SAGE BENCHMARKS IN DISCOURSE STUDIES
DISCOURSE STUDIES VOLUME I
EDITED BY TEUN A . VAN DIJK
Introduction and editorial arrangement © Teun A. van Dijk 2007 First published 2 0 0 7 Reprinted 2008 Apart from any fair dealing for the purposes of research or private study, or criticism or review, as permitted under the Copyright, Designs and Patents Act, 1988, this publication m a y b e reproduced, stored or transmitted i n any form, or by any means, o n l y with the prior permission in writing of the publishers, or in the case of reprographic reproduction, in accordance with the terms of licences issued b y the Copyright Licensing Agency. Enquiries concerning reproduction outside those terms should be sent to the publishers. Every effort has b e e n m a d e to trace and acknowledge all the copyright owners of the material reprinted herein. However, if any copyright owners have not b e e n located and contacted at the time of publication, the publishers will b e pleased to make the necessary arrangements at the first opportunity. i . "
E
S A G E Publications Ltd | 1 Oliver's Yard I 5 5 City Road London EC1Y ISP
.
S A G E Publications Inc. 2 4 5 5 Teller R o a d Thousand Oaks, California 91320 S A G E Publications India Pvt Ltd B 1 / 1 1 M o h a n Cooperative Industrial Area Mathura R o a d N e w Delhi 110 0 4 4 S A G E Publications Asia-Pacific Pte Ltd 3 3 Pekin Street #02-01 Far East Square Singapore 048763 British Library Cataloguing i n Publication data A catalogue record for this b o o k is available from the British Library I S B N : 978-1-4129-3617-0 (set of five volumes) L i b r a r y of C o n g r e s s C o n t r o l N u m b e r : 2006936728 Typeset b y Star Compugraphics Private Limited, D e l h i Printed o n paper from sustainable resources Printed and b o u n d in Great Britain b y TJI Digital, Padstow, Cornwall
CONTENTS
VOLUME I
A p p e n d i x of sources Editor's Introduction: T h e Study of Discourse: A n Introduction
xi xix
1. Interviewing Persons with a Learning Disability: H o w Setting Lower Standards M a y Inflate Weil-Being Scores Charles Antaki 1 2. Discourse Studies a n d the Ideology of 'Liberalism' Robert de Beaugrande 21 3. T h e Discourse Structure of News Stories Allan Bell 58 4. Crossing Genders, Mixing Languages: T h e Linguistic Construction of Transgenderism in Tonga Niko Besnier 90 5. T h e Power a n d Politics of G e n r e Vijay K. Bhatia < 111 6. O n the Complexity of Discourse Complexity: A Multidimensional Analysis Douglas Biber 127 7. T h e Dialogic Unconscious: Psychoanalysis, Discursive Psychology a n d the Nature of Repression Michael Billig '. 158 8. Investigating Narrative Inequality: African Asylum Seekers' Stories in Belgium Jan Blommaert 182 9. "You Gotta K n o w H o w to Tell a Story": Telling, Tales, a n d Tellers in A m e r i c a n and Israeli Narrative Events at D i n n e r Shoshana Blum-Kulka 221 10. Dialogue a n d Confrontation in Venezuelan Political Interaction Adriana Bolivar 261 11. Personal Web Pages a n d the Semiotic Construction of Academic Identities Carmen Rosa Caldas-Coulthard 275 12. Role and Position of Scientific Voices: R e p o r t e d Speech in the M e d i a Helena Cabamiglia and Carmen Lopez Ferrero 295 13. Relativity a n d Its Discontents: Language, Gender, a n d Pragmatics Deborah Cameron 322 14. T h e Analysis of Discourse Flow Wallace Chafe 15. Missing Links in Mainstream C D A : Modules, Blends a n d the Critical Instinct Paul Chilton
334 349
CONTENTS
V O L U M E II
16. T h e Interpenetration of Communicative Contexts: Examples from Medical Encounters Aaron V. Cicourel 17. Answers a n d Evasions Steven E. dayman 18. Author Identification, Idiolect, a n d Linguistic Uniqueness Malcolm Coulthard 19. Bodies, Beaches a n d Burn-Times: 'Environmentalism' a n d Its Discursive Competitors Nikolas Coupland andJustine Coupland 2 0 . Figures of Speech: Figurative Expressions a n d the M a n a g e m e n t of Topic Transition in Conversation Paul Drew and Elizabeth Holt 21. Universal a n d Culture-Specific Properties of Greetings Alessandro Duranti 22. Moaning, Whinging a n d Laughing: T h e Subjective Side of Complaints Derek Edwards 2 3 . Trial Discourse a n d Judicial Decision-Making: Constraining the Boundaries of G e n d e r e d Identities Susan Ehrlich 2 4 . T h e Occasioning a n d Structure of Conversational Stories Susan M. Ervin-Tripp and Aylin Kiintay 2 5 . Discourse, Social T h e o r y , a n d Social Research: T h e Discourse of Welfare Reform Norman Fairclough 2 6 . C o h e r e n c e in Text vs. C o h e r e n c e in M i n d T. Givon 27. Action a n d E m b o d i m e n t within Situated H u m a n Interaction Charles Goodwin 2 8 . Constructing Inferences during Narrative Text C o m p r e h e n s i o n Arthur C. Graesser, Murray Singer and Tom Trabasso 29. Space: Irrealis Objects in Technology Policy a n d T h e i r Role in a New Political E c o n o m y Philip Graham
1 20 62 78 96 126 159 183 200 229 258 304 338 390
V O L U M E III 30. Medical Discourse: Sociohistorical Construction 31.
B.L. Gunnarsson
1
Tactical Uses of Stories: Participation Frameworks within Girls' a n d Boys' Disputes Marjorie Harness Goodwin
14
32. T h e Terms of Agreement: I n d e x i n g Epistemic Authority and Subordination in Talk-in-Interaction John Heritage and Geoffrey Raymond
46
3 3 . Politeness, Power a n d Provocation: H o w H u m o u r Functions in the Workplace Janet Holmes
76
CONTENTS
vii
34. Stance a n d Engagement: A M o d e l of Interaction in Academic Discourse Ken Hyland
102
3 5 . W h e n Talk isn't C h e a p : Language a n d Political E c o n o m y Judith T. Irvine
122
36. Political Discourse: T h e Language of Right and Left in G e r m a n y Siegfriedjager
147
37.
T h e Uses a n d Representations of Local Languages in Tourist Destinations: A View from British T V Holiday P r o g r a m m e s Adamjaworski, Crispin Thurlow, Sarah Lawson and Virpi Ylanne-McEwen
3 8 . Notes o n 'Latency' in Overlap O n s e t
GailJefferson
168 193
39. Sociolinguistic Resources, Individual Identities, and Public Speech Styles of Texas W o m e n Barbara Johnstone
219
40. T h e Role of Knowledge in Discourse C o m p r e h e n s i o n : A Construction-Integration M o d e l Walter Kintsch
238
41.
Heteronormativity in Action: Reproducing the Heterosexual Nuclear Family in After-hours Medical Calls Celia Kitzynger
279
42. Oral Genres of H u m o r : O n the Dialectic of Genre Knowledge a n d Creative Authoring Helga Kotthoff
306
4 3 . Against Arbitrariness: T h e Social Production of the Sign as a Foundational Issue in Critical Discourse Analysis Gunther Kress
337
44. Narrative Analysis: O r a l Versions of Personal Experience William Labov and Joshua WaletzJsy
359
4 5 . G e n d e r , Discourse a n d Semiotics: T h e Politics of P a r e n t h o o d Representations Michelle M. Lazflr 46. Reconstructing Topical Sensitivity: Aspects of Face-Work in Talks between Midwives a n d Expectant Mothers Per Linell and Margareta Bredmar
391
418
VOLUME IV
47.
T h e Body Literate: Discourse a n d Inscription in Early Literacy Training Allan Luke 48. You K n o w , It D e p e n d s Ronald Macaulay 49. Analysing Self-Constituting Discourses Dominique Maingueneau 50. Grace: T h e Logogenesis of F r e e d o m J.R. Martin 51. Division a n d Rejection: F r o m the Personification of the Gulf Conflict to the Demonization of S a d d a m Hussein Luisa Martin Rojo
1 23 42 67
96
viii
CONTENTS
52.
Multimodal Resources for Turn-Taking: Pointing a n d the E m e r g e n c e of Possible Next Speakers Lorenzo. Mondada 5 3 . T h e Political Function of Narrative in Organizations Dennis K. Mumby 54. Talking to Children in Western S a m o a Elinor Ochs
n •
55. Nominal and Temporal A n a p h o r a Barbara H. Partee 56. T h e Organization of Ideological Diversity in Discourse: M o d e r n a n d Neotraditional Visions of the Tongan State Susan U. Philips 57. Argumentation Studies a n d Discourse Analysis: T h e French Situation a n d Global Perspectives Christian Plantin 58. E x t r e m e Case Formulations: A W a y of Legitimizing Claims. Anita Pomerantz 5 9 . Discursive Social Psychology: F r o m Attitudes to , Evaluative Practices Jonathan Potter 60. Were You Ever in a Situation W h e r e You Were in Serious D a n g e r of Being Killed? Narrator-Listener Interaction in L a b o v and Waletzky's Narratives Uta M. Quasthoff 61. T h e Inference-Making Machine: Notes on Observability Harvey Sacks 62. Confirming Allusions: Toward a n Empirical Account of Action Emanuel A. Schegloff 6 3 . Narrative as Self-Portrait: Sociolinguistic Constructions of Identity Deborah Schiffrin
126 158 176 203 242 277 302 311
342 349 359 406
VOLUME V
64. T h e Discourses of Food in the World System: Toward a Nexus Analysis of a World Problem Ron Scollon 6 5 . A Discourse-Centered A p p r o a c h to Language a n d Culture Joel Sherzer 66. Linguistics as Metaphor: Analysing the Discursive Ontology of the Object of Linguistic Inquiry Shi-Xu 67. T h e Writing of Research Article Introductions John Swales and Hazem Najjar 68. T h e Relativity of Linguistic Strategies: Rethinking Power a n d Solidarity in G e n d e r a n d D o m i n a n c e Deborah Tannen 69. Talk a n d Institution: A Reconsideration of the "Asymmetry" of Doctor-Patient Interaction Paul ten Have
1 21 39 63 76 99
70. T h e Construction of Conflicting Accounts in 101
CONTENTS
71. 72. 73. 74. 75. 76. 77. 78. 79.
T h e Clause as a Locus of G r a m m a r a n d Interaction Sandra A. Thompson and Elizabeth Couper-Kuhlen T h e O J . Simpson Case as a n Exercise in Narrative Analysis Robin Tolmach Lakoff T h e Identity Work of Questioning in Intellectual Discussion Karen Tracy andJulie Naughton Discourse a n d the Denial of Racism Teun A. van Dijk Strategic M a n e u v e r i n g : Maintaining a Delicate Balance Frans H. van Eemeren and Peter Houtlosser T h e Representation of Social Actors Theo van Leeuwen Predicaments of Criticism JefVerschueren Positioning a n d Interpretative Repertoires: Conversation Analysis a n d Post-Structuralism in Dialogue Margaret Wetherell Turning the Tables: Antisemitic Discourse in Post-War Austria Ruth Wodak
14 16 18 20 23; 261 301 32; 35(
Appendix of Sources
Grateful acknowledgement is m a d e to the following sources for permission ta reproduce material for the book. 1. "Interviewing Persons with a L e a r n i n g Disability: H o w Setting Lower Standards M a y Inflate Weil-Being Scores," Charles Antaki Qualitative Health Research vol. 9, n o . 4 , 1 9 9 9 , p p . 4 3 7 - 4 5 4 . Published b y Sage Publications, Inc. Reprinted with permission.
