.

The Social Cognitive Neuroscience of Infancy

365

Bentin, S., Allison, T., Puce, A., Perez, A., & McCarthy, A. (1996). Electrophysiological studies of face perception in humans. Journal of Cognitive Neuroscience, 8, 551–565. Bloom, P. (2000). How children learn the meanings of words. Cambridge: MIT Press. Breiter, H. C., Etcoff, N. L., Whalen, P. J., Kennedy, W. A., Rauch, S. L., Buckner, R. L., Strauss, M. M., Hyman, S. E., & Rosen, B. R. (1996). Response and habituation of the human amygdala during visual processing of facial expression. Neuron, 17, 875–887. Brothers, L. (1990). The social brain: A project for integrating primate behavior and neurophysiology in a new domain. Concepts Neuroscience, 1, 27–51. Bruyer, R., Laterre, C., Seron, X., Feyereisen, P., Strypstein, E., Pierrard, E., & Rectem, D. (1983). A case of prosopagnosia with some preserved covert remembrance of familiar faces. Brain and Cognition, 2, 257–284. Bush, G., Luu, P., & Posner, M. I. (2000). Cognitive and emotional inﬂuences in anterior cingulate cortex. Trends in Cognitive Science, 4, 215–222. Bushnell, I. W. R., Sai, F., & Mullin, J. T. (1989). Neonatal recognition of the mother’s face. British Journal of Developmental Psychology, 7, 3–15. Buzsaki, G., & Draguhn, A. (2004). Neuronal oscillations in cortical networks. Science, 304, 1926–1929. Cacioppo, J. T., & Berntson, G. G. (1999). The affect system: Architecture and operating characteristics. Current Directions in Psychological Science, 8, 133–137. Campos, J. J., Anderson, D. I., Barbu-Roth, M. A., Hubbard, E. M., Hertenstein, M. J., & Witherington, D. (2000). Travel broadens the mind. Infancy, 1, 149–219. Chugani, H. T., & Phelps, M. E. (1986). Maturational changes in cerebral function in infants determined by 18FDG positron emission tomography. Science, 231, 840–843. Chugani, H. T., Phelps, M. E., & Mazziotta, J. C. (1987). Positron emission tomography study of human brain functional development. Annals of Neurology, 22, 487–497. Cohen-Kadosh, K., & Johnson, M. H. (2007). Developing a cortex specialized for face perception. Trends in Cognitive Sciences, 11, 367–369. Courchesne, E., Ganz, L., & Norcia, M. E. (1981). Event-related brain potentials to human faces in infants. Child Development, 52, 804–811. Csibra, G., & Gergely, G. (2006). Social learning and social cognition: The case for pedagogy. In Y. Munakata & M. H. Johnson (Eds.), Processes of change in brain and cognitive development. Attention and Performance XXI (pp. 249–274). Oxford: Oxford University Press. Csibra, G., Davis, G., Spratling, M. W., & Johnson, M. H. (2000). Gamma oscillations and object processing in the infant brain. Science, 290, 1582–1585. Damasio, A. R., Grabowski, T. J., Bechara, A., Damasio, H., Ponto, L. L. B., Parvizi, J., & Hichwa, R. D. (2000). Subcortical and cortical brain activity during the feeling of selfgenerated emotions. Nature Neuroscience, 3, 1049–1056. Decety, J., & Sommerville, J. A. (2003). Shared representations between self and others: A social cognitive neuroscience view. Trends in Cognitive Sciences, 7, 527–533. de Haan, M., & Nelson, C. A. (1997). Recognition of the mother’s face by six-month old infants: A neurobehavioral study. Child Development, 68, 187–210. de Haan, M., & Thomas, K. M. (2002). Applications of ERP and fMRI techniques to developmental science. Developmental Science, 5, 335–343. de Haan, M., Pascalis, O., & Johnson, M. H. (2002). Specialization of neural mechanisms underlying face recognition in human infants. Journal of Cognitive Neuroscience, 14, 199–209. de Haan, M., Johnson, M. H., & Halit, H. (2003). Development of face-sensitive event-related potentials during infancy: A review. International Journal Psychophysiology, 51, 45–58.

366

Mark H. Johnson et al.

de Haan, M., Belsky, J., Reid, V., Volein, A., & Johnson, M. H. (2004). Maternal personality and infants’ neural and visual responsitivity to facial expressions of emotion. Journal of Child Psychology and Psychiatry, 45, 1209–1218. Dehaene-Lambertz, G., Dehaene, S., & Hertz-Pannier, L. (2002). Functional neuroimaging of speech perception in infants. Science, 298, 2013–2015. de Schonen, S., & Mathivet, E. (1990). Hemispheric asymmetry in a face discrimination task in infants. Child Development, 61, 1192–1205. Diamond, A. (1991). Frontal lobe involvement in cognitive changes during the ﬁrst year of life. In K. R. Gibson & A. C. Peterson (Eds.), Brain maturation and cognitive development: Comparative and cross-cultural perspectives (pp. 127–180). New York: Aldine de Gruyter. Diamond, A. (2000). Close interrelation of motor development and cognitive development and of the cerebellum and prefrontal cortex. Child Development, 71, 44–56. Driver, J., Davis, G., Ricciardelli, P., Kidd, P., Maxwell, E., & Baron-Cohen, S. (1999). Gaze perception in triggers reﬂexive visuospatial orienting. Visual Cognition, 6, 509–540. Easterbrook, M., Hains, S., Muir, D., & Kisilevsky, B. (1999). Faceness or complexity: Evidence from newborn visual tracking of facelike stimuli. Infant Behavior and Development, 22, 17–35. Eastwood, J. D., Smilek, D., & Merikle, P. (2001). Differential attentional guidance by unattended faces expressing positive and negative emotion. Perception & Psychophysics, 63, 1004–1013. Eimer, M., & Holmes, A. (2002). An ERP study on the time course of emotional face processing. NeuroReport, 13, 427–431. Engel, A. K., Fries, P., & Singer, W. (2001). Dynamic predictions: Oscillations and synchrony in top–down processing. Nature Reviews Neuroscience, 2, 704–716. Farroni, T., Johnson, M. H., Brockbank, M., & Simion, F. (2000). Infants’ use of gaze direction to cue attention: The importance of perceived motion. Visual Cognition, 7, 705–718. Farroni, T., Csibra, G., Simion, F., & Johnson, M. H. (2002). Eye contact detection inhumans from birth. Proceedings of the National Academy of Science USA, 99, 9602–9605. Farroni, T., Mansﬁeld, E. M., Lai, C., & Johnson, M. H. (2003). Infants perceiving and acting on the eyes: Tests of an evolutionary hypothesis. Journal of Experimental Child Psychology, 85, 199–212. Farroni, T., Pividori, D., Simion, F., Massaccesi, S., & Johnson, M. H. (2004a). Eye gaze cueing of attention in newborns. Infancy, 5, 39–60. Farroni, T., Johnson, M. H., & Csibra, G. (2004b). Mechanisms of eye gaze perception during infancy. Journal of Cognitive Neuroscience, 16, 1320–1326. Farroni, T., Johnson, M. H., Menon, E., Zulian, L., Faraguna, D., & Csibra, G. (2005). Newborn’s preference for face-relevant stimuli: Effects of contrast polarity. Proceedings of the National Academy of Sciences USA, 102, 17245–17250. Farroni, T., Menon, E., & Johnson, M. H. (2006). Factors inﬂuencing newborns’ face preferences for faces with eye contact. Journal of Experimental Child Psychology, 95, 298–308. Farroni, T., Massaccesi, S., Menon, E., & Johnson, M. H. (2007). Direct gaze modulates face recognition in young infants. Cognition, 102, 396–404. Fiebach, C. J., Gruber, T., & Supp, G. (2005). Neuronal mechanisms of repetition priming in occipitotemporal cortex: Spatiotemporal evidence from functional magnetic imaging and electroencephalography. Journal of Neuroscience, 25, 3414–3422. Foucher, J. R., Otzenberger, H., & Gounot, D. (2003). The BOLD response and gamma oscillations respond differently than evoked potentials: An interleaved EEG-fMRI study. BMC Neuroscience, 4, 22.

The Social Cognitive Neuroscience of Infancy

367

Fox, E., Russo, R., & Dutton, K. (2002). Attentional bias for threat: Evidence for delayed disengagement from emotional faces. Cognition & Emotion, 16, 355–379. Friedman, D. (1991). The endogenous scalp-recorded brain potentials and their relation to cognitive development. In J. R. Jennings & M. G. H. Coles (Eds.), Handbook of cognitive psychophysiology: Central and autonomic nervous system approaches (pp. 621–656). New York: Wiley. Friedman, D., Putnam, L., Ritter, W., Hamberger, M., & Berman, S. (1992). A developmental study of picture matching in children, adolescents, and young-adults: A replication and extension. Psychophysiology, 29, 593–610. Friesen, C. K., & Kingstone, A. (1998). The eyes have it! Reﬂexive orienting is triggered by nonpredictive gaze. Psychonomic Bulletin & Review, 5, 490–495. Friesen, C. K., & Kingstone, A. (2003). Covert and overt orienting to gaze direction cues and the effects of ﬁxation offset. NeuroReport, 14, 489–493. Gauthier, I., & Nelson, C. A. (2001). The development of face expertise. Current Opinion in Neurobiology, 11, 219–224. Gauthier, I., Tarr, M. J., Anderson, A. W., Skudlarski, P., & Gore, J. C. (1999). Activation of the middle fusiform ‘‘face area’’ increases with expertise in recognizing novel objects. Nature Neuroscience, 2, 568–573. Gauthier, I., Skudlarski, P., Gore, J. C., & Anderson, A. W. (2000). Expertise for cars and birds recruits brain areas involved in face recognition. Nature Neuroscience, 3, 191–197. George, N., Dolan, R. J., Fink, G. R., Baylis, G. C., Russel, C., & Driver, J. (1999). Contrast polarity and face recognition in the human fusiform gyrus. Nature Neuroscience, 2, 574–580. George, N., Driver, J., & Dolan, R. J. (2001). Seen gaze-direction modulates fusiform activity and its coupling with other brain areas during face processing. NeuroImage, 13, 1102–1112. Gorno-Tempini, M. L., Pradelli, S., Seraﬁni, M., Pagnoni, G., Baraldi, P., Porro, C., Nicoletti, R., Umita, C., & Nichelli, P. (2001). Explicit and incidental facial expression processing: An fMRI study. NeuroImage, 14, 465–473. Grandjean, D., Sander, D., Pourtois, G., Schwartz, S., Seghier, M. L., Scherer, K. R., & Vuilleumier, P. (2005). The voices of wrath: Brain responses to angry prosody in meaningless speech. Nature Neuroscience, 8, 145–146. Grice, S. J., Halit, H., Farroni, T., Baron-Cohen, S., Bolton, P., & Johnson, M. H. (2005). Neural correlates of eye-gaze detection in young children with autism. Cortex, 41, 342–353. Grossmann, T., & Farroni, T. (in press). Decoding social signals in the infant brain: A look at eye gaze perception. In M. de Haan, & M. Gunnar (Eds.), Handbook of Developmental Social Neuroscience. New York: Guilford Press. Grossmann, T., Striano, T., & Friederici, A. D. (2005). Infants’ electric brain responses to emotional prosody. NeuroReport, 16, 1825–1828. Grossmann, T., Striano, T., & Friederici, A. D. (2006). Crossmodal integration of emotional information from face and voice in the infant brain. Developmental Science, 9, 309–315. Grossmann, T., Striano, T., & Friederici, A. D. (2007). Developmental changes in infants’ processing of happy and angry facial expressions: A neurobehavioral study. Brain and Cognition, 64, 30–41. Grossmann, T., Johnson, M. H., Farroni, T., & Csibra, G. (2007). Social perception in the infant brain: Gamma oscillatory activity in response to eye gaze. Social Cognitive & Affective Neuroscience, 2, 284–291. Grossmann, T., Johnson, M. H., Lloyd-Fox, S., Blasi, A., Deligianni, F., Elwell, C., & Csibra, G. (submitted). Early cortical specialization for face-to-face communication in human infants.

368

Mark H. Johnson et al.

Halit, H., de Haan, M., & Johnson, M. H. (2003). Cortical specialisation for face processing: Face-sensitive event-related potential components in 3- and 12-month-old infants. NeuroImage, 19, 1180–1193. Haxby, J. V., Hoffman, E. A., & Gobbini, M. I. (2000). The distributed human neural system for face perception. Trends in Cognitive Sciences, 4, 223–233. Hermann, C. S., Lenz, D., Junge, S., Busch, N. A., & Maess, B. (2003). Memory-matches evoke human gamma responses. BMC Neuroscience, 5, 13. Hermann, C. S., Munk, M. H. J., & Engel, A. K. (2004). Cognitive functions of gamma band activity: Memory match and utilization. Trends in Cognitive Sciences, 8, 347–355. Hoffman, E. A., & Haxby, J. V. (2000). Distinct representations of eye gaze and identity in the distributed human neural system for face perception. Nature Neuroscience, 3, 80–84. Hood, B. M., Willen, J. D., & Driver, J. (1998). Adult’s eyes trigger shifts of visual attention in human infants. Psychological Science, 9, 131–134. Horn, G. (2004). Pathways of the past: The imprint of memory. Nature Reviews Neuroscience, 5, 108–120. Huttenlocher, P. R. (2002). Neural plasticity. Cambridge: Harvard University Press. Huttenlocher, P. R., & Dabholkar, A. S. (1997). Regional differences in synaptogenesis in human cerebral cortex. Journal of Comparative Neurology, 387, 167–178. Izard, C. E. (1991). The psychology of emotions. London: Plenum. Johnson, M. H. (1990). Cortical maturation and the development of visual attention in early infancy. Journal of Cognitive Neuroscience, 2, 81–95. Johnson, M. H. (2001). Functional brain development in humans. Nature Reviews Neuroscience, 2, 475–483. Johnson, M. H. (2005a). Subcortical face processing. Nature Reviews Neuroscience, 6, 766–774. Johnson, M. H. (2005b). Developmental cognitive neuroscience (2nd ed.). Oxford: Blackwell. Johnson, M. H., & Morton, J. (1991). Biology and cognitive development: The case of face recognition. Oxford: Blackwell. Johnson, M. H., & Munakata, Y. (2005). Processes of change in brain and cognitive development. Trends in Cognitive Sciences, 9, 152–158. Johnson, M. H., Dziurawiec, S., Ellis, H., & Morton, J. (1991). Newborns’ preferential tracking of face-like stimuli and its subsequent decline. Cognition, 40, 1–19. Johnson, M. H., Grifﬁn, R., Csibra, G., Halit, H., Farroni, T., de Haan, M., Baron-Cohen, S., & Richards, J. (2005). The emergence of the social brain network: Evidence from typical and atypical development. Development and Psychopathology, 17, 599–619. Kampe, K., Frith, C. D., & Frith, U. (2003). ‘‘Hey John’’: Signals conveying communicative intention toward self activate brain regions associated with ‘‘mentalizing,’’ regardless of modality. Journal of Neuroscience, 12, 5258–5263. Kanwisher, N. (2000). Domain speciﬁcity in face processing. Nature Neuroscience, 3, 759–763. Kanwisher, N., McDermott, J., & Chun, M. M. (1997). The fusiform face area: A module in human extrastriate cortex specialized for face perception. Journal of Neuroscience, 17, 4302–4311. Kesler-West, M. L., Andersen, A. H., Smith, C. D., Avison, M. J., Davis, C. E., Kryscio, R. J., & Blonder, L. X. (2001). Neural substrates of facial emotion processing using fMRI. Cognitive Brain Research, 11, 213–226. Kilner, J. M., Mattout, J., Henson, R., & Friston, K. J. (2005). Hemodynamic correlates of EEG: A heuristic. NeuroImage, 280, 280–286. Kimble, C. E., & Olszewski, D. A. (1980). Gaze and emotional expression: The effects of message positivity-negativity and emotional intensity. Journal of Research in Personality, 14, 60–69.

The Social Cognitive Neuroscience of Infancy

369

Kleiner, K. A. (1993). Speciﬁc versus non-speciﬁc face recognition device. In B. de BoyssonBardies, S. de Schonen, P. Jusczyk, P. MacNeilage, & J. Morton (Eds.), Developmental neurocognition: Speech and face processing in the first year of life (pp. 103–108). Dordrecht: Kluwer Academic Press. Kleiner, K. A., & Banks, M. S. (1987). Stimulus energy does not account for 2-month-olds’ face preferences. Journal of Experimental Psychology: Human Perception and Performance, 13, 594–600. Kleinke, C. L. (1986). Gaze and eye contact: A research review. Psychological Bulletin, 100, 68–72. Kobiella, A., Grossmann, T., Reid, V. M., & Striano, T. (2008). The discrimination of angry and fearful facial expressions in 7-month-old infants: An event-related potential study. Cognition & Emotion, 22, 134–146. Kuhl, P. (2004). Early language acquisition: Cracking the speech code. Nature Reviews Neuroscience, 5, 831–841. Kurucz, J., & Feldmar, G. (1979). Prosopo-affective agnosia as a symptom of cerebral organic disease. Journal of the American Geriatics Society, 27, 225–230. Lane, R. D., Reiman, E. M., Bradley, M. M., Lang, P. J., Ahern, G. L., Davidson, R. J., & Schwartz, G. E. (1997). Neuroanatomical correlates of pleasant and unpleasant emotion. Neuropsychologica, 35, 1437–1444. Langlois, J. H., & Roggman, L. A. (1990). Attractive faces are only average. Psychological Science, 1, 115–121. Leppa¨nen, J. M., Moulson, M. C., Vogel-Farley, V. K., & Nelson, C. A. (2007). An ERP study of emotional face processing in the adult and infant brain. Child Development, 78, 232–245. Macchi Cassia, V., Valenza E., Simion, F., & Leo I. (2007). Congruency as a non-specific perceptual property contributing to newborn’s face preference. Unpublished manuscript. Meltzoff, A. N. (2002). Imitation as a mechanism of social cognition: Origins of empathy, theory of mind, and the representation of action. In U. Goswami (Ed.), Handbook of childhood cognitive development (pp. 6–25). Oxford: Blackwell. Meltzoff, A. N. (2005). Imitation and other minds: The ‘‘Like Me’’ hypothesis. In S. Hurley & N. Chater (Eds.), Perspectives on imitation: From cognitive neuroscience to social science (pp. 55–77). Cambridge: MIT Press. Meltzoff, A. N., & Decety, J. (2003). What imitation tells us about social cognition: A rapprochement between developmental psychology and cognitive neuroscience. Philosophical Transactions of the Royal Society London Series B, 358, 491–500. Mitchell, R. L. C., Elliott, R., Barry, M., Cruttenden, A., & Woodruff, P. W. R. (2003). The neural response to emotional prosody as revealed by functional magnetic resonance imaging. Neuropsychologia, 41, 1410–1421. Morton, J., & Johnson, M. H. (1991). CONSPEC and CONLERN: A two-process theory of infant face recognition. Psychological Review, 98, 164–181. Morton, J., Johnson, M. H., & Maurer, D. (1990). On the reasons for newborns’ responses to faces. Infant Behavior and Development, 13, 99–103. Munte, T. F., Brack, M., Groother, O., Wieringa, B. M., Matzke, M., & Johannes, S. (1998). Event-related brain potentials to unfamiliar faces in explicit and implicit memory tasks. Neuroscience Research, 28, 223–233. Nelson, C. A., & Collins, P. F. (1991). Event-related potential and looking time analysis of infants’ responses to familiar and novel events: Implications for visual recognition memory. Developmental Psychology, 27, 50–58. Nelson, C. A., & de Haan, M. (1996). Neural correlates of infants’ visual responsiveness to facial expression of emotion. Developmental Psychobiology, 29, 577–595.

370

Mark H. Johnson et al.

Nelson, C. A., Thomas, K. M., de Haan, M., & Wewerka, S. (1998). Delayed recognition memory in infants and adults as revealed by event-related potentials. International Journal of Psychophysiology, 29, 145–165. Niessing, J., Ebisch, B., Schmidt, K. E., Niessing, M., Singer, W., & Galuske, R. A. W. (2005). Hemodynamic signals correlate tightly with synchronized gamma oscillations. Science, 309, 948–951. O¨hman, A., Flykt, A., & Esteves, F. (2001). Emotion drives attention: Detecting the snake in the grass. Journal of Experimental Psychology: General, 130, 466–478. Onishi, K., & Baillargeon, R. (2005). Do 15-month-old infants understand false beliefs? Science, 308, 255–258. Parker, S.W., Nelson, C.A., & The Bucharest Early Intervention Project Core Group. (2005). The impact of early institutional rearing on the ability to discriminate facial expressions of emotion: An event-related potential study. Child Development, 76, 54–72. Pelphrey, K. A., Singerman, J. D., Allison, T., & McCarthy, G. (2003). Brain activation evoked by perception of gaze shifts: The inﬂuence of context. Neuropsychologia, 41, 156–170. Pelphrey, K. A., Viola, R. J., & McCarthy, G. (2004). When strangers pass: Processing of mutual and averted gaze in the superior temporal sulcus. Psychological Science, 15, 598–603. Perrett, D. I., Hietanen, J. K., Oran, M. W., & Benson, P. J. (1992). Organization and functions of cells responsive to faces in the temporal cortex. Philosophical Transactions of the Royal Society London B, 29, 23–30. Posner, M. I. (1980). Orienting of attention. Quarterly Journal of Experimental Psychology, 32, 3–25. Posner, M. I., & Rothbart, M. K. (2000). Developing mechanisms of self regulation. Development and Psychopathology, 12, 427–441. Pourtois, G., Sander, D., Andres, M., Grandjean, D., Reveret, L., Olivier, E., & Vuilleumier, P. (2004). Dissociable roles of the human somatosensory and superior temporal cortices for processing social face signals. European Journal of Neuroscience, 20, 3507–3515. Quinn, P. C., & Slater, A. (2003). Face perception at birth and beyond. In O. Pascalis & A. Slater (Eds.), Face perception in infancy and early childhood: Current perspectives (pp. 3–11). New York: NOVA Science Publishers. Reid, V. M., Striano, T., Kaufman, J., & Johnson, M. H. (2004). Eye gaze cueing facilitates neural processing of objects in 4 month old infants. NeuroReport, 15, 2553–2556. Reynolds, G. D., & Richards, J. E. (2005). Familiarization, attention, and recognition memory in infancy: An ERP and cortical source analysis study. Developmental Psychology, 41, 598–615. Richards, J. E. (2004). Recovering dipole sources from scalp-recorded event-related-potentials using component analysis: Principal component analysis and independent component analysis. International Journal of Psychophysiology, 54, 201–220. Rodriguez, E., George, N., Lachaux, J. P., Martinere, J., Renault, B., & Varela, F. (1999). Perception’s shadow: Long distance synchronization of human brain activity. Nature, 397, 430–433. Rugg, M. D., & Coles, M. G. H. (1995). Electrophysiology of mind: Event-related brain potentials and cognition. Oxford: Oxford University Press. Scherer, K. R. (1989). Vocal correlates of emotion. In A. Manstead & H. Wagner (Eds.), Handbook of psychophysiology: Emotion and social behavior (pp. 165–197). London: Wiley. Scherer, K. R., Banse, R., & Wallbott, H. G. (2001). Emotion inferences from vocal expression correlate across languages and cultures. Journal of Cross-Cultural Psychology, 32, 76–92.

The Social Cognitive Neuroscience of Infancy

371

Schilbach, L., Wohlschlager, A. M., Newen, A., Shah, N. J., Fink, G. R., & Vogeley, K. (2006). Being with virtual others: Neural correlates of social interaction. Neuropsychologica, 44, 718–730. Sejnowski, T. J., & Paulsen, O. (2006). Network oscillations: Emerging computational principles. Journal of Neuroscience, 26, 1673–1676. Senju, A., Johnson, M. H., & Csibra, G. (2006). The development and neural basis of referential gaze perception. Social Neuroscience, 1, 220–234. Sergent, J., Ohta, S., & MacDonald, B. (1992). Functional neuroanatomy of face and object processing: A positron emission tomography study. Brain, 1, 15–36. Sergent, J., Ohta, S., MacDonald, B., & Zuck, E. (1994). Segregated processing of facial identity and emotion in the human brain: A PET study. Visual Cognition, 1, 349–369. Serrano, J. M., Iglesias, J., & Loeches, A. (1995). Infants’ responses to adult static facial expressions. Infant Behavior and Development, 18, 477–482. Simion, F., Valenza, E., Macchi Cassia, V., Turati, C., & Umilta`, C. (2002). Newborns’ preference for up-down asymmetrical conﬁgurations. Developmental Science, 5, 427–434. Simion, F., Macchi Cassia, V., Turati, C., & Valenza, E. (2003). Non-speciﬁc perceptual biases at the origins of face processing. In O. Pascalis & A. Slater (Eds.), The development of face processing in infancy and early childhood: Current perspectives (pp. 13–26). New York: Nova Science Publisher. Slater, A., Von der Schulenburg, C., Brown, E., & Badenoch, M. (1998). Newborn infants prefer attractive faces. Infant Behavior and Developement, 21, 345–354. Slater, A., Bremner, G., Johnson, S. P., Sherwood, P., Hayes, R., & Brown, E. (2000). Newborn infants’ preference for attractive faces: The role of internal and external facial features. Infancy, 1, 265–274. Snyder, K., Webb, S. J., & Nelson, C. A. (2002). Theoretical and methodological implications of variability in infant brain response during recognition memory paradigm. Infant Behavior and Development, 25, 466–494. Soken, N. H., & Pick, A. D. (1992). Intermodal perception of happy and angry expressive behaviors by seven-month-old infants. Child Development, 63, 787–795. Southgate, V., Senju, A., & Csibra, G. (2007). Action anticipation through attribution of false belief by two-year-olds. Psychological Science, 18, 587–592. Stern, D. (1985). The interpersonal world of the infant. New York: Basic Books. Streit, M., Ioannides, A. A., Liu, L., Wo¨lwer, W., Dammers, J., Gross, J., Gaebel, W., & Mu¨ller-Ga¨rtner, H. W. (1999). Neurophysiological correlates of the recognition of facial expressions of emotion as revealed by magnetoencephalography. Cognitive Brain Research, 7, 481–491. Striano, T., Kopp, F., Grossmann, T., & Reid, V. M. (2006). Eye contact inﬂuences neural processing of emotional expressions in 4-month-old infants. Social Cognitive and Affective Neuroscience, 1, 87–95. Surian, L., Caldi, S., & Sperber, D. (2007). Attribution of beliefs by 13-month-old infants. Psychological Science, 18, 580–586. Tallon-Baudry, C., & Bertrand, O. (1999). Oscillatory gamma activity in humans and its role in object representation. Trends in Cognitive Science, 3, 151–162. Taylor, M. J., Itier, R. J., Allison, T., & Edmonds, G. E. (2001). Direction of gaze effects on early face processing: Eyes-only vs full faces. Cognitive Brain Research, 10, 333–340. Taylor, M. J., Batty, M., & Itier, R. J. (2004). The faces of development: A review of early face processing over childhood. Journal of Cognitive Neuroscience, 16, 1426–1442. Tomasello, M., Carpenter, M., Call, J., Behne, T., & Moll, H. (2005). Understanding and sharing intentions: The origins of cultural cognition. Behavioral and Brain Sciences, 28, 675–691.

372

Mark H. Johnson et al.

Turati, C., Sangrioli, S., Ruel, J., & de Schonen, S. (2004). Evidence of the face inversion effect in 4-month-old infants. Infancy, 6, 275–297. Tzourio-Mazoyer, N., De Schonen, S., Crivello, F., Reutter, B., Aujard, Y., & Mazoyer, B. (2002). Neural correlates of woman face processing by 2-month-old infants. NeuroImage, 15, 454–461. Valenza, E., Simion, F., Macchi Cassia, V., & Umilta, C. (1996). Face preference at birth. Journal of Experimental Psychology: Human Perception and Performance, 27, 892–903. Vuilleumier, P. (2000). Faces call for attention: Evidence from patients with visual extinction. Neuropsychologia, 38, 693–700. Vuilleumier, P., & Sagiv, N. (2001). Two eyes make a pair: Facial organization and perceptual learning reduce visual extinction. Neuropsychologia, 39, 1144–1149. Walker, A. S. (1982). Intermodal perception of expressive behaviors by human infants. Journal of Experimental Child Psychology, 33, 514–535. Walker-Andrews, A. S. (1986). Intermodal perception of expressive behaviors: Relation of eye and voice? Developmental Psychology, 22, 373–377. Walker-Andrews, A. S. (1997). Infants’ perception of expressive behaviors: Differentiation of multimodal information. Psychological Bulletin, 121, 1–20. Whalen, P. J., Rauch, S. L., Etcoff, N. L., McInerney, S. C., Lee, M. B., & Jenike, M. A. (1998). Masked presentations of emotional facial expressions modulate amygdala activity without explicit knowledge. Journal of Neuroscience, 18, 411–418. Whalen, P. J., Shin, L. M., McInerney, S. C., Fischer, H., Wright, C. I., & Rauch, S. L. (2001). A functional MRI study of human amygdala responses to facial expressions of fear and anger. Emotion, 1, 70–83. Wicker, B., Keysers, C., Plailly, J., Royet, J. P., Gallese, V., & Rizzolatti, G. (2003). Both of us disgusted in my insula: The common neural basis of seeing and feeling disgust. Neuron, 40, 655–664. Wojciulik, E., Kanwisher, N., & Driver, J. (1998). Covert visual attention modulates facespeciﬁc activity in the human fusiform gyrus: fMRI study. Journal of Neurophysiology, 79, 1574–1578.

CHILDREN’S THINKING IS NOT JUST ABOUT WHAT IS IN THE HEAD: UNDERSTANDING THE ORGANISM AND ENVIRONMENT AS A UNIFIED SYSTEM

Jodie M. Plumert DEPARTMENT OF PSYCHOLOGY, UNIVERSITY OF IOWA, IOWA CITY, IA 52242, USA

I. INTRODUCTION II. THE GENERAL THEORETICAL FRAMEWORK III. THE DEVELOPMENT OF SPATIAL CATEGORIZATION A . USING SPATIAL CATEGORIES TO ORGANIZE RECALL B . USING SPATIAL CATEGORIES TO REMEMBER OBJECT POSITIONS IV. EXPLAINING THE EMERGENCE OF SPATIAL CATEGORIZATION SKILLS A . THE EMERGENCE OF SPATIAL CATEGORIZATION SKILLS IN THE MOMENT B . THE EMERGENCE OF SPATIAL CATEGORIZATION SKILLS OVER DEVELOPMENT V. LIMITS AND CONCLUSIONS ACKNOWLEDGEMENTS REFERENCES

I. Introduction Traditional approaches to cognition view thinking as something that sits exclusively within a person’s head. An implicit assumption of this view is that thinking can occur independent of the physical, social, or task environment. Yet this view is at odds with the plethora of data showing wide variations in thinking across different environmental contexts, from controlled laboratory experiments to naturalistic observation studies (e.g., Correa-Chavez, Rogoff, & Arauz, 2005; Oakes, Plumert, Lansink, & Merryman, 1996; Plumert & Strahan, 1997; Samuelson, Schutte, & Horst, 2007). Even what appear to be small variations in task instructions (e.g., Plumert & Strahan) or minor differences in response measures (e.g., Huttenlocher et al., 2008) can lead to rather large differences in performance. 373 Advances in Child Development and Behavior R.V. Kail : Editor

Copyright r 2008 Elsevier B.V. All rights reserved.

374

Jodie M. Plumert

One response to this dilemma is to conclude that certain performance demands or experimental contexts ‘‘mask’’ children’s true abilities (Baillargeon, 2001; Sophian, 1997; Spelke, 2000; Spelke & Newport, 1998). The problem with this view is that there is a never-ending search for the ‘‘right’’ task to tap underlying cognitive competencies (i.e., the one that results in the best performance at the youngest age). Another response to the problem is to characterize children’s abilities as emerging in pieces or steps over development, with rudimentary aspects of skills emerging in simpler task contexts at earlier points in development (e.g., Huttenlocher et al., 2008; Woolley, 2006). This approach takes seriously the problem of identifying the task components that lead to different response patterns in different test situations. At its core, however, this approach shares the assumption that the environment sits outside of the thinking process. The goal of this chapter is to attempt (once again) to move the ﬁeld away from the separation of the organism and the environment by focusing on the idea of emergence. The central idea is that thinking is not just about what is in the child’s head. Rather, thinking emerges out of the interaction of the cognitive system and environmental structure (i.e., the physical, social, or task environment). From this perspective, the environment is not just an ‘‘inﬂuence’’ on thinking or development. Rather, the child and the environment are part of a uniﬁed system. In this chapter, I illustrate these ideas by reviewing two lines of research on the development of spatial categorization skills. These two programs of research focus on children’s use of spatial categories to organize their recall and to remember object positions. The goals of the chapter are twofold. One is to describe how age changes in children’s spatial categorization skills depend both on the child and the environment. The other, more difﬁcult goal is to speculate about how these changes come about.

II. The General Theoretical Framework The notion that thinking emerges from the interaction of the organism and the environment has been central to several major theories over the last century, including those of Piaget (1954), Vygotsky (1978), Thelen and Smith (1994), Gottlieb (Gottlieb & Lickliter, 2007), and the Gibsons (J. J. Gibson, 1979; E. J. Gibson, 1988). In this chapter, I focus primarily on ideas about organism–environment interaction from J. J. and E. J. Gibson. A critical concept introduced by the Gibsons is the complementarity of the organism and environment. In other words, possibilities for action (i.e., affordances) depend on both the characteristics of the organism and the structure of the environment (e.g., water offers a surface of support for

Understanding the Organism and Environment as a Unified System

375

a water bug but not for a human). Within the ecological perspective, this concept of the mutuality between the organism and the environment has mainly been applied to understanding perception and action (e.g., Adolph, 2000; Gibson & Pick, 2000; Lockman, 2000; Plumert, Kearney, & Cremer, 2004, 2007b; Rieser et al., 1995; Warren, 1984). Thus, changes in the environment and changes in the organism (or both) lead to changes in possibilities for action. For example, Adolph and her colleagues (Adolph, 1997, 2000; Adolph, Eppler, & Gibson, 1993; Eppler, Adolph, & Weiner, 1996) have shown that toddlers’ decisions about whether to descend a slope depend both on walking skill (a characteristic of the organism) and on the steepness of the slope (a property of the environment). Changes in walking skill and changes in the steepness of the slope fundamentally alter the interaction between the perceiver and the environment, leading to changes in possibilities for action. We can see this experimentally by engineering either the organism (e.g., adding heavy or light weights to the child’s backpack) or the environment (e.g., changing the steepness or slipperiness of the slope). In this chapter, I expand this view of perception to the domain of cognition. In other words, perceiving, acting, and thinking emerge out of the interaction of the characteristics of the organism and the characteristics of the environment. Moreover, I argue that this view of organism– environment interaction provides a particularly good framework for conceptualizing how spatial thinking emerges over time. What does it mean to say that thinking is a joint function of the characteristics of the organism and the structure of the environment? Put simply, thinking emerges out of interactions between the organism and the environment that take place in the context of solving problems. Thus, to fully understand any behavior both in the moment and over development, we cannot simply examine the characteristics of the organism or what the environment offers the organism. Rather, we must understand how the two interact at any given point in time and how these organism–environment interactions change over time. This view necessarily implies that thinking (like perceiving and acting) is a dynamic process in which changes in the organism or the environment (or both) alter the nature of the interaction, resulting in changes in thinking. From this perspective, cognition is not something that sits in the head of the organism. Rather, thinking is an emergent product of a system that includes both the organism and the environment. An important consequence of this view is that neither the organism nor the environment has causal priority for explaining behavior either in the moment or over development. Organisms cannot perceive, move, or think independent of environmental structure, and environmental structure has no meaning independent of the characteristics of the organism.

376

Jodie M. Plumert

In ecological terms, organisms use the available information in the environment to guide thinking, but what is ‘‘available’’ is constrained by the characteristics of the organism. Thus, the functional value of environmental structure is constrained by the cognitive system (e.g., informationprocessing skills and background knowledge). Likewise, informationprocessing skills and background knowledge can only function in the context of environmental structure. In other words, thinking can only happen as the organism and the environment work together as a uniﬁed system. Like possibilities for action, possibilities for thought (e.g., solutions to problems) are created in the moment based on what the cognitive system and the environment bring to the table. This necessarily means that we need to understand both the characteristics of the cognitive system (an endeavor traditionally left up to the ﬁeld of information processing) and the structure available in the environment (the physical, social, and task environment) for guiding thinking. What are the implications of this view for understanding changes in spatial thinking over longer developmental time scales? From an ecological perspective, the key to understanding developmental change is to specify how experience leads to changes in the organism–environment interaction (E. J. Gibson, 1988; Gibson & Pick, 2000). Like Piaget’s concepts of assimilation and accommodation or Vygotsky’s ideas about scaffolding and the zone of proximal development, this view suggests that there is a cyclical quality to organism–environment interaction over both shorter and longer time scales. That is, changes in the organism lead to changes in the information that is available, thereby allowing the organism to experience the environment in a new way. In turn, these new experiences lead to further changes in the organism at both neural and behavioral levels. Thus, interaction with environmental structure is necessary to produce changes in the organism, but the structure that is ‘‘available’’ (i.e., can be experienced) is constrained by the characteristics of the organism. In the past, research from an ecological perspective has focused on how changes in the action capabilities of the organism lead to changes in the amount or type of perceptual information that is ‘‘available,’’ and how experiences with using new perceptual information to guide action lead to further changes in the organism (Adolph, 1997; Gibson & Gibson, 1955). I argue that this developmental framework is also relevant for thinking about how cognitive change occurs. In particular, changes in cognitive skills (e.g., attention, memory, or strategy use) lead to changes in the amount or type of information that is available for solving speciﬁc problems. Experience with using new information to solve speciﬁc problems leads to further changes in cognitive skills. For example, experience using salient environmental structure (e.g., physical barriers that separate locations

Understanding the Organism and Environment as a Unified System

377

into regions) to organize searches for objects may lead to improvements in children’s spatial clustering strategies. In turn, increased use of spatial clustering strategies might sensitize children to more subtle environmental structure (e.g., perceptual boundaries that separate locations into regions) to organize their searches for objects. As this example illustrates, the developmental changes we see in children’s spatial thinking come about through recurrent organism–environment interactions that alter how the cognitive system interacts with environmental structure to solve everyday problems. Note that this framework for understanding developmental change in children’s spatial thinking has much in common with Vygotsky’s notions about scaffolding and the zone of proximal development. The basic premise underlying this approach to cognitive development is that children often acquire knowledge and skills through social interaction with more skilled individuals. Adult guidance of cognitive performance is thought to be particularly important during times of transition, sometimes referred to as the zone of proximal development (Vygotsky, 1978). During such times, children are sensitive to experiences that allow them to try out new ways of thinking and acting. More speciﬁcally, children are in a state of readiness to beneﬁt from guidance that provides them with the necessary support to use their skills in novel ways. Over time, scaffolding can be modiﬁed or withdrawn as the child becomes increasingly competent at executing the skill. Developmental change results as responsibility for structuring cognitive performance shifts from the adult to the child. Subsequent reformulations of Vygotsky’s (1978) contextual approach to cognitive development stress the notion of ‘‘guided participation’’ as a vehicle for cognitive change (Gauvain, 2001; Rogoff, 1990). Guided participation emphasizes both active participation of children and guidance from others as contributors to the process of change. According to Rogoff (1990), older, more experienced individuals such as parents capitalize on children’s eagerness to learn by providing guidance that will advance children’s skills and understanding. From this perspective, the primary task for the adult is to provide guidance that is appropriately geared to the developmental level of the child. That is, adults must provide guidance that supports, and yet challenges children’s skills and understanding. The concept of scaffolding has been almost exclusively applied to the role of social interaction in promoting cognitive change. However, these notions about scaffolding can be expanded to encompass to the physical context and the task context. In other words, highly supportive structure in the physical or task environment provides young children with the scaffolding necessary to execute particular skills. Repeated experience with supportive physical or task environments allows young children to reﬁne their skills.