2. "Discourse Studies a n d the Ideology of 'Liberalism'," Robert de Beaugrande Discourse Studies vol. 1, n o . 3 , 1 9 9 9 , p p . 2 5 9 - 2 9 5 . Published b y Sage Publications Ltd. Reprinted with permission.
3. "The Discourse Structure of News Stories," Allan Bell Allan Bell and Peter Garrett (eds), Approaches to Media Discourse, 1996, pp. 6 4 - 1 0 4 . Published by Blackwell Publishing. Reprinted with permission.
4. "Crossing G e n d e r s , M i x i n g Languages: T h e Linguistic Construction of Transgenderism in Tonga," Niko Besnier J a n e t H o l m e s a n d M i r i a m M e y e r h o f f (eds), Handbook of Language pp. 2 7 9 - 3 0 1 . Published b y Blackwell Publishing. Reprinted with permission.
and Gender, 2 0 0 3 ,
5. " T h e Power a n d Politics of Genre," Vijay K. Bhatia World Englishes v o l 16, n o . 3 , 1997, pp. 3 5 9 - 3 7 1 . Published b y Blackwell Publishing. Reprinted with permission.
6. " O n the C o m p l e x i t y of D i s c o u r s e C o m p l e x i t y : A M u l t i d i m e n s i o n a l Analysis," Douglas Biber Discourse Processes vol. 1 5 , 1 9 9 2 , pp. 133-163. Published b y Lawrence Erlbaum Associates. Reprinted with permission.
7. "The Dialogic Unconscious: Psychoanalysis, Discursive Psychology a n d the Nature of Repression," Michael Billig British Journal of'Social Psychologyvol. 3 6 , 1 9 9 7 , pp. 139-159. Published b y T h e British Psychological Society. Reprinted with permission.
8. "Investigating Narrative Inequality: African Asylum Seekers' Stories in Belgium," Jan Blommaert Discourse & Society vol. 12, no. 4 , 2 0 0 1 , pp. 4 1 3 - 4 4 9 . Published b y Sage Publications Ltd. Reprinted with permission.
APPENDIX OF SOURCES
xii
9. "'You G o t t a K n o w H o w to Tell a Story': Telling, Tales, a n d Tellers in A m e r i c a n a n d Israeli Narrative Events at Dinner," Shoshana Blum-Kulka Language in Society vol. 2 2 , 1 9 9 3 , pp. 3 6 1 - 4 0 2 . © Cambridge University Press reprinted with permission of the publisher.
10. "Dialogue and Confrontation in Venezuelan Political Interaction," Adriana Bolivar AILA Review vol. 1 8 , 2 0 0 5 , pp. 3-17. Reprinted with kind permission by J o h n Benjamins Publishing C o m p a n y , A m s t e r d a m / Philadelphia, www.benjamins.com, and the foundation 'Foundation of Language'.
11. "Personal W e b Pages a n d the Semiotic Construction of Academic Identities," Carmen Rosa Caldas-Coulthard C.R. Caldas-Coulthard and M. Toolan (eds), The Writer's Craft, The Culture's Technology, 2 0 0 5 , pp. 2 3 - 4 6 . Published by Rodopi. Reprinted with permission.
12. "Role a n d Position of Scientific Voices: R e p o r t e d Speech in the Media," Helena Calsamiglia and Carmen Lopez Ferrero Discourse Studies vol. 5, n o . 2, 2 0 0 3 , pp. 147-173. Published b y Sage Publications Ltd. Reprinted with permission.
13. "Relativity a n d Its Discontents: Language, G e n d e r , and Pragmatics," Deborah Cameron Intercultural Pragmatics vol. 2, n o . 3, 2 0 0 5 , p p . 3 2 1 - 3 3 4 . Published b y Walter de Gruyter. Reprinted with permission.
14. " T h e Analysis of Discourse Flow," Wallace Chafe D e b o r a h Schiffrin, D e b o r a h Tannen and Heidi E. Hamilton (eds), The Handbook of Discourse Analysis, 2 0 0 1 , pp. 673-687. Published b y Blackwell Publishing. Reprinted with permission.
15. "Missing Links in M a i n s t r e a m C D A : Modules, Blends a n d the Critical Instinct," Paul Chilton Ruth Wodak and Paul Chilton (eds), A New Agenda in (Critical) Discourse Analysis, 2 0 0 5 , pp. 19-51. Reprinted with kind permission by J o h n Benjamins Publishing C o m p a n y , A m s t e r d a m / Philadelphia, www.benjamins.com, and the foundation 'Foundation of Language'.
16. "The Interpenetration of Communicative Contexts: Examples from Medical Encounters," Aaron V. Cicourel Alessandro Duranti and Charles G o o d w i n (eds), Rethinking Context: Languages as an Interactive Phenomenon, 1992, p p . 2 9 1 - 3 1 0 . © Cambridge University Press reprinted with permission of the publisher.
17. "Answers and Evasions," Steven E.
dayman
Language in Society vol. 3 0 , n o . 3, 2001, pp. 4 0 3 - 4 4 2 . © Cambridge University Press reprinted with permission of the publisher.
1
18. "Author Identification, Idiolect, a n d Linguistic U n i q u e n e s s , " Malcolm Coulthard Applied Linguistics vol. 2 5 , n o . 4, 2 0 0 4 , pp. 431-447. Published b y permission of Oxford University Press. Reproduced with permission of Oxford University Press via Copyright Clearance Center's Rightslink Service.
APPENDIX O F SOURCES
xiii
19. "Bodies, Beaches a n d Burn-Times: 'Environmentalism' a n d Its Discursive Competitors," Nikolas Coupland andJustine Coupland Discourse & Society vol. 8, n o . 1, 1997, p p . 7 - 2 5 . Published b y Sage Publications Ltd. Reprinted with permission.
20. "Figures of Speech: Figurative Expressions a n d the M a n a g e m e n t of Topic Transition in Conversation," Paul Drew and Elizabeth Holt Language in Society vol. 27, n o . 4, 1998, pp. 4 9 5 - 5 2 2 . © Cambridge University Press reprinted with permission of the publisher.
21. "Universal and Culture-Specific Properties of Greetings," Alessandro Duranti Journal of'Linguistic Anthropologyvol.
7, n o . 1, 1997, pp. 63-97.
Copyright © 1997, American Anthropological Association. Reproduced with permission of American Anthropological Association via Copyright Clearance Center's Rightslink Service.
22. "Moaning, W h i n g i n g a n d Laughing: T h e Subjective Side of Complaints," Derek Edwards Discourse Studies vol. 7, n o . 1 , 2 0 0 5 , pp. 5 - 2 8 . Published b y Sage Publications Ltd. Reprinted with permission.
2 3 . "Trial Discourse a n dJudicial Decision-Making: Constraining the Boundaries of G e n d e r e d Identities," Susan Ehrlich Judith Baxter (ed.), Speaking Out: The Female Voice in Public Contexts, 2 0 0 6 , pp. 139-158. Reproduced with permission of Palgrave Macmillan.
24. "The Occasioning a n d Structure of Conversational Stories," Susan M. Ervin-Tripp andAylin Kiintay Talmy G i v o n (ed.), Cognitive Communicative and Social Perspectives, 1997, pp. 1 3 3 - 1 6 6 . Reprinted with kind permission b y J o h n Benjamins Publishing C o m p a n y , A m s t e r d a m / Philadelphia, www.benjamins.com, and the foundation 'Foundation of Language'.
2 5 . "Discourse, Social T h e o r y , a n d Social Research: T h e Discourse of Welfare Reform," Norman Fairclough Journal of Sociolinguistics vol. 4, n o . 2, 2 0 0 0 , pp. 1 6 3 - 1 9 5 . Published b y Blackwell Publishing. Reprinted with permission.
26. " C o h e r e n c e in Text vs. C o h e r e n c e in M i n d , " T. Givon Morton A n n Gernsbacher and T. G i v o n (eds), Coherence in Spontaneous Text, 1995, pp. 5 9 - 1 1 5 . Reprinted with kind permission b y J o h n Benjamins Publishing C o m p a n y , A m s t e r d a m / Philadelphia, www.benjamins.com, and the foundation 'Foundation of Language'.
27. "Action a n d E m b o d i m e n t within Situated H u m a n Interaction," Charles Goodwin Journal of Pragmatics vol. 3 2 , 2 0 0 0 , p p . 1 4 8 9 - 1 5 2 2 . Copyright (2000), reprinted with permission from Elsevier.
2 8 . "Constructing Inferences during Narrative Text C o m p r e h e n s i o n , " Arthur C. Graesser, Murray Singer and Tom Trabasso Psychological Review vol. 101, n o . 3 , 1 9 9 4 , pp. 3 7 1 - 3 9 5 . Copyright 1994 b y the American Psychological Association Inc. Reprinted with permission of American Psychological Association.
29. "Space: Irrealis Objects in Technology Policy a n d T h e i r Role in a N e w Political E c o n o m y , " Philip Graham Discourse & Society vol. 12, n o . 6 , 2 0 0 1 , pp. 761-788. Published b y Sage Publications Ltd. Reprinted with permission.
APPENDIX OF SOURCES
xiv
30. "Medical Discourse: Sociohistorical Construction," B.L. Gunnarsson Keith Brown (ed.), Encyclopedia of Language and Linguistics, vol. 7, Article 2 3 6 0 . 2 n d Edition. 2 0 0 6 , p p . 709-717. Copyright (2006), reprinted with permission from Elsevier.
31. "Tactical Uses of Stories: Participation Frameworks within Girls' a n d Boys' Disputes," Marjorie Harness Goodwin D e b o r a h Tannen (ed.), Gender and Conversational Interaction, 1993, pp. 110-143. Published b y permission of Oxford University Press, Inc.
3 2 . "The Terms of Agreement: Indexing Epistemic Authority and Subordination in Talk-in-Interaction," John Heritage and Geoffrey Raymond Social Psychology Quarterly vol. 68, n o . 1, 2 0 0 5 , pp. 15-38. Published b y American Sociological Association. Reprinted with permission.
3 3 . "Politeness, P o w e r a n d P r o v o c a t i o n : H o w H u m o u r F u n c t i o n s in t h e Workplace," Janet Holmes Discourse Studies vol. 2, n o . 2 , 2 0 0 0 , pp. 159-185. Published b y Sage Publications Ltd. Reprinted with permission.
34. "Stance a n d Engagement: A M o d e l of Interaction in Academic Discourse," Ken Hyland Discourse Studies vol. 7, n o . 2, 2 0 0 5 , pp. 173-192. Published by Sage Publications Ltd. Reprinted with permission.
3 5 . " W h e n Talk isn't C h e a p : Language a n d Political E c o n o m y , " Judith T. Irvine American Ethnologist vol. 16, n o . 2 , 1 9 8 9 , pp. 248-267. Copyright © 1989 b y the American Anthropological Association. Reproduced with permission of American Anthropological Association via Copyright Clearance Center's Rightslink Service.
36. "Political Discourse: T h e Language of Right a n d Left in G e r m a n y , " Siegfriedjager Patrick Stevenson (ed.), The German Language and the Real World: Sociolinguistic, Cultural, and Pragmatic Perspectives on Contemporary German, 1995, pp. 2 3 1 - 2 5 5 . Published b y permission of Oxford University Press, Inc.
37. " T h e Uses a n d Representations of Local Languages in Tourist Destinations: A View from British T V H o l i d a y P r o g r a m m e s , " Adam Jaworski, Crispin Thurlow, Sarah Lawson and Virpi Ylanne-McEwen Language Awareness-vol. 12, n o . 1, 2001, p p . 1-25. © 2 0 0 3 AJaworski et al. Published b y Multilingual Matters Ltd. Reprinted with permission.
3 8 . "Notes o n 'Latency' in Overlap Onset," Gailjefferson Human Studies vol. 9 , 1 9 8 6 , p p . 1 5 3 - 1 8 3 . © Martinus Nijhoff Publishers, Dordrecht. Reprinted with kind permission from Springer Science and Business Media.
39. "Sociolinguistic Resources, Individual Identities, a n d Public Speech Styles of Texas W o m e n , " Barbara Johnstone Journal of'Linguistic Anthropology vol. 5, n o . 2 , 1 9 9 5 , pp. 1 8 3 - 2 0 2 . Copyright © 1995, American Anthropological Association. Reproduced with permission of American Anthropological Association via Copyright Clearance Center's Rightslink Service.
APPENDIX O F SOURCES
xv
40. " T h e Role of Knowledge in Discourse C o m p r e h e n s i o n : A ConstructionIntegration M o d e l , " Walter Kintsch Psychological Review vol. 9 5 , n o . 2 , 1988, p p . 163-182. Copyright 1988 b y the American Psychological Association, Inc. Reprinted with permission of American Psychological Association.
41. " H e t e r o n o r m a t i v i t y i n Action: R e p r o d u c i n g t h e H e t e r o s e x u a l N u c l e a r Family i n After-hours Medical Calls," Celia Kit&nger Social Problems vol. 5 2 , n o . 4 , 2 0 0 5 , p p . 4 7 7 - 4 9 8 . Published b y University of California Press. Reproduced with permission of University of California Press via Copyright Clearance Center's Rightslink Service.
42. "Oral Genres of H u m o r : O n the Dialectic of Genre Knowledge a n d Creative Authoring," Helga Kotthoff Pragmatics (forthcoming, 2007). Published with permission o f the author.
4 3 . "Against Arbitrariness: T h e Social Production of the Sign as a Foundational Issue in Critical Discourse Analysis," Gunther Kress Discourse & Society vol. 4, n o . 2, 1993, p p . 169-191. Published b y Sage Publications Ltd. Reprinted with permission.
44. "Narrative Analysis: O r a l Versions of Personal Experience," WilliamLabov and Joshua Waletzky Journal of 'Narrative and Life History vol. 7, no. 1 - 4 , 1 9 9 7 , p p . 3 - 3 8 . Published b y Lawrence Erlbaum Associates.
4 5 . "Gender, Discourse a n d Semiotics: T h e Politics of P a r e n t h o o d Representations," Michelle M. Lazar Discourse & Society vol. 11, n o . 3 , 2 0 0 0 , p p . 3 7 3 - 4 0 0 . Published b y Sage Publications Ltd. Reprinted with permission.
46. "Reconstructing Topical Sensitivity: Aspects of Face-Work i n Talks b e t w e e n Midwives a n d Expectant Mothers," Per Linell and Margareta Bredmar Research on Language and Social Interaction vol. 2 9 , n o . 4 , 1 9 9 6 , p p . 3 4 7 - 3 7 9 . Copyright © 1996, Lawrence Erlbaum Associates, Inc. Published b y Lawrence Erlbaum Associates. Reprinted with permission.
47. " T h e B o d y Literate: Discourse a n d Inscription in Early Literacy Training," Allan Luke Linguistics and Education vol. 4 , n o . 1 , 1 9 9 2 , p p . 107-129. Copyright (1991), reprinted with permission from Elsevier.
48. "You K n o w , It D e p e n d s , " Ronald Macaulay Journal of Pragmatics vol. 3 4 , 2 0 0 2 , p p . 749-767. Copyright (2002), reprinted with permission from Elsevier.
49. "Analysing Self-Constituting Discourses," Dominique Maingueneau Discourse Studies vol. 1, n o . 2, 1999, p p . 175-199. Published b y Sage Publications Ltd. Reprinted with permission.
50. "Grace: T h e Logogenesis of Freedom," J.R. Martin Discourse Studies vol. 1, n o . 1 , 1 9 9 9 , p p . 2 9 - 5 6 . Published b y Sage Publications Ltd. Reprinted with permission.
xvi
APPENDIX O F SOURCES
51. "Division a n d Rejection: F r o m the Personification of the Gulf Conflict to the Demonization of S a d d a m Hussein," Luisa Martin Rojo Discourse & Society vol. 6, n o . 1, 1995, pp. 4 9 - 8 0 . Published b y Sage Publications Ltd. Reprinted with permission.
52. "Multimodal Resources for Turn-Taking: Pointing a n d the E m e r g e n c e of Possible N e x t Speakers," Lorenza Mondada (Translated in English) Cahiers de Linguistique Francaise vol. 2 6 , 2 0 0 4 , pp. 2 6 9 - 2 9 2 . Published b y Sage Publications Ltd. Reprinted with permission.
5 3 . " T h e Political Function of Narrative in Organizations," Dennis K. Mumby Communication Monographs vol. 5 4 , 1 9 8 7 , pp. 113-127. Reproduced b y permission of Taylor & Francis Books U K .
54. "Talking to Children in Western Samoa," Elinor Ochs Language in Society vol. 1 1 , 1 9 8 2 , pp. 7 7 - 1 0 4 . © Cambridge University Press reprinted with permission of the publisher.
,
55. "Nominal a n d Temporal A n a p h o r a " , Barbara H. Partee Linguistics and Philosophy vol. 7, 1984, pp. 2 4 3 - 2 8 6 . Published by Springer Science and Business Media.
56. " T h e Organization of Ideological Diversity i n Discourse: M o d e r n a n d Neotraditional Visions of the Tongan State," Susan U. Philips American Ethnologist vol. 31, n o . 2, 2 0 0 4 , p p . 2 3 1 - 2 5 0 . © 2 0 0 4 American Anthropological Association. Reproduced with permission of American Anthropological Association via Copyright Clearance Center's Rightslink Service.
57. "Argumentation Studies and Discourse Analysis: T h e French Situation and Global Perspectives," Christian Plantin Discourse Studies vol. 4, n o . 3, 2 0 0 2 , pp. 3 4 3 - 3 6 8 . Published by Sage Publications Ltd. Reprinted with permission.
5 8 . "Extreme Case Formulations: A Way of Legitimizing Claims," Anita Pomerantz Human Studies vol. 9, n o . 2 / 3 , 1 9 8 6 , pp. 2 1 9 - 2 2 9 . © Martinus Nijhoff Publishers, Dordrecht. Reprinted with kind permission from Springer Science and Business Media.
59. "Discursive Social Psychology: F r o m Attitudes to Evaluative Practices," Jonathan Potter European Review of Social Psychology vol. 9, 1998, pp. 2 3 3 - 2 6 6 . © 1998, J o h n Wiley & Sons, Ltd. Reproduced b y permission o f j o h n Wiley & Sons l i m i t e d .
60. "Were You Ever in a Situation W h e r e You Were in Serious D a n g e r of Being Killed? Narrator-Listener Interaction in L a b o v a n d Waletzky's Narratives," UtaM. Quasthoff Journal of Narrative and Life History vol. 7 n o . 1-4, 1997, pp. 121-128. Copyright © 1997, Lawrence Erlbaum Associates, Inc. Published by Lawrence Erlbaum Associates.
61. " T h e Inference-Making M a c h i n e : Notes o n Observability," Harvey Sacks Human Studies vol. 12, n o . 3 - 4 , 1989, pp. 1 3 - 2 3 . Reprinted with kind permission from Springer Science and Business Media.
APPENDIX OF SOURCES
xvii
62. "Confirming Allusions: Toward an Empirical Account of Action," Emanuel A. Schegloff American Journal ofSociology vol. 102, n o . 1,1996, p p . 161-216. ti 1996 by T h e University of Chicago. Reprinted with permission.
63. "Narrative as Self-Portrait: Sociolinguistic Constructions of Identity," Deborah Schiffiin Language in Society vol. 2 5 , n o . 2, 1996, p p . 1 6 7 - 2 0 3 . © Cambridge University Press reprinted with permission of the publisher.
64. "The Discourses of Food in the World System: Toward a N e x u s Analysis of a World Problem," Ron Scollon Journal of Language and Politics vol. 4, n o . 3, 2 0 0 5 , pp. 4 6 7 - 4 8 9 . Reprinted with kind permission b y J o h n Benjamins Publishing C o m p a n y , A m s t e r d a m / Philadelphia, www.benjamins.com, and the foundation 'Foundation of Language'.
65. "A Discourse-Centered A p p r o a c h to Language a n d Culture," Joel Sherzer American Anthropologist vol. 8 9 , n o . 2, 1987, p p . 2 9 5 - 3 0 9 . Copyright © 1987, American Anthropological Association. Reproduced with permission of American Anthropological Association via Copyright Clearance Center's Rightslink Service.
66. "Linguistics as M e t a p h o r : Analysing t h e Discursive O n t o l o g y of the Object of Linguistic Inquiry," Shi-Xu Language Sciences vol 2 2 , 2 0 0 0 , pp. 4 2 3 - 4 4 6 . Copyright (2000), reprinted with permission from Elsevier.
67. "The Writing of Research Article Introductions,"./e/bz Swales and Hazem Najjar Written Communication vol. 4, n o . 2 , 1 9 8 7 , pp. 175-190. Published b y Sage Publications, Inc. Reprinted with permission.
68. "The Relativity of Linguistic Strategies: Rethinking Power a n d Solidarity in Gender a n d D o m i n a n c e , " Deborah Tannen Deborah Tannen, Gender and Discourse, 1994, pp. 1 9 - 5 2 . Published b y permission of Oxford University Press, Inc.
69. "Talk a n d Institution: A Reconsideration of t h e 'Asymmetry' of D o c t o r Patient Interaction," Paul ten Have Deirdre B o d e n a n d D o n a l d H . Z i m m e r m a n (eds), Talk and Social Structure: Ethnomethodology and Conversation Analysis, 1991, pp. 138-163. Published b y Polity. Reprinted with permission.
Studies
in
70. " T h e Construction of Conflicting Accounts in Public Participation T V , " Joanna Thornborrow Language in Society vol. 2 9 , n o . 3 , 2 0 0 0 , pp. 357-377. © Cambridge University Press reprinted with permission of the publisher.
71. "The Clause as a Locus of G r a m m a r a n d Interaction," Sandra A. Thompson and Elizflbeth Couper-Kuhlen Discourse Studies vol. 7, n o . 4 - 5 , 2 0 0 5 , pp. 4 8 1 - 5 0 5 . Published b y Sage Publications Ltd. Reprinted with permission.
72. "The O J . Simpson Case as an Exercise in Narrative Analysis," Robin Tohnach Lakoff Discourse Processes vol. 2 3 , n o . 3 , 1 9 9 7 , pp. 5 4 7 - 5 6 6 . Published by A l b e x Publication Corporation.
APPENDIX OF SOURCES
xviii
73. " T h e Identity Work of Questioning in Intellectual Discussion," Karen Tracy andJulie Naughton Communication Monographs \o\. 61, 1994, p p . 2 8 1 - 3 0 2 . Reprinted b y permission of Taylor & Francis Ltd, http://www.tandf.co.uk/journals
74. "Discourse a n d t h e Denial of Racism," Teun A. van Dijk Discourse & Society vol. 3, n o . 1 , 1 9 9 2 , p p . 87-118. Published b y Sage Publications Ltd. Reprinted with permission.
75. "Strategic Maneuvering: Maintaining a Delicate Balance," Frans H. van Eemeren and Peter Houtlosser Frans H . v a n E e m e r e n and Peter Houtlosser, Dialectic and Rhetoric: The Warp and Woof of Argumentation Analysis, 2 0 0 2 , p p . 131-159. Reprinted with kind permission from Springer Science and Business M e d i a .
76. " T h e Representation of Social Actors," Theo van Leeuwen C a r m e n R o s a Caldas-Coulthard and Malcolm Coulthard (eds), Texts andPractices-R*»din$in Critical Discourse Analysis, 1996, p p . 3 2 - 7 0 . Reproduced b y permission of Taylor & Francis B o o k s U K .