378

Jodie M. Plumert

Over time, children begin to use these same skills in less supportive physical and task environments. Again, it is important to point out that scaffolding from the social, physical, or task environment alone does not produce developmental change. Environmental structure cannot inﬂuence the child’s thinking unless the child has the necessary background knowledge or processing capabilities to notice this structure. Likewise, the child cannot develop new background knowledge or reﬁne her processing capabilities without exposure to environmental structure. Importantly, this perspective on developmental change shifts the focus from only considering the changes that occur within the child or the environmental factors that impact performance to considering how organism–environment interactions change over development. I now turn to considering empirical examples of organism–environment interaction drawn from our work on the development of spatial cognition. The goal here is to provide examples of how one must simultaneously consider both what the child (or adult) brings to the situation and what the environment provides in order to construct a coherent account of the processes underlying thinking in the moment and changes in thinking over time. Note that I draw both on concepts about the characteristics of the cognitive system from mainstream information-processing approaches and on ideas about the structure of the environment from a traditional ecological approach. Although the two approaches are rarely considered together, ideas from both lead to a much richer picture of spatial thinking and development than does either one alone.

III. The Development of Spatial Categorization The ability to organize locations within some kind of spatial structure is fundamental to the process of learning and remembering locations. Without this structure, remembering the locations of the countless objects and places encountered everyday would become insurmountable. By organizing knowledge of location, memory demands are lessened because locations are not treated as isolated pieces of spatial information. In this section, I review two lines of research on the development of spatial categorization. The ﬁrst concerns how children come to use spatial categorization as an organizational device to enhance recall of objects and locations. The second line of research concerns the role that spatial categorization plays in remembering object positions. Both lines of research serve as illustrations of how thinking emerges out of the interaction of the cognitive system and environmental structure.

Understanding the Organism and Environment as a Unified System

379

A. USING SPATIAL CATEGORIES TO ORGANIZE RECALL In general, spatial clustering refers to grouping objects or locations on the basis of common membership within a spatial region. Typically, spatial regions are deﬁned by physical or perceptual boundaries, or by proximity to salient landmarks (Huttenlocher & Lourenco, 2007; McNamara, 1986). For example, one might think of a table, stool, and refrigerator as belonging together because they are all located in the kitchen, or one might think of a couch, rug, and rocking chair as belonging together because they are all located near the ﬁreplace. We might also think of a region as a cluster of nearby objects or places such as pieces of play equipment on a schoolyard or a group of university buildings on campus. In the real world, cues for forming spatial groups typically overlap. For example, a spatial group might be deﬁned by perceptual boundaries, salient landmarks, and close proximity of objects. These cues can function as the basis for spatial organizational strategies for retrieving or recalling sets of objects. When recalling objects from home, for example, one might retrieve or recall objects in the kitchen, then the objects in the living room, then the objects in the laundry room, and so on.

1. The Development of Spatial Organizational Strategies A large body of research indicates that spatial clustering strategies do not develop in an all-or-none fashion. Rather, more sophisticated use of spatial clustering strategies emerges gradually over childhood, starting during the preschool years. A major part of this development is using spatial clustering strategies in increasingly complex tasks. One of the ﬁrst manifestations of spatial clustering is young children’s tendency to retrieve the items in one spatial region before retrieving items in another spatial region (Cornell & Heth, 1986; Haake, Somerville, & Wellman, 1980; Plumert et al., 1994; Wellman et al., 1984). For example, Wellman et al. (1984) found that 4- and 5-year-olds minimized the number of traverses they made between two clusters of locations while retrieving Easter eggs they had previously seen hidden in ﬁve buckets on a playground. Similarly, Cornell and Heth (1986) found that both 5- and 7-year-olds hid objects in spatial clusters and tended to search those clusters exhaustively when later retrieving the objects. These results suggest that the ability to use spatial organization to guide physical activity emerges fairly early in development. Given young children’s use of spatial clustering to guide the organization of their own movements, to what extent do they use spatial clustering to guide the organization of another person’s movements? We examined this question by comparing the order in which 6-year-olds and adults searched

380

Jodie M. Plumert

for objects with the order in which they referred to object locations in their route directions (Plumert et al., 1994). The basic paradigm involved having children and adults help an experimenter hide nine small tokens along an inefﬁcient random route on the three levels of their home. After all the tokens were hidden, participants were asked to go ﬁnd them all again or to tell another experimenter how to go ﬁnd all those pieces through a walkietalkie. Comparison of the routes children produced in their searches and verbal directions showed that they used an efﬁcient order to retrieve the objects themselves (i.e., organized their searches by ﬂoor), but directed the other person to the locations in an order that resulted in a very inefﬁcient search. Adults, in contrast, both searched efﬁciently and gave efﬁcient directions. Interestingly, although children did not spontaneously order the locations by ﬂoor in their directions, they demonstrated that they had the necessary spatial knowledge to do so because they were able to give spatially efﬁcient directions if prompted after each direction to tell the listener where the next closest object was. This suggests that young children were able to use visual cues and the actual physical structure of the layout to help them organize their searches, but were unable to mentally access that structure to produce spatially organized directions unless their listener provided an explicit organizational framework. In sum, 6-year-old children exhibited a marked difference in their use of spatial clustering to organize their searches and directions, even when all other aspects of the situation were equated. These results naturally raise the question of when do children begin to use spatial clustering in their verbal directions? We investigated this question in a study comparing the organization of children’s free recall and tour plans (Plumert & Strahan, 1997). Note that although tour planning and direction giving are not the same tasks, both require the person to think about the order in which another person will visit or search a set of locations. In this study, 6-, 8-, and 10-year-old children ﬁrst helped an experimenter hide 16 unrelated objects in a 4-room dollhouse (see Figure 1). After hiding the objects the ﬁrst time, children were asked to turn around so

Fig. 1. The four-room dollhouse used as the experimental space.

Understanding the Organism and Environment as a Unified System

381

that they faced away from the dollhouse. The experimenter then removed the objects from the dollhouse. After children turned to face the dollhouse again, the experimenter handed them the objects in a different random order and asked them to put the objects back exactly where they were before. The experimenter immediately corrected any placement mistakes. This procedure was repeated until children reached the criterion of correctly replacing all 16 objects in a single trial. After children completed the location learning procedure, the experimenter placed an opaque cover over the entire dollhouse. Children in the free recall condition then were instructed to name as many of the hidden objects as they could remember. Children in the tour-planning condition were then asked to plan a tour of the hidden objects for a doll ﬁgure. The experimenter instructed children to ‘‘tell me which one you would show him ﬁrst, and second, and so on.’’ Thus, all aspects of the experiment were the same except for the task used to get children to name the objects. We calculated Adjusted Ratio of Clustering (ARC) scores to assess the degree of spatial clustering (i.e., ordering the objects by room) in children’s tour plans and free recall (Roenker, Thompson, & Brown, 1971). As shown in Figure 2, 6-year-olds showed equally low levels of spatial clustering in both their tour plans and free recall. The difference between 8-year-olds’ ARC scores in the tour plan and free recall conditions approached conventional levels of signiﬁcance. Thus, 8-year-olds in the tour-planning condition showed somewhat higher levels of spatial clustering than did their counterparts in the free recall condition. Finally, 10-year-olds in the tourplanning condition had signiﬁcantly higher spatial clustering scores than did 10-year-olds in the free recall condition. The results of this investigation clearly show that the task context plays a major role in children’s use of spatial clustering. Speciﬁcally, children’s use of spatial clustering in the tour-planning task increased gradually between the ages of 6 and 10. At none of the ages tested, however, did children spontaneously use spatial clustering in the standard free recall task. Again, it is important to point out that all aspects of the experiment were the same up to the point children were given the task instructions, indicating that spatial clustering emerged out of the interaction of the child and the task. When do children use spatial clustering to organize their free recall of object names? I addressed this question by asking 10-, 12-, 14-, 16-year-olds, and adults to recall the furniture from their home (Plumert, 1994). Making a furniture inventory from memory is a particularly interesting task for investigating organizational strategies because furniture items can be grouped either by spatial region (e.g., kitchen, living room, bedroom, laundry) or by object category (e.g., tables, chairs, beds, dressers). Furthermore, because adults and children have repeated experience with their

382

Jodie M. Plumert

Fig. 2. Mean spatial clustering scores in the tour-planning and free recall conditions by age.

furniture, both the locations and types of furniture in their home are very well known to them. Quite surprisingly, analysis of spatial and categorical ARC scores revealed that 10-year-olds signiﬁcantly preferred to organize their furniture by category than by room (see Figure 3). In contrast, the 16-year-olds and adults showed a strong preference for spatial over categorical organization. The 12- and 14-year-olds showed about equal use of both organizations. This pattern of results clearly shows that children’s use of spatial clustering strategies to organize their recall of object names changes between 10 and 16 years of age. Speciﬁcally, 10-year-olds seem to have difﬁculty using their spatial clustering strategies to structure their recall of object names. Twelve- and 14-year-olds may be moving into a more transitional age in which there are large individual differences in spatial and categorical organization. This idea is supported by the fact that there were strong correlations between spatial organization and number of furniture items recalled for these ages. By 16 years of age, adolescents are clearly able to use what according to adult standards appears to be the most appropriate strategy for the task of recalling furniture from their home.

Understanding the Organism and Environment as a Unified System

383

Fig. 3. Mean spatial and categorical clustering scores in the furniture inventory task by age.

This unanticipated developmental change in preferences for categorical and spatial organization raised the question of whether these preferences could be pushed around by manipulating the nature of the task. In a second experiment, 10- and 12-year-olds learned the locations of 16 toy objects that were divided into four object categories (vehicles, animals, clothing, and furniture). The objects were hidden in four rooms, with one of each type of object in each room. After hiding all the objects, children were taken to each location again in the same order as they hid the objects to give them a second chance to learn the object-location pairings. After seeing the objects the second time, children were taken to a separate testing room and given two recall tasks. The ﬁrst task was to recall the toys they had hidden. The second task was to recall each object and where it was hidden. As expected, both 10- and 12-year-olds exhibited more categorical than spatial clustering when recalling just the objects. However, the 12-year-olds exhibited more spatial than categorical clustering when recalling the objects and their locations. Thus, 12-year-olds were more sensitive to the changing nature of the recall task than were 10-year-olds. Again, these results underscore the

384

Jodie M. Plumert

point that thinking emerges out of the interaction of the child’s information-processing skills and the structure of the task. 2. Summary The work described here clearly shows that children’s use of spatial organizational strategies differs by age and task. At the youngest ages, children use spatial clustering strategies to organize their searches for objects, but not their directions for ﬁnding objects or tour plans for viewing objects. At intermediate ages, children use spatial clustering to organize their directions and tour plans, but they do not use spatial clustering to organize their free recall of objects. At the oldest ages, children use spatial clustering in all tasks, including free recall. Together, these studies clearly illustrate that spatial clustering strategies do not reside in the head. Rather, they are an emergent property of a system that includes both the cognitive system and the environment.

B. USING SPATIAL CATEGORIES TO REMEMBER OBJECT POSITIONS The work described in the previous section clearly shows that spatial categories play an important role in organizing recall of objects and locations. As with other memory strategies, spatial clustering serves an important function of enhancing the number of items one can recall. But does forming a spatial category (e.g., grouping furniture items by room or ﬂoor) also play a role in remembering object positions? In particular, do people think that locations from the same spatial category are closer together than they really are? 1. Making Judgments about Locations One of the ﬁrst hints that spatial categories play a role in memory for location came from adult priming studies in which people read the names of two objects presented one after the other and then tried to make a judgment about the second object as quickly as possible. For example, people might be asked to judge whether or not the second object was present in the collection of objects in a layout they had previously learned. The rationale behind this approach is that if locations from the same spatial region are closely associated in memory, then the time required to respond to an object should be faster if it is preceded by the name of an object from the same region than from a different region. Indeed, a number of these spatial priming studies have shown that adults are faster to respond to an object if

Understanding the Organism and Environment as a Unified System

385

it is preceded by the name of an object from the same region than a different region (e.g., McNamara, 1986; McNamara, Hardy, & Hirtle, 1989). This occurs even when the object from the different region is physically closer to the target object than is the one from the same region as the target object. These ﬁndings suggest that people remember objects from the same region (i.e., spatial category) as closer together than they really are. The errors adults make when judging spatial relations also underscore the importance of spatial categories in memory for location. For example, Seattle is usually judged to be farther south of Montreal, when in fact it is farther north (Friedman & Brown, 2000a; Montello, 2003; Stevens & Coupe, 1978). Presumably, this error occurs because individuals rely on the north– south relations between the larger geographic regions to judge spatial relations between locations contained within those regions. Similar studies with children have also shown that spatial categories exert a powerful inﬂuence on their memory for locations (e.g., Acredolo & Boulter, 1984; Allen, 1981; Kosslyn, Pick, & Fariello, 1974; Newcombe & Liben, 1982). When asked to make spatial judgments about individual locations belonging to different spatial regions, for example, even 6-year-olds tend to rely on the overall spatial relations between regions rather than on the actual spatial relations between the individual locations (Acredolo & Boulter, 1984). Similarly, Allen (1981) found that 7- and 10-year-olds and adults tend to partition routes into subdivisions, and use these subdivisions to make distance judgments about locations along the route. In particular, children and adults often judged locations from two adjoining subdivisions as more distant than locations within the same subdivision even when the locations within the same subdivision were more physically distant than the locations from adjoining subdivisions. Together, these studies show that spatial categories play an important role in judgments about spatial relations.

2. Reproducing Previously Seen Locations Subsequent work focused on whether children and adults show similar kinds of biases when physically reproducing previously seen locations (e.g., Engebretson & Huttenlocher, 1996; Huttenlocher, Hedges, & Duncan, 1991; Huttenlocher, Newcombe, & Sandberg, 1994; Sandberg, Huttenlocher, & Newcombe, 1996). Most of these studies involve brieﬂy showing children and adults single locations in a homogenous space and then asking them to reproduce those locations in the same space. For example, in the sandbox task developed by Huttenlocher, Newcombe, and colleagues, children between the ages of two and 10 years watched an experimenter hide a toy in a long, narrow sandbox (Huttenlocher et al., 1994). After a short delay in which children were turned away from the sandbox, they were allowed to

386

Jodie M. Plumert

search for the toy. Analysis of search patterns indicated that all age groups exhibited systematic biases toward the region centers. Speciﬁcally, 2- and 6-year-olds’ searches were biased toward the center of the entire sandbox, and 10-year-olds’ searches were biased toward the centers of the two halves of the sandbox. Older children and adults show similar biases in tasks where they are asked to reproduce the location of a dot inside a circle or a line inside a 901 angle. In particular, older children and adults place locations closer to the centers of the circle quadrants and to the halves of the 901 angle than they actually are (Engebretson & Huttenlocher, 1996; Sandberg et al., 1996). Together, these studies have repeatedly shown that children and adults alike show systematic biases toward the centers of geometric regions, supporting the idea that spatial categories play an important role in memory for location.

3. Categorical Bias as an Emergent Product of the Organism–Environment System Our work on spatial category effects in memory for location focuses on understanding how ‘‘decisions’’ about where to place objects emerge out of the interaction of available environmental structure and the cognitive processes involved in coding, maintaining, and retrieving spatial information. Our basic task involves a learning phase and a test phase. Participants ﬁrst learn the locations of 20 miniature objects marked by dots on the ﬂoor of an open, square box (approximately 3 ft long 3 ft wide 12 in. high) placed on the ﬂoor of a laboratory room (see Figure 4). We typically

Fig. 4. Box used as the experimental space. The left panel shows the floor of the box with boundaries during learning. The right panel shows the floor of the box at test.

Understanding the Organism and Environment as a Unified System

387

provide structure during learning (e.g., boundaries subdividing the box into quadrants) so that the locations are organized into four groups of ﬁve locations. Participants ﬁrst watch while the experimenter names the objects and places them one at a time on the dots until all 20 objects have been placed. The experimenter then gives the objects to participants one at a time and asks them to try to place them on the correct dots. The test phase begins after participants reach a learning criterion of placing all the objects correctly in a single trial. During test, participants attempt to place the objects in the correct locations without the aid of the dots marking the locations and other structure organizing the locations (e.g., boundaries). It is important to note that participants are given no foreknowledge of the test prior to this point in the session. We record the x and y coordinates of each object to obtain a precise measure of where participants placed the objects. Our primary measures are mean and variable error (computed based on the absolute distance from the correct locations) and center displacement (the degree to which people place the objects belonging to the same spatial group closer together than they actually are). a. Where Does Bias Come From? Systematic bias in memory for location is seen as an important signature of the underlying processes involved in reproducing previously seen locations (Huttenlocher et al., 1991; Plumert, Hund, & Recker, 2007a; Spencer et al., 2007). A key question is where does this bias come from? According to the categoryadjustment model proposed by Huttenlocher et al. (1991), retrieval of locations from memory is a hierarchical process involving the use of both ﬁne-grained and categorical information. When trying to remember a previously learned location, people make estimates based on their memory of ﬁne-grained, metric information such as distance and direction from an edge. However, because memory for ﬁne-grained information is inexact, people adjust these estimates based on categorical information about the location represented by a prototype located at the center of the spatial region or group. Hence, adjustments based on categorical information lead to systematic distortions toward the centers of spatial categories. According to this model, the magnitude of distortion toward category centers depends on the certainty of the ﬁne-grained, metric information. When memory for ﬁne-grained information is relatively certain, categorical information receives a low weight, resulting in only small distortions toward category centers. Conversely, when memory for ﬁne-grained information is relatively uncertain, categorical information receives a high weight, resulting in large distortions toward category centers. The end result of such systematic bias is that responses are less variable, leading to greater overall accuracy.

388

Jodie M. Plumert

Subsequently, Spencer, Scho¨ner, and colleagues developed the dynamic ﬁeld theory of spatial memory to account for spatial biases that often occur when people attempt to reproduce single locations after short delays (Johnson, Spencer, & Scho¨ner, in press; Simmering, Schutte, & Spencer, in press; Spencer et al., 2007). The dynamic ﬁeld theory is a neural network model that captures how location-related activation in a network of neurons can be sustained from moment-to-moment and drift over short time periods. Brieﬂy, the model consists of several interconnected layers (i.e., ﬁelds). These layers include perceptual, working memory, and longterm memory ﬁelds, as well as inhibitory interneurons. The perceptual ﬁeld forms peaks of activation generated by input from perception of visible reference frames and the current target’s (visible) location. The perceptual ﬁeld passes activation about both the reference frame (e.g., an axis) and the target location to the working memory ﬁeld. The working memory ﬁeld passes self-sustained activation on to an associated long-term memory ﬁeld. This ﬁeld accumulates traces of activation representing the locations of other previously seen targets, with stronger traces associated with more frequently seen targets. The long-term memory ﬁeld also passes activation back to the working memory ﬁeld. Drift over time (i.e., bias) can occur through the interaction of the working memory and long-term memory ﬁelds, producing bias toward frequently remembered targets (Spencer & Hund, 2002, 2003). According to the dynamic ﬁeld model, spatial biases emerge in a primarily bottom-up fashion out of moment-by-moment interactions among multiple components of the system. The fact that we ask participants to remember the locations of 20 objects at one time makes our task considerably more complex than the tasks used to test the category adjustment and dynamic ﬁeld theory models. As in both of these models, we assume that children and adults code ﬁne-grained, metric information about the precise location of each object. Remembering the precise location of each object is necessary for distinguishing nearby locations from each other and for reproducing locations in an accurate manner. We also assume that children and adults code coarse-grained, categorical information about the group or region to which each location belongs. Clearly, remembering the group to which each location belongs is useful for reducing the demands of remembering 20 individual locations. Quite likely, coding spatial groups in our task involves the contribution of top-down, strategic processes that results in forming strong associations among the members of the spatial groups. We assume that categorical bias in placements reﬂects the ‘‘pull’’ of memory for the spatial groups. Speciﬁcally, when memory for the spatial groups (i.e., associations among locations in the spatial groups) is strong relative to memory for the individual locations, people place the objects closer together than they

Understanding the Organism and Environment as a Unified System

389

really are. Conversely, when memory for the individual locations is strong relative to memory for the spatial groups, people exhibit little or no categorical bias in their placements. A major question is what governs the strength of memory for ﬁnegrained and categorical information? From a traditional perspective, patterns of bias depend solely on how the cognitive system codes, maintains, and retrieves ﬁne-grained and categorical information. At most, the environment plays a supporting role in providing cues for encoding and retrieving information. From an ecological perspective, however, environmental structure and the cognitive system are inextricably linked as part of a complete system. That is, patterns of bias emerge out of the interaction of structure available in the task and the characteristics of the cognitive system. Hence, both differences in the cognitive system and differences in the available perceptual structure can alter the interaction, leading to changes in the pattern of categorical bias. For example, we might expect to see more categorical bias when multiple cues are available to code the spatial groups during learning. Paralleling perception-action research (e.g., Adolph, 1997; Plumert et al., 2004), experimental manipulations of either environmental structure (e.g., imposing boundaries that divide locations into groups) or the cognitive system (e.g., strengthening ﬁne-grained memory through repeated opportunities for learning) can reveal the nature of these underlying interactions that govern object placements. Across multiple experiments, our goal has been to examine how bias in placements varies in response to manipulations of environmental structure while children and adults are coding and reproducing sets of locations. We have examined how categorical bias emerges out of interactions of task structure and coding processes by providing cues for organizing the locations into groups during learning. In particular, we have examined how children and adults use visible boundaries subdividing the space, experience with visiting nearby locations close together in time, and categorical relatedness between objects occupying the same region to organize the locations into groups, leading to systematic variations in categorical bias at test (Hund & Plumert, 2003, 2005; Hund, Plumert, & Benney, 2002; Plumert & Hund, 2001). Likewise, we have examined how categorical bias varies in response to interactions of task structure and retrieval processes by varying the available perceptual structure at test (Plumert & Hund, 2001; Recker, Plumert, & Stevens, 2007a). Again, our focus is on using experimental manipulations of task structure to understand how interactions between the cognitive system and task structure produce systematic changes in decisions about where to place objects (i.e., categorical bias). This contrasts with a more traditional focus on using experimental

390

Jodie M. Plumert

manipulations of task structure to understand aspects of the cognitive system itself (e.g., using precues to understand how attention operates). We have chosen to study 7-, 9-, 11-year-old children and adults because we hypothesize that important developmental changes are occurring in the cognitive system during late childhood and early adulthood. These developmental changes fundamentally alter the interaction between the cognitive system and the task structure because they lead to differences in the amount and kind of information that is ‘‘available’’ for use. First, we hypothesize that the precision of ﬁne-grained, metric coding is improving between the ages of 7 and 11 years (and possibly between 11 years and adulthood). In virtually every study that we have conducted to date, younger children exhibit greater mean and variable error than do older children and adults (see also Hund & Spencer, 2003; Spencer & Hund, 2003). The hypothesized increase in the precision of ﬁne-grained coding likely depends on recurrent organism–environment interactions that occur as children repeatedly face the problem of localizing objects, thereby leading to increasing sensitivity to information about distance and direction (for related ideas, see Schutte & Spencer, 2002; Schutte, Spencer, & Scho¨ner, 2003; Spencer & Hund, 2002, 2003). Second, the use of spatial clustering strategies is also increasing across this age range. As discussed previously, adults readily use spatial clustering strategies to help them recall both objects and locations (Plumert, 1994). Children’s use of spatial clustering undergoes signiﬁcant change across childhood, not reaching adultlike performance until early to mid adolescence. As a result, adults may form much stronger associations among the locations within a spatial group than do children. These stronger associations increase the ‘‘pull’’ from the spatial groups, thereby increasing the likelihood of bias in placements. In the next sections, I illustrate how categorical bias in placements emerges out of the interaction of the organism and the environment in the context of our location memory task. b. Coding Locations: How do Cues for Forming Spatial Groups Influence Categorical Bias? We have carried out several studies examining how the availability of cues for forming spatial groups during learning affects categorical bias at test (Hund & Plumert, 2003, 2005; Hund et al., 2002; Plumert & Hund, 2001). We are especially interested in how the structure available for organizing the locations into groups (e.g., visible boundaries, spatiotemporal experience) interacts with characteristics of the cognitive system (i.e., age-related changes in the coding of ﬁne-grained and categorical information) to produce particular patterns of categorical bias. Thus far, we have looked at three types of cues for forming spatial groups: visible boundaries subdividing the space, experience with visiting

Understanding the Organism and Environment as a Unified System

391

nearby locations close together in time, and categorical relations between objects occupying the same region. I. VISIBLE BOUNDARIES. Visible boundaries that divide locations into groups are perhaps the most obvious source of perceptual structure for forming spatial groups. As discussed previously, numerous studies have shown that both children and adults tend to underestimate distances between locations within the same region and overestimate distances between locations in different regions.(Allen, 1981; Kosslyn et al., 1974; McNamara, 1986; Newcombe & Liben, 1982). In our initial work on categorical bias in estimates of location, we examined how boundary salience during learning inﬂuenced categorical bias at test (Plumert & Hund, 2001). Seven-, 9-, 11-year-olds and adults learned the locations of 20 unrelated objects in a random order. In the walls condition, interior walls the same height as the exterior walls divided the box into four quadrants. In the lines condition, lines on the ﬂoor divided the box into four quadrants. In the no boundaries condition, no visible boundaries were present. After participants reached the learning criterion, the test phase began. The experimenter removed the dots marking the locations and any boundaries subdividing the space. Participants then attempted to place the objects in the correct locations. To what extent did children and adults in each boundary condition place the objects belonging to each group closer together than they actually were? As expected, participants exhibited greater categorical bias when boundaries were present during learning than when they were not present. In addition, they exhibited more categorical bias when more salient boundaries were present during learning than when less salient boundaries were present. Adults and 11-year-olds in the walls condition and adults in the lines condition placed the objects signiﬁcantly closer together than they really were (see Figure 5). In the no boundaries condition, however, children and adults showed very little categorical bias. Thus, when no cues were available to organize the locations into groups, even adults had difﬁculty forming strong associations among the locations within each spatial group. What do these results tell us about how the cognitive system and environmental structure interact to produce patterns of categorical bias? As Figure 5 shows, all age groups responded to boundary salience. Categorical bias was always highest in the walls condition, intermediate in the lines condition, and lowest in the no boundaries condition. This clearly shows that the salience of perceptual structure during learning affected categorical bias at test. We hypothesize that more salient boundaries helped children and adults create stronger associations among the locations in the spatial

Jodie M. Plumert

392

0.8 No Boundaries Lines Walls

0.6

*

* *

0.4 0.2 0 -0.2 -0.4 * -0.6

*

*

-0.8 7 Years

9 Years

11 Years

Adults

Age Fig. 5. Mean categorical bias in lines, walls, and no boundaries conditions by age. Asterisks denote significant results (po.05) of one-sample t-tests comparing the displacement score to the expected score with no displacement. Positive scores reflect bias toward the category centers.

groups as they were learning the locations. Stronger associations led to greater ‘‘pull’’ from the spatial groups when participants placed the objects at test. Note, however, that the magnitude of categorical bias in the three boundary conditions differed across the four age groups. This indicates that there were developmental differences in how the cognitive system interacted with the structure in the task. Unlike adults, children (with the exception of the 11-year-olds in the walls condition) did not place the objects in the spatial groups signiﬁcantly closer together than they really were. Subsequent studies also revealed that children often do not show signiﬁcant levels of categorical bias when only lines or walls divide the locations into groups (Hund & Plumert, 2002, 2003; Hund et al., 2002). Apparently, boundaries alone often are not sufﬁciently salient to help children form strong connections among the locations within the spatial groups. Without strong connections, children do not place objects closer together than they really are at test. These differences between children and adults underscore the idea that the extent to which children and adults make use of environmental structure is constrained by the characteristics of the

Understanding the Organism and Environment as a Unified System

393

organism. Even though children and adults were provided with the same perceptual structure during learning, adults were more able to make use of the organization than were children. Together, these ﬁndings highlight that understanding how the cognitive system and environmental structure operate as a complete system is necessary to fully explain behavior. II. EXPERIENCE WITH VISITING NEARBY LOCATIONS CLOSE TOGETHER IN TIME. Another cue that people can use to form spatial groups is spatiotemporal experience (Clayton & Habibi, 1991; Curiel & Radvansky, 1998; McNamara, Halpin, & Hardy, 1992; Sherman & Lim, 1991). Speciﬁcally, experience with visiting several locations close together in time may lead children and adults to form associations among those locations. For example, suppose a child has a paper route that involves delivering papers to the same set of houses on one side of town. This spatiotemporally contiguous experience may strengthen the relations among this particular set of houses. Not surprisingly, the child may come to think that these houses are much closer together than they really are. In many cases, such experiences with temporal contiguity are inﬂuenced by visible boundaries. In particular, physical barriers or perceptual boundaries guide locomotion (or decisions about locomotion) so that people usually visit sites on one side of a boundary before visiting sites on the opposite side. For example, the set of houses along the paper route may also be bordered by salient boundaries such as railroad tracks or major streets. We examined how children and adults use spatiotemporal experience and visible boundaries to remember locations by manipulating the order in which children and adults learned the locations in our spatial memory task (Hund et al., 2002). Seven-, 9-, 11-year-olds and adults learned 20 locations with lines subdividing the box into four quadrants. In the random learning condition, participants learned the locations in a random order (i.e., our standard learning procedure). In the contiguous learning condition, participants experienced the locations belonging to each quadrant together in time during learning. Participants ﬁrst watched while the experimenter placed all ﬁve objects in one quadrant, then placed ﬁve objects in another quadrant, and so on. During the subsequent learning trials, the experimenter handed participants the objects from one quadrant before moving on to another quadrant. Thus, participants placed the objects quadrant by quadrant during the learning phase of the experiment. The order of quadrants and the order of locations within quadrants were randomized for each learning trial. For both conditions, the experimenter removed the dots marking the locations and the boundaries subdividing the box prior to test.

Jodie M. Plumert

394

The primary question of interest was whether children and adults in each learning condition placed the objects belonging to each group closer together than they actually were. As shown in Figure 6, adults placed objects belonging to the same spatial group signiﬁcantly closer together than they really were in both the random and contiguous learning condition. In contrast, following random experience with the locations during learning, at no age did children place objects signiﬁcantly closer together than they actually were. In the contiguous learning condition, however, 9- and 11-year-olds placed objects belonging to the same spatial group signiﬁcantly closer together than they really were. Seven-year-olds showed a very similar pattern, but their center displacement scores in the contiguous learning condition did not differ signiﬁcantly from 0 due to high variability in their placements. The ﬁnding that adults exhibited categorical bias in both learning conditions whereas children only exhibited categorical bias in the contiguous learning condition again supports the idea that categorical bias

1 Random 0.9 Contiguous 0.8 *

0.7 0.6

* *

0.5

*

0.4 0.3 0.2 0.1 0 7 Years

9 Years

11 Years

Adults

Age Fig. 6. Mean categorical bias in random and contiguous conditions by age. Asterisks denote significant results (po.05) of one-sample t-tests comparing the displacement score to the expected score with no displacement. Positive scores reflect bias toward the category centers.

Understanding the Organism and Environment as a Unified System

395

emerges out of the interaction of task structure (e.g., spatiotemporal experience and visible boundaries) and the cognitive system. Adults easily formed strong associations among the locations within each group even when only a single cue (visible boundaries) organized the locations into groups. In contrast, children only formed strong associations among the locations within each group when two cues (visible boundaries and spatiotemporal contiguity) organized the locations into groups. Thus, age differences in the coding of ﬁne-grained and categorical information interacted with the structure provided in the task to produce different patterns of categorical bias. III. CATEGORICAL RELATIONS AMONG OBJECTS OCCUPYING NEARBY LOCATIONS. Another type of cue that people might use to form spatial groups is categorical relations among objects occupying nearby locations. In everyday environments, thematically or categorically related objects often are found together. For example, shoes, housewares, clothing, and cosmetics are typically located in different areas of a department store. Quite likely, this kind of structure helps people organize locations into groups. Such groupings are useful both for organizing retrieval (e.g., one is likely to retrieve items from one section before moving on to another section) and for cueing recall (e.g., one might try to recall needed items by thinking about which items are in each section). We examined whether categorical relations among nearby objects play a role in memory for location by manipulating whether the objects in each quadrant of the box were from the same object category (Hund & Plumert, 2003). In Experiment 1, children and adults learned the locations of 20 objects belonging to four categories: animals, vehicles, food, and clothing. In the related condition, objects belonging to the same object category were located in the same quadrant of the box. In the unrelated condition, the same objects and locations were used, but they were randomly paired. In both conditions, the experimenter gave the objects to participants in a random order on each learning trial. After participants reached the learning criterion, they attempted to place the objects in the correct locations without the aid of the dots marking the locations. Of particular interest was whether participants in the related condition would place the objects belonging to the same group closer together than would participants in the unrelated condition, suggesting that children and adults use information about objects to organize memory for locations. Overall, participants in the related condition placed the objects signiﬁcantly closer to the centers of the spatial groups than did participants in the unrelated condition. As shown in Figure 7, however, categorical bias in the related condition followed a U-shaped developmental pattern.

Jodie M. Plumert

396

0.9

Unrelated Related *

0.7

*

* 0.5

0.3

0.1

-0.1

-0.3 * -0.5 7 Years

9 Years

11 Years

Adults

Age Fig. 7. Mean categorical bias by age in related and unrelated conditions when no boundaries were present during learning. Asterisks denote significant results (po.05) of one-sample t-tests comparing the displacement score to the expected score with no displacement. Positive scores reflect bias toward the category centers.

Seven- and 9-year-olds and adults in the related condition placed the objects belonging to the same spatial group signiﬁcantly closer together than they actually were. In contrast, 11-year-olds in the related condition did not place the objects signiﬁcantly closer together than they actually were. In the unrelated condition, both children and adults showed very little categorical bias. In fact, 7-year-olds placed the objects signiﬁcantly further from the category centers than they actually were (they showed bias toward the corners of the box). Again, this shows that children and adults have trouble forming strong spatial groups in our task when no visible cues are available to organize the locations into groups. Why did the 11-year-olds in the related condition show only minimal categorical bias? One possibility is that their strong memory for the individual locations effectively counteracted the ‘‘pull’’ from their memory for the spatial groups. To test this possibility, we conducted a second experiment in which two categorical cues were present (i.e., object

Understanding the Organism and Environment as a Unified System

397

relatedness and visible boundaries), thereby increasing the strength of the spatial groups. All aspects of Experiment 2 were the same as in Experiment 1 except that visible boundaries divided the box into four quadrants during learning. We expected that 11-year-olds in the related condition would place objects belonging to the same group closer together than would their counterparts in the unrelated condition, suggesting that coincident cues (i.e., visible boundaries and object relatedness) lead to stronger associations among the locations in the spatial groups. As shown in Figure 8, the pattern of categorical bias in the unrelated condition followed a U-shaped pattern. Thus, when unrelated groups of objects were separated by boundaries, the magnitude of categorical bias followed a U-shaped developmental pattern similar to that seen when related objects were not separated by visible boundaries. In contrast, the pattern of categorical bias in the related condition no longer followed a U-shaped pattern. Instead, all age groups placed the objects belonging to

1 Unrelated Related 0.8

*

+ *

*

*

0.6 * 0.4

0.2

0

-0.2

-0.4 7 Years

9 Years

11 Years

Adults

Age Fig. 8. Mean categorical bias by age in related and unrelated conditions when boundaries were present during learning. Asterisks denote significant results (po.05) of one-sample t-tests comparing the displacement score to the expected score with no displacement. Positive scores reflect bias toward the category centers.

398

Jodie M. Plumert

the same spatial groups signiﬁcantly closer together than they really were. The ﬁnding that providing two coincident cues for coding the spatial groups (i.e., visible boundaries and object relatedness) erased the U-shaped pattern in categorical bias supports the claim that boosting the associations among the locations in the spatial groups changed the dynamics of the interaction. That is, strengthening the associations increased the ‘‘pull’’ of memory for spatial groups relative to memory for the individual locations, leading to increased categorical bias in 11-year-olds’ placements. The U-shaped developmental patterns of categorical bias seen in these experiments provide particularly compelling examples of organism– environment interaction because they illustrate how differences in the cognitive system and differences in environmental structure alter the interaction between the cognitive system and the environmental structure, leading to changes in the pattern of categorical bias. On the side of the cognitive system, there are age-related changes both in the coding of ﬁnegrained, metric information and in the coding of coarse-grained, categorical information. In all of our studies, adults exhibit signiﬁcantly less mean and variable error than do the younger children. By 11 years of age, coding of ﬁne-grained, metric information is nearly as good as that of adults. In contrast, strategic coding of the spatial groups appears to be undergoing change between 11 years of age and adulthood. Unlike children, adults form very strong associations among the locations in the spatial groups because they rely heavily on spatial clustering strategies to learn the locations. We hypothesize that adults exhibit strong categorical bias in their placements because their memory for the individual locations (though very good) cannot counteract the strong ‘‘pull’’ of the spatial groups. Elevenyear-olds often do not exhibit categorical bias in their placements because their strong memory for the individual locations effectively counteracts the weaker ‘‘pull’’ of the spatial groups. In contrast, 7- and 9-year-olds exhibit categorical bias in their placements because their relatively weak memory for the individual locations cannot counteract the ‘‘pull’’ from the spatial groups. Thus, the younger age groups exhibit categorical bias because their coding of the individual locations is relatively weak, whereas the adults exhibit categorical bias because their coding of the spatial groups is relatively strong. Together, these ﬁndings illustrate how characteristics of the cognitive system (e.g., age-related differences in the coding of ﬁnegrained and categorical information) and structure available in the task (e.g., types of cues available for coding the spatial groups) jointly determine patterns of categorical bias. c. Reproducing Locations: How Does the Available Perceptual Structure at Test Influence Categorical Bias? Thus far, I have discussed experimental

Understanding the Organism and Environment as a Unified System

399

manipulations designed to alter the interaction of the cognitive system and the task structure during learning. These ﬁndings leave open the question of how the cognitive system and the task structure interact when children and adults are in the process of replacing the objects during the test phase. We addressed this question by examining whether changing the available perceptual structure during the test phase inﬂuences categorical bias (Plumert & Hund, 2001). (Note that some of these data were presented earlier in the chapter in the discussion of how the salience of boundaries during learning inﬂuences categorical bias.) In particular, do children and adults exhibit more categorical bias when boundaries are present during learning but not during test than when boundaries are present during both learning and test? We reasoned that taking away perceptual structure at test that was available at learning would be more disruptive to memory for ﬁnegrained, metric information than to memory for coarse-grained, categorical information. Speciﬁcally, people likely rely on boundaries and other landmarks to retrieve precise information about individual locations at test, whereas people may not need boundaries to retrieve memory for the spatial groups at test. Greater uncertainty about the individual locations (i.e., in the absence of boundaries) should lead to greater ‘‘pull’’ from the spatial groups. Hence, children and adults should exhibit more categorical bias when boundaries are absent than present during test. Participants learned 20 locations with either walls or lines subdividing the box into four quadrants. During the test phase, boundaries were either present or absent while participants attempted to replace the objects without the aid of the dots. How was the tendency to place objects closer together than they really were inﬂuenced by the presence or absence of boundaries at test? As expected, participants exhibited more categorical bias when no boundaries were present at test than when boundaries were present at test. In fact, not even the adults exhibited signiﬁcant categorical bias when boundaries were present during test (see Figure 9). In contrast, when the boundaries were not present during test, 11-year-olds and adults in the walls condition and adults in the lines condition placed the objects closer together than they really were, exhibiting signiﬁcant categorical bias. Given that all aspects of the procedure were the same up to the moment participants began placing the objects at test, these results demonstrate that decisions about where to place the objects during the test phase emerged out of the interaction of memory for the locations and perceptual structure available at the time of test. In particular, we propose that during learning, adults coded the distance and direction of the locations relative to the boundaries and formed strong connections among the locations within each group. When the boundaries were present at test, adults could rely on their memory for the precise locations of the objects relative to the boundaries.

Jodie M. Plumert

400

1 Lines at Test 0.8

Walls at Test No Lines at Test

0.6

*

*

No Walls at Test

*

0.4 0.2 0 -0.2 -0.4 -0.6 -0.8

* *

* *

-1 7 Years

8 Years

11 Years

Adults

Age Fig. 9. Mean categorical bias by age when lines were present versus absent at test. Asterisks denote significant results (po.05) of one-sample t-tests comparing the displacement score to the expected score with no displacement. Positive scores reflect bias toward the category centers.

When perceptual structure was absent at test, however, adults could not readily use their memory for the precise locations of the objects relative to the boundaries. (This idea is supported by better placement accuracy when boundaries were present than absent during test.) In the absence of boundaries during test, adults relied more heavily on their memory for the spatial groups, leading to greater categorical bias. Children also exhibited greater bias when boundaries were absent than when they were present at test, but with the exception of the 11-year-olds in the more salient boundary condition, the level of categorical bias was not signiﬁcantly greater than 0. These ﬁndings suggest that children formed weaker connections among the locations within each group than did the adults. As a consequence, the ‘‘pull’’ from the spatial groups was not strong enough to offset their memory for the individual locations even when there was less perceptual support during test. Together, these results provide an intriguing example of how decisions about where to place the objects are not solely about what

Understanding the Organism and Environment as a Unified System

401

is in the head. Rather, placements emerge in the moment out of the interaction of the memory representation and the available perceptual structure. d. Summary. Taken together, the results of these studies illustrate the contention that neither the cognitive system nor environmental structure has causal priority in explaining behavior. We cannot explain patterns of categorical bias by referring only to task structure (e.g., presence or absence of boundaries) or by referring only to developmental differences in the cognitive system (e.g., strategic encoding of spatial groups). Our studies on categorical bias in estimates of location have repeatedly shown that all age groups exhibit categorical bias under some task conditions but not under others. For example, adults always show signiﬁcant categorical bias when at least one cue is available during learning, but they do not show bias when no cues are available during learning. Thus, it is impossible to predict categorical bias by referring to age alone. Likewise, our studies have repeatedly shown that the four age groups frequently differ in how they respond to the same task structure. For example, children and adults often differ in how they respond to cues for organizing the locations into groups, such as visible boundaries, spatiotemporal experience, or object relations. Clearly, children and adults extract different things from their experience with these tasks even though the task structure is identical for all participants. These variations in how the same age group responds to different task structure and how different age groups respond to the same task structure support the idea that categorical bias emerges out of the interaction of the cognitive system and the task structure.

IV. Explaining the Emergence of Spatial Categorization Skills In this section, I bring together these two lines of research on children’s spatial categorization skills by addressing the question of how these skills emerge in the moment and over time. Clearly, ordering items based on common membership in a spatial group and placing objects from the same spatial group closer together than they really are depend critically on the ability to form and use spatial groups. And yet, forming and using spatial groups are not unitary skills that develop in an all-or-none fashion. As amply demonstrated throughout this chapter, such skills emerge in the moment depending on both age and task. With respect to the use of spatial categorization as an organizational device, children progress from using

402

Jodie M. Plumert

spatial clustering strategies ﬁrst to organize their searches, then to organize their directions and tour plans, and ﬁnally to organize their free recall. With respect to using spatial groups to remember object positions, adults almost always exhibit categorical bias at test when even a single cue (e.g., boundaries) is available to organize the locations into groups during learning. In contrast, children often do not exhibit categorical bias at test unless two cues were available during learning (e.g., boundaries and spatiotemporal contiguity). In sum, the overall picture from these two programs of research is one of increasing reliance over development on spatial groups to organize object recall and remember object positions. It is especially noteworthy that adults consistently show more categorical bias in their estimates of location, despite the fact that adults almost always exhibit less absolute error in their placements than do children. This indicates that the ‘‘pull’’ from the spatial groups exerts an even stronger inﬂuence on adults’ placements than on children’s placements. Thus, adultlike performance in both recall organization and location memory is marked by strong reliance on spatial categories. Two key questions these ﬁndings raise are (1) What accounts for the age differences in spatial categorization skills observed across task contexts? and (2) How do developmental changes in spatial categorization skills come about?

A. THE EMERGENCE OF SPATIAL CATEGORIZATION SKILLS IN THE MOMENT I start with the question of how age differences in spatial categorization skills emerge in the moment because this question sets the stage for thinking about how spatial categorization skills emerge over development. I propose that age differences in the moment emerge through the interaction of how well children represent and access spatial relational information (characteristics of the organism) and how explicitly tasks cue spatial relations among locations (characteristics of the environment). To form and use spatial groups, children must focus on the spatial relations among the locations. In other words, children must attend to which objects belong to the same spatial region or are in close proximity to each other. The ease with which children represent and access spatial relational information and the extent to which tasks cue spatial relations among locations impact the likelihood that children will form and use associations among locations based on common membership in a spatial group. I discuss these ideas in more detail below using examples from the research presented previously. Note that

Understanding the Organism and Environment as a Unified System

403

organism and environment characteristics are considered separately in this discussion, with later commentary on how the two interact in the moment.

1. Age Differences in Representing Spatial Relational Information On the side of the organism, the ability to mentally represent information about the spatial relations among locations is critical for forming and using spatial groups. At a coarse-grained level, this involves classifying locations into groups based on spatial proximity and region membership. Such classiﬁcation of locations into groups allows children and adults to form associations among locations within a group. Note that in both our studies of categorical clustering and categorical bias, children must notice which objects belong to the same spatial group despite the fact that they usually learn the objects in a random order. This is no small feat given the myriad of other relations children might attend to in these tasks. At a ﬁne-grained level, the ability to mentally represent spatial relational information involves coding the distance and direction of locations relative to one other. Such coding of metric information about locations allows children and adults to remember groups of locations as spatial conﬁgurations. What evidence is there that the ability to represent and access spatial relational information impacts categorical bias in children’s placements? In a recent study, we examined how categorical bias at test was affected by opportunities to view objects together in time during learning (Recker & Plumert, 2007). Our previous work has consistently provided children and adults with many opportunities to simultaneously view the objects during learning. That is, participants cumulatively place the 20 objects on the ﬂoor of the box until they are all fully visible. We reasoned that it should be much more difﬁcult to notice that particular objects belong to the same spatial group when they are seen in isolation as opposed to when they are seen together. If so, this would suggest that children have more difﬁculty than adults with mentally representing spatial relations among locations. As in our previous studies, children and adults learned the locations of 20 objects marked by dots on the ﬂoor of an open, square box. The objects were always placed in a random order. In the simultaneous viewing condition, the objects were cumulatively placed on the ﬂoor of the box until all 20 objects were in their correct locations. Thus, after all objects had been placed, participants were able to simultaneously view the objects on the ﬂoor of the box. In the isolated viewing condition, participants were shown only one object in its correct location at a time. An object was placed in its location for approximately ﬁve seconds and was then removed by the

404

Jodie M. Plumert

experimenter. Thus, the objects were never seen together in time. For both conditions, the dots marking the locations remained visible at all times during learning. Thus, opportunities to view the locations together in time remained constant across conditions, but opportunities to view the objects together in time varied across conditions. At test, participants replaced the objects without the aid of the dots. In Experiment 1, we examined categorical bias in 7-, 9-, 11-year-olds’ and adults’ placements when two converging cues for forming spatial groups were present (i.e., lines divided the box into quadrants and the objects in each quadrant were categorically related). We found that 7-, 9-, and 11-year-olds and adults in the simultaneous viewing condition exhibited categorical bias. However, only the 11-year-olds and adults in the isolated viewing condition exhibited categorical bias, suggesting that younger children had much more difﬁculty forming strong associations among the locations when they viewed the objects in isolation. In Experiment 2, we examined categorical bias under simultaneous and isolated viewing conditions when only a single cue for forming spatial groups was present (i.e., the objects were categorically related but no boundaries were present). Eleven-year-olds and adults in the simultaneous viewing condition again exhibited categorical bias, but only the adults showed bias in the isolated viewing condition. Thus, 11-year-olds had more difﬁculty than did adults in forming strong spatial groups when they viewed the objects in isolation and only one cue deﬁned the spatial groups. Finally, in Experiment 3 we examined categorical bias in adults’ placements when boundaries were present but the objects were not related. Adults again exhibited bias in both the simultaneous and isolated viewing conditions. Together, these results indicate that the ability to represent spatial relational information is undergoing signiﬁcant change over late childhood and early adolescence. We can see how the ability to represent spatial relational information impacts children’s use of spatial clustering strategies by comparing performance in tasks with differing representational demands. The study showing that 6-year-olds’ directions were far less spatially organized than their searches even when all other aspects of the situation are equated is a good case in point (Plumert et al., 1994). How might the representational demands of searching and direction giving affect the use of spatial organizational strategies? Although searching and direction giving both involve ordering locations, searching places far fewer representational demands on children than does direction giving. First, children can rely on the visible structure of the physical environment when searching for objects, whereas children have to mentally represent the structure of the physical environment when giving directions. In particular, children can use the

Understanding the Organism and Environment as a Unified System

405

structure of the physical environment to constrain their movement while searching, but have to use their representation of the physical environment to constrain their imagined movement while giving directions. Clearly, relying on visually present environmental structure is far less taxing than relying on mentally represented environmental structure. Second, children can use physical movement during searching to generate visual information about locations, whereas children have to use imagined movement during direction giving to retrieve information about locations. There is evidence indicating younger children have difﬁculty with representing imagined movement. For example, Gauvain and Rogoff (1989) found that not until nine years of age did children’s descriptions of spatial layouts contain characteristics of a mental tour. Studies of developmental changes in children’s use of elaboration also have shown that older elementary school children tend to generate dynamic images to help them remember information, but younger elementary school children tend to general static images (Reese, 1977). The idea that using imagined movement to access spatial information plays an important role in efﬁcient direction giving is further supported by the high frequency of movement references in adults’ directions, suggesting that they spontaneously organized their directions as if they were guiding their listener on a walk through the house (e.g., ‘‘walk into the kitchen, pick up the toaster, and you’ll ﬁnd a piece hidden there’’). In short, they used imagined movement as a mechanism for accessing their memory for the locations and for organizing their directions for ﬁnding the locations. In sum, the ability to represent and access spatial relational information appears to play a key role in the age differences we see in children’s spatial clustering strategies and categorical bias in estimates of location. To date, most of our work has focused on the ability to form associations among locations within a group rather than the ability to remember locations as spatial conﬁgurations. Clearly, research by Uttal and his colleagues has shown that children’s ability to code and reproduce relative distance and direction in spatial conﬁgurations undergoes signiﬁcant change from the preschool years through the elementary school years (Uttal, 1994, 1996; Uttal, Fisher, & Taylor, 2006; Uttal & Wellman, 1989). One interesting possibility that we are just beginning to consider is that coding the associations among locations in a spatial group and coding conﬁgural information about locations within a spatial group may work hand-in-hand to produce categorical bias in estimates of location. That is, coding spatial groups as conﬁgurations (i.e., creating a holistic representation) about may actually strengthen the associations among locations within a group, thereby leading to more, rather than less categorical bias. Preliminary work with adults suggests that this may be the case.

406

Jodie M. Plumert

2. Task Differences in Cueing Spatial Relations among Locations On the side of the environment, the extent to which the task cues spatial relations among locations also plays a critical role in forming and using spatial groups. Two basic principles are at work here. One is that some kind of environmental structure is necessary to group locations or subdivide spaces (even for adults), and the other is that more salient environmental structure makes it easier to group locations and subdivide spaces. This is clearly seen in our work on how cues for forming spatial groups during learning impact categorical bias in placements at test (Hund & Plumert, 2003, 2005; Hund et al., 2002; Plumert & Hund, 2001; Recker & Plumert, 2007). First, it is noteworthy that even adults do not show signiﬁcant categorical bias in their placements at test if no cues are available to group locations during learning (Plumert & Hund, 2001). In subsequent work, we further conﬁrmed that categorical bias at test only occurs if cues are available for forming spatial groups during learning by having children and adults learn the locations without boundaries and then testing them either with or without boundaries (Recker, Plumert, & Stevens, 2007a). Both conditions resulted in the same low level of categorical bias. Thus, providing structure for grouping locations after the fact did not lead to categorical bias in people’s object placements. Related work by Simmering and Spencer (2007) has also shown that adults are unable to mentally impose category boundaries unless they are tied to perceptually available environmental structure (e.g., axes of symmetry). Together, these studies provide strong evidence that thinking is not just about what is in the head. Our work also shows that the salience of cues for forming spatial groups impacts spatial categorization skills. One indication that cue salience matters is the fact that children often require two converging cues for forming spatial groups in order to show signiﬁcant categorical bias at test (Hund et al., 2002; Recker & Plumert, 2007). The work described earlier on boundary salience also showed that children and adults exhibited more categorical bias when more salient boundaries (walls) were present during learning than when less salient boundaries (lines) were present. In other work, we found that 7- to 11-year-old children used spatiotemporal contiguity to form spatial groups when they experienced all of the locations in one group together before moving on to the next group, but not when they experienced only 75% of the locations in one group together before moving on to another group (Hund & Plumert, 2005). In contrast, adults used spatiotemporal contiguity to form spatial groups in both conditions. There is also evidence indicating that boundary salience inﬂuences the order in which children organize their searches for objects (Nichols-Whitehead & Plumert, 2001). In particular, 3- and 4-year-olds’ object retrieval was more spatially organized when a tall opaque or short opaque boundary divided a

Understanding the Organism and Environment as a Unified System

407

small dollhouse in half than when a tall transparent boundary divided the dollhouse in half. Together, these results clearly show that the salience of environmental cues for forming spatial groups has a major impact on categorical bias in placements and use of spatial clustering as an organizational device. We can also see how the extent to which the task cues spatial relations among locations plays a role in forming spatial groups by comparing spatial clustering in tasks that have similar representational demands but differ in how easily they draw children’s attention to the spatial relations among locations. A good case in point is the study described previously showing that 10-year-olds exhibit much more spatial clustering when planning a tour of a set of objects than when recalling the names of a set of objects (Plumert & Strahan, 1997). An interesting feature of this study is that everything was the same in the tour-planning and free recall conditions until the moment children were given the memory test. Thus, the differences in 10-year-olds’ tour plans and free recall can only be attributed to differences in how easily the two tasks drew children’s attention to the spatial relations among the objects. Why might this be the case? A task such as planning a tour of a set of objects may readily draw children’s attention to the spatial connections among objects by making the listener’s movement through space more salient. Speciﬁcally, imagining the listener in the space may prime them to think about locations nearby the listener (Morrow, Greenspan, & Bower, 1987). When faced with an unstructured task such as free recall, however, children may have difﬁculty focusing on the spatial connections among the objects because the explicitly stated goal of the task is to remember what the objects are, not where they are located. As seen earlier, in situations in which both categorical and spatial organization are available (e.g., recalling the furniture from one’s home), younger children attend more to the categorical than to the spatial relations among the items (Plumert, 1994). The tour-planning task may also have drawn children’s attention to the spatial connections among objects because it activated children’s metacognitive knowledge about tours. Although tours can be organized in many ways, they often involve taking the viewer from each location to the next closest location. Older children are more likely than younger children to have had experience with tours and hence may have a better understanding of the goals of tour planning. Other research has shown that children are more likely to deploy strategies to enhance recall or exhibit more sophisticated reasoning in tasks that contain a goal that is meaningful and familiar to them (Gauvain, 1993; Gauvain & Rogoff, 1986; Woody-Ramsey & Miller, 1988; Woolley, 2006). For example, Woody-Ramsey and Miller (1988) found that 4- and 5-year-olds were much more likely to use a

408

Jodie M. Plumert

selective attention strategy when the task was embedded in the context of a meaningful story.

3. Organism–Environment Interactions in the Moment The preceding discussion necessarily separates what the child and the environment contribute to the emergence of spatial categorization skills. Yet explaining the patterns of recall organization and categorical bias observed in these studies clearly requires a consideration of how the two interact at any moment in time. In fact, just knowing the extent of the child’s ability to represent spatial relational information or the extent to which the task cues spatial relations among locations will not allow one to accurately predict patterns of recall organization or categorical bias. This is clearly illustrated in the many studies showing that age interacts with the task. For example, our recent work on age differences in children’s ability to form spatial groups under isolated versus simultaneous viewing conditions has shown that the ability to form strong spatial groups under isolated viewing conditions is not an all-or-none ability (Recker & Plumert, 2007). Rather, 11-year-olds form strong spatial groups when converging cues deﬁne the spatial groups (i.e., visible boundaries and categorical relatedness), but not when only a single cue deﬁnes the spatial group (visible boundaries). Likewise, our work comparing spatial clustering in children’s tour plans versus free recall shows that the tour-planning task is effective in eliciting strong spatial clustering in 10-year-olds, but not in younger children (Plumert & Strahan, 1997). Again, these ﬁndings underscore the idea that age differences in thinking can only be understood as emerging out of a system that includes both the child and the environment. One general theme that has emerged from both lines of work is that children are able to exploit more subtle cues for forming and using spatial groups as they grow older. What implications does this have for the general argument made throughout this chapter? One possibility is that thinking resides more and more ﬁrmly ‘‘in the head’’ as development proceeds. In other words, the environment becomes less and less important for children’s thinking as they grow older. Interestingly, many theories either implicitly or explicitly accept this idea about cognitive development. For example, Piaget’s entire theory of cognitive development was based on the idea that children progress from strong to minimal reliance on the environment to guide their thinking as they move from the sensorimotor period to the formal operations period. Intuitively, this characterization of children’s cognitive development seems right. But I would contend that the environment is always a part of thinking, even for adults. One way to think about this is that the coupling between the cognitive system and the

Understanding the Organism and Environment as a Unified System

409

physical, task, or social environment becomes more ﬁnely tuned with development (and experience). Although speculative at this point, this tighter coupling might result from reduced variability and greater stability of bottom-up processes such as memory and attention, and from greater top-down control from executive centers and long-term memory. The end result is a greater sensitivity to more subtle environmental structure over learning and development.

B. THE EMERGENCE OF SPATIAL CATEGORIZATION SKILLS OVER DEVELOPMENT I turn now to thinking about how spatial categorization skills might emerge over longer time scales. As outlined at the beginning of this chapter, we need to think about cyclical organism–environment interactions over time in order to understand how skills emerge over development. In particular, we need to think about how the organism and environment characteristics proposed to play a role in the emergence of spatial categorization skills in the moment interact over time through the medium of experience. This way of thinking about developmental changes rests on a fundamental continuity between real-time and developmental processes (see also Elman et al., 1996; Thelen & Smith, 1994). The same aspects of the cognitive system and the environment that work together in real time to produce age differences in thinking also work together over longer time scales to produce developmental changes in thinking (for an example of real-time change during shortterm learning experiences, see Recker et al., 2007b). I propose that two complementary processes are at work to produce developmental change in use of spatial clustering strategies: (1) exposure to salient cues that highlight the spatial relations among locations leads to increases in children’s ability to represent spatial relational information; and (2) increases in children’s ability to represent spatial relational information allows children to notice less salient cues that highlight spatial relations among locations. What evidence is there that these two processes are actually at work? At present, the kinds of changes hypothesized to occur have only been investigated over very short time scales (i.e., within the course of an experiment). The general approach that my colleagues and I have taken is to examine whether experiences with using spatial clustering in a more supportive task lead to increases in the use of spatial clustering in a less supportive task. By supportive, I mean here tasks that highlight the spatial connections among objects (i.e., common membership in a spatial region and/or distance and direction of locations relative to one another). For example, we found that when 6-year-olds were allowed to search for

Jodie M. Plumert

410

objects before giving directions for ﬁnding them, they exhibited high levels of spatial clustering in their subsequent directions (Plumert et al., 1994). These results suggest that although children apply their spatial clustering skills to searching before they apply those same skills to direction giving, experience with using spatial clustering during searching facilitates 6-year-olds’ ability to apply their spatial clustering skills to the more difﬁcult task of direction giving. Likewise, Plumert and Strahan (1997) found that 10-year-olds could be induced to use spatial clustering in a free recall task if given experience with using spatial clustering in a tour-planning task ﬁrst (see Figure 10). In contrast, 8-year-olds exhibited relatively low levels of spatial clustering in their subsequent free recall regardless of whether they performed the tour-planning or the free recall task ﬁrst. These results suggest that experience with the more supportive tour-planning task cued 10-year-olds about the spatial connections among the objects. Once cued, 10-year-olds could apply a spatial clustering strategy to the less supportive free recall task. Thus, over the short-term time scale of an experiment, we see that experience with using spatial clustering in a more supportive task leads to transfer of spatial clustering to a less supportive task. 1 0.9 0.8

First Task Second Task

0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 Tour-planning first

Free recall first

Tour-planning first

8-year-olds

Free recall first

10-year-olds Age Group and Condition

Fig. 10. Mean spatial clustering scores by age and condition in the first and second tasks.

Understanding the Organism and Environment as a Unified System

411

Research in other domains of cognitive development also has shown that short-term experience with using a skill in a simpler task facilitates children’s ability to use that skill in a more difﬁcult task. Marzolf and DeLoache (1994), for example, found that 2½-year-olds were more likely to succeed on a difﬁcult scale model task if given a simpler scale model task ﬁrst. They argue that experience with the simpler task sensitized young children to the symbolic relations between the scale model and the real space. These ﬁndings are also consistent with the literature on analogical reasoning showing that young children are capable of transferring a solution to a more difﬁcult problem if given experience with solving a simpler problem ﬁrst (Brown, 1989; Gentner, 1989). In a similar vein, Newcombe and colleagues have shown that 3-year-old children integrate features (e.g., colored walls or large objects) with room geometry (e.g., length of walls) to reorient in larger, but not smaller spaces (Learmonth et al., 2007). However, 3-year-olds do integrate features and geometry to reorient in a small space if given experience with reorienting in a larger scale space ﬁrst. Similar ﬁndings have been recently reported by Twyman, Friedman, and Spetch (2007). Together, these results suggest that shortterm experiences with solving speciﬁc problems in supportive environments sensitizes children to critical features of the problem, thereby affecting how children solve those same problems in less supportive environments. Although these transfer studies are promising, our understanding of how change occurs over longer time scales is clearly limited. One missing link here is that changes in children’s ability to represent spatial relational information are inferred, not directly tested. If experiences with using spatial clustering in more supportive task contexts sensitizes children to the spatial connections among objects, this should be evident in tests of children’s spatial relational knowledge as well. A second problem is that most of these studies document near transfer rather than far transfer. For example, in the study showing that 10-year-old children were more likely to use spatial clustering to organize their free recall if given experience using spatial clustering to organize their tour plans ﬁrst, the same dollhouse and objects were used in both tasks. Thus, we do not know whether the experience of using spatial clustering in the more supportive task would transfer to the less supportive task if children were learning the locations of a new set of objects in a new space. A third problem is that these studies are looking at transfer over very short time periods. More work is needed to determine whether the changes we see in the short term are the same kinds of changes we see over longer time periods. Microgenetic studies of change would be a useful step in this direction. Finally, we might ask to what extent does using spatial clustering strategies in supportive tasks act as mechanisms for change in everyday life?

412

Jodie M. Plumert

In the laboratory, we can carefully control the order in which children experience cues or tasks. Children’s everyday experiences with cues for forming spatial groups and tasks calling for spatial clustering strategies are likely to be considerably less orderly. For example, children may be exposed to less salient cues for forming spatial groups before they are exposed to highly salient cues. Likewise, they may encounter more difﬁcult tasks before they encounter less difﬁcult ones. However, the sensitivity of the cognitive system to environmental structure may provide a built-in mechanism for ensuring that children’s everyday experiences are more orderly than they may seem at ﬁrst glance. With an immature cognitive system, young children’s ‘‘experiences’’ may well be limited to noticing only salient cues for forming spatial groups and using spatial clustering in highly supportive tasks. Thus, young children do not experience a bewildering array of inputs simply because they are not sensitive to these inputs. This constraint on experience imposed by the cognitive system may be critical for ensuring that the child’s experience of environmental structure proceeds in an orderly fashion (see also Newport, 1990).

V. Limits and Conclusions What are the limits of this approach to understanding cognition and cognitive development? One is that in order to study the interaction of the child and the environment, we often end up separating the two. Although this reductionist approach may be necessary on a practical level for doing scientiﬁc research, one need not appeal to reductionism at a theoretical level for explaining cognitive development. Another potential problem is accurately identifying the kinds of environmental structure that matter in children’s everyday lives (Oakes, Newcombe, & Plumert, in press). Given that children could attend to many different aspects of the physical, social, and task environment to in the real world, how do we determine which ones actually matter? Although we can make some theoretically informed guesses, progress on this front probably requires more back-and-forth between laboratory-based and real-world studies than is typically the case (Liben & Myers, 2007). Finally, if all thinking emerges out of the interaction of some aspect of the cognitive system and some aspect of environmental structure, doesn’t this lead to an inﬁnite regress? For example, if spatial categorization skills depend in part on the child’s ability to mentally represent spatial relational information, where does this ability come from? Though inﬁnite regress is theoretically a problem, a pragmatically sensible solution is to identify a behavior of interest and then determine how the components of the cognitive system and the

Understanding the Organism and Environment as a Unified System

413

environment work together to produce the behavior in the moment and over development, at the same time acknowledging that the components themselves are emergent products of organism–environment interactions occurring at other levels of the system. In closing, I have argued that variations in children’s thinking across different contexts arise out of the real-time interaction of the cognitive system and the available environmental structure. From this perspective, thinking does not reside only in the head. Rather, thinking is an activity that emerges out of a uniﬁed system that includes both the child and the environment. Development proceeds in a cyclical fashion with changes in the cognitive system opening up new sensitivities to environmental structure, and new sensitivities to environmental structure leading to the further reﬁnement of cognitive skills. Ultimately, a better understanding of these interactions both in the moment and over development should allow us to better predict the course of cognitive development for individual children and to successfully intervene when aspects of cognitive development go awry.

Acknowledgements The research reported in this chapter was supported by grants awarded to the author from the National Institutes of Health (R03-HD36761) and the National Science Foundation (BCS-0343034). I thank the members of the Iowa Center for Developmental and Learning Sciences (ICDLS) for their insightful comments about the ideas presented in this chapter.

REFERENCES Acredolo, L. P., & Boulter, L. T. (1984). Effects of hierarchical organization on children’s judgments of distance and direction. Journal of Experimental Child Psychology, 37, 409–425. Adolph, K. E. (1997). Learning in the development of infant locomotion. Monographs of the Society for Research in Child Development, 62(3), 1–140. Adolph, K. E. (2000). Speciﬁcity of learning: Why infants fall over a veritable cliff. Psychological Science, 11, 290–295. Adolph, K. E., Eppler, M. A., & Gibson, E. J. (1993). Crawling versus walking infants’ perception of affordances for locomotion over sloping surfaces. Child Development, 64, 1158–1174. Allen, G. L. (1981). A developmental perspective on the effects of ‘‘subdividing’’ macrospatial experience. Journal of Experimental Psychology: Human Learning and Memory, 7, 120–132.

414

Jodie M. Plumert

Baillargeon, R. (2001). Infants’ physical knowledge: Of acquired expectations and core principles. In E. Dupoux (Ed.), Language, brain, and cognitive development: Essays in honor of Jacques Mehler (pp. 341–361). Cambridge, MA: The MIT Press. Brown, A. L. (1989). Analogical learning and transfer: What develops? In S. Vosniadou & A. Ortony (Eds.), Similarity and analogical reasoning (pp. 369–412). New York: Cambridge University Press. Clayton, K., & Habibi, A. (1991). Contribution of temporal contiguity to the spatial priming effect. Journal of Experimental Psychology: Learning, Memory, and Cognition, 17, 263–271. Cornell, E. H., & Heth, C. D. (1986). The spatial organization of hiding and recovering of objects by children. Child Development, 57, 603–615. Correa-Chavez, M., Rogoff, B., & Arauz, R. M. (2005). Cultural patterns in attending to two events at once. Child Development, 76, 664–678. Curiel, J. M., & Radvansky, G. A. (1998). Mental organization of maps. Journal of Experimental Psychology: Learning, Memory, and Cognition, 24, 202–214. Elman, J. L., Bates, E. A., Johnson, M. H., Karmiloff-Smith, A., Parisi, D., & Plunkett, K. (1996). Rethinking innateness: A connectionist perspective on development. Cambridge, MA: MIT Press. Engebretson, P. H., & Huttenlocher, J. (1996). Bias in spatial location due to categorization: Comment on Tversky and Schiano. Journal of Experimental Psychology: General, 125, 96–108. Eppler, M. A., Adolph, K. E., & Weiner, T. (1996). The developmental relationship between infant’s exploration and action on slanted surfaces. Infant Behavior and Development, 19, 259–264. Friedman, A., & Brown, N. (2000a). Reasoning about geography. Journal of Experimental Psychology: General, 129, 193–219. Gauvain, M. (1993). The development of spatial thinking in everyday activity. Developmental Review, 13, 92–121. Gauvain, M. (2001). The social context of cognitive development. New York: The Guilford Press. Gauvain, M., & Rogoff, B. (1986). Inﬂuence of the goal on children’s exploration and memory of large-scale space. Developmental Psychology, 22, 72–77. Gauvain, M., & Rogoff, B. (1989). Ways of speaking about space: The development of children’s skill in communicating spatial knowledge. Cognitive Development, 4, 295–307. Gentner, D. (1989). The mechanisms of analogical learning. In S. Vosniadou & A. Ortony (Eds.), Similarity and analogical reasoning (pp. 199–241). New York: Cambridge University Press. Gibson, J. J. (1979). The ecological approach to visual perception. Hillsdale, NJ: Lawrence Erlbaum Associates. Gibson, E. J. (1988). Exploratory behavior in the development of perceiving, acting, and the acquiring of knowledge. Annual Review of Psychology, 39, 1–41. Gibson, J. J., & Gibson, E. J. (1955). Perceptual learning: Differentiation or enrichment? Psychological Review, 62, 32–41. Gibson, E. J., & Pick, A. D. (2000). An ecological approach to perceptual learning and development. New York: Oxford University Press. Gottlieb, G., & Lickliter, R. (2007). Probabilistic epigenesis. Developmental Science, 10, 1–11. Haake, R. J., Somerville, S. C., & Wellman, H. M. (1980). Logical ability of young children in searching a large-scale environment. Child Development, 51, 1299–1302. Hund, A. M., & Plumert, J. M. (2002). Delay-induced bias in children’s memory for location. Child Development, 73, 829–840. Hund, A. M., & Plumert, J. M. (2003). Does information about what things are inﬂuence children’s memory for where things are? Developmental Psychology, 39, 938–948.

Understanding the Organism and Environment as a Unified System

415

Hund, A. M., & Plumert, J. M. (2005). The stability and ﬂexibility of spatial categories. Cognitive Psychology, 50, 1–44. Hund, A. M., & Spencer, J. P. (2003). Developmental changes in the relative weighting of geometric and experience-dependent location cues. Journal of Cognition and Development, 4, 3–38. Hund, A. M., Plumert, J. M., & Benney, C. J. (2002). Experiencing nearby locations together in time: The role of spatiotemporal contiguity in children’s memory for location. Journal of Experimental Child Psychology, 82, 200–225. Huttenlocher, J., & Lourenco, S. (2007). Using spatial categories to reason about location. In J. M. Plumert & J. Spencer (Eds.), The emerging spatial mind (pp. 3–24). New York: Oxford University Press. Huttenlocher, J., Hedges, L. V., & Duncan, S. (1991). Categories and particulars: Prototype effects in estimating spatial location. Psychological Review, 98, 352–376. Huttenlocher, J., Newcombe, N., & Sandberg, E. H. (1994). The coding of spatial location in young children. Cognitive Psychology, 27, 115–147. Huttenlocher, J., Vasilyeva, M., Newcombe, N., & Duffy, S. (2008). Developing symbolic capacity one step at a time. Cognition, 106, 1–12. Johnson, J. S., Spencer, J. P., & Scho¨ner, G. (in press). Moving to higher ground: The dynamic ﬁeld theory and the dynamics of visual cognition. New Ideas in Psychology. Kosslyn, S. M., Pick, H. L., Jr., & Fariello, G. R. (1974). Cognitive maps in children and men. Child Development, 45, 707–716. Learmonth, A. E., Newcombe, N. S., Sheridan, N., & Jones, M. (in press). Why size counts: Children’s spatial reorientation in large and small enclosures. Developmental Science. Liben, L. S., & Myers, L. J. (2007). Developmental changes in children’s understanding of maps: What? When? and How? In J. M. Plumert & J. Spencer (Eds.), The emerging spatial mind (pp. 193–218). New York: Oxford University Press. Lockman, J. J. (2000). A perception-action perspective on tool use development. Child Development, 71, 137–144. Marzolf, D. P., & DeLoache, J. S. (1994). Transfer in young children’s understanding of spatial representations. Child Development, 65, 1–15. McNamara, T. P. (1986). Mental representation of spatial relations. Cognitive Psychology, 18, 87–121. McNamara, T. P., Hardy, J. K., & Hirtle, S. C. (1989). Subjective hierarchies in spatial memory. Journal of Experimental Psychology: Learning, Memory, and Cognition, 15, 211–227. McNamara, T. P., Halpin, J. A., & Hardy, J. K. (1992). The representation and integration in memory of spatial and nonspatial information. Memory & Cognition, 20, 519–532. Montello, D. R. (2003). Regions in geography: Process and content. In M. Duckham, M. F. Goodchild, & M. F. Worboys (Eds.), Foundations of geographic information science (pp. 173–189). London: Taylor & Francis. Morrow, D. G., Greenspan, S. L., & Bower, G. H. (1987). Accessibility and situation models in narrative comprehension. Journal of Memory and Language, 26, 165–187. Newcombe, N., & Liben, L. S. (1982). Barrier effects in the cognitive maps of children and adults. Journal of Experimental Child Psychology, 34, 46–58. Newport, E. L. (1990). Maturational constraints on language learning. Cognitive Science, 14, 11–28. Nichols-Whitehead, P., & Plumert, J. M. (2001). The inﬂuence of boundaries on young children’s searching and gathering. Journal of Cognition and Development, 2, 367–388. Oakes, L. M., Plumert, J. M., Lansink, J. M., & Merryman, J. D. (1996). Evidence for taskdependent categorization in infancy. Infant Behavior and Development, 19, 425–440.

416

Jodie M. Plumert

Oakes, L. M., Newcombe, N. S., & Plumert, J. M. (in press). Are dynamic systems and connectionist approaches an alternative to ‘‘Good Old Fashioned Cognitive Development?’’. In J. P. Spencer, M. Thomas, & J. McClelland (Eds), Toward a new grand theory of development: Connectionism and dynamic systems theory re-considered. New York: Oxford University Press. Piaget, J. (1954). The construction of reality in the child. New York: Basic Books. Plumert, J. M. (1994). Flexibility in children’s use of spatial and categorical organizational strategies in recall. Developmental Psychology, 30, 738–747. Plumert, J. M., & Hund, A. M. (2001). The development of memory for object location: What role do spatial prototypes play? Child Development, 72, 370–384. Plumert, J. M., & Strahan, D. (1997). Relations between task structure and developmental changes in children’s use of spatial clustering strategies. British Journal of Developmental Psychology, 15, 495–514. Plumert, J. M., Hund, A. M., & Recker, K. M. (2007a). Organism-environment interaction in spatial development: Explaining categorical bias in memory for location. In J. M. Plumert & J. Spencer (Eds.), The emerging spatial mind (pp. 25–52). New York: Oxford University Press. Plumert, J. M., Kearney, J. K., & Cremer, J. F. (2004). Children’s perception of gap affordances: Bicycling across trafﬁc-ﬁlled intersections in an immersive virtual environment. Child Development, 75, 1243–1253. Plumert, J. M., Kearney, J. K., & Cremer, J. F. (2007b). Children’s road crossing: A window into perceptual-motor development. Current Directions in Psychological Science, 16, 255–258. Plumert, J. M., Pick, H. L., Jr., Marks, R. A., Kintsch, A. S., & Wegesin, D. (1994). Locating objects and communicating about locations: Organizational differences in children’s searching and direction-giving. Developmental Psychology, 30, 443–453. Recker, K. M., & Plumert, J. M. (2007). How do opportunities to view objects together in time influence children’s memory for location? Unpublished manuscript. Recker, K. M., Plumert, J. M., & Stevens, E. N. (2007a). Children’s memory for location is not only about what’s in the head. Poster presented at the 2007 Biennial Meeting of the Society for Research in Child Development, Boston, MA. Recker, K. M., Plumert, J. M., Hund, A. M., & Reimer, R. (2007b). How do biases in spatial memory change as children and adults are learning locations?. Journal of Experimental Child Psychology, 98, 217–232. Reese, H. W. (1977). Imagery and associative memory. In R. V. Kail & J. W. Hagen (Eds.), Perspectives on the development of memory and cognition (pp. 113–175). Hillsdale, NJ: Erlbaum. Rieser, J. J., Pick, H. L., Ashmead, D. H., & Garing, A. E. (1995). Calibration of human locomotion and models of perceptual-motor organization. Journal of Experimental Psychology: Human Perception and Performance, 21, 480–497. Roenker, D. L., Thompson, C. P., & Brown, S. C. (1971). Comparison of measures for the estimation of clustering in free recall. Psychological Bulletin, 76, 45–48. Rogoff, B. (1990). Apprenticeship in thinking. New York: Oxford University Press. Samuelson, L. K., Schutte, A. R., & Horst, J. S. (2007). The dynamic nature of knowledge: Insights from a dynamic field model of children’s novel noun generalizations. Unpublished manuscript. Sandberg, E. H., Huttenlocher, J., & Newcombe, N. (1996). The development of hierarchical representation of two-dimensional space. Child Development, 67, 721–739. Schutte, A. R., & Spencer, J. P. (2002). Generalizing the dynamic ﬁeld theory of the A-not-B error beyond infancy: Three-year-olds’ delay- and experience-dependent location memory biases. Child Development, 73, 377–404.