77. "Predicaments of Criticism," JefVerschueren Critique of Anthropology vol. 21, n o . 1, 2001, pp. 5 9 - 8 1 . Published b y Sage Publications Ltd. Reprinted with permission.
78. "Positioning a n d Interpretative Repertoires: Conversation Analysis a n d Post-Structuralism in Dialogue," Margaret Wetherell Discourse & Society vol. 9, n o . 3 , 1 9 9 8 , p p . 3 8 7 - 4 1 2 . Published b y Sage Publications Ltd. Reprinted with permission.
79. "Turning the Tables: Antisemitic Discourse in Post-War Austria,'' Ruth Wodak Discourse & Society voL 2 , n o . 1,1991, p p . 6 5 - 8 3 . Published b y Sage Publications Ltd. Reprinted with permission.
Editor's Introduction: The Study of Discourse: An Introduction T h e E m e r g e n c e of a N e w Cross-Discipline
n t h e m i d - 1 9 6 0 s , t h e h u m a n i t i e s a n d t h e social sciences witnessed a remarkably synchronous p a r a d i g m shift with the birth of several n e w b u t JL mutually related 'interdisciplines' such as semiotics, psycholinguistics, sociolinguistics, pragmatics as well as the study of discourse. T h e articles collected in the volumes of this b o o k are prominent contributions to the latter field of study, traditionally called Discourse Analysis, b u t m o r e appro priately labeled Discourse Studies, because it is not limited to a m e t h o d of analysis, as is the case for Content Analysis in the social sciences, b u t also has important theoretical objectives. A n d because the study of discourse manifests itself in virtually all disciplines in the humanities a n d social sciences, it is appropriate to speak of a n e w 'cross-discipline' or even a 'trans-discipline'. T h e story of this n e w field of study has b e e n told m a n y times, a n d is as excit ing as the contributions it has m a d e to our insights into language use, c o m m u n i cation a n d social interaction. For some of the m o t h e r a n d sister disciplines, this story of the emergence of the study of discourse m a y b e summarized as follows. Anthropology Already in the early 1960s, a m o n g the first to recognize the relevance of the study of discourse anthropologists such as Dell H y m e s (1972) b e c a m e interested in the ethnographic study of communicative events (beyond the traditional study of myths and folklore), a direction of research followed b y m a n y other anthro pologists u n d e r the label of the 'ethnography of speaking' (or the 'ethnography of communication'; B a u m a n & Sherzer, 1974; Saville-Troike, 1982) a n d t h e n more broadly within linguistic anthropology (Duranti, 2001). 1
Linguistics Linguists w e r e n o t lagging far b e h i n d during the late 1960s, w h e n some of t h e m realized that the use of language obviously was not r e d u c e d to the structures of isolated, abstract, invented sentences - as was the case in structural and generative grammars - b u t n e e d e d analyses of structures ' b e v o n d the sentence' a n d n f
EDITOR'S I N T R O D U C T I O N
XX
whole 'texts', for instance to account for a n a p h o r a and coherence. W h e r e a s initially still largely within the formal p a r a d i g m of 'text g r a m m a r s ' , also this linguistic approach soon m e r g e d with the other approaches to a m o r e empirical analysis of actual language use. T h e names associated with these early attempts at text a n d discourse g r a m m a r s are J a n o s Petofi (1971), Wolfgang Dressier (1972), and Teun A. v a n Dijk (1972, 1977), in E u r o p e , a n d J o s e p h Grimes (1975), Tom Givon (1979), Sandra T h o m p s o n a n d Bill M a n n in the U S A , the latter two u n d e r the label of Rhetorical Structure Theory (Mann & Thompson, 1988). T h e roots of the E u r o p e a n text grammars, apart from the obvious influence of N o a m Chomsky's generative grammar, are however diverse a n d range from literary theory and semiotics to Russian formalism a n d Czech and French structuralism. Although not u n d e r the label of 'text g r a m m a r ' , also early studies in Func tional Systemic G r a m m a r , founded b y Michael Halliday in the U K (and then Australia), p a i d m u c h attention to discourse, for instance in the account of 'cohesion', the grammatical expression of semantic coherence (Halliday & Hasan, 1976). This work was later followed b y a large n u m b e r of other studies o n the grammatical a n d semiotic aspects of discourse in the same SF-paradigm (among m a n y other studies, see, e.g., Martin, 1992). Formal Grammar O n the other h a n d , increasingly formal a n d explicit studies of language use for discourse participants, coreference, deictic expressions a n d tenses, continued to b e engaged in, from the 1970s b o t h b y logicians and philosophers, such as H a n s K a m p (1981) a n d his Discourse Representation Theory a n d others influenced b y the mathematician a n d formal philosopher Richard Montague. This approach states that discourse semantics is dynamic a n d d e p e n d s o n context. Pragmatics Within the tradition of British analytical philosophy, the 1960s also saw the very successful birth of another n e w interdiscipline, n a m e l y pragmatics. Based o n the work of Austin (1962) o n How to Do Things with Words, it is especially the study of J o h n Searle (1969) o n speech acts a n d an influential essay of H . P. Grice (1975) o n conversational m a x i m s that sparked a flow of studies o n language use extend ing the traditional focus o n syntax a n d semantics with a pragmatic component, accounting for the illocutive functions of language in terms of speech acts, implicatures a n d other aspects of contextually based language use. M o r e gener ally, Pragmatics has b e c o m e the discipline that houses m a n y of the studies of language use b e y o n d g r a m m a r , such as the influential work o n politeness b y Penelope Brown a n d Stephen Levinson (Brown & Levinson, 1987). Semiotics Within the study of literature a n d the arts, the mid-1960s also witnessed the emergence of semiotics, the general study of signs and symbol systems. Originally
xxi
EDITOR'S I N T R O D U C T I O N
based on the work of philosopher Charles Sanders Peirce, a n d structural linguists such as Louis Hjelmslev a n d A n d r e Martinet, this n e w discipline b e c a m e popular especially due to the work of U m b e r t o Eco (1976) in Italy, a n d R o l a n d Barthes (1964) and m a n y others in France. Semiotics was n o t limited to language, stories and other forms of discourse, b u t was also studied in other semiotic codes, such as images, film, d a n c e a n d architecture. Within discourse studies, semiotics has especially b e e n propagated, in a rather different paradigm, b y the work of Gunther Kress, a n d T h e o v a n L e e u w e n (Kress & v a n Leeuwen, 1990; v a n Leeuwen, 2005). Conversation Analysis
1
In sociology, the interest in discourse e m e r g e d within the b r o a d e r framework of 'ethnomethodology', a direction in microsociology focusing o n the ways people understand and m a n a g e their everyday life. U n d e r the influence of H a r o l d Garfinkel (1967), o n the one h a n d , a n d of Erving Gofiman (1959, 1961), o n the other h a n d , this interest in m u n d a n e interaction b e c a m e very popular with the study of conversation, p i o n e e r e d b y H a r v e y Sacks, M a n n y Schegloff, Gail Jefferson in a very influential article in Language (Sacks, Schegloff & Jefferson, 1974), followed b y m a n y other studies in several disciplines. W h e r e a s discourse grammars studied sequences of sentences, Conversation Analysis (CA) closely analyzed interactional sequences a n d p h e n o m e n a such as turn taking, inter ruptions, pauses, laughter, o p e n i n g a n d closing conversations, a n d m a n y other properties and strategic m o v e s of spontaneous talk n o w b e i n g accessible d u e to meticulous transcriptions of audio a n d video recordings (the influential collections by Atkinson & Heritage, 1984; a n d D r e w & Heritage, 1992). Sociolinguistics At the e n d of the 1960s, a p p e a r e d another n e w discipline at the boundaries of linguistics a n d the social sciences, Sociolinguistics. Although initially studying variation of g r a m m a r , especially pronunciation, due to variables as class, age or gender, some of these studies, also b y the founders of sociolinguistics, Bill L a b o v (1972a, 1972b) a n d Susan Ervin Tripp (1972), focused o n naturally occurring discourse, such as child discourse, storytelling about everyday experiences or the verbal play b y African-American adolescents (Gumperz & H y m e s , 1972). From a different perspective, later work in 'interactional sociolinguistics' provided more insight into details of interaction a n d their relation to the social context (Gumperz, 1982a, 1982b). The Psychology of Text Processing and Artificial Intelligence A few years later, at the beginning of the 1970s, also cognitive psychology (such as the work of Walter Kintsch, 1974) went b e y o n d the self-imposed limitations of the study of the mental processing of words a n d isolated sentences, a n d b e g a n to study the production, comprehension a n d m e m n r v o f r l i s m n r c o ir, ™ ~ n i n
xxii
EDITOR'S I N T R O D U C T I O N
and of stories in particular. T h u s , it could be shown that the notion of macrostructure, first developed in text g r a m m a r (van Dijk, 1972,1977,1980), also h a d a cognitive basis, for instance in the production a n d comprehension of discourse topics (van Dijk & Kintsch, 1983). This direction of research soon b e c a m e very popular in cognitive psychology, also because of its m a n y obvious applications, for instance in education a n d the mass media. O n e of the m a n y influential notions introduced in this research is that of a mental model - a representation of events and situations in 'episodic m e m o r y ' (the record of all our personal experiences) - as the basis of all discourse produc tion a n d understanding (Johnson-Laird, 1983; v a n Dijk & Kintsch, 1983). A n o t h e r important contribution c a m e from the closely related field of Arti ficial Intelligence (AI), n a m e l y the fundamental role of knowledge in discourse processing, for instance in the form of mental 'scripts' of prototypical episodes (Schank & Abelson, 1977). Although m u c h of this work was (and is) carried out in the various domains of cognitive science, it also has h a d m u c h influence in linguistics, sociolinguistics, psycholinguistics and the study of literature, such as studies on the comprehension of radio messages (Lutz & Wodak, 1987). Methodological Common Ground We see that m o r e or less at the same time, n a m e l y between the mid-1960s and the early-1970s, we witness closely related n e w disciplines emerging in the humanities a n d the social sciences. Despite their backgrounds in different mother disciplines, a n d despite a large diversity of m e t h o d s a n d objects of study, these n e w disciplines of semiotics, pragmatics, psycho- a n d sociolinguistics, ethnog r a p h y of speaking as well as conversation analysis a n d discourse studies h a d several things in c o m m o n . We m a y summarize this methodological c o m m o n ground as follows: • Interest in properties of 'naturally occurring' language use b y real lan guage users, instead of a study of abstract language systems a n d invented examples. • A study of larger units t h a n isolated words a n d sentences, and n e w basic units of analysis: texts, discourses, conversations or communicative events. • Extension of linguistics b e y o n d g r a m m a r towards a study of action and interaction. • Extension to non-verbal (semiotic) aspects of interaction a n d communi cation: gestures, images, film a n d multimedia. • Focus o n dynamic cognitive or interactional m o v e s a n d strategies. • Study of the role of the social, cultural a n d cognitive contexts of language use. • Analysis of a vast n u m b e r of hitherto largely ignored p h e n o m e n a of lan guage use: coherence, anaphora, topics, macrostructures, speech acts, inter actions, turn-taking, signs, politeness, mental models, a n d m a n y other asnects of discourse.