Understanding the Organism and Environment as a Unified System

417

Schutte, A. R., Spencer, J. P., & Scho¨ner, G. (2003). Testing the dynamic ﬁeld theory: Working memory for locations becomes more spatially precise over development. Child Development, 74, 1393–1417. Sherman, R. C., & Lim, K. M. (1991). Determinants of spatial priming in environmental memory. Memory & Cognition, 19, 283–292. Simmering, V. R., & Spencer, J. P. (2007). Carving up space at imaginary joints: Can people mentally impose spatial category boundaries?. Journal of Experimental Psychology: Human Perception and Performance, 33, 871–894. Simmering, V. R., Schutte, A. R., & Spencer, J. P. (in press). Generalizing the dynamic ﬁeld theory of spatial working memory across real and developmental time scales. Brain Research. Sophian, C. (1997). Beyond competence: The signiﬁcance of performance for conceptual development. Cognitive Development, 12, 281–303. Spelke, E. S. (2000). Core knowledge. American Psychologist, 55, 1233–1243. Spelke, E. S., & Newport, E. L. (1998). Nativism, empiricism, and the development of knowledge. In W. Damon & R. M. Lerner (Eds.), Handbook of child psychology, volume 1: Theoretical models of human development (pp. 275–340). Hoboken, NJ: Wiley. Spencer, J. P., & Hund, A. M. (2002). Prototypes and particulars: Geometric and experiencedependent spatial categories. Journal of Experimental Psychology: General, 131, 16–37. Spencer, J. P., & Hund, A. M. (2003). Developmental continuity in the processes that underlie spatial recall. Cognitive Psychology, 47, 432–480. Spencer, J. P., Simmering, V. R., Schutte, A. R., & Scho¨ner, G. (2007). What does theoretical neuroscience have to offer the study of behavioral development? Insights from a dynamic ﬁeld theory of spatial cognition. In J. M. Plumert & J. P. Spencer (Eds.), The emerging spatial mind (pp. 320–361). New York: Oxford University Press. Stevens, A., & Coupe, P. (1978). Distortions in judged spatial relations. Cognitive Psychology, 10, 422–437. Thelen, E., & Smith, L. B. (1994). A dynamic systems approach to the development of cognition and action. Cambridge, MA: MIT Press. Twyman, A., Friedman, A., & Spetch, M. L. (2007). Penetrating the geometric module: Catalyzing children’s use of landmarks. Developmental Psychology, 43, 1523–1530. Uttal, D. H. (1994). Preschoolers’ and adults’ scale translation and reconstruction of spatial information acquired from maps. British Journal of Developmental Psychology, 12, 259–275. Uttal, D. H. (1996). Angles and distances: Children’s and adults’ reconstruction and scaling of spatial conﬁgurations. Child Development, 67, 2763–2779. Uttal, D. H., & Wellman, H. M. (1989). Young children’s representation of spatial information acquired from maps. Developmental Psychology, 25, 128–138. Uttal, D. H., Fisher, J. A., & Taylor, H. A. (2006). Words and maps: Developmental changes in mental models of spatial information acquired from descriptions and depictions. Developmental Science, 9, 221–235. Vygotsky, L. S. (1978). Mind in society. Cambridge, MA: Harvard University Press. Warren, W. H. (1984). Perceiving affordances: Visual guidance of stair climbing. Journal of Experimental Psychology: Human Perception and Performance, 10, 683–703. Wellman, H. M., Somerville, S. C., Revelle, G. L., Haake, R. J., & Sophian, C. (1984). The development of comprehensive search skills. Child Development, 55, 472–481. Woody-Ramsey, J., & Miller, P. H. (1988). The facilitation of selective attention in preschoolers. Child Development, 59, 1497–1503. Woolley, J. D. (2006). Verbal-behavioral dissociations in development. Child Development, 77, 1539–1553.

This page intentionally left blank

REMOTE TRANSFER OF SCIENTIFIC-REASONING AND PROBLEM-SOLVING STRATEGIES IN CHILDREN

Zhe Chena and David Klahrb a

HUMAN DEVELOPMENT AND FAMILY STUDIES, UNIVERSITY OF CALIFORNIA, DAVIS, CALIFORNIA 95616, USA b DEPARTMENT OF PSYCHOLOGY, CARNEGIE-MELLON UNIVERSITY, PITTSBURGH, PENNSYLVANIA 15213, USA

I. INTRODUCTION: THEORETICAL, EMPIRICAL, AND PRACTICAL IMPORTANCE OF RESEARCH IN REMOTE TRANSFER II. PARADOX, TAXONOMY, AND PARADIGMS III. EVIDENCE OF REMOTE TRANSFER I: EXPERIMENTAL APPROACH A . SEVEN- TO 10-YEAR-OLD CHILDREN’S LEARNING OF THE CONTROL OF VARIABLE STRATEGY IN DESIGNING EXPERIMENTS AND TRANSFERRING THE STRATEGY B . FOUR- AND 5-YEAR-OLDS’ UNDERSTANDING OF INDETERMINACY AND TRANSFER OF STRATEGIES AFTER A 7-MONTH DELAY C . SIX- TO 8-YEAR-OLDS’ TRANSFER OF A HYPOTHESIS-TESTING STRATEGY AFTER A 24-MONTH DELAY IV. EVIDENCE OF REMOTE TRANSFER II: NATURALISTIC, CROSS-CULTURAL APPROACH V. PROCESSES INVOLVED IN REMOTE TRANSFER A . ENCODING FEATURES OF SOURCE ANALOGUES B . ACCESSING SOURCE INFORMATION C . MAPPING STRUCTURAL ELEMENTS BETWEEN PROBLEMS D . EXECUTING A LEARNED SOLUTION VI. DEVELOPMENTAL DIFFERENCES VII. EDUCATIONAL IMPLICATIONS A . ROLE OF ANALOGY B . EXPLICIT INSTRUCTION C . IMPLICIT FEEDBACK AND SELF-EXPLANATIONS D . LEARNING IN NATURALISTIC AND CULTURAL SETTINGS VIII. CONCLUSIONS AND FUTURE DIRECTIONS REFERENCES

419 Advances in Child Development and Behavior R.V. Kail : Editor

Copyright r 2008 Elsevier B.V. All rights reserved.

420

Zhe Chen and David Klahr

I. Introduction: Theoretical, Empirical, and Practical Importance of Research in Remote Transfer Transfer, the application of information acquired in original learning situations to new, relevant problems, is a fundamental and long-enduring issue in human cognition, children’s learning, and education (e.g., Barnett & Ceci, 2002; Detterman & Sternberg, 1993; Singley & Anderson, 1989; Thorndike & Woodworth, 1901). The basic questions concern how, when, and how well people retrieve relevant information from long-term memory and use examples acquired in the past to solve analogous problems. Despite the unarguable centrality of learning and transfer in children’s development, most studies in our ﬁeld over the past decades ‘‘have focused on skills and knowledge at particular ages rather than on the processes through which children acquired the skills and knowledge’’ (Siegler, 2006). Consequently, the topic of transfer in children’s learning and thinking has been rather peripheral in developmental psychology. Among the tens of thousands of articles appearing in Child Development, Developmental Psychology, and the Journal of Experimental Child Psychology over the past half century, only about 1% address the topic of transfer, and even these few have been limited to studies focusing on the application of acquired information to fairly similar problem situations only after a short delay and within the same lab setting. Children’s remote transfer—that is, the application of concepts and strategies across substantially different contexts and after a long delay—has been virtually unexplored. The domains or tasks in which children’s transfer has been examined include memory strategies (Blo¨te et al., 1999; Coyle & Bjorklund, 1997), mathematical reasoning (e.g., Alibali, 1999; Goldin-Meadow & Alibali, 2002; Rittle-Johnson, 2006; Siegler, 2002), number estimation (Siegler & Opfer, 2003), conservation (e.g., Gelman, 1969; Siegler, 1995), understanding of physical rules (Siegler & Chen, 1998), tool use and causal reasoning (Brown & Kane, 1988; Chen & Siegler, 2000; Chen, Sanchez, & Campbell, 1997), scientiﬁc-reasoning strategies (Klahr & Nigam, 2004; Kuhn, Schauble, & Garcia-Mila, 1992; Schauble, 1990, 1996), computer programming (Klahr & Carver, 1988), analogical mapping (Honomichl & Chen, 2006; Kotovsky & Gentner, 1996; Siegler & Svetina, 2002), transitive inference (e.g., Goswami, 1995), symbolic understanding (Chen, 2007; DeLoache, 2004; Loewenstein & Gentner, 2001; Marzolf & DeLoache, 1994), and theory of mind (Flynn, O’Malley, & Wood, 2004). The distance or gap in terms of context and time interval between the original learning and transfer tasks in these studies is quite narrow. Although this research on near transfer has provided critical insights into important aspects of children’s thinking, we still know little about how broadly, ﬂexibly, and

Scientific-Reasoning and Problem-Solving Strategies in Children

421

effectively children transfer acquired strategies to remote situations, or about the processes involved in remote transfer. But remote transfer, when it does occur, is a critical measure of children’s learning, because it reﬂects how deeply children understand, how broadly they generalize acquired concepts and strategies to different situations, and how ﬂexibly they think and reason. Thus, the examination of the processes involved in children’s ability to transfer of strategies to distant situations should be a pivotal part of the research in children’s thinking because it has theoretical, empirical, and practical implications. With respect to theory, remote transfer studies provide critical tests for theories and models related to representation, learning, and intelligence. The process of remote transfer involves encoding information about the original learning situation, retrieving relevant information when encountering tasks that require similar strategies, establishing a mapping between the original and the new tasks, and applying the strategies to solve the current problems. The central questions are: (a) How do children embed source information in the original learning situations? (b) How do they represent information in a more general and ﬂexible fashion? (c) How do they retrieve relevant information from long-term memory when encountering a novel problem? (d) How do they map the structures between problems despite differences in superﬁcial features? and (e) How do they execute a solution in solving a target problem? Examination of children’s remote transfer thus address key issues that are closely related to the nature of representations and the retrieval of source analogues, and to theories of human learning and cognition. The empirical challenge, given the debates and mixed ﬁndings about the evidence of short-term transfer, is to demonstrate the existence of far transfer. Thus far, our knowledge of children’s ability to transfer from original to remote and novel situations is only intuitive and speculative, and there is little empirical evidence that children retrieve and use acquired structural information or solution strategies to solve relevant problems in different contexts after a long delay. Finally, there are practical educational implications that may be derived from an increased focus on far transfer, because the primary goal of education is to promote ﬂexible and broad transfer of concepts or strategies to other contexts and situations (e.g., Halpern, 1998). Given the pivotal role of transfer in education, it is critical to explore the issues of how students learn and to what extent they apply what they learn, and to examine factors that facilitate broad and ﬂexible generalization of knowledge across wide gaps. The issue bears upon the phenomenon of ‘‘inert’’ knowledge (Bransford et al., 1986; Bransford et al., 1989; Brown, 1989; Brown & Campione, 1981; Whitehead, 1929), deﬁned as ‘‘knowledge that is accessed

422

Zhe Chen and David Klahr

only in a restricted set of contexts even though it is applicable to a wide variety of domains’’ (Bransford et al., 1989, p. 472). The primary purpose of this chapter, therefore, is to demonstrate the extent to which children can demonstrate long-term transfer of higher-order cognitive skills over months and even years, and the ways in which such remote transfer depends on different acquisition conditions. In the following sections, we ﬁrst discuss the paradox about the existence of remote transfer. The paucity of empirical studies on remote transfer might be a consequence of conceptual obstacles, or of methodological challenges, or both. We thus describe a conceptual taxonomy of transfer for measuring transfer distance. Then we present extensive new evidence of remote transfer in children. These studies were conducted using two promising methods: an experimental paradigm in which children learned a problemsolving strategy and then were given the opportunity to use that strategy in solving isomorphic problems after a long delay, and the naturalistic, a cross-cultural method in which children learned a problem-solving solution in various cultural settings and their problem-solving strategies were assessed in a laboratory context many years later. Studies with these two approaches yielded new insights about developmental differences in remote transfer and revealed key processes/components involved in remote transfer. We then address questions concerning the relation between the acquisition and the application of strategies, and ﬁnally we address the educational implications of these ﬁndings and suggest some future directions for research on children’s remote transfer.

II. Paradox, Taxonomy, and Paradigms The intensive research on near transfer over the past century has generated more questions than answers to key issues, and has presented in particular a paradox—dubbed by Dunbar (2001) as ‘‘the analogical paradox’’—between intuitive plausibility and empirical evidence of remote transfer. Human thinking and problem solving is unquestionably inﬂuenced by experience, and thus it would be quite implausible to suggest that prior knowledge plays no role in solving novel problems. Indeed, cognitive development would be impossible unless children were able to use what they learn at one point in time and in one context when encountering relevant tasks at another point in time and in another relevant context. However, signiﬁcant transfer has proven difﬁcult to demonstrate. The research in mostly short-term transfer over the past century has yielded conﬂicting ﬁndings, showing both transfer successes and failures, and leading to extensive debates about the existence of transfer. Although

Scientific-Reasoning and Problem-Solving Strategies in Children

423

studies have demonstrated successful transfer even in young children (e.g., Brown & Kane, 1988; Brown, Kane, & Echols, 1986; Chen & Siegler, 2000), laboratory experimental results are often counterintuitive, showing that even older children and adults often do not make use of highly relevant information to which they are likely to have easy access. The narrowness of children’s learning and a lack of transfer has been reported often (e.g., Bransford, Brown, & Cocking, 1999; Cognition and Technology Group at Vanderbilt, 1997; Lave, 1988), and consequently, the existence of transfer has been debated extensively, as evident in the book ‘‘Transfer on Trial ’’ (Detterman & Sternberg, 1993). The paucity of empirical studies of remote transfer is partially due to the fact that a clear conceptualization of the underlying construct of ‘‘transfer distance’’ remains elusive. Without an articulation of the dimensions of transfer, it is impossible to compare the different distances of transfer in a meaningful way. Barnett and Ceci (2002) formulated a taxonomy suggesting that transfer distance involves two taxonomic factors of transfer. The ﬁrst general factor involves content, which includes the speciﬁcity– generality of the learned skills (e.g., transferring procedures or principles). The second factor involves context, which includes knowledge domain (e.g., biology vs. economics), physical context (school vs. lab), temporal context (time interval), functional context (e.g., academic vs. play), social context (individual vs. group), and modality (e.g., multiple choice vs. essay test). We adapt Barnett and Ceci’s taxonomy by separating ‘‘temporal context’’ from other aspects of context and combining content and knowledge domain. The three-dimensional transfer framework thus involves task similarity between the original source and target problems, context similarity between original learning and target problem solving, and time interval. The three key dimensions in deﬁning transfer distance are as follows: 1.

Task similarity: presence/absence of shared superficial and structural features between the source and target problems. The ﬁrst dimension is whether the source and target tasks share structural similarity and overlap in superﬁcial features. Superﬁcial similarity refers to objects and their properties, story protagonists, story lines, format of the problems (e.g., verbal vs. hands-on manipulation), or task domain (e.g., mathematical, physics, or social domains) common to both problems, whereas structural similarity refers to the underlying relations among the key objects shared by the source and target problems (Gentner, Rattermann, & Forbus, 1993). Extensive research in children’s analogical problem solving has shown that overlapped superﬁcial features and shared common structures provide cues for

424

2.

3.

Zhe Chen and David Klahr

spontaneous retrieval of source information and mapping of structural relations between problems (e.g., Brown, 1989; Chen, 1996; Daehler & Chen, 1993; Gentner et al., 1993; Goswami, 1996). Context similarity: same/different contexts in which the solutions or strategies are acquired and used. The context in which the original or ‘‘source’’ task is acquired or a target problem is solved involves physical and/or social aspects. Physical context refers to the location where the source analogues are acquired and the target problems/tasks are encountered, whereas social context involves the people and activities associated with the target problems/task. Either of these two types of contextual similarities may provide cues for retrieval of the source information when encountering a relevant problem. In examining analogical transfer across context, Spencer and Weisberg (1986) showed that college students experienced difﬁculty in solving a target problem in a setting (laboratory or classroom) that was different from the context (classroom or laboratory) in which a source problem was learned, demonstrating the important role of contextual similarity in transfer. Still, only a few studies have addressed the issue of immediate physical/perceptual context in infants and young children’s learning (e.g., DeLoache, 2004; Rovee-Collier, 1999), and we know little about the effects of this dimension on children’s transfer of problem-solving strategies. Time interval: short/long time gap between tasks, ranging from minutes to decades. Research on the development of memory reveals that, with an increasing time gap, it becomes increasingly challenging for infants, toddlers, and older children to recognize or recall the events to which they were exposed previously (e.g., Bauer, 1997; Ceci & Bruck, 1998; Rovee-Collier, 1999). However, despite the fact that the time gap between original learning and testing undoubtedly inﬂuences the retrievability and applicability of acquired information, this dimension has almost never been considered in investigating children’s transfer.

Figure 1 depicts a three-dimensional transfer distance space. In this space, the circle at the left-bottom corner represents a source problem, and the cubes represent target problems that differ on different degrees in these three dimensions. Transfer distance between the source and target problems can thus be represented in different regions of the space. The distance between the source problem and Target Problem A depicts a typical relation between source and target problems, in which source and target differ somewhat in superﬁcial task features, but are similar in context (e.g., the lab setting) with short-term delay. The relations between the source and

Scientific-Reasoning and Problem-Solving Strategies in Children

425

Fig. 1. A three-dimensional transfer distance space. The x-axis corresponds to the temporal interval between the learning of the source task and the application of that knowledge to the target task. The y-axis corresponds to the task similarity between the source and target. The z-axis corresponds to the contextual similarity between the initial learning situation and the transfer situation. Each of the labeled cubes in the space corresponds to a different transfer distance between source and the target. Point A corresponds to the most common form of transfer assessment: high contextual similarity, medium task similarity, and a short temporal interval between training and transfer assessment, whereas point E represents a transfer task with a long time interval, and substantial differences between source and target tasks and contexts. The other points in the space are described in the text.

other targets (Problems B, C, D, and E in Figure 1) represent the transfer distance in the remote transfer studies to be described in the following sections. There is little doubt that the degree of transfer is (a) inversely proportional to the size of the temporal interval between the original learning experience and subsequent target problem solving and (b) directly proportional to the degree of semantic and contextual similarity between problems. However, theorizing and modeling in analogical problem solving have been almost entirely based on experimental results of short-term transfer within the same contexts. Previous research in transfer has typically manipulated levels of superﬁcial similarity between learning and transfer problems while ignoring the last two dimensions. Results from such studies may thus exaggerate children’s transfer ability, given that transfer distance is deﬁned mainly only by the ﬁrst dimension. Conversely, without considering contextual similarity and time delay, it is difﬁcult to assess

426

Zhe Chen and David Klahr

children’s ﬂexibility and effectiveness in using learned concepts and strategies. As mentioned earlier, given the host of practical constraints involved in attempts to separate the presentation of source and target problems with a substantial time gap, most traditional studies of learning and transfer assess transfer performance within the same setting (i.e., laboratory context) and within a short interval. In this chapter, we describe studies in which we attempted to address these limitations by using two approaches to assess the transfer of problem-solving strategies after substantial delays and across distinct contexts. In one, an experimental approach, children acquired solutions or strategies in an original learning context, and their strategy use was then assessed using analogous tasks in different contexts with substantial time delays. This approach resembles the standard lab approach to studying transfer, except that it involves much longer durations between original learning and transfer assessment. Instead of delays of only a few hours or days, as in typical transfer studies, we assessed transfer many months after the initial acquisition of strategies and solutions and did so using a wide variety of test contexts. In the second, naturalistic, cross-cultural approach, problem-solving strategies were assessed in a laboratory context, years after children acquired relevant solutions from stories, tales, and events in various naturalistic, cultural settings. This approach relies on naturally occurring differences between stories and events to which children in different cultures were exposed. More speciﬁcally, exposure to a culturally speciﬁc ‘‘source analogue’’ was presumed to be nearly universal and repetitive in one culture, but extremely rare in another. The comparison of children’s performance in solving problems that are analogous to the speciﬁc experiences in one culture but not in another enabled us to examine remote transfer of analogous strategies acquired from speciﬁc cultural contexts. The use of this approach helps overcome the inherent obstacles in testing transfer across longer time gaps and contexts.

III. Evidence of Remote Transfer I: Experimental Approach The experimental approach to examining remote transfer described in this chapter focuses on transfer of scientiﬁc-reasoning strategies (or processing skills) in children. Scientiﬁc-reasoning involves basic inference processes in forming hypotheses, in designing experiments to test hypotheses, in distinguishing determinate from indeterminate evidence, and in interpreting results as evidence that support or refutes the hypotheses. Although children and even lay adults often have difﬁculty in

Scientific-Reasoning and Problem-Solving Strategies in Children

427

engaging in scientiﬁc thinking, studies conducted since the 1990s indicate that—when given many repeated opportunities to reason about patterns of evidence—children and adults are capable of acquiring scientiﬁc processing skills (e.g., Klahr, 2000; Kuhn et al., 1992; Kuhn et al., 1995; Schauble, 1990, 1996). Most of these studies used a microgenetic approach in order to examine late elementary-school childrens’ and adults’ acquisition of scientiﬁc experimentation skills during many sessions spread over a period of 10 weeks or more. Although these studies are signiﬁcant for demonstrating scientiﬁcreasoning strategies in both children and adults over an extended intervention period, they did not focus on the fundamental questions of whether and how younger children acquire and transfer scientiﬁc-reasoning strategies. Furthermore, the gap in time interval and contexts between the initial learning and transfer situations remained relatively narrow. A fundamental issue addressed in the studies described here concerns whether preschool and elementary-school children are capable of maintaining the strategies acquired in an initial learning situation and applying them to solve problems with long-term delays—of 7 months to 2 years— across contexts.

A. SEVEN- TO 10-YEAR-OLD CHILDREN’S LEARNING OF THE CONTROL OF VARIABLE STRATEGY IN DESIGNING EXPERIMENTS AND TRANSFERRING THE STRATEGY Our ﬁrst study (Chen & Klahr, 1999) addressed an important issue in scientiﬁc-reasoning and cognitive development: how children learn a domain-general processing strategy of designing valid experiments—the control of variables strategy (CVS)—and generalize it across various tasks in different contexts. We deﬁne CVS in both logical and procedural terms. The logical aspects of CVS include the ability to make appropriate inferences from the outcomes of unconfounded experiments as well as an understanding of the inherent indeterminacy of confounded experiments. Procedurally, CVS is a method for creating experiments in which a single contrast is made between experimental conditions. The full strategy involves not only creating such contrasts, but also being able to distinguish between confounded and unconfounded experiments. Previous studies (Bullock & Ziegler, 1999; Kuhn, 1995; Kuhn et al., 1995) indicated that elementary-school children had very limited understanding of CVS and demonstrated that the majority of ﬁfth and sixth graders produced mainly confounded experimental designs.

428

Zhe Chen and David Klahr

Given that CVS is a fundamental scientiﬁc-reasoning skill and that few elementary-school children spontaneously use it when they should, it is important to identify effective ways to teach CVS. One aim of this study was to examine which approaches work best in facilitating children’s acquisition of CVS, and to determine the extent to which children transfer CVS to situations beyond original learning context and after a long delay. For example, after learning how to design unconfounded experiments to determine various factors in the stretching of springs, are children able to utilize CVS in creating valid experiments dealing with balls rolling down ramps? Or do they apply the strategy only to situations that share very similar surface features? This study consisted of two parts. Part I examined children’s acquisition and relatively near transfer of CVS with the hands-on tasks. Part II was a paper-and-pencil posttest given several months after the initial learning phase. In Part I, with the hands-on tasks in each domain (task), children were asked to design experiments to test the possible effects of different variables. Several isomorphic, hand-on tasks were designed to help children master CVS. The tasks were physics based and involved springs, slopes, and sinking designs. In each task there were four variables that could assume either of two values. In each task, children were asked to focus on a single outcome that was affected by all four variables. For example, in the slope task (Figure 2), children had to generate test comparisons to determine how different variables affected the distance that balls traveled after rolling down a ramp. Children could set the steepness of the downhill ramps (either steep or low) using wooden blocks that ﬁt under the ramps. Children could also set the surface of the ramps (either rough or smooth) by placing inserts on the downhill ramps either carpet-side up or smooth wood-side up and they could set determine how far the balls rolled on the downhill ramp by placing gates at either of two positions at different distances from the top of the ramp (long or short run). Finally, children could choose from two kinds of balls: either rubber squash balls or golf balls. To set up a comparison, participants constructed two ramps, setting the steepness, surface, and length of run for each and then placing a ball behind the gate on each ramp. To execute a comparison, participants removed the gates and watched as the balls rolled down the ramps and came to a stop. The outcome measured was how far the balls traveled into the serrated stopping area. Figure 2 depicts a completely confounded comparison, as all four variables differ between the two ramps. The key question in this study was whether, after learning how to design unconfounded experiments to determine the inﬂuence of various factors in the ball’s rolling down ramps, children would be able to utilize CVS to create valid experiments dealing with different physical contexts, such as

Scientific-Reasoning and Problem-Solving Strategies in Children

429

Fig. 2. The Slope Domain. On each of the two slopes, children can vary the angle of the slope, the surface of the ramp, the length of the ramp, and the type of ball to determine the effect of one of these four variables on how far a ball rolls along on the serrated stopping area at the end of the slope. The confounded experiment depicted here contrasts (a) a golf ball on a steep, smooth, short ramp with (b) a rubber ball on a shallow, rough, long ramp.

springs stretching or objects sinking in water. That is, would they transfer the underlying CVS procedure from one physical context to a very different physical context having the same deep structure or would they only apply the strategy to the contexts in which they were trained? To test children’s learning and transfer, in Part I children were assigned to different instructional conditions, which differed in whether children received explicit instruction in CVS and whether they received systematic probe questions concerning why they designed the tests as they did and what they learned from the tests. Children in the training-probe condition received explicit training instructions for the ﬁrst comparison and were asked systematic probe questions about why they designed the test they did for each trial. Training included an explanation of the rationale behind controlling variables as well as examples of how to make unconfounded comparisons. After the test was executed, children were asked if they could ‘‘tell for sure’’ from the test whether the variable they were testing made a difference and also why they were sure or not sure. During training, the children were given both confounded and unconfounded examples and were asked to judge whether each example was a

430

Zhe Chen and David Klahr

good or bad comparison and to explain why. The experimenter then provided rationales for why it was not a good comparison. In the no training-probe condition, children received no explicit training, but they did receive the same series of probe questions surrounding each comparison as were used in the training-probe condition. Children in the no training–no probe condition received neither training nor probes. Using three isomorphic tasks, which shared the same structure but superﬁcial features, children were asked to make a series of four paired comparisons to test particular variables of each problem in four phases of the study: Exploration (e.g., the Slope Task), Assessment (e.g., different variables of the Slope Task), Transfer-1 (e.g., the Sinking Task), and Transfer-2 (e.g., the Spring Task). (Note that each of these physical domains involves ‘‘hands-on’’ set up and execution of children’s experimental designs.) The order of the tasks was counterbalanced. Children were presented with materials in a source task in which their initial CVS strategy was examined. Then their learning was assessed on the same initial task in the Assessment phase. This assessment measured children’s very near transfer, deﬁned as the application of CVS to test a new aspect (i.e., testing different variables) of the same materials used in the original learning problem. Children’s performance on Transfer-1 and Transfer-2 phases, with one novel task on each phase, was measured to assess their near transfer, deﬁned as the use of CVS to solve novel problems using a set of different materials that are still in the same general domain as the original problem within the same context after a short delay. This type of transfer resembles the traditional assessment of transfer, as Target Problem A illustrated in Figure 1. Children’s remote transfer was examined in Part II. Remote transfer refers to the application of CVS to solve problems with domains, formats, and contexts different from the original training task after a long delay, as illustrated in the relation between the source and Target Problem B in Figure 1. To determine whether and how children at different age levels and in different conditions change their strategies in designing experiments in Part I, the frequency of CVS use in each phase was examined. Children’s performance in designing valid experiments in each phase is illustrated in Figure 3. The analyses indicated that only children in the training-probe condition increased their performance over phases: Training-probe children did better in the Assessment, Transfer-1, and Transfer-2 phases than in the Exploration phase. They increased the use of CVS from about one-third of the trials in the Exploration phase (before training) to nearly two-thirds of the trials at the end of the hands-on phase. In contrast, children’s performance in the no training-probe and no training–no probe conditions did not signiﬁcantly improve over phase.

Scientific-Reasoning and Problem-Solving Strategies in Children

431

Training Tr 70 Training-Probe

% Correct CVS Use

No Training-Probe 60

No Training-No Probe

50

40

Transfer2

Transfer1

Assessment

Exploration

30

Phase Fig. 3. Children’s initial learning and near transfer: percentage of trials with correct use of CVS by phase and condition. The between-subjects training treatment occurred between the Exploration phase and the Assessment phase. Each line represents the mean of 2nd, 3rd, and 4th graders.

As Figure 4 illustrates, further analyses indicated that older children outperformed younger children and that children’s performance patterns in different grades differed in each training condition. Only third and fourth graders in the training-probe condition improved their performance over phases. Second graders’ performance improved somewhat between the Assessment and Exploration phases; however, their transfer performance was not better than the exploration performance. A similar pattern in age differences was also found in the no training-probe condition in that the third and fourth graders but not second graders showed some improvement over phases. Children in the no training–no probe participants showed no performance increases over phase for any grade level. Elementary-school children thus demonstrated an impressive ability to apply learned strategies across problems in Part I. With appropriate instruction, elementary-school children are capable of understanding, learning, and generalizing the strategy when designing and evaluating simple tests. At the conclusion of Part I, all children received explicit training so that their remote transfer could be assessed at a later date. Given the scope of this chapter, we focus on children’s remote transfer.

Zhe Chen and David Klahr

432

% Correct CVS Use

100

Training − Probe

80 60 40 20 0

% Correct CVS Use

100

No Training − No Probe

80 60

Fourth graders Third graders Second graders

40 20 0

% Correct CVS Use

100

No Training − No Probe

80 60 40 20 0 Exploration Assessment Transfer 1 Transfer 2

Fig. 4. Children’s initial learning and near transfer by grade, phase, and condition: percentage of trials with correct use of CVS.

Scientific-Reasoning and Problem-Solving Strategies in Children

433

Part II was designed to examine children’s ability to transfer CVS from the hands-on phases to remote situations after a 7-month delay. Third and fourth graders who had participated in Part I (except the second graders, who did not learn optimally in the earlier phases) and their classmates who had not previously participated were given a posttest in which they were asked to solve problems in various novel domains. The posttest tasks involved evaluating whether each of a series of paired comparisons was a good test of the effect of a speciﬁc variable (Figure 5). All children who had participated previously were considered the experimental group in the remote transfer test, while their classmates who had not participated made up the control group. We consider the application of the general strategy, CVS, in the posttest as ‘‘remote transfer’’ for several reasons: (1) the tasks differed in several aspects, including the format (generating tests in the hands-on tasks vs. evaluating tests on paper in the posttest), content (mechanical vs. other types of domains; involving four variables vs. three variables of each problem); (2) the hands-on phase and posttest differed substantially in the testing context (working with an experimenter in the lab vs. their science Does the amount of water make a difference in how well a plant grows? Plant A

Plant B

Lots of Sunlight

No Sunlight

Lots of Water Lots of Plant Food

A Little Plant Food

A Little Water

Good Test Bad Test Fig. 5. An example of the CVS remote transfer task. The figure depicts one of the items in the 15-page test booklet. Children had to indicate whether they thought the depicted comparison was a ‘‘good test’’ or a ‘‘bad test’’ for the focal question. This example is a ‘‘bad test’’ because it is confounded.

Zhe Chen and David Klahr

434

teacher in the science classroom); and (3) 7 months elapsed between the hands-on experiences and the posttest. The difference between this posttest and the original learning represents a far region of the transfer distance space (Target Problem B in Figure 1). The main dependent measure in Part II was the number of correct responses to the 15 posttest problems. Fourth graders in the experimental conditions outperformed those in the control condition, but conditions did not differ signiﬁcantly for third graders. Another measure of remote transfer involved the percentage of ‘‘good reasoners’’ in the experimental and control conditions. Children who made 13 or more correct judgments out of a total of 15 problems (over 80% instead of 100% of correct strategy use to allow random errors) were considered good reasoners. As shown in Figure 6, 62% of children in the experimental group were categorized as good reasoners, compared to only 19% in the control group. Separate analyses at each grade level revealed that the experimental and control conditions differed in the percentage of good reasoners at grade 4, but not grade 3. This study thus reveals developmental differences in learning and transfer of scientiﬁc-reasoning skills. Second graders, like older children, learned 100 Control Experimental

% Good Reasoners

80

60

40

20

0 Third Grade

Fourth Grade

Fig. 6. Children’s remote transfer performance: percentage of Good Reasoners (13 out of 15 correct) on the posttest administered after a 7-month delay.

Scientific-Reasoning and Problem-Solving Strategies in Children

435

CVS when the transfer tasks were within the same domain as the initial training. Third graders demonstrated the ability to transfer CVS across problems within the domain of mechanics (i.e., when reasoning about the springs, slopes, and sinking tasks) and with a short-term delay (one week). Only fourth graders displayed remote transfer. With age, children increasingly improve their ability to transfer learned strategies to remote situations; and children’s transfer became increasingly ﬂexible and broad.

B. FOUR- AND 5-YEAR-OLDS’ UNDERSTANDING OF INDETERMINACY AND TRANSFER OF STRATEGIES AFTER A 7-MONTH DELAY Children’s ability to transfer learned strategies to remote situations undoubtedly is inﬂuenced by the complexity of the tasks that they face. Although second graders were ineffective in relatively near transfer (e.g., in transferring CVS from the Slope Task to the Spring Task) and third grader’s performed poorly on a remote transfer task (from the original learning to posttest problems) perhaps even younger children could demonstrate transfer if given age-appropriate tasks. In a subsequent study (Klahr & Chen, 2003), we further examined remote transfer of scientiﬁc strategies in younger children. In this study, we focused on the time gap between the learning and transfer phases while minimizing the transfer differences between the tasks and contexts. The transfer distance is illustrated by point C in Figure 1: only the temporal interval between initial learning and posttest was substantial. We examined 4- and 5-year-olds’ learning and transfer using an age-appropriate task involving children’s assessment of the informativeness of different patterns of evidence. The ability to distinguish determinate from indeterminate evidence is a prerequisite to appreciating the logic of evidence-theory relations and confounded vs. unconfounded experimental designs. We addressed the issue of children’s acquisition and transfer of the concept of indeterminacy by presenting 4- and 5-year-olds with various isomorphic tasks (adapted from Fay & Klahr, 1996) as depicted in Figure 7. For example, in the marker task (Figure 7b), children saw a target object (e.g., a sketch of an animal) and a set of covered boxes each of which contained one marker. At the outset of each trial, all boxes were closed. Then they were opened sequentially, revealing the color of the marker contained in the box. Prior to the opening of each box, children were asked whether and why they ‘‘knew for sure,’’ or ‘‘would have to guess,’’ about which box contained the marker that was used to draw the animal. In the speciﬁc example shown in Figure 7b, the target is a sketch of an elephant drawn in black, and there are

Zhe Chen and David Klahr

436

(a)

(b)

Fig. 7. Examples of materials used for determinacy/indeterminacy tasks: (a) ‘‘Stamps’’ problem. The task is to determine whether one can know for sure which stamp was used to make the bicycle image at the bottom. In Fig. 7a the identity of two of the stamps has been revealed: the first stamp (bicycle) matches the target and the second (soccer ball) does not, but the identity of the third stamp remains unknown because it is still hidden under the while tag. Thus, this is an example of a positive-capture problem. (b) ‘‘Marker’’ problem. The task is to determine whether one can know for sure which marker was used to make the sketch of the animal at the bottom. The identity (color) of two of the markers has been revealed: the leftmost marker is orange and does not match the color of the drawing (black) and the second one (black) does match, but the color of the third marker remains unknown because the marker cover has not yet been removed. Thus, this is also an example of a positive-capture problem.

three boxes, containing, respectively, orange, black, and black markers. We use the symbols ‘‘?,’’ ‘‘+,’’ and ‘‘-’’ to denote, respectively, a closed box, an open box that matches the target, and an open box that does not match. For the problem just described, the sequence of evidence patterns is as follows: before any of the boxes are opened: ???; after the ﬁrst (leftmost) box, an orange, non-matching marker, is opened: -??; after the second box, containing a black, matching marker, is opened: -+?; and ﬁnally, after all boxes are opened: -++. This problem is indeterminate throughout. That is, there is never sufﬁcient evidence to eliminate the uncertainty about which marker was used to draw the elephant. In contrast, although all problems start out with indeterminate evidence patterns, some become fully determinate after the ﬁnal box is opened (e.g., -+- ). There was one evidence type for which only about 10% of 4-year-olds’ and 15% of 5-year-olds’ initial responses on the pretest were correct across conditions (Klahr & Chen, 2003). Such patterns included at least one closed

Scientific-Reasoning and Problem-Solving Strategies in Children

437

box and a single positive instance (e.g.,+- ?). On such problems, children incorrectly responded ‘‘know’’ approximately 80% of the time when they should have said ‘‘guess.’’ Fay and Klahr (1996) called this the positive-capture strategy because the single positive instance seemed to capture children’s attention and, in effect, blind them to the fact that the unexplored option might yet render the problem indeterminate. The positive-capture strategy resembles inclusion errors in scientiﬁc-reasoning (e.g., Kuhn et al., 1995), which involve attribution of causal relations to a variable that covaries with the outcome on only a single occasion. For example, in a task in which a set of manipulable features inﬂuence the speed with which boats are pulled down a towing tank by a weight and pulley system, children often make co-occurrence inclusion errors when they concluded that a feature (e.g., sail size) played a causal role in the speed based on a single instance (large sail co-occurred with fast speed), although there is no effect of this variable in the causal structure of the task. Younger children are particularly likely to generate this kind of error, but 11- and 14-year-olds and even adults often use incorrect single-instance inclusion inferences (Kuhn et al., 1995). Thus, a form of the positive-capture strategy appears to cause difﬁculty in scientiﬁc-reasoning well into adulthood. Given the consequences of failing to recognize indeterminate situations in scientiﬁc-reasoning, it is important to investigate whether and how, with experience, young children eventually replace it with a more advanced strategy. For investigating children’s performance on a particularly difﬁcult acquisition, we used a microgenetic training approach in which we provided explicit and systematic feedback on all pattern types over several days in several phases, including a pretest (e.g., the marker task on Day 1), two learning phases (e.g., the stamp task shown in Figure 7a for learning phases I & II on Day 2 and Day 3), a posttest (e.g., an analogous task in which children were shown a necklace of colored beads as well as boxes containing one color of bead, and were asked to identify the box that was used to make the necklace), and a delayed transfer test, with 5 problems on each phase. The ﬁrst four phases took place approximately a week apart, and the delayed transfer test took place 7 months later. Children were assigned to two learning conditions. In the learning phases, for the children in the training condition, correct and incorrect responses were pointed out, and the experimenter provided the rationale for the correct answer immediately after the child responded to each question. That is, children received immediate, explicit, and consistent feedback after each box was opened. In the control condition, children received no explicit feedback from the experimenter about whether their response was correct. Various isomorphic tasks with different materials were used to explore whether and how young children at different ages transfer the learned strategy from one

Zhe Chen and David Klahr

438

task to another. The delayed transfer test took place approximately 7 months after the earlier learning phases. It was similar to the markers task described previously, except in the task presentation format, in that the experimenter and participants traveled throughout the room to pre-made pictures set on little tables in front of the marker boxes. Not surprisingly, children at different ages responded differently to problem-solving experience in the control condition and the explicit feedback in the training condition. As Figure 8 shows, the analyses of learning on the positive-capture problems (a) children at both ages and in both conditions started with similar inaccurate performance; (b) children in the training condition improved much more than their peers in the control condition; (c) 5-year-olds improved more than 4-year-olds throughout the learning phase; and (d) by the follow-up phase 7 months later, 4-year-olds’ performance decreased almost to pretest levels whereas 5-year-olds’ performance remained quite accurate. Developmental differences in strategy acquisition and generalization were therefore evident in several aspects: older children learned more readily from training, improved their performance to a greater extent, and transferred the learned correct strategy across tasks (e.g., from boxes to markers) more effectively, whereas for younger children the short-term effectiveness of training was not long lasting. This study shows that although the positive-capture strategy is a robust phenomenon in young children’s reasoning, 5-year-olds were able to learn from both problem-solving experience and especially from explicit feedback 100

5-year-olds, training

% Correct Responses

4-year-olds, training

75

5-year-olds, control 4-year-olds, control

50

25

0 Pretest

Learning I Learning II Posttest

Follow-up

Phase Fig. 8. Children’s learning and transfer of correct strategies: percentage of correct responses to positive-capture problems over phase by condition and age. Learning I and II took place on Days 2 and 4, respectively. Posttest took place on Day 4, and Follow-up took place 7 months later.

Scientific-Reasoning and Problem-Solving Strategies in Children

439

but 4-year-olds’ performance on indeterminacy problems beneﬁted somewhat from explicit feedback but not from implicit feedback. Older children’s learning proved more effective in magnitude, rate, and generalization. Although 5-year-olds were initially unable to process positivecapture problems, they learned readily from feedback to recognize and correctly explain both determinate and indeterminate situations, and the acquired strategy was effectively generalized across different tasks with a 7-month delay. Thus, compared to 4-year-olds, 5-year-olds improved their performance to a greater extent, learned faster, and generalized the newly acquired strategy to a broader range of indeterminacy tasks and to more remote situations. In both the CVS and indeterminacy studies just described, the time gap between the initial learning and remote transfer phases, although substantially longer than the intervals reported in most previous studies with children and adults, was still relatively short (7 months). To further explore young children’s remote transfer of scientiﬁc-reasoning strategies, we next examined young children’s strategy transfer after a substantially longer time gap, in order to compare more directly children’s shorter- and longer-term transfer.

C. SIX- TO 8-YEAR-OLDS’ TRANSFER OF A HYPOTHESISTESTING STRATEGY AFTER A 24-MONTH DELAY The ability to generate a conclusive test for a hypothesis is a critical component of scientiﬁc-reasoning processing (e.g., Klahr & Dunbar, 1988; Klayman & Ha, 1987). Early investigations of children’s ability to understand hypotheses and evidence suggest that only by 11 or 12 years of age do children start to master the conceptual and procedural strategies required to seek and evaluate evidence that can either conﬁrm or disconﬁrm hypotheses (e.g., Kuhn et al., 1988). In contrast, some studies indicate that when presented with two relatively simple and mutually exclusive hypotheses, even early elementary-school children can correctly choose the ‘‘experiments’’ that conclusively determine which hypothesis is correct (e.g., Ruffman et al., 1993; Sodian, Zaitchik, & Carey, 1991). Sodian et al. (1991) told ﬁrst and second graders a story in which they had to determine whether a big mouse or a small mouse was eating food in the kitchen each night. Children were told that they could place food in either of two houses: one with a big door that could accommodate either a large or small mouse, and one with a small door that could accommodate only a small mouse. Most children correctly reasoned that they should put the food in the house with the small door: if the food was gone in the

440

Zhe Chen and David Klahr

morning, it must be a small mouse. Sodian et al. also demonstrated that children’s responses were not simply due to a preference for the small house, because when the goal was changed from determining the size of the mouse to being sure that the mouse would be fed, they chose to use the house with the large door. In a second task, children needed to ﬁgure out how to determine whether a pet aardvark had a good or poor sense of smell. A good portion of children in both grades generated spontaneous tests (e.g., placing food far away) for a hypothesis, suggesting that they had some nascent understanding of the fundamental logic of hypothesis testing. However, young children’s understanding of hypothesis testing involving choosing from two alternatives appeared very limited as shown in this study; only a little over a quarter of these children demonstrated the ability. Furthermore, the two tasks were presented one after another, and thus children might have learned from the experience after encountering the ﬁrst task. The Sodian et al. (1991) study provides an important ‘‘snapshot’’ of young children’s hypotheses-testing strategies at a given point in time. In order to extend their ﬁndings, we focused on young children’s acquisition and transfer of such hypothesis-testing strategies (Chen, Mo, & Klahr, in preparation). Challenging tasks were designed to examine the ability of kindergartners and ﬁrst and second graders to acquire and transfer the fundamental logic of hypothesis testing. The basic task involved designing an adequate test for a hypothesis by choosing a correct item from three options. For example, in one of the isomorphic problems, the ‘‘Who sank the boat?’’ task (Figure 9), children were presented a story in which a ﬁsherman needed to ﬁgure out whether a big or small bear messed up his boats at night. The solution involved leaving one of his three different-sized boats in the water (the one that a big bear, but not a small bear, could sink). If it was sunk by morning, then the big bear must be responsible. The general approach involved presenting three isomorphic tasks which involved testing hypotheses related to the size, weight, volume, or height of the target objects. Figure 9 illustrates an example of the ‘‘Who sank the boat?’’ problem (a weight task). The story was read to the child while the props were presented: ‘‘A ﬁsherman has three boats of different sizes. One morning, he discovered some bear messed up his boats, but he did not know whether it was a small or big bear. He thought it might be a big bear, and thought of a way to ﬁgure out if it was a big or small bear by using his boats.’’ In the ﬁrst trial, the child was told: ‘‘Here are the three boats that the ﬁsherman had. The ﬁsherman knew that a big bear could sink all the boats (A, B, and C), but a small bear could sink only these two (B and C).’’ After conﬁrming that the child understood and remembered the relations between the bear and the three boats, the child was asked to help the

Scientific-Reasoning and Problem-Solving Strategies in Children

A

B

441

C

Fig. 9. An example of the hypothesis testing tasks: Illustration of the toy props used to accompany and illustrate the ‘‘who sank the boat’’ task (a weight task). The child was told that, for example, a big bear would sink all three boats, but a small bear could only sink the two smaller boats. Which size boat should be left out overnight to determine which of the two bears was the culprit?

ﬁsherman ﬁgure out a way to test the hypothesis, and his/her strategies were assessed. The other three analogous tasks were an area task (‘‘Who ate the food?’’ which was adapted from Sodian et al., 1991), a volume task (‘‘Who spilt the water?’’), and a height task (‘‘Who took the gift?’’). The height task was used during the posttest; the toy props used in this task are illustrated in Figure 10. The four tasks shared parallel problem structure and logic (all involving designing an adequate test for the hypothesis by selecting an appropriate item from alternatives based on the given relations between items and the target protagonist), but involved different objects, protagonist, and story line. The early task(s) served as analogue(s) for later one(s). Children’s relatively near and remote transfer was assessed by comparing their initial performance, short-term transfer (learning within the initial phase), and remote transfer (12-month or 24-month-delay posttests). To examine how various types of feedback affects remote transfer, a total of 110 kindergartners, ﬁrst graders, and second graders from China were assigned to one of the three conditions distinguished by the type of feedback provided. On each of two trials for each task, depending on

Zhe Chen and David Klahr

442

A

B

C

D

Fig. 10. Materials used for the posttest task: ‘‘who took the gift’’ (a height task). The child was told that, for example, the tall giraffe could remove a gift from any of the trees, but the small giraffe could only reach the two smaller trees. Where should the gift be placed?

different conditions, questions were asked and/or prop demonstrations were provided. These questions and demonstrations served as feedback as well as providing an assessment of children’s strategies in solving the problems. The children at each age level were assigned to one of the three conditions. At the end of each trial, the children in the verbal and physical feedback (explicit) condition received verbal instruction and physical demonstration with props that a big, but not a small bear would sink a speciﬁc boat. The verbal instruction illustrated how and why a correct choice would allow one to conclusively test the hypothesis: ‘‘If the ﬁsherman chooses the biggest boat that the big bear could sink but the small bear could not, and if it was sunken by the morning, then he would be able to ﬁnd out whether it was the big bear sank it. If he chooses a boat that both bears could sink, or a boat that either could sink, the ﬁsherman would not be able to ﬁnd out.’’ In the physical feedback (intermediate) condition, only physical demonstration, but not verbal explanation was provided. Physical demonstration involved showing a correct choice and then an incorrect choice with the props and asking questions concerning why it was a good or bad choice. For example, in the case of a correct choice, the experimenter showed how

Scientific-Reasoning and Problem-Solving Strategies in Children

443

an appropriate boat works: ‘‘Let’s see what happens if the ﬁsherman chooses this boat (Boat A). See, if the big bear steps on, the boat sinks. If the small bear steps on, it does not sink the boat. If the ﬁsherman ﬁnds that the boat has been sunken in the morning, would he be able to tell whether it was the big bear that did so? And why?’’ In the implicit feedback condition, children received no explicit feedback, but the experimenter’s very speciﬁc questions served as implicit feedback. During the learning phase, children were tested individually with three analogous tasks, with two trials within each task. The trials within a task differed in the relations between the bear and boats. For example, on one trial, a big bear could sink the medium boat but not the big boat, and a small bear could only sink the small boat on the ﬁrst trial. On another trial, a big bear could only sink the big boat, and a small bear could sink the medium and small boats. To ensure that children remember the relations, on each trial, after the relations were presented and before the inference questions were asked, the child was asked to remind the experimenter which boat (or box) a big bear (mouse) could sink (enter), and which boat a small bear could sink. In order to examine remote transfer, at the end of the learning and near transfer phase, all children were trained with the explicit instruction: the verbal and physical feedback at the end of the third (last) trial. During the posttest—remote transfer one or two years later—only kindergartners and ﬁrst graders participating in the verbal and physical and physical only conditions were tested in three trials of the posttest task. The posttest task is illustrated as point D in Figure 1, which represents a substantial distance between the initial learning phase and the posttest in task and context similarity, and in temporal interval (i.e., with a 12- or a 24-month time gap). Children in the implicit feedback condition, although ultimately receiving explicit feedback, did not learn to the same degree as those in the experimental conditions and thus were not included on the posttest. The posttest task would be less challenging when the second graders become fourth graders two years after they participated in the learning phase, and the second graders were thus not included on the posttests. A portion of the kindergartners and ﬁrst graders participating in the two experimental conditions during the learning phase were tested one year later (ﬁrst and second graders when the one-year-delay posttest was conducted). Classmates who never participated in the study served as a control group. Other kindergartners and ﬁrst graders (second and third graders when the two-year-delay posttest was conducted) were tested 24 months later. Again, classmates who never participated in the study served as a control group. Neither verbal nor physical feedback was provided at the 12- or 24-month posttest. Children’s understanding of

444

Zhe Chen and David Klahr

hypothesis testing during each trial was again measured at three levels. The task used in this posttest was ‘‘Who took the gift?’’ (Figure 10). The experimenter’s series of questions served as implicit feedback as well as assessment of children’s strategies. Children’s strategies for testing a hypothesis were assessed at three levels at each trial. The most sophisticated and challenging level is the Spontaneously Generating Hypothesis Test. The hypothesis that a big bear was sinking the boats was provided, and children had to generate a conclusive test which involved leaving one of the boats in the water that only a big bear, but not a small bear, could sink. The experimenter ﬁrst formed a hypothesis and asked the child to generate a test for the hypothesis: ‘‘The ﬁsherman thought that the bear might be a big one. Now what could he do to ﬁnd out whether it was a big bear that has sunken the boat overnight?’’ An example of a correct answer would be: ‘‘He would leave the biggest boat in the water, and if it was sunken by the morning, it must be a big bear that did it.’’ Given the scope of this chapter, we report results only concerning the most general and challenging level of assessment of children’s strategies in testing hypotheses. Overall, children acquired hypothesis-testing strategies with experience. As Figure 11 shows, during the learning phase, second and ﬁrst graders learned more effectively than kindergartners to spontaneously generate a hypothesis-testing strategy. Few children generated an appropriate test for the hypothesis on the ﬁrst trial of the ﬁrst task. However, even kindergartners improved their performance on the second trial and the subsequent tasks. Older children learned more effectively: across conditions, over four-ﬁfth of the second graders and nearly two-thirds of the ﬁrst graders generated an appropriate test, but only about one-fourth of kindergartners did so on the last task. Different types of feedback from implicit to explicit (from implicit feedback from the speciﬁc questions that an experimenter asked, to more explicit feedback from physical demonstration, to very explicit feedback of verbal rationale) seem to have differential effects on learning at different age levels. Effects of condition were evident especially in kindergarten. Kindergartners in the verbal and physical feedback condition performed more effectively than those in the physical feedback condition, who in turn, performed better than those in the no feedback condition for Trial B of Task 1, and for Tasks 2 and 3. This indicated that explicit feedback (verbal and physical feedback) was more effective than the intermediate feedback (physical feedback), which was more effective than implicit feedback, especially for younger children. Nevertheless, even implicit feedback (either encountering the experimenters’ speciﬁc questions or physical feedback) facilitated ﬁrst graders in all three levels of understanding.

% Correct

Scientific-Reasoning and Problem-Solving Strategies in Children

100

100

80

80

60

60

40

40

20

20

0

445

0 1A

1B

2A 2B Trial

3A

3B

a. Kindergarten

1A

1B

2A 2B Trial

3A

3B

b. First Grade

100

% Correct

80 No Feedback

60

Physical Feedback 40

Verbal & Physical Feedback

20 0 1A

1B

2A

2B

3A

3B

Trial c. Second Grade Fig. 11. Children’s performance in generating a correct hypothesis test during the learning phase. Percent correct responses by trial and condition. (a) Kindergarten; (b) First grade; (c) Second grade.

One primary aim of this study was to examine whether children in the experimental conditions of the initial learning phase outperformed their peers who were not exposed to the initial tasks in the control condition. As shown in Panel a of Figure 12, when encountering the posttest task, few ﬁrst graders in either condition on the 12-month-delay posttest spontaneously came up with correct hypothesis tests on the ﬁrst trial. In contrast, about one-third of the second graders in the experimental condition used a correct hypothesis-testing strategy on the ﬁrst trial, as compared to only a few of their peers in the control group. Converging results were also obtained on the 24-month-delay posttest (Panel b of Figure 12). These ﬁndings suggest that kindergartners and ﬁrst graders learned

Zhe Chen and David Klahr

446

First Grade, Control

100

First Grade, Experimental Second Grade, Control

(a) 12-monthdelay

% Correct

75

Second Grade, Experimental 50

25

0 Trial A

Trial B

Trial C

Trial Second Grade, Control

100

Second Grade, Experimental Third Grade, Control

(b) 24-monthdelay

% Correct

75

Third Grade, Experimental 50

25

0 Trial A

Trial B

Trial C

Trial Fig. 12. Children’s performance in generating a correct hypothesis test on the posttest after a 12-month or 24-month delay. Percent correct by grade and condition. (a) 12-month delay; (b) 24-month delay. Note: A portion of the kindergartners and first graders participated in a posttest after a 12-month or 24-month delay. Thus, the participants in the 12-month delay group (and the control group) were first and second graders when the posttest was performed, whereas the participants of the 24-month delay group (and the control group) were second and third graders.

hypothesis-testing strategies from the analogous tasks in the learning phase, and transferred the strategies when they encountered an isomorphic problem in a different context even after one or two years later. Another aim was to compare children’s performances between initial learning and transfer after a one- or two-year delay. The comparison focused on second graders’ performance between the last (third) task (a total of two trials) during learning phase, the ﬁrst two trials of the oneyear-delay posttest, and the ﬁrst two trials of the two-year-delay posttest. Second graders’ strategies used to solve the third task during the learning phase were considered as near transfer because they had learned the

Scientific-Reasoning and Problem-Solving Strategies in Children

447

hypothesis-testing strategy when solving the ﬁrst two tasks. For the second graders (in the verbal and physical feedback and the physical feedback conditions when solving the third task), 85% of their strategies generated to test the hypotheses were correct (Figure 11, Panel c, combined over the two trials of the task), indicating that they readily transferred the hypothesistesting strategy from the ﬁrst and second to the third task during the learning phase. In contrast, second graders (who were ﬁrst graders when participating in the learning phase) generated 58% correct strategies in solving the one-year-delay posttest task (Figure 12, Panel a, combined over the ﬁrst two trials of the task), and second graders (who were kindergartners when participating in the learning phase) generated 38% correct strategies in solving the two-year-delay posttest task (Figure 12, Panel b, combined over the ﬁrst two trials of the task). This pattern of performance could be interpreted as differences between near and remote transfer, and between the one-year and two-year-delay posttests. The differences in transfer performance might have also been due to their differential initial learning, because the second graders whose performance was assessed on the ﬁrst posttests were in ﬁrst grade when they participated in the initial learning phase one year earlier, and the second graders whose performance was assessed on the second posttest were kindergartners when they participated in the initial learning phase two years earlier. This dilemma has not been resolved. Nevertheless, the performance pattern suggests that when time elapses, transfer performance declined. Moreover, although their long-term transfer performance was less impressive than their short-term transfer performance, young children nevertheless demonstrated the ability to generalize the learned hypothesistesting strategy to different contexts even after a two-year delay. The studies with an experimental approach described herein were designed to assess young children’s remote transfer of scientiﬁc-reasoning strategies (e.g., CVS, hypothesis testing, and distinguishing determinacy/ indeterminacy strategies) after long delays. The ﬁrst experiment demonstrated that fourth graders readily transferred the strategy to design valid experiments from the original learning situation in the laboratory to the science classroom context 7 months later. The transfer between the initial learning and the posttest tasks in the ﬁrst experiment involved substantial distance in all three dimensions: task and context similarity, and long temporal interval, as illustrated by point B in Figure 1. Third graders failed to transfer presumably because they might have failed to notice the similarities between the early hands-on tasks and the paper-and-pencil problems and/or because they experienced difﬁculties in executing the acquired strategies. We further examined even younger children’s remote transfer by adapting an age-appropriate scientiﬁc-reasoning task in the

448

Zhe Chen and David Klahr

study of young children’s evaluation of indeterminate evidence. As illustrated by point C in Figure 1, the posttest problem was similar to one of the initial learning tasks, and encountered in a similar context. Kindergarteners proved capable of transferring a determinacy/indeterminacyreasoning strategy learned 7 months earlier whereas 4-year-olds failed to do so even when the learning and testing tasks are very similar and administered in similar contexts. In the ﬁnal experiment, on children’s hypothesis testing, we extended these ﬁndings and tested more distant transfer in all three dimensions in kindergartners’ and early elementaryschool children. The transfer distance in this experiment is illustrated by point D, which represents substantial distance in task and contextual similarity, and in temporal interval (i.e., a 2-year time gap) between the learning phase and the posttest. Even kindergartners successfully transferred a hypothesis-testing strategy to structurally similar but superﬁcially dissimilar task in a different context two years later. It was also evident that transfer distance in time gap inﬂuenced transfer performance. Overall, these ﬁndings demonstrate that young children are able to transfer scientiﬁc-reasoning strategies acquired in one context to novel situations superﬁcially different from the original context. Young children were able to retrieve relevant information and to use it to solve problems that were structurally similar but superﬁcially dissimilar to those they encountered 6, 12, or even 24 months earlier in different contexts. These studies suggest that children’s learning is broader and more ﬂexible than previously assumed. The strategies that they learned do not automatically become constrained to the original learning situation, and they do not decay rapidly. Instead, children as young as ﬁve years of age proved capable of transferring acquired strategies to solve problems with different perceptual and contextual features with a time gap as long as two years.

IV. Evidence of Remote Transfer II: Naturalistic, Cross-Cultural Approach Despite evidence of remote transfer from the studies described in Section III, the time gap between the learning and transfer was about only one or two years, and the contexts in which the strategies or concepts were learned and transferred typically involved in school settings, and thus were not substantially diverse after all. To overcome these limitations, we adopted a naturalistic, cross-cultural approach to address the issues of whether and how individuals retrieve and use source analogues that were acquired in naturalistic, speciﬁc cultural settings many years ago.

Scientific-Reasoning and Problem-Solving Strategies in Children

449

The naturalistic, cross-cultural studies were thus designed to explore longterm analogical transfer of problem solutions presented in folk tales that middle- and high-school students heard during their childhood. The transfer distance involved in the naturalistic cross-cultural study is therefore vastly more substantial than that involved in the experiments reported above in all three key dimensions, as illustrated by point E of Figure 1. As mentioned previously and illustrated in Figure 1, the typical experimental paradigm for examining analogical problem solving involves presenting a source analogue and then a target problem shortly afterward within a lab setting. One reason for the predominant use of this paradigm is that it is impractical and unrealistic to provide a speciﬁc source problem and its solution, and then present analogous problems for participants to solve with an extended time gap in a different setting. It thus seems obvious that extending previous ﬁndings regarding analogical transfer to more naturalistic contexts and with more extensive time gaps between problems would have both theoretical and practical implications. Dunbar and colleagues (e.g., Dunbar, 2001) have explored how analogy is used ‘‘in vivo,’’ that is, in naturalistic settings, such as scientiﬁc laboratory meetings and political speeches. Blanchette and Dunbar (2001) observed use of analogies by leading molecular biologists and immunologists during their lab meetings and found that—unlike the typical scenarios manipulated by experimental psychologists, in which participants typically use source information to solve problems superﬁcially similar to the source—real scientists in real working contexts often use structural features and higherorder relations in analogizing during the discovery process. In one example taken from their study, a scientist is investigating the way that HIV works and obtains a very strange result. To explain what happened, the scientist spontaneously draws an analogy to a genetic mechanism found in heatresistant bacteria. Structural analogies thus are a frequent, rather than a rare, phenomenon when human reasoning and problem solving are investigated in naturalistic settings. The difﬁculties of demonstrating analogical transfer in psychological experiments, along with the ﬁndings of rich use of analogy by Dunbar and colleagues, point to the importance of developing effective approaches to examine how analogical transfer occurs across contexts and with long delays. A naturalistic approach to explore analogical problem solving seems promising for overcoming these methodological limitations. Such a naturalistic approach allows the examination of analogical problem solving with more extensive time gaps and across contexts. Toward this end, Chen, Mo, and Honomichl (2004) conducted a series of studies using a novel method, which entails cross-cultural comparisons of problem solving. Speciﬁcally, middle-school, high-school, and college students from China

450

Zhe Chen and David Klahr

were asked to solve a problem that was analogous to a widely known tale introduced in naturalistic settings, such as reading from a book or listening to a story during childhood many years ago. The target problem-solving performance of Chinese students was compared to those from a different culture (U.S.) who had never heard of the source story. The source story problem was adopted from a Chinese tale, entitled ‘‘Weigh the Elephant,’’ which describes a scenario in which a Chinese emperor was faced with a difﬁcult task of weighing a large elephant without the beneﬁt of a proper scale. However, the emperor’s son solved the problem by putting the elephant in a boat, marking the water level on the boat, and then replacing the elephant with small stones so that the water level reached the same mark (a compression strategy). The son then weighed the stones separately with a small scale and then totaled all the weights of the stones. Thus, this solution involves a principle of equalizing the weights of the smaller items and a large object by compressing the boat to the same degree. A preliminary study indicated that over 90% of Chinese students could recall the key elements (goal and solution) of the story, suggesting the availability of the source tale. In an experiment with college students to determine the effects of culturally speciﬁc experience on long-term transfer across contexts, folk tales from China and the U.S. were chosen as source analogues. Target problems, which were isomorphic to the source tales, were created. One target problem, the ‘‘statue’’ problem, described a scenario in which a chief of a riverside village needed to ﬁnd a way to measure an amount of gold commensurate in weight to a statue, without the beneﬁt of a conventional balance scale. This problem was designed to be isomorphic to the wellknown ‘‘Weigh the Elephant’’ tale. The solution involved a principle of equalizing the weights of the smaller items and a large object by compressing the boat to the same degree. Another target problem was the ‘‘Cave’’ problem, which described a scenario in which a treasure hunter needed to travel into a cave and then ﬁnd his way out. The solution involved leaving a trail of sand while traveling through the cave and following this trail out to exit. The Cave Problem was created to be analogous to the tale ‘‘Hansel and Gretel,’’ by the Brothers Grimm, which is commonly read or heard by children in the U.S. but virtually unheard of by Chinese children. In the Hansel and Gretel tale, a brother and sister tried to ﬁnd their way out of a forest by creating a trail with pebbles and bread crumbs. The target problems and the source analogues involved similar solutions, but they shared few contextual and superﬁcial features, such as similar objects and characters. Chinese college students, who had experienced the elephant tale as children, were predicted to be more likely than U.S.

Scientific-Reasoning and Problem-Solving Strategies in Children

451

students to come up with the compression solution to the Statue problem. Similarly, U.S. students, who almost certainly had heard the Hansel and Gretel story as children, were predicted to be more likely than Chinese students to solve the Cave Problem by creating a trail of small objects. Additional insight problems were chosen as ‘‘neutral’’ tasks, because of no known analogues to these problems in either culture. Examples of these problems included the radiation problem (Duncker, 1945; Gick & Holyoak, 1980) and the string problem (Maier, 1931). The radiation problem involved using a number of less-intense rays from different directions to converge on a tumor such that the rays effectively destroy the tumor without damaging the surrounding healthy tissue. The string problem involved grabbing two strings hanging apart from the ceiling by tying an available object to one string and swinging it. These control problems were used to test the hypothesis that performance on insight problem solving would be equivalent between the two cultures. Substantial culture-speciﬁc analogical transfer was found when American and Chinese participants’ performance was compared on isomorphs of problems based on European vs. Chinese folk tales. U.S. participants typically remembered the Hansel and Gretel story and solved the Cave Problem more effectively than Chinese students. Chinese participants usually remembered the elephant tales and outperformed U.S. students in solving the Statue problem (Figure 13). U.S. and Chinese participants were equally successful in solving the control problems, showing that differences in solving the target problems were not due to the cultural differences in general ability to solve insight problems. Instead, the complementary pattern of performance on the Statue and Cave Problems across the two samples provides clear and compelling evidence that participants are capable of drawing on culturally speciﬁc experience in solving analogous problems. This study demonstrates college students’ ability to access to and use of remote analogy in solving problems. Yet, despite that previous studies have shown that younger children are less likely to solve analogical problems, little is known concerning age differences in retrieving source information from long-term memory and in solving remote analogical problems. To address the issues of how middle- and high-school students differ in remote transfer, a developmental study was designed (Chen et al., 2004). The experiment examined three issues: (1) the ability to analogize with a substantial time interval, (2) factors that inﬂuence long-term transfer processes (i.e., accessing the source information, mapping the key objects, and executing the solution strategy), and (3) age differences in long-term transfer among middle-school vs. high-school students.

Zhe Chen and David Klahr

452

80 USA CHINA

% Solving Target Tasks

60

40

20

Water Jar

String

Radiation

Pool

Cave

Statue

0

Problem Fig. 13. Percentage of U.S. and Chinese participants solving the target tasks (the Cave and Statue Problems) and control problems.

Several versions of the target problem were designed by systematically varying the similarity in key objects between the source and target problems, speciﬁcally, the object needing to be weighed (elephant vs. asteroid) and the tool available to facilitate the compression strategy (boat vs. spring platform). This manipulation resulted in four versions of the target problems, with the elephant/boat version being the most similar to the original tale and the asteroid/spring version being the most dissimilar. These four conditions were similar goal object and similar solution tool, dissimilar goal object but similar solution tool, similar goal object but dissimilar solution tool, and dissimilar goal object and dissimilar solution tool. Participants read a target problem and then viewed a set of illustrated objects that could potentially be used to solve the problem. Figure 14 illustrates the dissimilar goal object and dissimilar solution tool for the asteroid/spring problem. After attempting to solve the problem, participants were asked to answer several questions designed to reveal component processes involved in remote analogical transfer.

Scientific-Reasoning and Problem-Solving Strategies in Children

Asteroid

453

Small Scale

Boxes

Wood Logs

Tools

Rope

Bucket Small Rocks

Tree

Spring Platform

Fig. 14. An example of the target problem in the cross-cultural study: The asteroid/spring target problem.

The representing component was assessed by determining whether students recalled the goal and solution of the source tale. The accessing component was assessed by whether participants reported being reminded of the tale as they solved the problem. The mapping process was measured by participants’ selection of both the goal object and the solution tool during problem solving. The executing component was measured by the participants’ accurate use of the weight compression and equivalence principles with the solution tool. Finally, the evaluating component was measured by the students’ judgment of how much the source analogue had helped them solve the target problem. It is important to note that the actual order of assessing the participants’ problem-solving activities does not correspond to the order of presenting these ﬁve measures. Participants were ﬁrst asked to attempt to solve the target problem before being asked to report any potentially relevant source stories, to evaluate the usefulness of the reported source analogue, and to recall the source tale. For theoretical reasons, data analyses are presented in the order of recalling, accessing, mapping, executing, and evaluating, which is the sequence in which these processes are hypothesized to occur.

454

Zhe Chen and David Klahr

The results revealed that that both age groups remembered the source story well (memory score of 1.84 out of 2), indicating that the source story was well represented in long-term memory. Overall, 10th graders (85%) were more likely than 7th graders (68%) to access the source story during problem solving, but children retrieved the source story differently in different conditions. At both grades, students in the dissimilar goals and dissimilar tools condition were less likely to access the source problem than the other conditions. That is, students were least likely to retrieve the folk tale spontaneously when the problem to be solved had a different goal and a different tool. Regarding mapping, at both grades students typically mapped the key items when problems involved similar goals and similar tools but were much less likely to map correctly when problems involved dissimilar goals and dissimilar solution tools. At both ages, students were most likely to solve problems involving similar goals similar tools but least likely to solve them when they involved different goals and tools. Similarly, students rated the source tale to be most helpful on problems with similar goals and tools but least helpful for problems with dissimilar goals and tools. These results indicate that both middle- and high-school students had little difﬁculty in recalling the source story. The pattern of accessing the source analogue across the four conditions was similar to that in college students: dissimilarity in both goal object and solution tool hindered the accessing process. Middle-school students performed as well as high-school students when the problems shared both a common goal object and solution tool. However, they were less likely to access the source story, to map the key objects, and to execute the solution when the objects, especially the solution tool, were different between the problems. Despite the developmental differences in these measures, even middle-school students showed the ability to transfer the solution from a source analogue acquired from the distant past in different contexts. The naturalistic, cross-cultural approach proved effective in exploring remote analogical transfer of problem solutions that participants had heard in folk tales during their childhood, many years before encountering the target problems. Substantial culture-speciﬁc analogical transfer was found when American and Chinese participants’ performance was compared on isomorphs of problems solved in European vs. Chinese folk tales. Comparisons of different versions of the target problems indicate that similarity of solution tool affected the accessing, mapping, and executing components of problem solving, whereas similarity of goal object had only a moderate effect on accessing. Developmental differences in remote transfer were also evidenced. These ﬁndings demonstrate that source analogues can be acquired in various cultural settings, the conceptual

Scientific-Reasoning and Problem-Solving Strategies in Children

455

information can be represented, and the solution strategies can be transferred to novel problems in different contexts after a substantial delay.

V. Processes Involved in Remote Transfer The studies described here not only demonstrated the existence of remote transfer, but also revealed the processes involved in transfer. Previous studies in analogical problem solving have identiﬁed, in separate contexts or tasks, three key components involved in transfer: accessing a source problem, mapping the structural relations, and executing a solution or strategy (e.g., Gentner & Toupin, 1986; Reed & Bolstad, 1991; Ross, 1989). In the cross-cultural studies, we examined the contributions of all three components involved in remote transfer in the same task by making ﬁner distinctions between types of object similarity and exploring the inﬂuence of different types of object similarity on various processes involved in analogical problem solving.

A. ENCODING FEATURES OF SOURCE ANALOGUES The ﬁrst component involved in remote transfer is representing the source analogues; the issue concerns what information is encoded from the original learning situations, and how it is stored in long-term memory. The effects of source representation quality on subsequent transfer have been demonstrated in earlier studies wherein children who formed an abstract schema of the source solutions showed greater transfer than children who encoded a problem’s speciﬁc details (Brown et al., 1986; Chen & Daehler, 1989). The ﬁndings summarized in Section IV indicate that children readily recalled source analogues and transfer learned strategies to remote situations after long delays, suggesting that they represented and stored the original learning tasks in such a way that the solutions/strategies could be retrieved and used when encountering structurally similar problems. For example, in the cross-cultural research, students in the similar object conditions outperformed those in the dissimilar object conditions, who outperformed those in the control condition (their peers in another culture who had never heard the source analogue). These results suggest that children’s representations of source analogues contain elements of both superﬁcial and structural features in long-term memory. Thus, either type of feature can serve as a retrieval cue to activate the representation.

456

Zhe Chen and David Klahr

B. ACCESSING SOURCE INFORMATION The initial process of transfer involves searching for relevant information, such as similar events, instances, or stories in long-term memory at the appropriate time. One cue for such experiences is object reminding. The presence of common object attributes between an earlier source problem and current target problem increases the likelihood of retrieving a source analogue. Previous work with short-term analogical transfer has yielded consistent ﬁndings concerning the effects of superﬁcial similarity on accessibility (e.g., Brown & Campione, 1981, 1984; Crisaﬁ & Brown, 1986; Daehler & Chen, 1993; Gentner et al., 1993; Wharton et al., 1994). In the cross-cultural study, sharing either goal object or solution tool proved sufﬁcient to increase the accessibility of the source tale, suggesting that both were equally effective in reminding. These results suggest that representations of source analogues contain object-speciﬁc information about a story or problem situation and, thus, object cues in a target problem may trigger the retrieval of a source analogue. Another avenue of accessing relevant information involves structural reminding, which involves mapping the higher-order relations of the key entities between the source and target problems. Models of artiﬁcial intelligence and case-based reasoning view the process of human memory retrieval as typically driven by higher-order causal or goal structures (e.g., Hammond, Seifert, & Gray, 1991; Kolodner, 1993; Schank, 1982), implying that individuals are capable of accessing knowledge in long-term memory based on higher-order structural similarity. Over two-thirds of the Chinese students in the cross-cultural study retrieved the elephant tale from longterm memory; the goal structure of the asteroid/spring problem (e.g., the scientists needed to ﬁgure out the weight of the asteroid) reminded the Chinese participants of elephant tale, even though the target shared few common object attributes with the source. The ﬁndings from the experiments on transfer of scientiﬁc-reasoning strategies also expand previous ﬁndings by demonstrating that common goal structures can play a central role in guiding accessing when solving analogical problems after long delays. Even young children spontaneously retrieved acquired strategies from long-term memory and used them to solve isomorphic tasks that shared few similar surface features in new contexts after a long delay.

C. MAPPING STRUCTURAL ELEMENTS BETWEEN PROBLEMS After a source analogue has been retrieved, solvers still need to match the corresponding elements or objects between problems. Mapping problems

Scientific-Reasoning and Problem-Solving Strategies in Children

457

has been found to be guided by the common relational structures between analogous tasks (e.g., Chen & Daehler, 1992; Clement & Gentner, 1991; Gentner et al., 1993; Goswami, 1995). In the cross-cultural research, mapping was assessed by the students’ choice of goal object and solution tool during problem solving. In analogical problem solving, the selection of the corresponding target objects ultimately reﬂects the mapping of the structures between problems. After the relevant source story or event is retrieved, the common structural relations of the entities or objects associated with the entities between the problems serve as guides for matching the problems (e.g., Gentner & Markman, 1997). The present research further demonstrates that different types of object similarity play distinctive roles in transfer. Although solution tool similarity played a critical role in the mapping process, goal object commonality did not appear to inﬂuence how the key elements were mapped between problems.