xxiii
EDITOR'S I N T R O D U C T I O N
Historical and Social Backgrounds It might b e speculative to relate this renewal in the humanities and social sciences to the m o r e general revolutionary spirit of the end of the 1960s, as we also k n o w it from the Civil Rights M o v e m e n t in the U S A , the feminist m o v e m e n t world wide, as well as the student m o v e m e n t s b o t h in the U S A a n d Europe. Yet, it is probably n o coincidence that against the b r o a d e r background of AfricanAmericans rebelling against racism, of w o m e n opposing patriarchy, a n d students protesting against traditional university hierarchies, the e n d of the 1960s also saw the emergence of n e w scholarly paradigms. Not seldom, these w e r e also defined b y y o u n g scholars rejecting the theories a n d m e t h o d s of their teachers and of the traditional orientations in literature, linguistics, sociology, anthro pology and psychology. Note though, that the social a n d political nature of the changes in society did not (yet) lead to similar changes in the disciplines. Indeed, initially m a n y of the changes of the n e w disciplines w e r e formal a n d only defined b y a broader object of study, rather that inspired b y a revolutionary spirit aiming to change society. There are (generally European) exceptions to this general 'apolitical' beginning of discourse analysis a n d sociolinguistics, especially in G e r m a n y a n d Austria, such as the w o r k b y Dittmar (1976) o n sociolinguistics, b y Leodolter (=Wodak) (1975) on language in the courtroom, W o d a k (1986) o n therapy groups a n d Wodak & Schulz (1986), o n the Language of love and guilt - discourses between mothers and daughters. Although in the social sciences especially frequent reference was m a d e to the 'Critical T h e o r y ' of the Frankfurt School, the n e w studies of signs, speech acts, language variation, stories, communicative events, conversation, film, text processing or discourse structures seldom took place within a b r o a d e r socio political m o v e m e n t of dissidence a n d opposition against social inequality. References to Adorno a n d Benjamin, for instance, initially were found in literature or in the philosophy of the social sciences, rather than in the n e w disciplines of language use a n d discourse. M o r e frequent references to the work of H a b e r m a s had to wait until H a b e r m a s himself h a d discovered pragmatic theory (Habermas, 1981). In that respect, at least in these disciplines, academic theory a n d social movement r e m a i n e d two distinct areas of social practice. It is only with such later developments as critical discourse studies that these two different forms of dissent merged.
I n t e g r a t i o n a n d Further D e v e l o p m e n t s Although most of the n e w (inter- or cross) disciplines m e n t i o n e d above h a d dif ferent backgrounds, objects, m e t h o d s a n d sometimes inconsistent philosophies, the last decades h a v e witnessed increasing integration in the b r o a d field of the study of language use. I n the following decades, m u c h work in pragmatics, socio linguistics, conversation analysis a n d the ethnography of speaking often over lapped, especially because of their shared interest in the study of conversational interaction in social contexts.
EDITOR'SINTRODUCTION
xxiv
Discursive Psychology These interdisciplines were soon j o i n e d b y other directions of research a n d new paradigms i n the 1980s, such as Discursive Psychology within social psychology, interested especially in the way 'psychological' notions (such as 'memory') are e n a c t e d i n discourse i n g e n e r a l , a n d c o n v e r s a t i o n in particular (Potter & Wetherell, 1987). Rejecting individual cognitive psychology, laboratory ex periments a n d traditional social psychological notions such as attitudes, Discursive Psychology, as pioneered b y Michael Billig (1987, 1988), J o n a t h a n Potter (1996), a n d D e r e k Edwards (1997), thus sought alignment with ethnomethodology and conversation analysis, o n the o n e h a n d , a n d the work of (the later) Wittgenstein, a n d the social constructivism of R o m H a r r e a n d K e n n e t h Gergen, o n the other h a n d . Interaction and Cognition W h e r e a s earlier studies in literature, semiotics a n d text g r a m m a r largely fo cused o n the analysis of written texts, nearly all w o r k i n the interdisciplines just mentioned focused o n spoken language in general, a n d o n 'talk in interaction' in particular. O n l y in cognitive science a n d some directions of discourse g r a m m a r , prag matics, sociology a n d anthropology do we find a m o r e cognitive orientation towards the study of language use a n d discourse - an orientation o n the 'mind' generally rejected or ignored b y interactional approaches to discourse (see the contribution in v a n Dijk, 2006). These developments h a v e their counterpart in cognitive linguistics, for instance in the work o n m e a n i n g a n d m e t a p h o r b y George Lakoff (Lakoff & J o h n s o n , 1980) - work that also has found application in (critical) discourse studies. T h e r e should b e litde d o u b t that before long also this b r o a d gap between interaction a n d cognition will b e bridged, especially since discourse obviously is b o t h a form of social interaction, as well as a form cognition of communication, and because there are m a n y 'mental' notions that are crucial in any kind of discourse study, such as meaning, coherence, topics, inference, presupposition, knowledge, belief, opinion, a n d so on. Indeed, the v e r y notions of action and interaction cannot properly b e defined without 'cognitive' notions such as plan, aim, goal, purpose, coordination, monitoring, a n d so on. It is within the study of discourse that such an integration of cognitive and interactional approaches is most fruitful. T h e last decades h a v e seen extraordinary advances in b o t h the study of interaction a n d that of cognition, a n d the time has c o m e to integrate these results. Insight into the detailed mental strategies (and their neurological basis) of discourse production a n d comprehension m a y thus b e c o m b i n e d with what we n o w k n o w o n the details of interactional m o v e s a n d strategies. Critical Discourse Studies At the e n d of the 1970s, another direction of research emerged in the study of discourse: Critical Discourse Analysis (CDA), originally introduced in a seminal
EDITOR'S I N T R O D U C T I O N
book by Roger Fowler, G u n t h e r Kress, B o b H o d g e a n d Tony Trew, Language and Control (1979), and later developed b y N o r m a n Fairclough (1989) in the U K , Ruth Wodak (1989) in Austria a n d Teun A. v a n Dijk (1993) in the Netherlands (for introduction, see, e.g., W o d a k & Meyer, 2001). Although s o m e isolated scholars earlier showed some interest in a m o r e critical approach (e.g., Mey, 1985), we have seen that most of the n e w studies of language use emerging in the 1960s were at first quite apolitical. Rather late w h e n considered against the revolutionary background of the e n d of the 1960s referred to above, Critical Discourse Analysis finally b e g a n to focus on issues of power, domination, and social inequality, a n d o n the relevance of gender, race a n d class in the study of text and talk. Much of this critical w o r k was paralleled b y similar w o r k d o n e mostly b y feminist w o m e n o n the relations between language, discourse a n d gender, and on the ways male domination is r e p r o d u c e d in text a n d talk (for an early study, see Kramarae, 1980; for the vast n u m b e r of later studies, see the references in Holmes & Meyerhoff, 2003). Although at first n o t explicitly carried out within a CDA-framework, m u c h work o n g e n d e r a n d discourse is in fact an excellent example of a C D A a p p r o a c h (for a n explicit C D A approach to gender studies see, e.g., Lazar, 2 0 0 5 ; Wodak, 1997). Other studies, such as b y W o d a k a n d v a n Dijk, especially focused o n the reproduction of racism a n d anti-Semitism i n discourse, for instance in political discourse, the press, a n d textbooks (among m a n y studies, see, e.g., v a n Dijk, 1993; Wodak, et al., 1990; W o d a k & van Dijk, 2000). In the same w a y as the study of discourse spread in most of the humanities and social sciences, also the m o r e critical approaches to language, discourse and interaction can n o w b e found in m a n y disciplines. T h e awareness has grown that with the increasingly powerful m e t h o d s of the explicit a n d systematic description of talk a n d text, and the m o r e sophisticated theories of cognition a n d interaction in their social and institutional contexts, discourse a n d conversation analysts are well p r e p a r e d to tackle m o r e complex and socially relevant issues. Though, still reluctantly b y the m o r e formally inclined scholars interested in m o r e 'autonomous' approaches to structures of text and interaction, it has become m o r e widely accepted that discourse is profoundly e m b e d d e d in society and culture, and hence, closely related also to all forms of power, p o w e r abuse and social inequality. It has b e e n shown h o w ethnic prejudices a n d ideologies are daily produced a n d r e p r o d u c e d b y political a n d m e d i a discourse of the elites, and thus contribute to the reproduction of racism. T h e same is true for everyday sexism in m a n y kinds of discourse - from conversation to advertising - and the reproduction of the system of m a l e domination. Obviously, such a critical focus o n the discursive reproduction of social inequality cannot b e limited to a n a r r o w discourse or conversation analytical approach, and requires further integration with the social sciences. Such an integration will also n e e d to go b e y o n d the gaps that still influence m u c h of scholarly practice, such as b e t w e e n micro a n d m a c r o approaches in sociology, or between cognitive and interactional approaches throughout the social sciences. This will also require further integration of the disrinlinp or> »o — J
xxvi
EDITOR'S I N T R O D U C T I O N
rather distant from the developments in discourse studies, n a m e l y political science - a strange p h e n o m e n o n w h e n we realize that policies a n d politics are ' d o n e ' virtually only in text a n d talk. Finally, besides such a critical perspective in the study of discourse, we face the challenge of a vast n u m b e r of urgently required applied studies. These fields of applications are m a n y , a n d fortunately m u c h work is already u n d e r way. Perhaps most relevant h e r e are the m a n y application in the field of education, from first a n d second language learning a n d literacy, to the d e v e l o p m e n t of cur ricula, textbook, classroom interaction, and teaching a n d learning in general m o s t of which is discursive (in the b r o a d semiotic sense: text, talk, sound and image). Similar observations h o l d for the applied studies of the mass media, of journalism education a n d of course the field of psychological intervention. I n d e e d , there are few applied fields of study a n d intervention that d o not have an important discursive dimension, a n d besides its critical perspective, discourse studies n e e d to increasingly also focus o n concrete practical issues of contem porary societies.
D i m e n s i o n s of Discourse a n d Fields of D i s c o u r s e Studies T h e historical sketch - given above - of the emergence of discourse studies in the various disciplines of the humanities of the social sciences, a n d its increasing integration with its sister disciplines, such as semiotics, pragmatics a n d sociolinguistics, already provides first insights into the various dimensions of discourse and the fields of the n e w cross-discipline. Let us n o w examine this configuration of the n e w discipline a n d its object of study m o r e closely. T h u s , for instance, w e o n the o n e h a n d h a v e cognitive studies of the mental processes involved in the production or comprehension of discourse, a n d inter actional studies of everyday conversation or institutional talk, o n the other hand. T h e same is true for m o r e cultural approaches in the ethnography of speaking, a n d - in a quite different, m o r e critical, perspective (focusing o n e.g., popular culture or youth culture) - i n the 'Cultural Studies' developed b y Stuart Hall and others in the U K (Hall, et al., 1978). Strangely, following the same general division of the mother disciplines, the historical approach to discourse - outside of the theory of historiography and oral history - so far has b e e n quite marginal in discourse studies, with the exception of, for instance, the work o n anti-Semitism b y R u t h W o d a k (Wodak, et al., 1990). Besides these interdisciplinary approaches in discourse studies, the 'core' of the n e w discipline remains the systematic and explicit analysis of the various structures and strategies of different levels of text and talk. Let us briefly review these, also in order to show h o w the vast field of discourse studies is organized in various subdisciplines that also have become m o r e or less independent, as also was the case for, say, lexicology, phonology, syntax a n d semantics within linguistics.
EDITOR'S I N T R O D U C T I O N
xxvii
Discourse Grammar The strong influence of linguistics on discourse studies and its d e v e l o p m e n t still shows in the p r o m i n e n t position of grammatical analysis in m a n y discourse studies. We have seen above that this influence of linguistics also played a central role in the development of discourse studies, n a m e l y in the first 'text grammars'. We also saw that various directions in formal g r a m m a r (as well as in logic a n d formal philosophy) continue to b e one of the productive areas of formal discourse analysis. Unfortunately, this formal direction of research is virtually u n k n o w n in other domains of discourse studies. Within less formal 'discourse g r a m m a r s ' , we continue to have studies of die sound structures of discourse (Bolinger, 1989; Brazil, 1975), for instance i n studies of intonation, as well as studies of discourse syntax (Givon, 1979) continuing for instance the early work o n anaphora, which also has links to formal dis course studies. Strangely, discourse semantics has r e m a i n e d an u n d e r d e v e l o p e d area of dis course grammar. Yet, if there is one level of discourse that contributes to the specific discursive nature of text and talk, it is the study of meaning, as we also know from the first studies of coherence in the 1960s and 1970s. Part of discourse semantics, and shared with work in cognitive linguistics, is of course the research on metaphor, already mentioned above. Also very relevant is the study of seman tic implication (entailment) a n d presupposition, for instance as one of the basic dimensions of coherence: I n order to establish coherence relations between the propositions of a discourse, w e often n e e d to spell out the 'missing links' of the propositions implied or presupposed b y the propositions explicitly expressed in discourse. There are m a n y m o r e aspects of discourse m e a n i n g that n e e d systematic an alysis and that cannot simply b e reduced to the semantics of words a n d sentences. For instance, discourse m a y describe (prescribe, account for, etc.,) events, actions and actors and m a y do so in m a n y ways: m o r e or less explicitly or implicitly, more or less generally or specifically, m o r e or less precisely or vaguely, with many or few details, as background or as foreground, a n d so on. T h e r e are m a n y constraints on sequences of descriptions, such as an increasing focus from broader to narrower objects of description (e.g., from a house to a r o o m in the house, from a r o o m in the house to furniture in the r o o m , and from furniture to an object on such furniture, a n d so o n - a n d in general n o t vice versa). T h e same is true for descriptions of time a n d tense sequences and, the way persons a n d social actors are described, a n d so on. O n e n e w line of research, carried out within the general framework of Functional-Systemic g r a m m a r , is that of Appraisal Theory, which examines the way opinions are expressed in discourse (Martin & White, 2005). Discourse meanings m a y b e characterized in terms of sequences of propos itions, but we k n o w that meanings are not limited to local or sequential structures, but also m a y characterize whole discourses. T h e classical example are the 'topics'
EDITOR'S INTRODUCTION
xxviii
of discourse, traditionally described in terms of 'semantic macrostructures', a n d typically expressed in headlines, leads, introductions, conclusions, initial 'thematic' sentences, a n d so o n (van Dijk, 1980). I n linguistic terms, topics are global meanings that dominate the local meanings of sequences of sentences or turns of talk. I n cognitive terms, topics represent the most important information of a discourse, as it is b e i n g assigned b y speakers/writers or recipients. Topics also represent the kind of information that is best recalled w h e n understanding discourse, a n d it is the kind of m e a n i n g we usually plan ahead before starting (or continuing) to speak or write. Despite the fundamental relevance of such global meanings in the organization a n d processing of discourse, it is strange that m a n y directions of discourse a n d conversation analysis ignore or d o not m a k e explicit such global semantic structures. Indeed, m u c h m o r e semantic research will b e necessary to examine in m u c h m o r e detail the relations between such 'macrostructures' a n d the 'microstructure' of local meanings of words and sentences. At the same time, these studies of local a n d global meanings of dis course of course n e e d to b e related to the cognitive analysis of discourse, also because they require an explicit account of the fundamental role of knowledge in the local a n d global coherence of text a n d talk. We see that b o t h at the local a n d the global level of discourse meaning, there is a vast area of discourse analysis that remains virtually unexplored, b u t that should form a n important element of future research o n discourse grammar. Stylistics Better explored, especially also in sociolinguistics a n d literary studies, has been the dimension of language a n d discourse 'style', for instance in w h a t has come to b e known as the subdiscipline of stylistics (Eckert & Rickford, 2001; Scherer & Giles, 1979). Notoriously difficult to define exactly, also because it has so many non-linguistic meanings (such as the style of clothes, houses, people, etc.,), style m a y briefly b e defined in terms of the variable expression of discourse as it is conditioned b y aspects of context. T h e most obvious manifestations of style m a y b e found, as we also know from sociolinguistics, in the various w a y people m a y p r o n o u n c e sounds, thus producing m o r e or less formal, m o r e or less casual, m o r e or less higher or lower class 'styles' of speech. Contrary to involuntary 'accents', such sound variation is called p a r t of the 'style' of a discourse if the speaker is able to control such variation of pronunciation, for instance to accommodate to the way the recipients speak, or to signal familiarity or a m o r e formal relationship. Similarly, also lexical variation has traditionally b e e n seen as one of the basic characteristics of discourse style, usually u n d e r the condition that the under lying meanings remain (more or less) the same. I n other words, 'saying the same thing in other w o r d s ' has often b e e n the rather informal definition of style. Again, such variation is stylistic if it has contextual conditions or consequences, as w h e n politicians or newspapers of different political ideologies use the lexical item 'freedom fighter' or 'rebel', rather than 'terrorist', or vice versa, or want to keep a balanced expression w h e n talking about 'insurgents'. T h a t is, w o r d choice
is one of the ways people betray their underlying opinions, social attitudes and ideologies, also because the use of lexical items is associated with underlying n^rms and values. Apart from such 'ideological' variation of lexical style, there is also a m o r e s; cial one. for instance in order to express or establish m o r e or less formal pos itions or relationships. T h u s , politicians in the U K will rather speak about 'expel ling economic immigrants' than about 'throwing scroungers out of the country' as some racist tabloids (and politicians) m a y d o , in which case popular styles ma}' combine with racist (ideological) style. Lexical styles typically c o m e in levels, such as high (very formal, official), m e d i u m (everyday public), a n d low colloquial, popular) or even very low (vulgar) levels of expression. M o r e in general, thus, lexical style signals important aspects of the context, such as the formality of the event, the social power, position a n d status of speakers or recipi ents, the relations b e t w e e n the participants, the opinions a n d ideologies of the speakers, and so on. Such is not only the case for lexical style (or pronunciation) in talk, but also "hows at other levels of discourse, as we k n o w from the stylistic difference between an English broadsheet, quality newspaper (now also often in smaller format) such as the Guardian, o n the o n e h a n d , and the popular style of the tabloid The Sun, o n the other h a n d , which also shows in size, type a n d color of headline, pictures, lay-out and m a n y other forms of multimodal expression. T h e same is true for the difference b e t w e e n the syntax of a Guardian editorial a n d than of a Sun editorial. Although usually limited to a study of context-dependent grammatical vari ation of expressions (sounds, lay-out, words, sentences), we might extend stylistic analysis also to other levels of discourse, as long as we maintain o n e (lower) level constant. For instance, elite and popular newspapers m a y write very dif ferent stories about the ' s a m e event' (that is, with the same underlying topic or semantic macrostructure), adding or omitting different details, a n d we might also call this a difference of 'style' between the newspapers. Rhetoric Discourse studies is often defined as the contemporary discipline of what used to b e called rhetoric since antiquity, that is, the practice a n d study of 'good' p u b tic speaking and writing, for instance in parliament, in court or in literature. Also today, and especially in the USA, the 'new' rhetoric is sometimes defined as a special (sub) discipline in the humanities that overlaps with discourse studies. As is the case for stylistics, rhetoric is often associated with the study of literature, rather than with the study of discourse m o r e generally (among a vast a m o u n t of studies of rhetoric, Sloane, 2001). In order to avoid collapsing rhetoric with discourse studies in general, w e narrowly) define rhetoric as the subdiscipline of discourse studies focusing o n the use of special 'rhetorical' structures of text and talk, such as metaphors, com parisons, irony, hyperboles, euphemisms, etc., that is, the kind of structures that were traditionally called 'figures of style' in classical rhetoric. Unlike other
EDITOR'S I N T R O D U C T I O N
XXX
structures of text a n d talk, these 'rhetorical' structures are optional, a n d used especially to convey or p r o d u c e specific effects, for instance as part of strategies of persuasion. T h e s e 'figures' emphasize or de-emphasize m e a n i n g a n d thus, d r a w special attention of recipients, which m a y lead to less/better m e m o r y of the meanings thus, [de] emphasized. For instance, if politicians or newspapers want to diminish the negative associations of the w o r d 'racism', t h e y m a y use the less harsh term 'popular discontent' instead. A n d vice versa, if they want to emphasize the negative aspect of t h e arrival of m a n y n e w immigrants, they might typically use the expressions 'wave' or 'invasion', which are at the same time m e t a p h o r s a n d hyperboles (van Dijk, 1993). Given these examples, it is n o t surprising that rhetoric is especially popular in the humanities - for instance i n the study of literature - a n d in t h e social sciences, for instance in studies of political rhetoric or advertising, although strictly speaking such studies should n o t b e limited to figures of speech, b u t also deal with the cognitive effects of such language use o n the recipients a n d the whole communicative context. Also, it should b e emphasized that discourse has m a n y other 'persuasive' dimensions apart from these special rhetoric structures, such as argumentation, the use of emotion words, a n d so on. Superstructures: Discourse Schemas
i
W h e r e a s stylistics a n d rhetoric were traditionally closely related to literature a n d g r a m m a r , there are other structures of text a n d talk m a t go far b e y o n d the grammatical characterization of discourse, a n d which m a y b e called 'super structures', because they are abstract form-sckemasthat globally organize discourse across sentence boundaries. A well-known e x a m p l e is the form-schema of argumentation, including such conventional categories as premises a n d conclusions. T h e s e h a v e b e e n further detailed in contemporary argumentation studies, a major subfield of discourse studies (van E e m e r e n , Grootendorst, H e n k e m a n s , 1996). Similarly, stories are often analyzed in terms of abstract narrative schemas, featuring such categories as Summary, Orientation, Complication, Resolution and Coda, m o r e or less in this order, as w e know from m u c h narrative studies, another large field of discourse studies (Labov & Waletzky, 1967; O c h s & Capps, 2001). I n the same way, m a n y other discourse genres h a v e 'canonical' structures that h a v e b e c o m e conventional a n d m o r e or less fixed 'forms' o r 'formats' of a genre. A scholarly article typically consist of such categories as Title, Abstract, Introduction, Theoretical Framework, Data/Subjects, Analysis, Conclusions, d e p e n d i n g o n the discipline a n d the subject matter. News reports in the press similarly h a v e o n e or m o r e Headlines, a Lead, M a i n Event Description, Context, Backgrounds, History a n d C o m m e n t s , as formal categories for t h e organization of specific kinds of information - typically obtained b y different news production strategies, sources or professionals (van Dijk, 1988). M a n y professional a n d insti tutional discourse types m a y h a v e such conventional formats. Even informal conversations h a v e such fixed formal categories, such as greetings a n d leave •
1
•
1
—
A
*—
1
-1- —
tYarvraB rloalo
with cnrli 'srripmatic'
EDITOR'S I N T R O D U C T I O N
xxxi
structures of text and talk (Bhatia, 1993; Gunnarsson, 1997; Swales, 2004; Ventola & Mauranen, 1996). Note that all these structures are global, a n d n o t local: J u s t like topics or semantic macrostructures, they characterize discourse as a whole, or apply to larger fragments of discourse. Also, even w h e n originally they might h a v e h a d specific meaning functions, they are formal categories defining abstract schemas. Tims, the headline of a news report is a fixed, obligatory category that applies to any news report, whatever its m e a n i n g or content. Yet the function of such a headline is semantic a n d cognitive: It expresses the m a i n topic of the text, which in turn organizes its local meanings, a n d signals the most important information about an event. Whereas most other structures of sentences a n d discourse correspond to various subdisciplines, there is n o subdiscipline that specifically deals with these schematic structures in general. Rather, different text types or genres m a y b e associated with such schemas, as w e h a v e seen for the conventional schema of an argumentation. Discourse Pragmatics We have seen that pragmatics is o n e of the overlapping sister-disciplines of dis course studies: M a n y studies of discourse are also called 'pragmatic' because they somehow h a v e to d o with the study of 'language use', rather t h a n with grammar. H e r e such a general use of the t e r m 'pragmatics' will b e avoided, because obviously the study of 'language u s e ' also takes place in socio- a n d psycholinguistics a n d other disciplines, a n d w e prefer to use the notion in a m o r e restricted, technical w a y than as s o m e kind as 'wastepaper basket' of linguistics as the philosopher Yehoshua Bar-Hillel used to say). Part of such a b r o a d e r study of language use, as w e also h a v e seen above, are for instance the ways lan guage users express or signal politeness a n d deference, a n d in general m a n a g e 'face'. Thus, whereas the study of grammar a n d style specifically focuses on form, and semantics focuses o n meaning, these pragmatic aspects rather are specific properties of interaction, such as the social relations between participants. Incidentally, although nearly all internationally influential studies referred to in this chapter are written in English, we should n o t forget that vast amounts of discourse studies h a v e b e e n published in French, G e r m a n , Spanish, Russian and other major languages. T h u s , the study of discourse pragmatics was carried out in G e r m a n y already since the early 1970s, for instance in the w o r k of Wunderlich, Ehlich, a n d R e h b e i n (see the papers in Wunderlich, 1972), scholars who later contributed m a n y other studies in the field of discourse analysis. T h e same is true, for instance, for the work on pragmatic discourse markers a n d argumentation b y D u c r o t (1972,1980,1984), in France. It is not feasible h e r e to review all relevant work in other languages t h a n English. More specifically, pragmatics will h e r e b e understood as the subdiscipline of discourse studies focusing o n speech acts or illocution, that is, the specific social acts accomplished b y language users a n d that typically are (only) accom plished b y text or talk, such as assertions, promises, questions, congratulations, and so on.