D. EXECUTING A LEARNED SOLUTION Accessing an analogue and mapping the corresponding elements between source and target problem structures do not ensure successful execution of a solution retrieved from long-term memory. Previous studies have shown that when source and target problems share a solution principle but differ in speciﬁc procedures, individuals are less likely to implement a source solution (e.g., Chen, 1996, 2002; Reed & Bolstad, 1991; Ross, 1989). The cross-cultural study results further reveal that participants can also suffer an execution deﬁciency when the solution tool for the target problem differs in attributes from the source tool. The likely reason why only the solution tool is associated with the executing component is that the solution tool, and not the goal object, serves as a crucial part of the causal path, whereas goal object is arbitrary: Any object can be weighed with compression, but only objects that can be compressed may serve as a solution tool. Furthermore, the goal object is selected automatically, whereas there are alternatives from which a solution tool can be chosen and utilized to generate a solution procedure. This conceptual model concerning the processes involved in remote transfer addresses the observation that individuals often experience difﬁculty with analogical transfer after a long delay and that younger children in particular often fail to solve problems by remote analogy. This model also addresses questions of why some analogues are more difﬁcult to use in problem solving than others. And as we demonstrate in the next section, the model helps pinpoint how children at different age levels differ

458

Zhe Chen and David Klahr

in solving remote, analogous tasks, and how various factors inﬂuence children’s performance on remote transfer.

VI. Developmental Differences One striking aspect of age differences in learning and thinking involves the ability to generalize acquired strategies to other tasks and situations. Earlier studies indicate noticeable developmental differences. Younger children’s learning is more perceptually bound and more greatly inﬂuenced by the superﬁcial features of problems than by their structural or causal properties (Chen, 1996; Chen, Yanowitz, & Daehler, 1995; Daehler & Chen, 1993; Gentner & Markman, 1997). Furthermore, to draw the analogy between source and target problems younger children tend to need explicit hints or aids that point out the usefulness of prior problems (Brown et al., 1986; Chen, Sanchez, & Campbell, 1997; Crisaﬁ & Brown, 1986; Holyoak, Junn, & Billman, 1984). With age, children become increasingly more effective in perceiving deep relations or causal structures (e.g., Gentner & Toupin, 1986), and hence rely less on surface commonalities or explicit hints as vehicles to draw analogies (e.g., Chen, 1996; Daehler & Chen, 1993; Vosniadou, 1987). Developmental differences are evident in both the experimental and cross-cultural studies summarized herein. With age, children’s transfer of strategies and concepts is increasingly broad, ﬂexible, and effective. In the research described in Section III, 5-year-old children who acquired the determinacy/indeterminacy expert rule from the learning phases continued to use it during the immediate posttest phase and the follow-up phase 7 months later. These results demonstrate kindergartners’ ability to generalize the acquired rule even after a long delay, and preschoolers’ difﬁculty in generalizing as the gap between the learning situation and the new problems became more distant. Thus, feedback proved effective for 5-year-olds but had no lasting effects on younger children. Likewise, for more complex tasks, third graders demonstrated the ability to transfer CVS across tasks within the domain of mechanics (i.e., when reasoning about springs, slopes, and sinking tasks) and after a short delay (within a week), but only fourth graders displayed remote transfer. One reason for developmental differences in the breadth of learning is related to differences in the depth of initial learning (e.g., Siegler, 2000, 2005). Older children’s more ‘‘robust learning’’ in terms of acquiring the effective strategies in solving the original problems might explain their better retention and wider generalization on relatively near transfer tasks. Older children show broader learning than younger ones (e.g., Bjorklund,

Scientific-Reasoning and Problem-Solving Strategies in Children

459

1988; Chen & Siegler, 2000; Dixon & Bangert, 2002; Schauble, 1996) because they master the original tasks to a greater extent. When younger and older children learn an initial task to the same extent, they demonstrate comparable transfer (e.g., Brown et al., 1986; Chletos & De Lisi, 1991; Crowley & Siegler, 1999). However, the research reported herein suggest that age differences in generalizing learned strategies across contexts after long delays are not due solely to differences in original learning. The cross-cultural studies showed that although middle- and high-school students recalled the source tale equally well, nevertheless high-school students were more successful than middle-school students in retrieving the analogue, mapping the key entities, and executing the solution strategy. The experimental studies described herein also indicate that even when children at different age levels acquired equally well the determinacy/indeterminacy rule, the hypothesis-testing strategies, or CVS during the original learning phase, older children nevertheless demonstrated higher efﬁciency in extending the strategies to other situations. When the transfer gap becomes increasingly distant, developmental differences in transfer become increasingly apparent. What are the processes involved in developmental differences in remote transfer? The cross-cultural research helps pinpoint how age differences may be associated with each component of remote transfer. We explored the contributions of all three key components on a single task of long-term analogical problem solving and found that age differences were evident with all the key processes: Middle-school students were more likely than highschool students to experience difﬁculties in accessing the source problem, in mapping the key objects between the problems, and in executing the source solution, especially in the dissimilar solution tool conditions. Despite the availability of the source tale in long-term memory, some young participants failed to come up with an appropriate solution because they failed to access the source story from long-term memory when they encountered the target problem. Others did not use the source analogue because they experienced difﬁculties in mapping the entities of the source story onto the target problem, thus failing to choose a novel tool that matched the boat in function, but not in attributes. Yet, others failed to execute the acquired analogous solution, due to their using the selected tool inaccurately. Despite the fact that high-school students had heard of the source analogue in the more distant past, they nevertheless were more likely to access and use it to solve a target problem than middle-school students. A likely explanation for this result is that the elements of the problem’s goal structure (i.e., the goal of weighing a large, heavy object, the obstacle of ﬁguring out the weight with a small scale, and the approach to overcoming

460

Zhe Chen and David Klahr

the obstacle) are strengthened, and the solution strategies (i.e., weight compression and equivalence principles) are further instantiated with relevant experience over time. With age and experience, the representation of the source analogue becomes increasingly accessible, and applicable, and the generalization of acquired source strategies becomes increasingly effective, ﬂexible, and broad.

VII. Educational Implications Age is not the only factor that inﬂuences remote transfer. It is evident in this series of studies that how children represented and processed the source analogues largely determined how long the learned strategies could be maintained and how widely the strategies could be generalized. The studies described here have generated ﬁndings concerning factors inﬂuencing remote transfer in children, and have important implications concerning the role of analogy, instruction, implicit feedback, self-explanation, and learning in naturalistic and cultural settings.

A. ROLE OF ANALOGY Analogy is a powerful heuristic for learning and transfer of scientiﬁcreasoning strategies. Children as young as ﬁve are capable of representing the underlying structures of the determinacy/indeterminacy problems during initial learning and of mapping the structures between the initial tasks and the target problems encountered several months later. Elementary-school children successfully applied CVS across problems with different formats and in different domains after a long delay. Likewise, kindergartners and ﬁrst graders also proved quite effective in learning to understand the fundamental principle of testing a hypothesis and in transferring the correct strategy by mapping the problems presented one or two years apart. The present results extend previous ﬁndings concerning children’s ability to solve problems by analogy (e.g., Brown, 1989; Brown & Kane, 1988; Chen & Daehler, 1989; Gentner & Markman, 1997; Goswami, 1991, 1996; Tunteler & Resing, 2002) and indicate that analogical reasoning plays a central role in remote transfer. By embedding the tasks and strategies in the different surface features and contexts, we helped children construct a schema that could be generalized to tasks with new features and settings (e.g., Brown et al., 1986; Butterﬁeld, Slocum, & Nelson, 1993; Chen & Daehler, 1989; Gentner, 1983, 1989; Gick & Holyoak, 1983).

Scientific-Reasoning and Problem-Solving Strategies in Children

461

B. EXPLICIT INSTRUCTION The extensive controversy about the beneﬁts and costs of instruction located at various points along the direct instruction and discovery spectrum (e.g., Kirschner, Sweller, & Clark, 2006; Mayer, 2004) raises the issue about what is learned and how the learned strategies are transferred with various instructional approaches. Some studies have claimed that students who discover new concepts or strategies are more likely to extend this knowledge to new tasks than those students who learn from direct instruction (e.g., McDaniel & Schlager, 1990; Stohr-Hunt, 1996). Many argue that children tend to acquire superﬁcial and short-lived knowledge when direct instruction is involved, as compared to discovery. However, few previous studies have examined the long-term effects of explicit instruction and discovery. In the remote transfer studies described herein, it is evident that explicit instruction has been shown to be an effective approach in facilitating the acquisition and transfer of a hypothesis-testing strategy, in teaching elementary-school students to design unconfounded experiments, and in helping 4- and 5-year-old’s learn the principles of determinacy/ indeterminacy. Elementary-school children can overcome what appear to be stubborn misconceptions when they receive instruction and explicit feedback (Chen & Klahr, 1999; Klahr & Carver, 1988; Klahr & Nigam, 2004; Siegler, 1996; Siegler & Chen, 1998). The present results also show that with appropriate instruction, elementary-school children are capable of understanding, learning, and transferring the basic strategy when designing and evaluating simple tests. Children in the training-probe condition increased their use of CVS from 34% of the trials in the Exploration phase (before training) to 65% in the Assessment phase within the task (after training), and to 61% and 64% of the trials across tasks in the Transfer I and II phases, respectively. In these cases, the nature of the CVS tasks made it very difﬁcult for self-directed and selfcorrection to take place, especially for younger children. The indeterminacy study also demonstrates that even younger children are capable of acquiring a difﬁcult reasoning strategy from explicit feedback and are able to transfer the learned strategy to different contexts even after a 7-month delay. Problem-solving experience alone did not greatly facilitate children’s understanding of indeterminate patterns, presumably because the outcome of subsequent box openings did not provide sufﬁciently consistent or salient feedback. In contrast, explicit training pinpointing the rationale was effective in facilitating the acquisition of understanding indeterminate patterns.

462

Zhe Chen and David Klahr

C. IMPLICIT FEEDBACK AND SELF-EXPLANATIONS Direct instruction is not the only effective way to facilitate learning and transfer in young children. In the example of overcoming the positivecapture strategy, 5-year-olds learned from instruction, but their learning and transfer across isomorphic tasks also beneﬁted from extended problemsolving experience, even in the absence of explicit instruction. The effects of implicit feedback were also evident in young children’s learning of hypothesis-testing strategies, as young children’s transfer of the strategies beneﬁted from demonstrations with toy props, and from implicit probe questions. The effects of implicit feedback on learning and transfer are likely due to the beneﬁt from children’s self-explanations. For example, when learning CVS, children in the no training-probe condition were asked systematic questions on each trial about why they designed the test they did. Children thus had the opportunity to generate explanations of their reasoning behind their choice of objects for their experiment and for the possible conclusions from their design. Children’s superior performance in learning and transferring CVS in the probe condition, as compared to the no probe condition, replicates a growing body of studies demonstrating that the opportunity to generate self-explanations enhances learning (e.g., Chi et al., 1994; Honomichl & Chen, 2006; Renkl, 2002; Rittle-Johnson, 2006; Siegler, 2002; Siegler & Chen, 1998). The power of self-explanation has not been demonstrated for long-term transfer; our studies extend prior research in indicating that asking children to explain their own responses beneﬁts remote transfer.

D. LEARNING IN NATURALISTIC AND CULTURAL SETTINGS Another implication of the present studies is that children learn effectively in naturalistic settings and can readily transfer their learning to remote situations. Our cross-cultural research provides compelling evidence for remote transfer in that the source and target problems shared few semantic and contextual features. Even with a signiﬁcant time gap between source and target problems, solvers can be reminded of the source analogue and can use it effectively when encountering an isomorphic problem in a different context. Although the most accurate performance occurred when common objects were shared between problems, the analogous target problem did not share key object attributes with the source analogue in long-term memory.

Scientific-Reasoning and Problem-Solving Strategies in Children

463

The lack of efﬁcient transfer found in previous laboratory studies even with adults (e.g., Perkins & Grotzer, 1997; Reed, Ernst, & Banerji, 1974) and the robust use of remote analogy evident in the present research seems to present a paradox. Yet, as Dunbar (2001) has pointed out, structural analogies are a frequent rather than a rare phenomenon when reasoning is investigated in naturalistic environments. For example, it was found that scientists often use structural features and higher-order relations in analogizing during the discovery process (Dunbar & Blanchette, 2001). Findings summarized here conﬁrm that source analogues can be acquired in various cultural settings, represented in rich ways, and transferred to novel problems in different contexts after a substantial delay. The implication about how cultural experiences inﬂuence thinking is that children often acquire solutions and strategies in naturalistic settings, and the acquired information is durable, accessible, and applicable.

VIII. Conclusions and Future Directions Whether and how children can ﬂexibly retrieve relevant information from long-term memory and generalize the acquired strategies to broad situations is a fundamental issue in human cognition, children’s learning, and education. And yet, there are virtually no studies addressing the issues of whether and how children transfer strategies across contexts after a prolonged delay. In this chapter we have reviewed a number of studies in which we attempted a multi-faceted exploration of remote transfer, using a traditional experimental approach and a naturalistic, cross-cultural approach. These studies represent an effort to bridge the gap between the central role of remote transfer in human cognition and children’s thinking and our as yet limited knowledge of children’s competencies and processes in remote transfer. These studies focus on the impact of time and context on children’s transfer of scientiﬁc-reasoning and problem-solving strategies from initial learning situations to target tasks that share few superﬁcial and contextual features. These studies suggest that despite the theoretical and methodological difﬁculties, remote transfer is documentable, and the ﬁndings of children’s transfer of strategies to remote situations are robust. Furthermore, the ﬁndings described here have yielded theoretical implications. The ﬁrst implication concerns the nature of representations. The issue is which information is encoded and stored in long-term memory and how. The results suggest that representations of the source examples contain both structural and surface features in long-term memory and that either type of feature can serve as retrieval cue to activate the representation. The second

464

Zhe Chen and David Klahr

implication concerns the nature of developmental differences. Age differences in remote transfer performance may be due to both the quality of the initial representations of the source analogue, and the ability to retrieve, map, and execute the source strategy when children encounter an isomorphic problem. With age and experience, children are increasingly able to encode analogues more deeply and to use the source information to solve target problems more ﬂexibly and effectively. The third implication concerns how cultural experiences inﬂuence thinking. Many studies have indicated that cultural values, systems, and practices inﬂuence cognitive styles and performance. The present ﬁndings suggest that differences in cognitive strategies and performance are traceable to speciﬁc cultural experiences, which may be richly represented, ﬂexibly retrieved, and effectively used in different contexts when appropriate. Finally, the fourth implication concerns whether and how the distant gap between source analogues and target problems can be crossed by encouraging children to represent deeper structural features and extract more general principles from the analogues. The studies described herein are only initial steps in exploring remote transfer in children. With the ‘‘rebirth’’ of research in children’s learning (Siegler, 2000, 2006), we expect to see more studies on children’s transfer and generalization of strategies, and research in issues related to remote analogical transfer is beginning to ﬂourish. Interesting avenues for further exploration of children’s remote transfer involve effects of dimensions of transfer distance on performance, early abilities of transfer, nature of remote transfer, and factors that promote distant transfer. One future direction involves ﬁner analyses and systematic manipulations of different dimensions of transfer distance. The proposed transfer distance space as illustrated in Figure 1 is only an initial step toward our understanding of remote transfer. It remains a critical task to reﬁne the effects of each dimension of transfer distance (task similarity, context similarity, and time gap) and the interaction between them on transfer performance in children at different ages. A second direction for future studies concerns the abilities of preschoolers, toddlers, and even infants to retrieve and use acquired strategies. Even infants and toddlers can demonstrate near analogical transfer, such as transfer of problem-solving strategies across analogous tasks within the same lab context and within the same hour or same day (e.g., Brown, 1989; Chen & Siegler, 2000; Chen et al., 1997), but the question of whether, with age-appropriate tasks, very young children are capable of demonstrating remote transfer remains unexplored. Even for very young children, experiences inevitably inﬂuence their thinking and learning; however, it remains to be seen how ﬂexibly and widely young

Scientific-Reasoning and Problem-Solving Strategies in Children

465

children’s experiences in problem solving is generalizable to tasks in different contexts and with long delays. A third direction for further exploration of remote transfer is related to the implicit and explicit nature of analogical transfer. Although both implicit and explicit use of analogies in solving problems have been evident in adults (e.g., Schunn & Dunbar, 1996), we do not yet fully understand (a) the extent to which children retrieve and use source analogues explicitly or only implicitly, (b) when and how they gain metacognitive and explicit understanding of analogical relations between problems, or (c) what factors (e.g., task and context similarity, time gap, and feedback during initial learning) inﬂuence the explicit retrieval and use of remote source analogues in children at different ages. A fourth avenue for further study is to explore how optimal remote transfer can be promoted. The precise roles of instruction, guided discovery, and self-explanation in remote transfer processes remain to be examined. Exploration of these issues of children’s remote transfer will yield signiﬁcant theoretical and educational implications for children’s thinking and learning. As we uncover ever more fully answers to these questions, our own theoretical, empirical, and pedagogical knowledge will become more complete, thereby enhancing our endeavors not only to understand but, more importantly, to enrich children’s thinking and learning experiences.

REFERENCES Alibali, M. W. (1999). How children change their minds: Strategy change can be gradual or abrupt. Developmental Psychology, 35, 127–145. Barnett, S. M., & Ceci, S. J. (2002). When and where do we apply what we learn? A taxonomy for far transfer. Psychological Bulletin, 128, 612–637. Bauer, P. J. (1997). Development of memory in early childhood. In N. Cowan (Ed.), The development of memory in childhood. Studies in developmental psychology (pp. 83–111). Hove, England: Psychology Press/Erlbaum. Bjorklund, D. F. (1988). Acquiring a mnemonic: Age and category knowledge effects. Journal of Experimental Child Psychology, 45, 71–87. Blanchette, I., & Dunbar, K. (2001). Analogy use in naturalistic settings: The inﬂuence of audience, emotion, and goals. Memory & Cognition, 29, 730–735. Blo¨te, A. W., Resing, W. C. M., Mazer, P., & Van Noort, D. A. (1999). Young children’s organizational strategies on a same–different task: A microgenetic study and a training study. Journal of Experimental Child Psychology, 74, 21–43. Bransford, J. D., Sherwood, R., Vye, N., & Rieser, J. (1986). Teaching thinking and problem solving: Research foundations. American Psychologist, 41, 1078–1089. Bransford, J. D., Franks, J. J., Vye, N. J., & Sherwood, R. D. (1989). New approaches to instruction: Because wisdom can’t be told. In S. Vosniadou & A. Ortony (Eds.), Similarity and analogical reasoning (pp. 470–497). London: Cambridge University Press.

466

Zhe Chen and David Klahr

Bransford, J. D., Brown, A. L., & Cocking, R. R. (1999). How people learn: Brain, mind, experience, and school. Washington, DC: National Academy Press. Brown, A. L. (1989). Analogical learning and transfer: What develops? In S. Vosniadou & A. Ortony (Eds.), Similarity and analogical reasoning (pp. 369–412). London: Cambridge University Press. Brown, A. L., & Campione, J. C. (1981). Inducing ﬂexible thinking: A problem of access. In M. Freidman, J. P. Das, & N. O’Connor (Eds.), Intelligence and learning (pp. 515–530). New York: Plenum. Brown, A. L., & Campione, J. C. (1984). Three faces of transfer: Implications for early competence, individual differences, and instruction. In M. Lamb, A. Brown, & B. Rogoff (Eds.), Advances in developmental psychology (Vol. 3, pp. 143–192). Hillsdale, NJ: Erlbaum. Brown, A. L., & Kane, M. J. (1988). Preschool children can learn to transfer: Learning to learn and learning from example. Cognitive Psychology, 20, 493–523. Brown, A. L., Kane, M. J., & Echols, C. H. (1986). Young children’s mental models determine analogical transfer across problems with a common goal structure. Cognitive Development, 1, 103–121. Bullock, M., & Ziegler, A. (1999). Scientiﬁc reasoning: Developmental and individual differences. In F. E. Weinert & W. Schneider (Eds.), Individual development from 3 to 12: Findings from the Munich Longitudinal Study (pp. 38–54). Munich: Max Plank Institute for Psychological Research. Butterﬁeld, E. C., Slocum, T. A., & Nelson, G. D. (1993). Cognitive and behavioral analyses of teaching and transfer: Are they different? In D. K. Detterman & R. J. Sternberg (Eds.), Transfer on trial: Intelligence, cognition, and instruction (pp. 192–257). Westport, CT: Ablex Publishing. Ceci, S. J., & Bruck, M. (1998). The ontogeny and durability of true and false memories: A Fuzzy Trace account. Journal of Experimental Child Psychology, 71, 165–169. Chen, Z. (1996). Children’s analogical problem solving: Effects of superﬁcial, structural, and procedural similarity. Journal of Experimental Child Psychology, 62, 410–431. Chen, Z. (2002). Analogical problem solving: A hierarchical analysis of procedural similarity. Journal of Experimental Psychology: Learning, Memory, & Cognition, 28, 81–98. Chen, Z. (2007). Learning to map: Strategy discovery and strategy change in young children. Developmental Psychology, 43, 386–403. Chen, Z., & Daehler, M. W. (1989). Positive and negative transfer in analogical problem solving by 6-year-old children. Cognitive Development, 4, 327–344. Chen, Z., & Daehler, M. W. (1992). Intention and outcome: Key components of causal structure facilitating mapping in children’s analogical transfer. Journal of Experimental Child Psychology, 53, 237–257. Chen, Z., & Klahr, D. (1999). All other things being equal: Acquisition of the control of variables strategy. Child Development, 70, 1098–1120. Chen, Z., & Siegler, R. S. (2000). Across the great divide: Bridging the gap between understanding of toddlers’ and older children’s thinking. Monographs of the Society for Research in Child Development, 65, v-96. Chen, Z., Yanowitz, K. L., & Daehler, M. W. (1995). Constraints on accessing abstract source information: Instantiation of principles facilitates children’s analogical transfer. Journal of Educational Psychology, 87, 445–454. Chen, Z., Sanchez, R. P., & Campbell, T. (1997). From beyond to within their grasp: The rudiments of analogical problem solving in 10- and 13-month-olds. Developmental Psychology, 33, 790–801.

Scientific-Reasoning and Problem-Solving Strategies in Children

467

Chen, Z., Mo, L., & Honomichl, R. (2004). Having the memory of an elephant: Long-term retrieval and the use of analogues in problem solving. Journal of Experimental Psychology: General, 133, 415–433. Chen, Z., Mo, L., & Klahr, D. (in preparation). Acquisition and transfer of a hypothesistesting strategy in children. Chi, M. T. H., de Leeuw, N., Chiu, M.-H., & LaVancher, C. (1994). Eliciting self-explanations improves understanding. Cognitive Science, 18, 439–477. Chletos, P. N., & De Lisi, R. (1991). A microgenetic study of proportional reasoning using balance scale problems. Journal of Applied Developmental Psychology, 12, 307–330. Clement, C. A., & Gentner, D. (1991). Systematicity as a selection constraint in analogical mapping. Cognitive Science, 15, 89–132. Cognition and Technology Group at Vanderbilt. (1997). The Jasper project: Lessons in curriculum instruction, assessment, and professional development. Mahwah, NJ: Erlbaum. Coyle, T. R., & Bjorklund, D. F. (1997). Age differences in, and consequences of, multipleand variable-strategy use on a multitrial sort-recall task. Developmental Psychology, 33, 372–380. Crisaﬁ, M. A., & Brown, A. L. (1986). Analogical transfer in very young children: Combining two separately learned solutions to reach a goal. Child Development, 57, 953–968. Crowley, K., & Siegler, R. S. (1999). Explanation and generalization in young children’s strategy learning. Child Development, 70, 304–316. Daehler, M. W., & Chen, Z. (1993). Protagonist, theme, and goal object: Effects of surface features on analogical transfer. Cognitive Development, 8, 211–229. DeLoache, J. S. (2004). Becoming symbol-minded. Trends in Cognitive Sciences, 8, 66–70. Detterman, D. K., & Sternberg, R. J. (1993). Transfer on trial: Intelligence, cognition, and instruction. Norwood, NJ: Ablex. Dixon, J. A., & Bangert, A. S. (2002). The prehistory of discovery: Precursors of representational change in solving gear system problems. Developmental Psychology, 38, 918–933. Dunbar, K. (2001). The analogical paradox: Why analogy is so easy in naturalistic settings yet so difﬁcult in the psychological laboratory. In D. Gentner & K. J. Holyoak (Eds.), The analogical mind: Perspectives from cognitive science (pp. 313–334). Cambridge, MA: MIT Press. Dunbar, K., & Blanchette, I. (2001). The in vivo/in vitro approach to cognition: The case of analogy. Trends in Cognitive Sciences, 5(8), 334–339. Duncker, K. (1945). On problem solving. Psychological Monographs, 58 (Whole No. 270). Fay, A., & Klahr, D. (1996). Knowing about guessing and guessing about knowing: Preschoolers’ understanding of indeterminacy. Child Development, 67, 689–716. Flynn, E., O’Malley, C., & Wood, D. (2004). A longitudinal, microgenetic study of the emergence of false belief understanding and inhibition skills. Developmental Science, 7, 103–115. Gelman, R. (1969). Conservation acquisition: A problem of learning to attend to relevant attributes. Journal of Experimental Child Psychology, 7, 167–187. Gentner, D. (1983). Structure-mapping: A theoretical framework for analogy. Cognitive Science, 7, 155–170. Gentner, D. (1989). The mechanisms of analogical learning. In S. Vosniadou & A. Ortony (Eds.), Similarity and analogical reasoning (pp. 199–241). London: Cambridge University Press. Gentner, D., & Markman, A. B. (1997). Structure mapping in analogy and similarity. American Psychologist, 52, 45–56. Gentner, D., & Toupin, C. (1986). Systematicity and surface similarity in the development of analogy. Cognitive Science, 10, 277–300.

468

Zhe Chen and David Klahr

Gentner, D., Rattermann, M. J., & Forbus, K. D. (1993). The roles of similarity in transfer: Separating retrievability from inferential soundness. Cognitive Psychology, 25, 524–575. Gick, M. L., & Holyoak, K. J. (1980). Analogical problem solving. Cognitive Psychology, 12, 306–355. Gick, M. L., & Holyoak, K. J. (1983). Schema induction and analogical transfer. Cognitive Psychology, 15, 1–38. Goldin-Meadow, S., & Alibali, M. W. (2002). Looking at the hands through time: A microgenetic perspective on learning and instruction. In N. Granott & J. Parziale (Eds.), Microdevelopment: Transition processes in development and learning (pp. 80–105). New York, NY: Cambridge University Press. Goswami, U. (1991). Analogical reasoning: What develops? A review of research and theory. Child Development, 62, 1–22. Goswami, U. (1995). Transitive relational mappings in three- and four-year-olds: The analogy of Goldilocks and the Three Bears. Child Development, 66, 877–892. Goswami, U. (1996). Analogical reasoning and cognitive development. In H. W. Reese (Ed.), Advances in child development and behavior (pp. 91–138). San Diego, CA: Academic Press. Halpern, D. F. (1998). Teaching critical thinking for transfer across domains. American Psychologist, 53, 449–455. Hammond, K. J., Seifert, C. M., & Gray, K. C. (1991). Functionality in analogical transfer: A hard match is good to ﬁnd. Journal of the Learning Sciences, 1, 111–152. Holyoak, K. J., Junn, E. N., & Billman, D. O. (1984). Developmental analogical problem solving skills. Child Development, 55, 2042–2055. Honomichl, R. D., & Chen, Z. (2006). Learning to align relations: The effects of feedback and self-explanation. Journal of Cognition and Development, 7, 527–550. Kirschner, P. A., Sweller, J., & Clark, R. E. (2006). Why minimal guidance during instruction does not work: An analysis of the failure of constructivist, discovery, problem-based, experiential, and inquiry-based teaching. Educational Psychologist, 41, 75–86. Klahr, D. (2000). Exploring science: The cognition and development of discovery processes. Cambridge, MA: MIT Press. Klahr, D., & Carver, S. M. (1988). Cognitive objectives in a LOGO debugging curriculum: Instruction, learning, and transfer. Cognitive Psychology, 20, 362–404. Klahr, D., & Chen, Z. (2003). Overcoming the positive-capture strategy in young children: Learning about indeterminacy. Child Development, 74, 1275–1296. Klahr, D., & Dunbar, K. (1988). Dual space search during scientiﬁc reasoning. Cognitive Science, 12, 1–48. Klahr, D., & Nigam, M. (2004). The equivalence of learning paths in early science instruction: Effects of direct instruction and discovery learning. Psychological Science, 15, 661–667. Klayman, J., & Ha, Y. W. (1987). Conﬁrmation, disconﬁrmation, and information in hypothesis testing. Psychological Review, 94, 211–228. Kolodner, J. (1993). Case-based reasoning. San Mateo, CA: Morgan Kaufmann. Kotovsky, L., & Gentner, D. (1996). Comparison and categorization in the development of relational similarity. Child Development, 67, 2797–2822. Kuhn, D. (1995). Microgenetic study of change: What has it told us? Psychological Science, 6, 133–139. Kuhn, D., Amsel, E., O’Loughlin, M., Schauble, L., Leadbeater, B., & Yotive, W. (1988). The development of scientific thinking skills. San Diego, CA: Academic Press. Kuhn, D., Schauble, L., & Garcia-Mila, M. (1992). Cross-domain development of scientiﬁc reasoning. Cognition & Instruction, 9, 285–327. Kuhn, D., Garcia-Mila, M., Zohar, A., & Andersen, C. (1995). Strategies of knowledge acquisition. Monographs of the Society for Research in Child Development, 60, v-128.

Scientific-Reasoning and Problem-Solving Strategies in Children

469

Lave, J. (1988). Cognition in practice: Mind, mathematics and culture in everyday life. New York, NY: Cambridge University Press. Loewenstein, J., & Gentner, D. (2001). Spatial mapping in preschoolers: Close comparisons facilitate far mappings. Journal of Cognition and Development, 2, 189–219. Maier, N. R. F. (1931). Reasoning in humans II. The solution of a problem and its appearance in consciousness. Journal of Comparative Psychology, 12, 181–194. Marzolf, D. P., & DeLoache, J. S. (1994). Transfer in young children’s understanding of spatial representations. Child Development, 65, 1–15. Mayer, R. E. (2004). Should there be a three-strikes rule against pure discovery learning? American Psychologist, 59, 14–19. McDaniel, M. A., & Schlager, M. S. (1990). Discovery learning and transfer of problemsolving skills. Cognition and Instruction, 7, 129–159. Perkins, D. N., & Grotzer, T. A. (1997). Teaching intelligence. American Psychologist, 52, 1125–1133. Reed, S. K., & Bolstad, C. A. (1991). Use of examples and procedures in problem solving. Journal of Experimental Psychology: Learning, Memory, & Cognition, 17, 753–766. Reed, S. K., Ernst, G. W., & Banerji, R. (1974). The role of analogy in transfer between similar problem states. Cognitive Psychology, 6, 436–450. Renkl, A. (2002). Worked-out examples: Instructional explanations support learning by selfexplanations. Learning and Instruction, 12, 529–556. Rittle-Johnson, B. (2006). Promoting transfer: Effects of self-explanation and direct instruction. Child Development, 77, 1–15. Ross, B. H. (1989). Distinguishing types of superﬁcial similarities: Different effects on the access and use of earlier problems. Journal of Experimental Psychology: Learning, Memory, & Cognition, 15, 456–468. Rovee-Collier, C. (1999). The development of infant memory. Current Directions in Psychological Science, 8, 80–85. Ruffman, T., Perner, J., Olson, D. R., & Doherty, M. (1993). Reﬂecting on scientiﬁc thinking: Children’s understanding of the hypothesis-evidence relation. Child Development, 64, 1617–1636. Schank, R. C. (1982). Dynamic memory. Cambridge, MA: Cambridge University Press. Schauble, L. (1990). Belief revision in children: The role of prior knowledge and strategies for generating evidence. Journal of Experimental Child Psychology, 49, 31–57. Schauble, L. (1996). The development of scientiﬁc reasoning in knowledge-rich contexts. Developmental Psychology, 32, 102–119. Schunn, C. D., & Dunbar, K. (1996). Priming, analogy, and awareness in complex reasoning. Memory & Cognition, 24, 271–284. Siegler, R. S. (1995). How does change occur: A microgenetic study of number conservation. Cognitive Psychology, 28, 225–273. Siegler, R. S. (1996). Emerging minds: The process of change in children’s thinking. New York, NY, US: Oxford University Press. Siegler, R. S. (2000). The rebirth of children’s learning. Child Development, 71, 26–35. Siegler, R. S. (2002). Microgenetic studies of self-explanation. In N. Granott & J. Parziale (Eds.), Microdevelopment: Transition processes in development and learning. Cambridge studies in cognitive perceptual development (pp. 31–58). New York, NY: Cambridge University Press. Siegler, R. S. (2005). Children’s learning. American Psychologist, 60, 769–778. Siegler, R. S. (2006). Microgenetic analyses of learning. In W. Damon & R. M. Lerner (Series Eds.) & D. Kuhn & R. S. Siegler (Vol. Eds.), Handbook of child psychology: Volume 2: Cognition, perception, and language (6th ed., pp. 464–510). Hoboken, NJ: Wiley.

470

Zhe Chen and David Klahr

Siegler, R. S., & Chen, Z. (1998). Developmental differences in rule learning: A microgenetic analysis. Cognitive Psychology, 36, 273–310. Siegler, R. S., & Opfer, J. E. (2003). The development of numerical estimation: Evidence for multiple representations of numerical quantity. Psychological Science, 14, 237–243. Siegler, R. S., & Svetina, M. (2002). A microgenetic/cross-sectional study of matrix completion: Comparing short-term and long-term change. Child Development, 73, 793–809. Singley, M. K., & Anderson, J. R. (1989). The transfer of cognitive skill. Cambridge, MA: Harvard University Press. Sodian, B., Zaitchik, D., & Carey, S. (1991). Young children’s differentiation of hypothetical beliefs from evidence. Child Development, 62, 753–766. Spencer, R. M., & Weisberg, R. W. (1986). Context-dependent effects on analogical transfer. Memory & Cognition, 14, 442–449. Stohr-Hunt, P. M. (1996). An analysis of frequency of hands-on experience and science achievement. Journal of Research in Science Teaching, 33, 101–109. Thorndike, E. L., & Woodworth, R. S. (1901). The inﬂuence of improvement in one mental function upon the efﬁciency of other functions: Functions involving attention, observation and discrimination. Psychological Review, 8, 553–564. Tunteler, E., & Resing, W. C. M. (2002). Spontaneous analogical transfer in 4-year-olds: A microgenetic study. Journal of Experimental Child Psychology, 83, 149–166. Vosniadou, S. (1987). Children and metaphors. Child Development, 58, 870–885. Wharton, C. M., Holyoak, K. J., Downing, P. E., Lange, T. E., Wickens, T. D., & Melz, E. R. (1994). Below the surface: Analogical similarity and retrieval competition in reminding. Cognitive Psychology, 26, 64–101. Whitehead, A. N. (1929). The function of reason. Oxford, England: Princeton University Press.

Author Index

A Aber, J. L., 314 Acerra, F., 336–7 Achenbach, T. M., 104 Acredolo, L. P., 385 Adams, M. J., 35, 38 Adams, R. B., 359, 361–2 Adolph, K. E., 375–6, 389 Adolphs, R., 331–2, 354–5, 359 Afﬂeck, G., 117 Ahn, W., 174 Ainslie, G., 265 Ainsworth, M. D. S., 85, 88 Akhtar, N., 5 Allen, G. L., 385, 391 Allen, J. P., 86, 87, 89, 90, 94, 96, 100, 190–1 Allison, T., 348, 359 Anderson, C. A., 301 Anderson, K. E., 207, 230 Anderson, M., 252 Anderson, S. W., 352 Angells, P., 52–3 Ans, B., 38, 40–1 Archer, N., 67, 68 Arkes, H. R., 267 Arterberry, M. E., 151 Arum, R., 306 Attar, B., 305

B Backer, T. E., 319 Baddeley, A. D., 66–7 Baillargeon, R., 374 Baldwin, D. A., 2, 153, 353, 355 Ball, L. J., 257 Bandura, A., 94, 193, 196, 222, 288, 301–3 Banse, R., 355 Barnett, S. M., 423 Baron, J., 38, 253, 265, 274–5 Baron-Cohen, S., 343 Barron, R. W., 36, 46 Barsalou, L. W., 138, 146, 149, 153, 173, 178

Bartholomew, K., 85, 88 Bates, E., 3, 6, 24 Bates, J. E., 300–1, 302, 303 Batki, A., 340, 343 Battistich, V., 316 Beauchaine, T. P., 99 Bednar, J. A., 337, 338 Beebe, B., 115 Behrend, D. A., 17 Bell, R. Q., 191, 193, 195, 202 Belsky, J., 101 Bentin, S., 341–2 Berger, R. D., 98 Berlin, L. J., 87, 92 Bernston, G. G., 105 Besevegis, E., 296 Biglan, A., 318, 319 Bisconti, T. L., 227 Blackmore, S., 262 Blakemore, S. J., 90 Blanchette, I., 449, 463 Bloch, M., 110–11 Block, J., 104 Bloom, L., 2–3, 18–19, 24 Bloom, P., 143, 148, 179, 343 Blo¨te, A. W., 420 Blum, P., 61 Bogartz, R. S., 193, 198 Booth, A. E., 136, 139, 140, 141, 142, 153, 171, 173 Bornstein, M., 10, 12 Botvin, G. J., 314 Bourgeois, K., 167 Bousha, D. M., 303 Bovard, E. W., 99 Bower, G. H., 407 Bowerman, M. F., 20 Bowers, P. G., 61 Bowey, J. A., 53–5, 60, 62 Bowlby, J., 83–4, 85, 90, 94, 97, 193, 225 Boxer, P., 293, 298, 311–13 Boyle, M. H., 207, 230 Bradley, R. H., 305 Brainerd, C. J., 254 Brandone, A., 11, 17, 21, 24

471

Bransford, J. D., 421–2 Brennan, K. A., 94 Bretherton, I., 193, 200 Bridges, L. J., 101 Brody, G. H., 207, 219, 232, 239 Bronfenbrenner, U., 24, 193, 197, 202, 289, 305 Brophy, M., 208, 230 Brosschot, J. F., 99 Brown, A. L., 381, 385, 411, 420, 421, 455, 458 Bruine de Bruin, W., 252, 271 Brunk, M. A., 208, 230, 242 Bruyer, R., 358 Buchanan, C. M., 100–1, 116–17 Bullock, M., 427 Bumpus, M. F., 220, 238 Bush, G., 355 Bushnell, I. W. R., 340 Buske-Kirschbaum, A., 100 Buss, D. M., 289, 294 Buxbaum, L. J., 136, 152, 172 Buzsaki, G., 349

C Cabrera, N. J., 116 Cacioppo, J. T., 97, 117, 118, 356 Cairns, R. B., 98 Calkins, S. D., 93, 98, 113, 115 Campbell, A., 294 Campione, J. C., 421 Campos, J. J., 361 Caplan, M., 295 Carey, S., 5 Carpenter, M., 4, 148, 151 Carter, C. S., 86 Carver, S. M., 461 Casasola, M., 13, 16 Casler, K., 137, 148 Cassar, M., 63 Cassidy, J., 85, 87, 92, 93, 96 Ceci, S. J., 423 Chaigneau, S. E., 138, 146, 149, 153, 173, 178 Chaiken, M. R., 291 Chao, L. L., 152, 172 Chen, Z., 419–70 Chomsky, N., 39–40 Chorpita, B. F., 100, 115 Chow, S.-M., 188 Chugani, H. T., 341 Cicchetti, D., 92, 193, 198, 289, 321

Clark, L. A., 208, 230 Clayton, K., 393 Cohen, L. B., 153, 155, 167, 168, 174, 175, 180 Cohen-Kadosh, K., 332 Coie, J. D., 289–90, 293, 297 Colder, C. R., 92 Cole, D. A., 189 Cole, P. M., 209, 226, 231 Collins, N. L., 88, 95 Collins, W. A., 193 Coltheart, M., 34–8, 42 Conger, R. D., 118 Connell, A. M., 192 Cook, T. D., 316 Cook, W. L., 89, 193, 209, 232, 239 Cooper, M. L., 86, 91, 92, 94–5, 96, 100, 109, 116 Corbetta, D., 167 Cornell, E. H., 379 Courchesne, E., 354 Covell, K., 209, 229, 230–1, 232 Cox, M. J., 189, 193, 200 Crick, N. R., 290, 296, 301 Crisaﬁ, M. A., 458 Croiset, G., 97 Csibra, G., 332, 347, 350 Cummings, E. M., 111, 187–250 Cunningham, A. E., 43, 51, 54, 60–1, 62, 67, 69

D Daehler, M. W., 455, 458 Dahlberg, L. L., 306 Damasio, A. R., 358 Damon, W., 320–1 Davidson, D., 269, 273–4 Davies, P. T., 188, 218, 236 Davila, J., 89 Davis, E., 215, 235 Dawes, R. M., 252, 258 Dawkins, R., 262 de Haan, M., 340, 341–2, 354, 355 de Jong, P. F., 56, 66 De Neys, W., 272 Deary, I. J., 252 Deci, E. L., 96, 108–9, 114 Dehaene-Lambertz, G., 342 Denes-Raj, V., 265, 271 DeQuervain, D. J. F., 294 Detterman, D. K., 423

deWaal, F., 288 Diamond, A., 355 Diesendruck, G., 137, 139, 143 DiTomasso, E., 104 Dodge, K. A., 113, 289–91, 293, 296–9, 301, 303–4, 307, 310–11 Driver, J., 359–60 Dunbar, K., 422, 449, 463 Duncker, K., 451 Dunn, J., 201, 220, 238

E Eargle, A., 309 Easterbrook, M., 335–6 Eastwood, J. D., 360 Eccles, J., 321 Edwards, W., 253 Ehri, L. C., 35, 42, 44, 51, 52, 54, 57, 64, 67, 68, 70 Ehrlinger, J., 262 Eiden, R. D., 215, 234 Eimer, M., 357 Eisenberg, N., 91, 93, 116, 209, 231 Elliott, D. S., 318–19 Emery, R. E., 193 Engebretson, P. H., 385–6 Engel, A. K., 349 Engfer, A., 215 Epstein, S., 265 Eron, L. D., 298 Evans, J., 253, 254, 255, 256, 257, 258–9, 265, 266–7, 274

F Fabes, R. A., 92, 93, 99 Fagan, J., 296 Fagot, B. I., 96 Farrington, D. P., 307, 310 Farroni, T., 338–51, 355, 361 Fay, A., 435, 437 Feeney, A., 258–9 Feeney, J. A., 88–9, 95, 96–7, 114 Fenson, L., 9, 18 Fiebach, C. J., 349, 351 Firth, I., 35, 36–7, 40–1 Flynn, E., 420 Fogel, A., 115

Fong, G. T., 267 Foorman, B. R., 38 Forbes, J. N., 20 Forbus, K. D., 423 Forster, K. I., 34 Fox, N. A., 91, 99, 113 Fraley, R. C., 87, 103 Fredrickson, B. L., 117 Fridja, N. H., 101 Friedman, A., 385, 411 Friedman, D., 356–7 Frith, U., 41–2, 70 Frosch, C. A., 215, 234 Frost, R., 62–3, 65 Furrer, S., 161

G Gable, S. L., 114–15 Galen, B. R., 290 Gauthier, I., 332, 334, 341, 359 Gauvain, M., 377, 405, 407 Gentner, D., 10, 12, 20–1, 423, 458 George, N., 359 Gergely, G., 4, 151 German, T. P., 150 Gershoff, E. T., 304 Gibson, E. J., 173, 374–6 Gibson, J. J., 374–6 Gick, M. L., 451 Gilbert, D. T., 258, 262, 266 Gillette, J., 12, 16, 18 Glaser, D., 92, 95 Gleitman, L., 16, 18–19, 20 Gohm, C. L., 101 Goldstein, A. P., 314 Goodale, M. A., 152, 165 Goodman, K. S., 67 Gough, P. B., 34 Graham, S., 311 Grandjean, D., 356 Granger, D. A., 102 Granic, I., 193, 198, 210, 222, 227, 233 Greenberg, M., 314, 320 Grice, S. J., 348 Gross, J. J., 92, 93, 101, 104 Grossman, D. C., 314 Grossman, K. E., 105 Grossman, T., 349–57 Grundy, A. M., 210, 236

Grusec, J. E., 195, 210 Grych, J. H., 192 Gutheil, G., 147

H de Haan, M., 340, 341–2, 354, 355 Halberstadt, A. G., 103 Halit, H., 342 Harach, L. D., 211, 231–2 Hardman, D., 257 Harm, M. W., 35, 46, 52 Hasher, L., 255 Hawkins, J. D., 306 Hawley, P. H., 289, 294 Haxby, J. V., 341, 351 Hazan, C., 84, 86–7, 88, 104–5, 106–7 Heilmann, M. F., 296 Henggeler, S. W., 316 Hennon, E., 21 Henry, D., 306 Hermann, C. S., 349, 350 Herschell, A., 315 Hinde, R. A., 199 Hofer, M., 115–16 Hoffman, E. A., 350, 352 Hogaboam, T. W., 40, 44, 52, 53, 57 Hollenstein, T., 211, 233 Hollich, G. J., 2–12, 21, 24 Holyoak, K. J., 451 Honomichl, R., 449 Hood, B. M., 345 Horn, G., 333 Horst, J. S., 136, 141, 152, 159, 160, 162, 176, 177 Howes, P., 215 Hsee, C. K., 266 Huesmann, L. R., 299, 301–2, 311 Huh, D., 211, 232 Hund, A. M., 387–93, 395–401, 406 Huttenlocher, J., 373–4, 379, 385–7 Huttenlocher, P. R., 341

I Imai, M., 16–17 Isen, A. M., 114, 117 Isotomina, Z. M., 407 Izard, C. E., 354

J Jacobs, J. E., 268–9, 273–4 Jaswal, V. K., 140 Jenkins, J. M., 189, 206, 236 Jessor, R., 298 Johnson: J. S., 388 Johnson, C. J., 48, 50–1, 59 Jorm, A. F., 35, 37–8, 40, 46–7, 51, 53

K Kahn, R. L., 191 Kahneman, D., 253, 254, 258, 259, 263, 265, 266, 269–70, 273 Kaldy, Z., 152, 172 Kampe, K., 343, 350–1, 352 Kanwisher, N., 332 Kashy, D. A., 239 Kazdin, A., 318 Keil, F. C., 137, 139, 179 Kelemen, D., 143 Kellenbach, M. L., 140, 153, 171 Kelley, H. H., 193, 197 Kellman, P. J., 148, 151 Kemler Nelson, D. G., 136, 140, 143, 171 Kerns, K. A., 116 Kerns, S. E. U., 291, 310 Kerr, M., 211, 232 Kilner, J. M., 349 Kim, T., 311–12 Kimble, C. E., 359 Kirkpatrick, L. A., 89 Kirschbaum, C., 100 Klaczynski, P. A., 225, 252, 271–2, 274–5 Klahr, D, 419–70 Kleiner, C. L., 336 Klimes-Dougan, B., 295 Knox, L., 288, 294 Kobak, R. R., 93, 115 Kobiella, A., 354–5 Kochanska, G., 212, 230 Kokis, J., 252, 268–9, 271–3 Kopp, C. B., 92 Kosslyn, S. M., 391 Kuczynski, L., 193, 196, 201 Kuhn, D., 427, 437 Kurucz, J., 358 Kyte, C. S., 48, 50–1, 59

L LaBerge, D., 38, 39 Lakey, B., 95 Lamey, A., 229 Landau, B., 144 Landerl, K., 65 Landi, N., 68 Lane, R. D., 117 Larsen, R. J., 91, 118 Larson, R. W., 94 Laursen, B., 86, 118 Learmonth, A. E., 411 Leslie, A. M., 153, 171, 255 Levesque, H. J., 256, 263 Levin, I., 49, 270 Lewis, M. D., 193, 212, 233 Liben, L. S., 412 Lipsey, M. W., 301 Little, T., 290 Lochman, J. E., 317–18 Logan, G. D., 34, 38, 39, 40, 52 Loney, B. R., 110 Lorenz, K., 288 Lukatela, K., 33 Luntz, B. K., 303–4 Lurie, S., 61, 66 Luthar, S. S., 297, 304 Lytton, H., 191, 193, 195, 212, 225, 231

Maguire, M. J., 12, 19 Maier, N. R. F., 451 Malamuth, N. M., 296 Mandler, J. M., 161, 179 Manis, F. R., 46, 57 Mareschal, D., 165, 172 Martin, A., 165 Martin-Chang, S. L., 68–70 Martini, T. S., 213, 230 Marzolf, D. P., 411 Masche, J. G., 213 Masterson, J., 19 Matan, A., 137, 140, 147, 150 Mervis, C. B., 144, 176 Mickelson, K. D., 95–6, 109 Mikulincer, M., 94–5, 116, 117 Miller, J. B., 100, 103 Miller, J. D., 300 Miller, P. H., 407-8 Miller, R., 56, 60, 62 Minuchin, P., 194 Mo, L., 440, 449 Mofﬁtt, T. E., 288, 295, 300, 307, 310 Moos, R. H., 104 Morrow, D. G., 407 Morton, J., 333, 335, 336 Mrazek, P. J., 288 Muller, D., 53–5, 60 Mungas, D., 300 Munte, T. F., 358 Murphy, G. L., 145

M N Ma, W., 18, 19, 21 Macchi Cassia, V., 338 Maccoby, E. E., 194, 200 McCormick, C. M., 111 McDonald, L., 316 McDonnough, C., 18, 21 McDonough, L., 179 McGuire, S., 220 McHale, J. P., 195, 220, 238 McKay, M., 309 McLoyd, V., 305, 306 McMurray, B., 4 McNamara, T. P., 379, 384–5, 391, 393 MacWhinney, B., 6 Madole, K. L., 135–85

Naigles, L., 12 Nation, K., 34, 52–4 Nation, M., 319 Needham, A., 163 Nelson, C. A., 353, 354 Nelson, K., 136, 139, 144, 147, 169, 172, 175 Newcombe, N., 385, 391, 411, 412 Newport, E. L., 412 Newson, E., 194 Newson, J., 194 Nichols-Whitehead, P., 406 Niessing, J., 349 Nigam, M., 461 Nisbett, R. E., 262, 263, 267

O Oakes, L. M., 135–85, 412 Ohman, A., 360 Olds, D. L., 315 Olson, D. R., 39, 60, 62 Olweus, D., 290, 316–17 Over, D. E., 257–8 Owen, M. T., 215, 234

P Paivio, A., 17 Parish, J., 3 Parker, S. W., 355 Patrick, D. L., 104 Patterson, G. R., 193, 194, 303, 315 Paulesu, E., 65 Pelphrey, K. A., 352, 353 Perfetti, C. A., 35, 38, 40, 42, 44, 52, 53, 57, 67 Perkins, D. N., 251, 259–61, 265 Perone, S., 136, 141, 150, 152, 159–60, 162–5, 169–70, 172, 174 Pettit, G. S., 304, 307, 311 Piaget, J., 374, 376 Pipp, S., 115 Porges, S. W., 98–9 Posner, M. I., 345 Potenza, M., 268–9, 273–4 Poulin-Dubois, D., 11, 20 Pourtois, G., 360 Powers, S. I., 194 Prasada, S., 140 Prinstein, M. J., 297 Pronin, E., 265 Pruden, K., 9–11, 15–16 Pulverman, R., 13–16

Q Quinn, P. C., 339, 340

R Rakison, D. H., 142, 148 Rasbash, J., 221, 239 Rattermann, M. J., 423 Recker, K. M., 387, 389, 403–4, 406, 408, 409 Reese, H. W., 405

Regier, T., 140 Reid, V. M., 353 Reis, H. T., 106, 118 Reitsma, P., 38, 40, 44–5, 46, 47, 48, 52, 54, 56–7, 64 Repetti, R. L., 91, 92, 96–8, 99, 108 Reyna, V. F., 252, 270–1 Reynolds, G. D., 354 Richards, J. E., 342 Rinaldi, C. M., 221, 238 Robinson, C. W., 158 Rodriguez, E., 350, 352 Roenker, D. L., 381 Rogers, H., 39 Rogoff, B., 377, 405, 407 Ross-Sheehy, S., 163 Rothbart, M. K., 104 Rousseau, J. J., 288 Ryff, C. D., 118

S Salmivalli, C., 290, 298 Salovey, P., 101–2, 104 Sameroff, A. J., 194, 197 Sampson, R. J., 305–6 Samuels, S. J., 70 Sanchez, R. P., 458 Sanders, M. R., 194 Scarpa, A., 99 Scarr, S., 194 Schaefer, E. S., 103 Schaffer, H. R., 194 Scherer, K. R., 355 Schermerhorn, A. C., 187–250 Schoppe, S. J., 216, 234 Schoppe-Sullivan, S. J., 201, 216, 235 Schore, A. N., 92, 97, 115 Sejnowski, T. J., 349 Senju, A., 353 Serrano, J. M., 355 Sexton, T. L., 315 Seymour, P. H. K., 47, 49 Shaddy, D. J., 158 Shahar-Yames, D., 70–2 Shalev, C., 58–59, 64–5 Sharon, T., 13 Shatil, E., 42 Shaver, P. R., 84, 88, 89, 91, 94, 95, 96

Shearer, C. L., 213, 231 Siegel, D. J., 113 Siegler, R. S., 420, 461, 464 Simion, F., 337–8 Simmering, V. R., 406 Singer, H., 70 Slater, A., 339–40 Slobin, D., 14 Smith, L. B., 5, 24, 146, 178, 194, 222, 374, 409 Smith, P. K., 296 Snedeker, J., 16–19 Sodian, B., 439–40 Soenens, B., 109 Soken, N. H., 357 Song, L., 15–16 Spear, L. P., 94 Spencer, J. P., 387–8, 390, 406 Spencer, R. M., 424 Spetch, M. L., 411 Sroufe, L. A., 85, 89 Stanovich, K. E., 69 Stattin, H., 232 Steinberg, L. D., 213, 231 Stevenson-Hinde, J., 199 Stice, E., 214 Strahan, D., 373, 380, 407, 408, 410 Striano, T., 361

T Talbot, J. A., 217, 234 Tallon-Baudry, C., 349, 352 Talmy, L., 12–13 Tardif, T., 12, 18–19 Tashiro, T., 319 Taylor, M. J., 342 Tellinghuisen, D. J., 162 Thelen, E., 189, 199, 227–8, 374, 409 Thompson, R. A., 192 Tolan, P. H., 289, 292, 300, 305, 307, 309, 313 Tooby, J., 289 Toplak, M. E., 253, 256, 260 Tremblay, R. E., 305, 311, 318 Trickett, P. K., 214, 231 Truxaw, D., 143 Tucker, C. J., 214, 230 Tunmer, W., 54, 69 Tversky, A., 253, 254, 255, 258, 265, 269–70, 273

Twyman, A., 411 Tzourio-Mazoyer, N., 340–3

U Uchino, B. N., 103 Uttal, D. H., 405

V Vaillancourt, T., 292, 297, 304 Valenza, E., 337 Vellutino, F. R., 37 Vuchinich, S., 221, 238 Vuilleumier, P., 333 Vygotsky, L. S., 374, 376–7

W Walker-Andrews, A. S., 355 Warren, S. L., 214, 232 Wason, P. C., 256–7 Webster-Stratton, C., 315 Weisberg, R. W., 424 Weiss, B., 291 Weiss, R. S., 107 Welder, A. N., 137, 171 Wellman, H. M., 379, 405 Wells, K. C., 317–18 Whitehead, A. N., 421 Wilcox, T., 139, 140, 141, 142, 153, 171, 173 Wilson, D. B., 313, 316, 317 Wojciulik, E., 359 Wolk, D. A., 152, 172 Woody-Ramsey, J. D., 407–8 Wright, G., 55

Y Yanowitz, K. L., 458 Yantis, S., 158 Yoon, E. Y., 153, 171 Yoshikawa, H., 311 Younger, B. A., 148, 161

Z Zeifman, D., 86, 104–5, 106–7 Ziegler, A., 427 Ziegler, J. C., 65

This page intentionally left blank

Subject Index

A aardvark tests: remote transfer, 440 accessing: problem-solving strategies, 453–9 accuracy: orthographic learning, 61–9 ACR see Adjusted Ratio Clustering action names, 3, 10–21 action tests, 169–71 actions, 12–19 alternative, 255 categorization, 15–16 characteristic, 136 Actor-Partner Interdependence Model (APIM), 239 adaptive function: aggression, 293–8 Adjusted Ratio Clustering (ACR), 381–2 adjustment, 104, 109–10 adolescence: aggression, 287–330 attachment theory, 87–94, 102–16 transactional family dynamics, 231–3 adults: attachment theory, 89–91 face processing, 358–60 object function, 179 affect regulation: attachment theory, 83–4, 90–114 affective forecasting, 266 affordance see object function age groups, 167–77, 403–5, 427–48, 458 agency: transactional family dynamics, 196 agent-produced movement, 143 aggression, 287–330 adaptive function, 293–8 calibration, 296–7 classiﬁcation, 290 cognitive behavior, 301, 308–14 developmental perspectives, 295–6 hierarchies, 294–7, 304 motivation for change, 297–8

predictability, 311–12 research, 318–21 Aggression Replacement Training, 314 alternative actions: rational thought, 255 Alternative Thinking Strategies Programs (provision), 314 amplitudes, 227 amygdala: brain function, 333, 359 analogies: problems, 448–9 relationships, 85–6 remote transfer, 460 see also naturalistic approaches animate object function, 143 anti-violence programming, 293 antisocial behavior see aggression anxiety, 88, 92–9, 103, 109–12 APA Presidential Task Force on Evidence-Based Practice, 291 APIM see Actor-Partner Interdependence Model appearance: infant attention, 155–75 articulation: orthographic learning, 50 Assessment, 430 association, 5–6, 9–10 bias, 263, 392 categorization, 405 orthography, 37–8 rational thought, 257–9, 269 assumptions: ECM model, 6–7 asteroid/spring problems, 452–6 attachment theory, 83–134, 232 attention, 104 bias, 338 cues, 6, 9, 22 function, 154–77 attractiveness, 339–40 attractors, 26–7 autism, 21–3 automaticity, 34, 38–9

479

autonomy: affect regulation, 93–6 attachment theory, 87, 108–9, 114 dual-process models, 254 orthography, 38 availability: organism-environment interactions, 376, 390, 398–401 aversive behavior, 303 see also aggression averted gaze, 343–5, 350–2, 361 avoidance, 88, 94–7, 103, 109–12

B backward priming, 34 ball experiments, 428–9 barriers, 393 see also boundaries baselines, 227 bear experiments, 440–8 behavior: aggression, 307–8 attachment theory, 84–5 safe haven, 85–7, 105–6 secure base behavior, 85–6, 105 self-regulation, 113–14 see also normative behavior; social behavior belief bias, 258, 265, 272 bias: hostility, 301, 307 organism-environment interactions, 386–412 rational thought, 251–85 bidirectionality, 195–6, 200–1 biobehavioral approaches: attachment theory/affect regulation, 100–2 bioecological theories, 197 biosocial approaches, 97 Blueprints projects, 318–19 blunted hypothalamic-pituitaryadrenocortical axis activity, 100 boat experiments, 440–8 BOLD responses, 349–51 bonding: attachment theory, 86 ‘‘bootstrapping’’, 20

borders: categorical bias, 338 see also boundaries bottom-up decoding, 66–7 boundaries, 338, 379, 391–401, 406–7 box experiments, 435–8 brain function, 331–71 amygdala, 333, 359 damage, 358 hypothalamic-pituitary-adrenocortical axis activity, 98–100, 105, 110–12 temporal gyrus, 340–3 see also neuroscience bridges: autonomy, 108–9 bridging traditions: adult-child relationships, 89–91 broaden-and-build theory: positive emotional experiences, 117 broadening attachment, 107 Bronfenbrenner, U., 197 building evidence-based programs, 318–21 bullying, 290–7, 304, 316–17 see also aggression

C calibration: understanding aggression, 296–7 card selection tasks, 256–7, 274 caregivers, 92–4, 355 see also attachment theory caricatures: word learning, 4–5 categorical bias: organism-environment interactions, 382–412 categorization, 12–16 functional features, 175–7 organism-environment interactions, 378–412 causality, 307–21 ‘‘Cave’’ problems, 450–2 CDI see Communicative Development Inventory ceiling tracking, 336 Center for the Study and Prevention of Violence, 318–19 cerebral cortex, 331, 340–3 see also neuroscience

certainty: problem-solving strategies, 429, 435–7 changing aggression, 297–8 characteristic actions: functional features, 136 child-to-parent inﬂuence processes, 191 choicefulness: self-determination theory, 96, 114 chronotropic control, 98 clarity: affect regulation, 104, 111–12 classiﬁcation: aggression, 290 bias, 263–7 classrooms, 308–9, 316–17 close interpersonal relationships, 300–5, 313–16 clustering, 297, 376–84, 398, 401–12 co-opting resources, 294 coactivation: affective states, 118 coarse-grained information, 388 coding: bias, 389–90 spatial categorization, 403–5 coercive styles: preventing aggression, 315 cognition, 61–3, 178 aggression, 301, 308–14 rational thought, 251–85 social neuroscience, 331–418 coincident conditions, 9–11 collective efﬁcacy: proximal contexts, 306 collective variables: family inﬂuence processes, 227 coming ‘‘on line’’: brain development, 334 Communicative Development Inventory (CDI), 18 companionship, 104 compensation, 58–61, 108 competencies: masking cognitive abilities, 374 competition models, 3, 6 complex face processing, 339–40, 347 compliance: aggression, 304 parent-child relationships, 231–2 compression solutions, 450–2

computational models, 36, 40 conceptualization, 137, 169 function, 145–53 concreteness, 18, 267–9 conduct-disorders: transactional family dynamics, 230 Conﬁg stimuli, 335–7 conﬁgurations: coding spatial groups, 405 face processing, 335 conﬁrmation bias, 266–7 conﬂict, 9–11, 223–6, 235–40 see also family relationships congruence, 69, 356–7 conjunction effects, 273–4 Conlern systems, 333 connectionist models, 52 conscious regulation, 101 consonant clusters, 64 Conspec systems, 333–5, 340, 345 constancy, 39 constraints theories, 4 construing functional features, 135–85 contaminated mindware, 260–4 content, 423, 433 context, 41, 300, 305–6, 313–18 remote transfer, 423–6 self-teaching, 67–70 contiguity, 393–4, 406 contingency, 230–1 contradiction, 196 contrast: face processing, 338–40 control, 58–9, 72, 226, 427–38, 458–62 control of variables strategy (CVS), 427–35, 458–62 convergence studies: orthographic learning, 46 coolness: parent-child relationships, 224 coparenting, 234–5 coping: stress, 117 corporal punishment, 303–4 correlates: affect regulation, 110–12 cortical structures, 331, 340–3 see also neuroscience cortisol levels, 100, 110–11

critical assessors: rational thought, 252 critical points: orthographic learning, 67 cross-cultural approaches, 426, 448–55 cross-modal priming, 45 crystallized orthography, 43 cueing spatial relations, 406–8 cues, 3–12, 16–23 aggression, 301 categorical bias, 390–8 context, 70 direct gaze, 352 face perception, 346–7 recall, 379–80 remote transfer, 456 spatial categorization skills, 409 cultural settings: problem-solving strategies, 462–3 cumulative natures, 306–7 CV syllables, 61, 64

D dance metaphor: family inﬂuence processes, 190–1 deactivating strategies, 95 decipherability, 39 decisions: categorical bias, 386 framing effects, 269–70 decoding, 31, 35–82 decoupling, 255–6, 261, 272 defense: understanding aggression, 294 deﬁciencies: executing components, 457 deﬁcit hypothesis, 37, 58–9 delay: transfer strategies, 435–48 delinquency, 232 see also aggression denominator neglect, 271–2 dependence, 104, 108–9, 114 depression, 92–4, 99, 232–3 descriptive invariance, 270 designing experiments: remote transfer, 427–35 detection: eye gaze, 343–5

deterring rivals, 294 developmental perspectives: aggression, 295–6 attachment theory/affect regulation, 83–134 developmental transitions: attachment theory, 85–8 deviant peer contagion, 298 dialectical models, 196 diary method experiments, 235–7 differentiation: affect regulation, 93–6 direct aggression, 290 direct gaze, 343–53, 361 direct-retrieval mechanisms, 39 direction: organizational strategies, 404–5 disabilities, 42–6 autism, 21–3 early onset, 65 orthographic learning, 57–65 disagreements: family relationships, 224–5, 237 discipline, 201, 303–4 discrimination, 12 dishabituation scores, 159–60, 169–71 disjunctive reasoning, 256 disorders, 42–6 autism, 21–3 early onset, 65 executing components, 457 orthographic learning, 57–65 parent-child relationships, 230 dissemination, 319–20 dissociable neural pathways, 358 distinctiveness: word speciﬁcness, 39 distress, 85, 92–4, 226 divergence studies: orthographic learning, 47, 66 divorce, 110–11 dollhouse experiments, 380–1, 406–7 drives: affect regulation, 111–13 dual-process theories, 253–9 dual-route theorizing, 36–9 durability: orthographic learning, 52–3 Durations, 224–5 dyads, 114–16, 227–30, 239 dynamic ﬁeld theories, 388

dynamic risk factors, 298–9 dynamic systems theories, 198–9, 237 dyslexia, 57–64 dysregulation, 110–11, 115–16, 236–7

E ‘‘earlier better than never’’ dictum, 310–11 early concept formation, 144–5 early onset hypothesis, 40–2, 63–7 early pioneers: orthographic studies, 44–6 early starter groups, 295 ECM see Emergentist Coalition Model ecological model of human development, 197 ecology, 298–309 education: classrooms, 308–9, 316–17 Kindergarten, 444–8 remote transfer, 460–3 schools, 305–9, 315–20, 448–55 training programs, 314–15 see also learning EEG studies, 341–3, 349–51 effectiveness: context, 69 providers, 93–4 ego-resilience, 104 egocentric mindware, 265–6 elephant tale experiments, 456 Emergentist Coalition Model (ECM), 2–11, 20–5 emotion, 90–112 attention, 333 intelligence, 101–2 parent-child relationships, 230 regulation, 84, 90–1 visual processing, 354–62 see also attachment theory empirical work: orthographic learning, 43–72 remote transfer, 420–2 transactional family dynamics, 228–39 encoding features, 455 enhanced oscillations, 350 environment, 91, 300–1, 333 episodes, 222–5 epistemic rationality, 252–3 ERPs see event-related brain potentials errors, 253, 259–63

eustress, 226 evaluation-disabling properties, 261–2 event sequences, 145–9, 152–3 event-related brain potentials (ERPs), 341–57, 361–2 evidence-based practices, 291, 318–21, 426–55 evidential rationality, 252–3 evoked responses, 350 evolutionary perspectives: aggression, 293–5 examination times, 166 executing: problem solving strategies, 453–9 experience: categorization, 409–12 visiting locations, 393–5, 409–12 experimental approaches: remote transfer, 426–48 experimentally tractable deﬁnitions: rational thought, 252–4 expertise: orthography, 33–46, 57–64, 68 spatial categorization, 401–12 see also orthographic learning explicit conditions, 442–4, 461 exploration: autonomy, 104, 108–9 problem solving strategies, 430 exposure: spatial categorization, 409 external affect regulation, 95 externalization, 96, 107–8, 113, 232–8 eye gaze, 343–62

F face processing, 334–43, 354–62 familiarization, 15, 38–40, 62, 158–9, 173–7 context, 68–9 early onset, 42 face processing, 340, 361 see also orthographic learning families: aggression, 302–16 transaction dynamics, 187–250 see also caregivers; parenting Families and Schools Together, 315–16 Fast Track (Conduct Problems Prevention Research Group), 291–2

feedback: problem-solving strategies, 442–7, 462 FFA see fusiform face area ﬁdelity, 320 ﬁgures in reserve, 86 ﬁne-grained information, 387–90 First grade, 445–6 ﬁshermen experiments, 440–8 ‘‘ﬂashlights’’, 22 ﬂexible adaptability, 233 ﬂuency, 61, 66–7 fMRI see functional magnetic resonance imaging focal bias, 258–75 folk tale experiments, 448–55 follow-up phases, 438, 458 following gaze, 345–7 forced-choice methods, 87 forecasting, 266 four-card selection tasks, 256–7, 274 frameworks: organism-environment interactions, 373–8 self-teaching, 46–57 transactional family dynamics, 228–39 framing effects, 269–71 free recall, 407–10 fronto-central negative components (Nc), 353–5 function: object representation, 135–85 function words, 44 Functional Family Therapy, 315 functional magnetic resonance imaging (fMRI), 341, 349–52, 356 fusiform face area (FFA), 332, 342 fusiform gyrus, 348 fusiform responses, 359

G gamma-band oscillations, 349–53 gaps: mindware, 260–5, 273–4 gaze: eye gaze processing, 343–62 general enhancement, 309 general intelligence, 62 general theoretical frameworks, 373–8 genetic factors: nature/nurture, 333

gift experiments, 442–4 glimmerings: orthographic learning, 65 goal objects, 454–7 good reasoners: measuring remote transfer, 434 graphemes, 35, 64 guided participation, 377

H ‘‘hands-on’’ tasks, 428–33 harmony, 226 heart rate, 98–9, 105 height tasks: problem-solving strategies, 440–2 heterogeneity, 289 heteronomy, 96 heuristics, 251–85 hierarchies: aggression, 294–7, 304 family transactional dynamics, 203–28 high-school, 448–55 HIPE model/theory, 138, 146–50 HIV see human immunodeﬁciency virus homeostatic regulation, 115 homonyms, 44 homophily, 305 homophonic spelling, 49–64 hostile attributional bias, 301, 307 HPA see hypothalamic-pituitaryadrenocortical axis human development: Bronfenbrenner’s ecological model, 197 human immunodeﬁciency virus (HIV), 449 Human Simulation paradigm, 16–17 hybrids: rational thought, 265, 274–5 hyperlexics, 34 hyperreactivity, 99 hypothalamic-pituitary-adrenocortical axis (HPA): brain function, 98–100, 105, 110–12 hypothesis-testing-strategies, 439–48

I iatrogenic effects, 313 imageability, 17–19 imagined movement, 405

immediate naming tasks, 46 implementation, 319–20 implicit feedback, 444, 462 impulsivity, 61–2, 263 in the moment behavior: organism-environment interactions, 402–9 Incredible Years Training for Parents Programs, 315 independence, 114 indeterminacy, 435–9 indexes, 103–4 indicated interventions, 288 indirect aggression, 290 individual differences: attachment theory, 88–9 self-teaching, 57–64 individual-environment interactions, 197–8 individuals, 84, 88–9, 300–5, 313–17 infants: attachment theory, 89–91 direct gaze, 347–53 function, 135–85 social brain functions, 331–71 inﬂicting costs, 294 inﬂuence, 222–31, 238–42, 374, 390–401 information-processing, 114, 162, 301 inheritance, 333 initial learning, 432 inﬂuence-processes, 240 inner resources, 85 input, 114 insecurity, 113 instance-based learning theories, 52 instantiations, 142 institutional rearing, 355 instrumental rationality, 252 integration, 113–14, 356–7 intelligence, 22, 62, 101–2, 251–2 intentionality, 22, 150–1, 241 interactions, 89, 111, 147, 300–7, 313–16, 357–62, 373–418 face identity/eye gaze/emotion, 357–62 interpersonal, 89, 111, 300–5, 313–16 organism-environment, 373–418 risk, 306–7 Interactive Intermodal Preferential Looking Paradigm (Interactive IPLP), 8–9 interactive natures: aggression, 306–7 Interactive Specialization, 334

intermediates, 442–4 Intermodal Preferential Looking Paradigm (IPLP), 7 internalization, 85, 93–6, 107, 113, 233–8 interparental relationships, 217–19, 234–8 interpersonal interactions, 89, 111, 300–5, 313–16 interpretation, 22 interventions, 288, 309–18 see also preventing aggression intimacy, 104–5 intraparietal sulcus (IPS), 350, 351 intuition, 171 inventories, 103–4 inverted faces, 342–4, 348–52 IPS see intraparietal sulcus irrelevant articulation, 50 isomorphic tasks: problem-solving strategies, 428–30, 446–50 item-based self-teaching, 40–1

J judgements, 384–5

K Kindergarten, 444–8 knowledge bases, 259

L labels, 10–12 landmarks, 379 language, 1–29, 254–9 latency, 342, 350 lateral motion, 346–7 lay theories, 266 leadership training programs, 312 learning, 1–82, 393–4, 406, 457–63 initial, 432 remote transfer, 427–8, 443–6 self-teaching, 31–72 task differences, 406 lessons learned: preventing aggression, 287–330 letter-by-letter decoding, 65–7

letter-sound knowledge, 42 Levels, 309–10, 317 lexicalization, 41, 49–50, 63–4 Life Skills Training, 314 lifespans: attachment theory/affect regulation, 83–134 linear systems models, 336 linguistic cues, 5–9, 20 linking prevention/positive youth development, 320–1 localist models, 34–5 localization, 348 locations, 384–408 locomotion, 393 loneliness, 104 long-term transfer processes, 451 looking times, 167–8, 173–7 loops, 66

M MACS see Metropolitan Area Child Study maladaptive regulation, 101 maladjustment, 92, 107, 232–3 manners, 10–16 mapping, 9, 11–13, 16–20 problem-solving strategies, 453–9 transactional family dynamics, 228–39 marital relationships, 195 infant social neuroscience, 333–42 interparental relationships, 235–8 markers: remote transfer experiments, 436 masking: cognitive competencies, 374 Matching Familiar Figures Test, 62 matching objects, 7–8 maternal attachment anxiety, 112 maturation, 94 ‘‘Maturational’’ perspectives: brain development, 334 maximal tracking: face processing, 336 maximizing utility: preference patterns, 253 memory: bias, 388 categorization, 402 context, 67 dyslexia, 62–3

mental development, 94, 95–6, 145 meta-cognitive knowledge, 407 metric coding, 390 Metropolitan Area Child Study (MACS), 291–2, 298, 307–12, 316–18 mice experiments, 439–48 middle-school, 448–55 mind organization, 113 mindware, 259–67, 273–4 miserliness, 260 mixed orientations, 106 models: APIM, 239 cognitive-ecological, 308–9 competition, 3, 6 computational, 36, 40 connectionist, 52 dialectical, 196 dual-process, 254–9 dynamic systems, 237 focal, 258–75 linear systems, 336 localist, 34–5 simple, 34–5 threshold, 52 transactional family dynamics, 228 triangle, 35 modiﬁable risk factors, 292 modularizing, 38 monitoring, 304 morphemes, 38–9, 41 morphology, 62, 64 most effective preventive measures, 313–17 most-at-risk children, 312–13 motivation for change: aggression, 297–8 motor development, 199 movement, 143, 405 multi-dimensional social-cognitive/social skills interventions, 314 multi-trace computational models, 40 multiple determinants, 307–8 multiply-determined behavior: aggression, 298–300 Multisystemic Family Therapy, 316 mutual inﬂuence processes, 197–8 myriads, 311 myside bias, 258, 262–5, 274–5

N N170 responses, 341–2, 347–8, 355 N290 responses, 342, 347–8, 354 naming tasks, 46 naturalistic approaches: problem-solving strategies, 426, 448–55 naturalistic face processing, 335 naturalistic settings: problem-solving strategies, 462–3 nature, 288, 333 Nc see fronto-central negative components Neale tests, 47 near transfer, 430–2, 446–7 nearby locations, 393–8 negative affect, 95, 116–18 negative components, 353–5 neighborhoods, 305–6 nested systems, 289 Network of Relationships Inventory (NRI), 104–6 neural pathways, 358 neural processing, 347–53 neuroscience, 331–71 neutral stimuli, 360 neutral tasks: problem-solving strategies, 451 newborns: face processing, 339–45 no probe conditions, 430, 462 nominal context, 69 non-face structural preferences, 337–9 non-localist connections, 35 non-matching objects, 7–8 non-modiﬁable risk factors, 292, 299 non-verbal intelligence, 22 nonsocial resources, 99 normative behavior, 84–5, 88, 107, 295 aggression, 302, 312 dyslexia, 57 focal bias, 272 nouns, 3, 10–12, 18–21 novelty, 15, 38, 156–9, 166 dyslexia, 62 face pathways, 361 preference, 157–8 remote transfer experiments, 433 self-teaching, 68–9 NRI see Network of Relationships Inventory

Nurse Home Visitation Programs, 315 nurture, 333

O objects, 3, 7–8, 18–21, 379–85, 440–51 bias, 391–412 examining times, 166 eye-gazing, 353–4 function, 145–9, 155 physical structures, 151–2 recall, 379–84 reminding, 456 representation, 135–85 spatial thinking, 384–401 surface features, 155, 158 obligating: orthographic learning, 47–8 offsetting negative affect, 117 Olweus Bullying Prevention Programs, 316–17 onset, 40–2, 64–7 optimizing input, 114 oral reading, 49, 55–6, 64 organism-environment interactions, 373–418 organization: recall, 379–84 regulatory processes, 113 spatial categorization, 401–7 orienting, 359–60 orthographic learning, 31–82 orthography-secondary/phonology-primary hypothesis, 42–3, 62–3 oscillations, 349–53 outcome: infant attention, 167–75 override failure, 260, 262–3, 264, 265 overwhelming emotions, 101