EDITOR'S INTRODUCTION
xxxii
l i k e sentences a n d their meanings (propositions), also speech acts usually c o m e in sequences, as is the case in conversations, parliamentary debates, and other types of discourse. A n d as we do for sequences of propositions, also se quences of speech acts can b e said to b e locally or globally coherent, for instance w h e n one speech acts provides reasons for the next o n e (such as in the sequence Assertion-Request 'It's stuffy in here. Could y o u please o p e n the window?'). Similarly, the global speech act performed b y this chapter is one of an assertion, whereas the global speech act of a n editorial in the press m a y b e an accusation or a r e c o m m e n d a t i o n a n d the weather forecast a prediction (van Dijk, 1981). Conversation Analysis Last b u t n o t least, the vast field of research c o m m o n l y labeled 'Conversation Analysis' (CA) specifically focuses o n the interactional nature of language use and discourse. Although early work in C A specifically dealt with informal, spon taneous everyday conversation, later studies also m o r e generally deal with 'talk in interaction', that is, also with institutional dialogues of m a n y kinds. Emerging from microsociology a n d e t h n o m e t h o d o l o g y , these studies are specifically interested in the 'local order' of social structure, and h o w also institutions and organizations are daily produced and reproduced b y talk (Boden & Zimmerman, 1991; D r e w & Heritage, 1992). Although often presented as a separate subdiscipline, the study of talk-ininteraction obviously belongs to the b r o a d e r study of discourse. M a n y of the interactional aspects of talk are closely related to grammar, semantics, pragmatics and other dimensions of discourse: Turn taking is based o n clues from intonation, syntactic structure or m e a n i n g units. Openings a n d closings of talk are schematic categories that h a v e similar functions as Introductions, Headlines, o n the one h a n d , or with Conclusions, o n the other h a n d , p r o m i n e n t in m a n y spoken or written types of discourse. M o v e s a n d strategies of interaction are organized also in terms of m e a n i n g , as is the case for agreements a n d disagreements. Selfpresentation strategies have b o t h interactional as well as semantic a n d formal characteristics, as we also know from such well-known disclaimers as "lam not a racist, but...". Indeed, most of the conditions of local a n d global coherence, of style a n d rhetoric, characterize b o t h spoken a n d written discourse, a n d it does not m a k e sense, therefore, to distinguish two disciplines of discourse studies. O n the other h a n d , studies of written discourse (for instance in argumentation), should n o t neglect the interactional dimension of such discourse. A n d in many forms of C o m p u t e r Mediated C o m m u n i c a t i o n (CMC) today, such as chatting, talk-in-interaction is written or multimodal rather than spoken. I n sum, the complex a n d subde structures and strategies of interaction are multiply related to all other levels a n d dimensions of discourse, a n d m a y be studied in a unified framework. Studies o n the formal or m e a n i n g aspects of discourse m a y b e c o m p l e m e n t e d b y a study of their interactional dimensions, a n d vice versa, as we h a v e seen above, the analysis of talk in interaction is in extricably related to other local, sequential and global dimensions of discourse, from intonation and syntax, local and global meanings, to schematic organization atirl sripprh arts.
xxxiii
EDITOR'S I N T R O D U C T I O N
Genre Analysis From the brief s u m m a r y given, w e see that the study of discourse is a vast field, consisting of m a n y subdisciplines, a n d at the same time overlapping with other uev interdisciplines, such as sociolinguistics, a n d with sister disciplines such as pragmatics a n d semiotics. It either is a part or overlaps with virtually all m o t h e r disciplines of the humanities a n d social sciences. This overlap with other disciplines also produces the kinds of studies that focus on different genres, such as the study of m a n y types a n d subtypes of text and talk in politics, the media, education, science, law, business, the bureaucracy and also parliamentary speeches, n e w s reports, editorials, textbooks, classroom lessons, laws, business letters, p h o n e calls, annual reports, meetings, bureaucratic forms along with a host of other genres (Bhatia, 1993; L e m k e , 1990). Note though that genres are n o t merely described i n terms of their structures at any of the dimensions m e n t i o n e d above, b u t especially also in terms of their contexts (van Dijk, 2007). T h u s , a parliamentary debate has very few exclusive itructures - its topics, its forms of rhetoric, its argumentation, a n d so o n m a y b e pan of any discourse about the same subject - a n d h e n c e needs to b e defined i n terms of specific context categories, such as M P s , political parties, g o v e r n m e n t m d opposition, constituents a n d voters, as well as in terms of political goals a n d processes, knowledge a n d ideologies. S o m e of these contextual elements m a y be accompanied b y specific discourse forms, as w h e n m e m b e r s of the same party in British parliament are traditionally addressed as ' m y h o n o u r a b l e friend'. Such contextual approaches m a y b e c o m b i n e d with the m o r e traditional descriptions of discourse genres in terms of their structural characteristics, for instance stories in terms of narrative schemas, style, topics or t h e perspective of the narrator, or news reports in terms of its canonical schema, featuring headlines and leads, a n d other categories - besides some special lexical items preferred i n news discourse (e.g., the short formal word 'bid' in English headlines, rather than the longer n o u n 'attempt'). Note that genre analysis is m e r e l y a collective label for w h a t in m a n y respects have become m o r e or less a u t o n o m o u s subdisciplines of discourse studies, such conversation analysis, narrative analysis, argumentation analysis, the study of classroom interaction, political discourse analysis, m e d i a discourse analysis, and so on. With the usual increasing specialization w e k n o w from other disciplines, it ib likely that in the future we'll h a v e discourse analysts specialized i n the study A news in the press, high school textbooks, schizophrenic talk, parliamentary debates, life stories, soaps (telenovelas), a n d so o n for m a n y h u n d r e d s of other discourse genres defined as discursive social practices. Applied Discourse Studies /dthough we m a y thus e x p a n d the field of discourse studies as far as the study of the h u m a n activity of text, talk a n d communication m a y bring us, w e n o w h a v e summarized at least some of its maior subdisdnlinps V.*rU nf „„,u J ; „ „ . - _ I ; —
EDITOR'S INTRODUCTION
xxxiv
h a s its o w n background, theories, terms, objects of analysis, methods, aims, introductions, handbooks, journals, conferences, and even associations of scholars. E a c h of these subdisciplines h a v e a m o r e applied dimension, w h e n n o longer m e r e theory or description is relevant, b u t concrete applications, interventions a n d the use of science in the solution of social problems are required. In linguistics, w e are familiar with the use of g r a m m a r in the study of first or second language learning, translation, a n d other aspects of language use. I n discourse studies, the n u m b e r of possible applications is so vast that they cannot even b e summar ized h e r e , because they pertain to any aspect of language use, interaction and communication, from literacy to the formation of journalists, peace negotiation and the critique of advertising and political manipulation. Pervasive a n d probably most relevant are all applications in education, such as curricula, the production of adequate (and non-racist, non-sexist, etc.,) textbooks, programs of classroom intervention, testing a n d so on. N o w w e h a v e some m o r e insight into the structures of talk a n d text, as well as their cognitive basis and social a n d cultural contexts, we in principle are also in a better position to engage in the treatment of the m a n y social issues that have a discursive dimension. People m a y n o t first think of text a n d talk w h e n dealing with racism, for instance, until it is shown that racist prejudices a n d ideologies that are the basis of racist discrimination are largely acquired b y discourse, especially the public discourses of the elites, e.g., in politics, the mass m e d i a and textbooks. M u c h critical a n d practical studies o n discourse combine theoretical, descriptive and 'applied' dimensions, and i n d e e d hardly differentiate between such dimensions of scholarly activity. Critical discourse analysis focuses on soc ial p r o b l e m s a n d n o t o n s c h o l a r l y p a r a d i g m s , a n d tries to u n d e r s t a n d a n d solve such problems with any kind of m e t h o d , theory or description thai m a y b e relevant - taking into account the experiences a n d perspectives of the participants.
L o o k i n g A h e a d : T h e Future o f D i s c o u r s e S t u d i e s I n the previous pages we have occasionally formulated speculations a n d wishes about the future developments of the study of discourse. If the 'logic' of the his torical development of the language sciences is followed, in which the nature a n d size of the object of study was gradually e x t e n d e d from sounds, words, syntax, meanings, speech acts, a n d interaction, to the contexts of society, culture a n d cognition, we m a y expect that research in the coming years will consolidate this further expansion as well as explore n e w extensions. More - and More Explicit -
'Traditional'Descriptions
Firstly, there are 'traditional' areas that n e e d (much) m o r e attention, such as the study of the m a n y dimensions of discourse semantics, and the further integra tion of micro- and macro-semantics. Secondly, we n e e d a m u c h m o r e explicit
EDITOR'S INTRODUCTION
xxxv
integration of rhetoric into the study of discourse, instead of the separate, m o r e traditional formulation of 'figures of speech'. Thirdly, t h e c o m p l e x notions of style and register n e e d to b e m a d e m o r e explicit t h a n h a s b e e n d o n e so far, thus also integrating linguistic, sociolinguistic a n d discourse analytical studies of lan guage use in its social context. A n d finally, we n e e d m o r e general insight into (he structures of conventional discourse formats, b e y o n d t h e well-known ones of narrative a n d argumentation. M a n y discourse types h a v e such canonical schemas. related also to their interactional a n d institutional functions, a n d m u c h more classical descriptive work will b e necessary within a m o r e general, typo logical approach of discourse. This is finally also the case for a m o r e general theory of genre and specific g e n r e descriptions. I n that respect, discourse analysis is still a young discipline. T h e n u m b e r of discourse types h a r d l y ever studied is probably m u c h bigger t h a n t h e discourse types that h a v e b e e n studied. Semiotic Extensions and Computer Mediated Communication It has already b e e n emphasized that discourse studies overlaps with its sister discipline of semiotics with w h i c h ideally it should m e r g e in t h e future - b o t h being involved in the study of communicative events. T h u s , e v e n w h e n limiting the study to discourse in a m o r e traditional sense, we n e e d to realize that such discourse has m a n y 'non-verbal' dimensions, such as intonation, gestures, ap plause, music and other aspects of oral performance, as well lay-out, printing types, color, pictures, drawings, film, a n d so o n for written discourse (van Leeuwen, 201)5). In sum, also u n d e r the influence of the t r e m e n d o u s variety of C o m p u t e r Mediated C o m m u n i c a t i o n (CMC) discourse a n d communication events h a v e become truly 'multimedia' events, in which sounds, music, m o v e m e n t and images combine, in which spoken a n d written m o d e s of discourse are closely integrated, for instance in 'chats', a n d in which 'interaction' (e.g., b e t w e e n h u m a n s a n d machines) has acquired n e w meanings (Danet, 2 0 0 1 ; Herring, 1996). Most of these forms of discourse still n e e d to b e studied, and in that respect also discourse studies is still a y o u n g discipline. Other Disciplines We have seen h o w discourse studies e m e r g e d from several disciplines in t h e humanities and social sciences, especially literary studies, linguistics, psychology, sociology and anthropology, later joined b y communication studies. T h e r e are many more disciplines in which discourse is a valid a n d important object of *tudy. We have m e n t i o n e d political science a n d history, b o t h disciplines that have regrettably played a less p r o m i n e n t role so far in discourse studies (and vice versa). There is n o d o u b t that also these disciplines will have a strong discourse analytical dimension in the future. Even in such disciplines as g e o g r a p h y a n d economics, discourses are objects of analysis, a n d discourse analysis m a y h e n c e also play a m o r e p r o m i n e n t role in those disciplines.