P ‘‘packaging’’ see mapping paper-and-pencil posttests, 428 paradigms: problem-solving strategies, 422–6 paradox: problem-solving strategies, 422–6 parasympathetic nervous system (PNS), 98–9, 105, 111–12 parenting, 103, 195, 302–3, 314–16

empirical contributions, 207–17 interparental relationships, 217–19, 234–8 transactional family dynamics, 200–1, 229–38 partner effects, 239 paths, 10–16 patterns: attachment transfer, 105–8 ‘‘peak-to-valley’’ method, 105 peers, 106–10, 293–8, 302–9 perception: emotion, 354–7 perception-action research, 389 perceptual: boundaries, 379, 393 cues, 5–7, 10–12, 16–21 ﬁelds, bias, 388 properties, 137 salience, 16–19 structures, 398–401 theories, 4 ‘‘personality’’, 200 PET see positron emission tomography phase shifts, 227–8 phase-locked oscillations, 350 phonemes, 45, 64 phonetic spellings, 58 phonological recoding, 31, 35–56, 57–82 phonology-primary/orthography-secondary hypothesis, 42–3, 62–3 physicality, 96–7, 405, 442–7 affection, 104 barriers/boundaries, 379, 393 physiological processes: attachment theory/affect regulation, 97–100 correlates, 110–12 stress regulation, 105 picture cues, 70 pioneering studies: orthography, 43–6 PNS see parasympathetic nervous system policies: aggression, 310–18 popularity: aggression, 292, 297, 304 positional regularities, 41 positions: spatial thinking, 384–401 positive affect, 83, 95, 117–18

positive youth development, 320–1 positive-capture, 437–8 positron emission tomography (PET), 341–2, 343 postnatal development, 332 posttests, 442–8 problem solving strategies, 428, 433–4, 437, 443–8, 458 poverty, 305–6, 310 power hierarchies, 294–7, 304 practice: aggression research, 318–21 context, 68–9 pragmatic theories, 4–5, 24 pre-self-teaching studies, 44–6 predictability: aggression, 311–12 cortisol levels, 111 orthographic learning, 61–3, 69–71 preferences: non-face structures, 337–9 preferential tracking, 335–9 premature transferring, 108 premises, 272 pretests, 437 preventing aggression, 287–330 previous locations, 385–6 primary attachment ﬁgures, 105–6, 116 priming, 45 prior knowledge, 272 proactivity: aggression, 290 stress coping, 117 probe conditions, 429–31, 462 problems: solving strategies, 419–70 violence, 294 processing, 60, 64, 99, 265–9, 332–64 attachment theory, 100–2, 114 programs, 269, 310–18 promotion, 114 pronunciation, 42, 50, 63–6 prosody, 355–6, 357 protective factors, 298–9 Providing Alternative Thinking Strategies Programs, 314 providing security, 90–4 proximal contexts, 300, 305–6, 313–17 proximal development: organism-environment interactions, 376–7

proximity maintenance, 86 proximity seeking, 85, 87 pseudowords, 32, 36–7 psycho-educational programs, 291 psychopathology, 198 puberty, 231 see also adolescence ‘‘pull’’: organism-environment interactions, 388–92, 396–402

Q qualitative differences, 225 ‘‘quick and dirty’’ neural pathways, 333–4

R radiation problems, 451 radical middles: word learning, 23–5 ramp experiments, 428–9 RAN see Rapid serial naming random learning conditions, 393–4 ranking, 107 Rapid serial naming (RAN), 60–3 rapidity, 52–3 rational thought, 251–85 reactive aggression, 290 reading: orthographic learning, 31–75 ‘‘real time’’: family transactional dynamics, 222 real words: orthographic learning, 54–5, 58 reasoning: remote transfer, 419–70 recall, 378–84, 407–10 RECAP program, 291 reciprocation, 230, 231, 239 recognition, 31–75 recovery, 104 reﬂection-impulsivity, 61–2 regression, 60, 62 regularities, 41 regulation: attachment theory, 83–4, 90–116 family systems theory, 200 transactional family dynamics, 236–7

relational aggression, 290 relational information, 403–5, 409 relations: appearance/action/outcome, 167–75 categorical bias, 395–8 eye-object, 353–4 relevant pronunciation, 50 reliable alliance, 104, 105 reminding, 456 remote transfer, 419–70 repair: affect regulation, 104 reports, 104 representations: orthographic learning, 34 problem-solving strategies, 453 see also function representation; object representation reproduction, 385–6, 389–90, 398–401, 405 research, 240–2, 310–21, 420–2 function, 154–77 reserve attachment ﬁgures, 107 reshufﬂing attachments, 86 Resolving Conﬂict Creatively Programs, 314 respiratory sinus arrhythmia (RSA), 98–9, 105, 111–12 retaliation, 295–6 retention, 68–9 rigidity, 233 rings of homeostasis, 97 risk, 93, 289, 298–321 rivalry see aggression robust learning, 458–9 romantic relationships, 85–6, 88–9, 90, 106–7, 112 rough and tumble play, 296 RSA see respiratory sinus arrhythmia

S safe haven behavior, 85–7, 105–6 salience, 11–19, 376–9, 391–2, 406–9 scaffolding, 376, 377–8 scales, 103–4 schematic faces, 335 schools, 305–9, 315–20, 448–55 scientiﬁc reasoning, 419–70 scrambled faces, 335 scripts, 302, 307–9 SCS see superior temporal sulcus

searching: organizational strategies, 404–5 Second grade, 447 Second Step Violence Prevention Programs, 314 secure base behavior, 85–6, 105 security, 85–94, 105–14, 232 see also attachment theory selected interventions, 288 selective attention, 408 see also attention self-control problems: rational thought, 265 self-efﬁcacy: aggression, 308 self-explanations: remote transfer, 462 self-expressiveness: attachment theory, 103 self-organizing: family relationships, 200, 227 self-perspectives: rational thought, 265–6 self-produced movement: functional features, 143 self-regulation, 96, 113–15, 200 self-reliance: attachment theory, 109 self-teaching, 31–72 context, 67–70 early onset, 64–7 self-understanding: aggression, 308 semantics, 57, 61, 62 sensitivity, 361 function, 149–53 sensory hypothesis, 336–7 separation distress, 85 serial processing, 254–9, 269 settings, 317–18, 462–3 problem-solving strategies, 462–3 sexual relationships, 90 see also romantic relationships shaping environments, 300–1 shifting attention, 155–65 short-latency, 350 short-term experience, 411 short-term memory: limitations, 172 sight vocabulary, 34, 40

silent reading, 44, 55–6, 60 similarities: cross-cultural approaches, 452 remote transfer, 423–4 simple models, 34–5 size tasks: problem-solving strategies, 440 skills: learning, 334 orthography, 33–46, 57–64, 68 spatial categorization, 401–12 snapshots, 440 SNS see sympathetic nervous system social behavior, 4–10, 20–4, 114, 290, 313–18 cognition, 331–71 hypothalamic-pituitary-adrenocortical axis activity, 99 transactional family dynamics, 199–200 solution tools, 454, 457 solutions, 457–8 sound, 42, 169–71 source information: problem-solving strategies, 424–6, 454–7 source separation: direct gaze, 348 spatial categorization, 350, 375–412 bias, 390–8 face processing, 336–7, 340, 347, 359–60 spatiotemporal experiences, 393, 406 speaker cues, 11 speciﬁc degrees, 104–5 speciﬁc risk factors, 313–17 speech melody, 355–6 speed, 61–2, 66 spelling, 32–6, 43–75 see also orthographic learning Spontaneously Generating Hypothesis Tests, 444 stability, 89 stage-based self-teaching, 40–1 stamp experiments, 436 state regulation, 91 static risk factors, 298 statue problems, 450–2 status, 294–7, 304 stereotypes, 274 strategies: problem solving, 419–70 streams, 358–9

stress, 99–100, 105, 111, 116, 226 structural processes, 337–9, 456–7 structure-action combination, 175 STS see superior temporal sulcus styles, 88–9, 109–10 sub-cortical structures, 331 see also neuroscience sub-syllabic CV, 61 subdivisions, 385, 386–7, 393 substantial time intervals, 451 superior temporal sulcus (STS), 342, 348–52, 359 support, 409–12 suppression, 104, 255 supra-lexical assistance, 42 surface features, 66–7, 155, 158 syllables, 61, 64 sympathetic nervous system (SNS), 98 syntactic cues, 10–12, 19–20 syntax, 61–3 systematic bias, 387 systematic prevention programming, 308, 310–12 Systems A-C: families, 222–42

thought: development, 251–85 organism-environment interactions, 373–418 threat: face processing, 361 threshold models, 52 time, 222–6, 393–5, 424–5, 451 see also posttests tonic levels, 100 ‘‘top-heavy’’ hypothesis, 338–40 toughness, 298 tour-planning, 410 tractable deﬁnitions: rational thought, 252–4 traditions, 89–91 training programs, 314–15 training-probe conditions, 429–31 transactional family dynamics, 187–250 transference, 86–7, 105–8, 410–11, 419–70 transfer, 430 transitions: attachment theory, 85–8 translating research, 318–21 triangle models, 35 tuning, 92, 97

T targeted interventions, 312–13 tasks: problem-solving strategies, 423–4 spatial relations, 406–8 taxonomy, 251–85, 422–6 problem-solving strategies, 422–6 temporal characteristics, 350, 425 temporal gyrus: brain function, 341–3 temporo-parietal regions, 352 tests: aardvark, 440 action, 169–71 categorical bias, 387, 398–401 Neale tests, 47 paper-and-pencil, 428 posttests, 428, 433–4, 437, 442–8, 458 pretests, 437 univariate tests, 107–8 theoretical frameworks, 373–8 theoretical rationality, 252–3 therapies, 315–16

U U-shaped patterns, 397–9 uncertainty, 399 unconscious regulation, 101 understanding aggression, 287–330 uniﬁed systems, 373–418 univariate tests, 107–8 universal interventions, 288 universalistic dualisms, 38 unpredictability: cortisol levels, 111 unregulated states, 101 unstructured tasks: spatial relations, 407

V vagal regulation see chronotropic control validating: ECM model, 7–21 variability, 60–3, 226–7, 306, 427–35

verbal feedback, 442–7 verbal protocols, 257–8 verbs, 3, 10–21 very near transfer, 430 violence, 293–4, 299, 305–8, 314–19 see also aggression visible boundaries, 391–3 visiting nearby locations, 393–5 visual cues, 380 visual ﬁelds, 338 visual inspection hypotheses, 48–51 visual-habituation, 161–2, 167–8 visual-orthographic processing, 64 vividness effects, 263, 267–9 vocabulary see language voices: emotion processing, 355–7 volition, 96, 109, 114 volume tasks, 440–1

W warped emotion processing, 99 ‘‘Weigh the Elephant’’ tale, 450 weight tasks: problem-solving strategies, 440–8 Woodcock Word Attack, 60–1 words: learning, 1–29 learning problem, 2–3 recognition, 31–75 working memory, 62–3, 172 writing: orthographic learning, 31–75

Y youth development, 320–1

Contents of Previous Volumes A Developmental Approach to Learning and Cognition Eugene S. Gollin Evidence for a Hierarchical Arrangement of Learning Processes Sheldon H. White Selected Anatomic Variables Analyzed for Interage Relationships of the Size-Size, Size-Gain, and Gain-Gain Varieties Howard V. Meredith

Volume 1 Responses of Infants and Children to Complex and Novel Stimulation Gordon N. Cantor Word Associations and Children’s Verbal Behavior David S. Palermo Change in the Stature and Body Weight of North American Boys during the Last 80 Years Howard V. Meredith Discrimination Learning Set in Children Hayne W. Reese Learning in the First Year of Life Lewis P. Lipsitt Some Methodological Contributions from a Functional Analysis of Child Development Sidney W. Bijou and Donald M. Baer The Hypothesis of Stimulus Interaction and an Explanation of Stimulus Compounding Charles C. Spiker The Development of ‘‘Overconstancy’’ in Space Perception Joachim F. Wohlwill Miniature Experiments in the Discrimination Learning of Retardates Betty J. House and David Zeaman

AUTHOR INDEX—SUBJECT INDEX Volume 3 Infant Sucking Behavior and Its Modiﬁcation Herbert Kaye The Study of Brain Electrical Activity in Infants Robert J. Ellingson Selective Auditory Attention in Children Eleanor E. Maccoby Stimulus Deﬁnition and Choice Michael D. Zeiler Experimental Analysis of Inferential Behavior in Children Tracy S. Kendler and Howard H. Kendler Perceptual Integration in Children Herbert L. Pick, Jr., Anne D. Pick, and Robert E. Klein Component Process Latencies in Reaction Times of Children and Adults Raymond H. Hohle

AUTHOR INDEX—SUBJECT INDEX

Volume 2 The Paired-Associates Method in the Study of Conﬂict Alfred Castaneda Transfer of Stimulus Pretraining to Motor Paired-Associate and Discrimination Learning Tasks Joan H. Cantor The Role of the Distance Receptors in the Development of Social Responsiveness Richard H. Walters and Ross D. Parke Social Reinforcement of Children’s Behavior Harold W. Stevenson Delayed Reinforcement Effects Glenn Terrell

AUTHOR INDEX—SUBJECT INDEX Volume 4 Developmental Studies of Figurative Perception David Elkind The Relations of Short-Term Memory to Development and Intelligence John M. Belmont and Earl C. Butterfield Learning, Developmental Research, and Individual Differences Frances Degen Horowitz

493

494

Contents of Previous Volumes

Psychophysiological Studies in Newborn Infants S. J. Hutt, H. G. Lenard, and H. F. R. Prechtl Development of the Sensory Analyzers during Infancy Yvonne Brackbill and Hiram E. Fitzgerald The Problem of Imitation Justin Aronfreed AUTHOR INDEX—SUBJECT INDEX

Volume 5 The Development of Human Fetal Activity and Its Relation to Postnatal Behavior Tryphena Humphrey Arousal Systems and Infant Heart Rate Responses Frances K. Graham and Jan C. Jackson Speciﬁc and Diversive Exploration Corinne Hutt Developmental Studies of Mediated Memory John H. Flavell Development and Choice Behavior in Probabilistic and Problem-Solving Tasks L. R. Goulet and Kathryn S. Goodwin AUTHOR INDEX—SUBJECT INDEX

Volume 6 Incentives and Learning in Children Sam L. Witryol Habituation in the Human Infant Wendell E. Jeffrey and Leslie B. Cohen Application of Hull–Spence Theory to the Discrimination Learning of Children Charles C. Spiker Growth in Body Size: A Compendium of Findings on Contemporary Children Living in Different Parts of the World Howard V. Meredith Imitation and Language Development James A. Sherman Conditional Responding as a Paradigm for Observational, Imitative Learning and Vicarious-Reinforcement Jacob L. Gewirtz AUTHOR INDEX—SUBJECT INDEX

Volume 7 Superstitious Behavior in Children: An Experimental Analysis Michael D. Zeiler Learning Strategies in Children from Different Socioeconomic Levels Jean L. Bresnahan and Martin M. Shapiro Time and Change in the Development of the Individual and Society Klaus F. Riegel The Nature and Development of Early Number Concepts Rochel Gelman Learning and Adaptation in Infancy: A Comparison of Models Arnold J. Sameroff AUTHOR INDEX—SUBJECT INDEX

Volume 8 Elaboration and Learning in Childhood and Adolescence William D. Rohwer, Jr. Exploratory Behavior and Human Development Jum C. Nunnally and L. Charles Lemond Operant Conditioning of Infant Behavior: A Review Robert C. Hulsebus Birth Order and Parental Experience in Monkeys and Man G. Mitchell and L. Schroers Fear of the Stranger: A Critical Examination Harriet L. Rheingold and Carol O. Eckerman Applications of Hull–Spence Theory to the Transfer of Discrimination Learning in Children Charles C. Spiker and Joan H. Cantor AUTHOR INDEX—SUBJECT INDEX

Volume 9 Children’s Discrimination Learning Based on Identity or Difference Betty J. House, Ann L. Brown, and Marcia S. Scott

Contents of Previous Volumes Two Aspects of Experience in Ontogeny: Development and Learning Hans G. Furth The Effects of Contextual Changes and Degree of Component Mastery on Transfer of Training Joseph C. Campione and Ann L. Brown Psychophysiological Functioning, Arousal, Attention, and Learning during the First Year of Life Richard Hirschman and Edward S. Katkin Self-Reinforcement Processes in Children John C. Masters and Janice R. Mokros AUTHOR INDEX—SUBJECT INDEX

Volume 10 Current Trends in Developmental Psychology Boyd R. McCandless and Mary Fulcher Geis The Development of Spatial Representations of Large-Scale Environments Alexander W. Siegel and Sheldon H. White Cognitive Perspectives on the Development of Memory John W. Hagen, Robert H. Jongeward, Jr., and Robert V. Kail, Jr. The Development of Memory: Knowing, Knowing About Knowing, and Knowing How to Know Ann L. Brown Developmental Trends in Visual Scanning Mary Carol Day The Development of Selective Attention: From Perceptual Exploration to Logical Search John C. Wright and Alice G. Vlietstra AUTHOR INDEX—SUBJECT INDEX

Volume 11 The Hyperactive Child: Characteristics, Treatment, and Evaluation of Research Design Gladys B. Baxley and Judith M. LeBlanc Peripheral and Neurochemical Parallels of Psychopathology: A Psychophysiological

495

Model Relating Autonomic Imbalance to Hyperactivity, Psychopathy, and Autism Stephen W. Porges Constructing Cognitive Operations Linguistically Harry Beilin Operant Acquisition of Social Behaviors in Infancy: Basic Problems and Constraints W. Stuart Millar Mother–Infant Interaction and Its Study Jacob L. Gewirtz and Elizabeth F. Boyd Symposium on Implications of Life-Span Developmental Psychology for Child Development: Introductory Remarks Paul B. Baltes Theory and Method in Life-Span Developmental Psychology: Implications for Child Development Aletha Huston-Stein and Paul B. Baltes The Development of Memory: Life-Span Perspectives Hayne W. Reese Cognitive Changes during the Adult Years: Implications for Developmental Theory and Research Nancy W. Denney and John C. Wright Social Cognition and Life-Span Approaches to the Study of Child Development Michael J. Chandler Life-Span Development of the Theory of Oneself: Implications for Child Development Orville G. Brim, Jr. Implication of Life-Span Developmental Psychology for Childhood Education Leo Montada and Sigrun-Heide Filipp AUTHOR INDEX—SUBJECT INDEX

Volume 12 Research between 1960 and 1970 on the Standing Height of Young Children in Different Parts of the World Howard V. Meredith The Representation of Children’s Knowledge David Klahr and Robert S. Siegler Chromatic Vision in Infancy Marc H. Bornstein

496

Contents of Previous Volumes

Developmental Memory Theories: Baldwin and Piaget Bruce M. Ross and Stephen M. Kerst Child Discipline and the Pursuit of Self: An Historical Interpretation Howard Gadlin Development of Time Concepts in Children William J. Friedman AUTHOR INDEX—SUBJECT INDEX

The Development of Understanding of the Spatial Terms Front and Back Lauren Julius Harris and Ellen A. Strommen The Organization and Control of Infant Sucking C. K. Crook Neurological Plasticity, Recovery from Brain Insult, and Child Development Ian St. James-Roberts

Volume 13

AUTHOR INDEX—SUBJECT INDEX

Coding of Spatial and Temporal Information in Episodic Memory Daniel B. Berch A Developmental Model of Human Learning Barry Gholson and Harry Beilin The Development of Discrimination Learning: A Levels-of-Functioning Explanation Tracy S. Kendler The Kendler Levels-of-Functioning Theory: Comments and an Alternative Schema Charles C. Spiker and Joan H. Cantor Commentary on Kendler’s Paper: An Alternative Perspective Barry Gholson and Therese Schuepfer Reply to Commentaries Tracy S. Kendler On the Development of Speech Perception: Mechanisms and Analogies Peter D. Eimas and Vivien C. Tartter The Economics of Infancy: A Review of Conjugate Reinforcement Carolyn Kent Rovee-Collier and Marcy J. Gekoski Human Facial Expressions in Response to Taste and Smell Stimulation Jacob E. Steiner AUTHOR INDEX—SUBJECT INDEX Volume 14 Development of Visual Memory in Infants John S. Werner and Marion Perlmutter Sibship-Constellation Effects on Psychosocial Development, Creativity, and Health Mazie Earle Wagner, Herman J. P. Schubert, and Daniel S. P. Schubert

Volume 15 Visual Development in Ontogenesis: Some Reevaluations Ju¨ri Allik and Jaan Valsiner Binocular Vision in Infants: A Review and a Theoretical Framework Richard N. Aslin and Susan T. Dumais Validating Theories of Intelligence Earl C. Butterfield, Dennis Siladi, and John M. Belmont Cognitive Differentiation and Developmental Learning William Fowler Children’s Clinical Syndromes and Generalized Expectations of Control Fred Rothbaum AUTHOR INDEX—SUBJECT INDEX

Volume 16 The History of the Boyd R. McCandless Young Scientist Awards: The First Recipients David S. Palermo Social Bases of Language Development: A Reassessment Elizabeth Bates, Inge Bretherton, Marjorie Beeghly-Smith, and Sandra McNew Perceptual Anisotropies in Infancy: Ontogenetic Origins and Implications of Inequalities in Spatial Vision Marc H. Bornstein Concept Development Martha J. Farah and Stephen M. Kosslyn

Contents of Previous Volumes Production and Perception of Facial Expressions in Infancy and Early Childhood Tiffany M. Field and Tedra A. Walden Individual Differences in Infant Sociability: Their Origins and Implications for Cognitive Development Michael E. Lamb The Development of Numerical Understandings Robert S. Siegler and Mitchell Robinson AUTHOR INDEX—SUBJECT INDEX

Volume 17 The Development of Problem-Solving Strategies Deanna Kuhn and Erin Phelps Information Processing and Cognitive Development Robert Kail and Jeffrey Bisanz Research between 1950 and 1980 on Urban–Rural Differences in Body Size and Growth Rate of Children and Youths Howard V. Meredith Word Meaning Acquisition in Young Children: A Review of Theory and Research Pamela Blewitt Language Play and Language Acquisition Stan A. Kuczaj II The Child Study Movement: Early Growth and Development of the Symbolized Child Alexander W. Siegel and Sheldon H. White AUTHOR INDEX—SUBJECT INDEX

Volume 18 The Development of Verbal Communicative Skills in Children Constance R. Schmidt and Scott G. Paris Auditory Feedback and Speech Development Gerald M. Siegel, Herbert L. Pick, Jr., and Sharon R. Garber Body Size of Infants and Children around the World in Relation to Socioeconomic Status Howard V. Meredith

497

Human Sexual Dimorphism: Its Cost and Beneﬁt James L. Mosley and Eileen A. Stan Symposium on Research Programs: Rational Alternatives to Kuhn’s Analysis of Scientiﬁc Progress—Introductory Remarks Hayne W. Reese, Chairman World Views and Their Inﬂuence on Psychological Theory and Research: Kuhn-Lakatos-Laudan Willis F. Overton The History of the Psychology of Learning as a Rational Process: Lakatos versus Kuhn Peter Barker and Barry Gholson Functionalist and Structuralist Research Programs in Developmental Psychology: Incommensurability or Synthesis? Harry Beilin In Defense of Kuhn: A Discussion of His Detractors David S. Palermo Comments on Beilin’s Epistemology and Palermo’s Defense of Kuhn Willis F. Overton From Kuhn to Lakatos to Laudan Peter Barker and Barry Gholson Overton’s and Palermo’s Relativism: One Step Forward, Two Steps Back Harry Beilin AUTHOR INDEX—SUBJECT INDEX

Volume 19 Response to Novelty: Continuity versus Discontinuity in the Developmental Course of Intelligence Cynthia A. Berg and Robert J. Sternberg Metaphoric Competence in Cognitive and Language Development Marc Marschark and Lynn Nall The Concept of Dimensions in Developmental Research Stuart I. Offenbach and Francine C. Blumberg Effects of the Knowledge Base on Children’s Memory Strategies Peter A. Ornstein and Mary J. Naus Effects of Sibling Spacing on Intelligence, Interfamilial Relations, Psychosocial

498

Contents of Previous Volumes

Characteristics, and Mental and Physical Health Mazie Earle Wagner, Herman J. P. Schubert, and Daniel S. P. Schubet Infant Visual Preferences: A Review and New Theoretical Treatment Martin S. Banks and Arthur P. Ginsburg AUTHOR INDEX—SUBJECT INDEX

Volume 20 Variation in Body Stockiness among and within Ethnic Groups at Ages from Birth to Adulthood Howard V. Meredith The Development of Conditional Reasoning: An Iffy Proposition David P. O’Brien Content Knowledge: Its Role, Representation, and Restructuring in Memory Development Michelene T. H. Chi and Stephen J. Ceci Descriptions: A Model of Nonstrategic Memory Development Brian P. Ackerman Reactivation of Infant Memory: Implications for Cognitive Development Carolyn Rovee-Collier and Harlene Hayne Gender Segregation in Childhood Eleanor E. Maccoby and Carol Nagy Jacklin Piaget, Attentional Capacity, and the Functional Implications of Formal Structure Michael Chapman INDEX

Volume 21 Social Development in Infancy: A 25-Year Perspective Ross D. Parke On the Uses of the Concept of Normality in Developmental Biology and Psychology Eugene S. Gollin, Gary Stahl, and Elyse Morgan Cognitive Psychology: Mentalistic or Behavioristic? Charles C. Spiker

Some Current Issues in Children’s Selective Attention Betty J. House Children’s Learning Revisited: The Contemporary Scope of the Modiﬁed Spence Discrimination Theory Joan H. Cantor and Charles C. Spiker Discrimination Learning Set in Children Hayne W. Reese A Developmental Analysis of Rule-Following Henry C. Riegler and Donald M. Baer Psychological Linguistics: Implications for a Theory of Initial Development and a Method for Research Sidney W. Bijou Psychic Conﬂict and Moral Development Gordon N. Cantor and David A. Parton Knowledge and the Child’s Developing Theory of the World David S. Palermo Childhood Events Recalled by Children and Adults David B. Pillemer and Sheldon H. White INDEX Volume 22 The Development of Representation in Young Children Judy S. DeLoache Children’s Understanding of Mental Phenomena David Estes, Henry M. Wellman, and Jacqueline D. Woolley Social Inﬂuences on Children’s Cognition: State of the Art and Future Directions Margarita Azmitia and Marion Perlmutter Understanding Maps as Symbols: The Development of Map Concepts Lynn S. Liben and Roger M. Downs The Development of Spatial Perspective Taking Nora Newcombe Developmental Studies of Alertness and Encoding Effects of Stimulus Repetition Daniel W. Smothergill and Alan G. Kraut Imitation in Infancy: A Critical Review Claire L. Poulson, Leila Regina de Paula Nunes, and Steven F. Warren AUTHOR INDEX—SUBJECT INDEX

Contents of Previous Volumes Volume 23 The Structure of Developmental Theory Willis F. Overton Questions a Satisfying Developmental Theory Would Answer: The Scope of a Complete Explanation of Development Phenomena Frank B. Murray The Development of World Views: Toward Future Synthesis? Ellin Kofsky Scholnick Metaphor, Recursive Systems, and Paradox in Science and Developmental Theory Willis F. Overton Children’s Iconic Realism: Object versus Property Realism Harry Beilin and Elise G. Pearlman The Role of Cognition in Understanding Gender Effects Carol Lynn Martin Development of Processing Speed in Childhood and Adolescence Robert Kail Contextualism and Developmental Psychology Hayne W. Reese Horizontality of Water Level: A Neo-Piagetian Developmental Review Juan Pascual-Leone and Sergio Morra AUTHOR INDEX—SUBJECT INDEX

Volume 24 Music and Speech Processing in the First Year of Life Sandra E. Trehub, Laurel J. Trainor, and Anna M. Unyk Effects of Feeding Method on Infant Temperament John Worobey The Development of Reading Linda S. Siegel Learning to Read: A Theoretical Synthesis John P. Rack, Charles Hulme, and Margaret J. Snowling Does Reading Make You Smarter? Literacy and the Development of Verbal Intelligence Keith E. Stanovich

499

Sex-of-Sibling Effects: Part I. Gender Role, Intelligence, Achievement, and Creativity Mazie Earle Wagner, Herman J. P. Schubert, and Daniel S. P. Schubert The Concept of Same Linda B. Smith Planning as Developmental Process Jacquelyn Baker-Sennett, Eugene Matusov, and Barbara Rogoff AUTHOR INDEX—SUBJECT INDEX

Volume 25 In Memoriam: Charles C. Spiker (1925–1993) Lewis P. Lipsitt Developmental Differences in Associative Memory: Strategy Use, Mental Effort, and Knowledge Access Interactions Daniel W. Kee A Unifying Framework for the Development of Children’s Activity Memory Hilary Horn Ratner and Mary Ann Foley Strategy Utilization Deﬁciencies in Children: When, Where, and Why Patricia H. Miller and Wendy L. Seier The Development of Children’s Ability to Use Spatial Representations Mark Blades and Christopher Spencer Fostering Metacognitive Development Linda Baker The HOME Inventory: Review and Reﬂections Robert H. Bradley Social Reasoning and the Varieties of Social Experiences in Cultural Contexts Elliot Turiel and Cecilia Wainryb Mechanisms in the Explanation of Developmental Change Harry Beilin AUTHOR INDEX—SUBJECT INDEX

Volume 26 Preparing to Read: The Foundations of Literacy Ellen Bialystok The Role of Schemata in Children’s Memory Denise Davidson

500

Contents of Previous Volumes

The Interaction of Knowledge, Aptitude, and Strategies in Children’s Memory Performance David F. Bjorklund and Wolfgang Schneider Analogical Reasoning and Cognitive Development Usha Goswami Sex-of-Sibling Effects: A Review Part II. Personality and Mental and Physical Health Mazie Earle Wagner, Herman J. P. Schubert, and Daniel S. P. Schubert Input and Learning Processes in First Language Acquisition Ernst L. Moerk

Fuzzy-Trace Theory: Dual Processes in Memory, Reasoning, and Cognitive Neuroscience C. J. Brainerd and V. F. Reyna Relational Frame Theory: A Post-Skinnerian Account of Human Language and Cognition Yvonne Barnes-Holmes, Steven C. Hayes, Dermot Barnes-Holmes, and Bryan Roche The Continuity of Depression across the Adolescent Transition Shelli Avenevoli and Laurence Steinberg The Time of Our Lives: Self-Continuity in Native and Non-Native Youth Michael J. Chandler AUTHOR INDEX—SUBJECT INDEX

AUTHOR INDEX—SUBJECT INDEX Volume 29 Volume 27 From Form to Meaning: A Role for Structural Alignment in the Acquisition of Language Cynthia Fisher The Role of Essentialism in Children’s Concepts Susan A. Gelman Infants’ Use of Prior Experiences with Objects in Object Segregation: Implications for Object Recognition in Infancy Amy Needham and Avani Modi Perseveration and Problem Solving in Infancy Andre´a Aguiar and Rene´e Baillargeon Temperament and Attachment: One Construct or Two? Sarah C. Mangelsdorf and Cynthia A. Frosch The Foundation of Piaget’s Theories: Mental and Physical Action Harry Beilin and Gary Fireman AUTHOR INDEX—SUBJECT INDEX

The Search for What is Fundamental in the Development of Working Memory Nelson Cowan, J. Scott Saults, and Emily M. Elliott Culture, Autonomy, and Personal Jurisdiction in Adolescent–Parent Relationships Judith G. Smetana Maternal Responsiveness and Early Language Acquisition Catherine S. Tamis-Lemonda and Marc H. Bornstein Schooling as Cultural Process: Working Together and Guidance by Children from Schools Differing in Collaborative Practices Eugene Matusov, Nancy Bell, and Barbara Rogoff Beyond Prototypes: Asymmetries in Infant Categorization and What They Teach Us about the Mechanisms Guiding Early Knowledge Acquisition Paul C. Quinn Peer Relations in the Transition to Adolescence Carollee Howes and Julie Wargo Aikins AUTHOR INDEX—SUBJECT INDEX

Volume 28 Volume 30 Variability in Children’s Reasoning Karl S. Rosengren and Gregory S. Braswell

Learning to Keep Balance Karen Adolph

Contents of Previous Volumes

501

Sexual Selection and Human Life History David C. Geary Developments in Early Recall Memory: Normative Trends and Individual Differences Patricia J. Bauer, Melissa M. Burch, and Erica E. Kleinknecht Intersensory Redundancy Guides Early Perceptual and Cognitive Development Lorraine E. Bahrick and Robert Lickliter Children’s Emotion-Related Regulation Nancy Eisenberg and Amanda Sheffield Morris Maternal Sensitivity and Attachment in Atypical Groups L. Beckwith, A. Rozga, and M. Sigman Inﬂuences of Friends and Friendships: Myths, Truths, and Research Recommendations Thomas J. Berndt and Lonna M. Murphy

A Bio-Social-Cognitive Approach to Understanding and Promoting the Outcomes of Children with Medical and Physical Disorders Daphne Blunt Bugental and David A. Beaulieu Expanding Our View of Context: The Bio-ecological Environment and Development Theodore D. Wachs Pathways to Early Literacy: The Complex Interplay of Child, Family, and Sociocultural Factors Megan M. McClelland, Maureen Kessenich, and Frederick J. Morrison

AUTHOR INDEX—SUBJECT INDEX

From the Innocent to the Intelligent Eye: The Early Development of Pictorial Competence Georgene L. Troseth, Sophia L. Pierroutsakos, and Judy S. DeLoache Bringing Culture into Relief: Cultural Contributions to the Development of Children’s Planning Skills Mary Gauvain A Dual-Process Model of Adolescent Development: Implications for Decision Making, Reasoning, and Identity Paul A. Klaczynski The High Price of Afﬂuence Suniya S. Luthar and Chris C. Sexton Attentional Inertia in Children’s Extended Looking at Television John E. Richards and Daniel R. Anderson Understanding Classroom Competence: The Role of Social-Motivational and Self-Processes Kathryn R. Wentzel Continuities and Discontinuities in Infants’ Representation of Objects and Events Rachel E. Keen and Neil E. Berthier The Mechanisms of Early Categorization and Induction: Smart or Dumb Infants? David H. Rakison and Erin R. Hahn

Volume 31 Beyond Point And Shoot: Children’s Developing Understanding of Photographs as Spatial and Expressive Representations Lynn S. Liben Probing the Adaptive Signiﬁcance of Children’s Behavior and Relationships in the School Context: A Child by Environment Perspective Gary W. Ladd The Role of Letter Names in the Acquisition of Literacy Rebecca Treiman and Brett Kessler Early Understandings of Emotion, Morality, and Self: Developing a Working Model Ross A. Thompson, Deborah J. Laible, and Lenna L. Ontai Working Memory in Infancy Kevin A. Pelphrey and J. Steven Reznick The Development of a Differentiated Sense of the Past and the Future William J. Friedman The Development of Cognitive Flexibility and Language Abilities Gedeon O. Dea´k

AUTHOR INDEX—SUBJECT INDEX

Volume 32

AUTHOR INDEX—SUBJECT INDEX

502

Contents of Previous Volumes

Volume 33 A Computational Model of Conscious and Unconscious Strategy Discovery Robert Siegler and Roberto Araya Out-of-School Settings as a Developmental Context for Children and Youth Deborah Lowe Vandell, Kim M. Pierce, and Kimberly Dadisman Mechanisms of Change in the Development of Mathematical Reasoning Martha W. Alibali A Social Identity Approach to Ethnic Differences in Family Relationships during Adolescence Andrew J. Fuligni and Lisa Flook What Develops in Language Development? LouAnn Gerken The Role of Children’s Competence Experiences in the Socialization Process: A Dynamic Process Framework for the Academic Arena Eva M. Pomerantz, Qian Wang, and Florrie Ng The Infant Origins of Intentional Understanding Amanda L. Woodward Analyzing Comorbidity Bruce F. Pennington, Erik Willcutt, and Soo Hyun Rhee Number Words and Number Concepts: The Interplay of Verbal and Nonverbal Quantiﬁcation in Early Childhood Kelly S. Mix, Catherine M. Sandhofer, and Arthur J. Baroody AUTHOR INDEX—SUBJECT INDEX

Volume 34 Mapping Sound to Meaning: Connections Between Learning About Sounds and Learning About Words Jenny R. Saffran and Katharine Graf Estes A Developmental Intergroup Theory of Social Stereotypes and Prejudice Rebecca S. Bigler and Lynn S. Liben

Income Poverty, Poverty Co-Factors, and the Adjustment of Children in Elementary School Brian P. Ackerman and Eleanor D. Brown I Thought She Knew That Would Hurt My Feelings: Developing Psychological Knowledge and Moral Thinking Cecilia Wainryb and Beverely A. Brehl Home Range and The Development of Children’s Way Finding Edward H. Cornel and C. Donald Heth The Development and Neural Bases of Facial Emotion Recognition Jukka M. Leppa¨nen and Charles A. Nelson Children’s Suggestibility: Characteristics and Mechanisms Stephen J. Ceci and Maggie Bruck The Emergence and Basis of Endogenous Attention in Infancy and Early Childhood John Colombo and Carol L. Cheatham The Probabilistic Epigenesis of Knowledge James A. Dixon and Elizabeth Kelley AUTHOR INDEX—SUBJECT INDEX

Volume 35 Evolved Probabilistic Cognitive Mechanisms: An Evolutionary Approach to Gene Environment Development Interactions David F. Bjorklund, Bruce J. Ellis, and Justin S. Rosenberg Development of Episodic and Autobiographical Memory: A Cognitive Neuroscience Perspective Nora S. Newcombe, Marianne E. Lloyd and Kristin R. Ratliff Advances in the Formulation of Emotional Security Theory: An Ethologically Based Perspective Patrick T. Davies and Melissa L. Sturge-Apple Processing Limitations and the Grammatical Proﬁle of Children with Speciﬁc Language Impairment Laurence B. Leonard

Contents of Previous Volumes Children’s Experiences and Judgments about Group Exclusion and Inclusion Melanie Killen, Stefanie Sinno, and Nancy Geyelin Margie Working Memory as the Interface between Processing and Retention: A Developmental Perspective John N. Towse, Graham J. Hitch, and Neil Horton Developmental Science and Education: The NICHD Study of Early Child Care

503

and Youth Development Findings from Elementary School Robert C. Pianta The Role of Morphology in Reading and Spelling Monique Se´ne´chal and Kyle Kearnan The Interactive Development of Social Smiling Daniel Messinger and Alan Fogel AUTHOR INDEX—SUBJECT INDEX