EDITOR'S INTRODUCTION
xxxvi
The Cognition-Interaction Gap We hardly n e e d to repeat what has already b e e n said about the regrettable gap between social a n d psychological studies, focusing o n interaction and cognition, respectively. Interaction and cognition mutually presuppose each other, and any integrated study of discourse in the future will n e e d to deal with both these aspects of language use. Cognition will n e e d to b e studied also in a social para digm because knowledge and beliefs are shared b y other m e m b e r s of collectivities a n d acquired in forms of social interaction w e call 'learning'. A n d vice versa, there are m a n y aspects of interaction that are in fact cognitive dimensions of action, such as the v e r y m e a n i n g or interpretation of social conduct, as well as notions such as intention, plan, goals, motivation, purpose, coordination, and so on. M o d e r n cognitive science has a lot to offer to c o n t e m p o r a r y interaction studies, whatever the relevance of m o r e autonomous approaches for the initial a d v a n c e m e n t of a n e w field. We n o t only n e e d m o r e cognitive a n d neurological studies of words a n d meanings, b u t also of the organization of m o r e complex interactional and discursive structures, such as text schemata or topics, as well as the understanding and m a n a g e m e n t of context. Context i
I n d e e d , while focusing o n text a n d talk, w e t e n d to forget the m a n y dimensions of a theory of context that is still in the m a k i n g a n d that needs to b e further de veloped in the future, a n d in several (sub) disciplines. It is crucial in any account of the structures of text a n d talk to describe a n d explain h o w they m a y vary with different structures of the communicative event as these are subjectively represented in context models b y the participants. Contexts a n d their impact o n discourse are v e r y complex a n d future studies n e e d to disentangle these dif ferent dimensions of contextual control a n d consequences. Most studied have b e e n gender, region a n d class as 'variables'. However, there are a vast number of other aspects of the social a n d communicative situation that n e e d to b e further studied in the future, such as time a n d space/place settings, the social and insti tutional roles a n d relations of participants, a n d the nature of intentions, plans a n d purposes as well as the fundamental role of knowledge in discourse produc tion a n d comprehension. We n e e d to k n o w h o w local (micro) situations may thus combine with m o r e global (meso and macro) levels of context interpretation a n d influence, for instance h o w daily practices of m e m b e r s of institutions are controlled both b y specific interaction types (e.g., interviews or meetings) as well as m o r e global (knowledge about the) structure of institutions or organizations. Methods and Theories C o m p a r e d to formal g r a m m a r , m o s t discourse studies are quite informal. This is as such n o p r o b l e m (also because informal studies are m o r e accessible to m o r e students a n d non-academic readers), b u t that does n o t m e a n that discourse analysis should b e less explicit a n d systematic. T h e description of structures
EDITOR'S I N T R O D U C T I O N
xxxvii
and strategies at all levels earlier m e n t i o n e d should take place in terms of explicit theories, and analyses of specific discourse types should b e systematic a n d rele vant, and not m e r e l y a personally-subjective interpretation, p a r a p h r a s e or comment on text or talk, as is still often the case in m a n y studies that p u r p o r t to do discourse analysis'. U should b e stressed though that unlike content analysis discourse analysis is not a m e t h o d but a discipline. T h e r e is n o such thing as psychological or sociological analysis either. Even w h e n studying the ' s a m e ' object, such as a news item or a parliamentary debate, we m a y h a v e recourse to m a n y 'methods', also depending on the kinds or structures w e want to focus on, a n d especially also depending on our aims of study.
Conclusion If we measure the success of a discipline b y its propagation in other disciplines, die study of discourse has b e e n very successful indeed. T h e r e are only a few dis ciplines and areas in the humanities a n d social sciences that h a v e n o t engaged in some form of discourse studies. This is n o t only true for the mother disciplines of discourse studies, such as linguistics, cognitive psychology, sociology a n d anthropology, b u t n o w also for social psychology, communication studies, pol itical science, history, a n d e v e n for such less discursive disciplines as economics and geography. T h e r e is n o t h i n g surprising about that, because humanity, soci ety and culture are primarily characterized b y their social interaction in general and language use or discourse in particular. I n fact, instead of w o n d e r i n g about the vast field a n d potential of discourse studies, one m a y ask w h y discourse studies as a general cross-discipline was not 'invented' before - although in a sense rhetoric h a d some of these functions since antiquity. in this introduction w e h a v e briefly sketched the history a n d the organization ->f the new field of discourse studies, as well as formulated some suggestions for iuture developments. It m a y h a v e b e c o m e clear that the study of discourse has made considerable progress in a few decades since its early studies of narrative, conversation and grammar in the 1960s a n d 1970s. It is n o w a vast cross-discipline vith equally vast sub-disciplines, each with its o w n handbooks, journals a n d congresses, as is the case for argumentation, narrative, media discourse a n d scien tific discourse analysis. Within the original m o t h e r disciplines also m a n y studies are associated with discourse analysis, as is obviously the case in literary studies, linguistics and anthropology. T h i s is even m o r e evident in sister-disciplines such as sociolinguistics. semiotics, ethnography, a n d especially pragmatics. All these new subdisciplines or sister disciplines of discourse studies also h a v e entered the stage of useful applications in the study of social p r o b l e m s a n d i m p l e m e n ration of m a n y social practices. In sum, discourse studies has c o m e of age, a n d has b e c o m e a major crossdiscipline within a n d r e l a t e d to o t h e r major disciplines in t h e h u m a n i t i e s and social sciences - a n d as o n e of the major disciplines accounting for the m o s t human of all p h e n o m e n a : language use.
Acknowledgement For comments and suggestions on an earlier version of this chapter, I am indebted to Ruth Wodak.
i
Notes
1. To limit a potentially vast bibliography, this introductory chapter only will cite some influential books on discourse in English. For detailed further references, see the articles collected in this book.
Further R e a d i n g Besides the basic books referred to above, here are some suggestions for further (espe cially also introductory) reading in various areas of discourse studies and its sisterdisciplines: General introductions to discourse studies: Georgakopoulou & Goutsos (1997)"Johnstone (2002); Renkema (2004); Schiffrin (1993); Van Dijk (1997); Wooffltt (2005). Readers of discourse studies: Jaworski & Coupland (1999); Wetherell, Taylor, & Yates (2001). Handbooks of discourse studies: SchifFrin, Tannen & Hamilton (2001); Van Dijk (1985). Conversation Analysis: Lemer (2004); Ten Have (1999); Ochs, Schegloff & Thompson (1996); Wooffltt (2005). Pragmatics: Mey (2001); Verschueren (1999). Sociolinguistics: Blommaert (2005); Bratt Poulston & Tucker (2003), Coulmas (2003); Holmes (1992); Sarangi & Coulthard (2000); Trudgill (1995); Wardhaugh (1998). Stylistics: Eckert & Rickford (2001); Toolan (1997). Rhetoric: Lucaites, Condit & Caudill (1999). Critical Discourse Studies: Caldas Coulthard & Coulthard (1995); Fairclough (1995); Lazar (2005); Toolan (2002); Wodak & Meyer (2001).
References
Atkinson,J. M., & Heritage,J. (Eds.). (1984). Structures of Social Action: Studies in Conversation Analysis. Cambridge: Cambridge University Press. Austin.J. L. (1962). How to Do Things with Words. Cambridge: Harvard University Press. Barthes, R. (1964). 'Elements de semiologie'. Communication 4, 91-135. Bauman, R., & Sherzer,J. (1974). Explorations in the Ethnography of Speaking. Cambridge: Cambridge University Press. Bhatia, V. K. (1993). Analysing Genre. Language Use in Professional Settings. London New York: Longman. Billig, M. (1987). Arguingand Thinking: A RhetoricalApproach to SocialPsychology. Cambridge Cambridge University Press. Billig, M. (1988). Ideological Dilemmas: A Social Psychology of Everyday Thinking. London Newbury Park: Sage Publications. Blommaert,J. (2005). Discourse. A Critical Introduction. Cambridge: Cambridge University Press.
TTOR'S INTRODUCTION
xxxix
Baden, D.. & Zimmerman, D. H. (Eds.). (1991). Talk and Social Structure: Studies in Ethnomethodology and Conversation Analysis. Berkeley: University of California Press. Bolinger. D. (1989). Intonation and Its Uses: Melody in Grammar and Discourse. Stanford, C'A: Stanford University Press. BTM Paulston, C , & Tucker, G. R. (Eds.). (2003). Sociolinguistics. The Essential Readings. f )xford: Blackwell. Brazil. D. (1975). Discourse Intonation. Birmingham: University of Birmingham, English Language Research. Bn>« n. P., & Levinson, S. (1987). Politeness: Some Universals in Language Usage. Cambridge: Cambridge University Press. Caidai-Coulthard, C. R., & Coulthard, M. (Eds.). (1995). Texts and Practices: Readings in Critical Discourse Analysis. London: Routledge. Ciiuimas, F. (Ed.). (1997). The Handbook ofSociolinguistics. Oxford, UK Cambridge, MA: Blackwell Publishers. Danci. B. (2001). Cyberpl@y. Oxford: Berg. Diumar. N. (1976). Sociolinguistics. A Critical Survey of Theory and Application. London: Edward Arnold. Drosler, W. (1972). Einfiihrung in die Textlinguistih Tubingen: M. Niemeyer. Druv^. P., & Heritage, J. (Eds.). (1992). Talk at Work. Interaction in Institutional Settings. Cambridge. Ducmi. O. (1972). Dire etnepas dire: Principes de semantique linguistique. Paris: Hermann. Dut-roi. O. (1980). Les ichelks argumentatives. Paris: Minuit. DutrtiL O. (1984). Le dire et le dit. Paris: Minuit. Duranti, A. (Ed.) (2001). Linguistic Anthropology: A Reader. Maiden, MA: Blackwell. Eckert, P., & Rickford, J. R. (Eds.). (2001). Style and Sociolinguistic Variation. Cambridge, UK New York, NY: Cambridge University Press. Ecu. U. (1976). A Theory ofSemiotics. Bloomington: Indiana University Press. Edwards, D. (1997). Discourse and Cognition. London: Sage. Eron-Tripp, S. (1972). 'On Sociolinguistic Rules: Alternation and Co-occurrence'. In j J. Gumperz and D. Hymes (Eds.), Directions in Sociolinguistics: The Ethnography of Communication (pp. 213-250). New York: Holt. Fairclough, N. (1989). Language and Power. Harlow: Longman. Fairciough, N. (1995). Critical Discourse Analysis. The Critical Study ofLanguage. London: Longman. Fowler. R.. Hodge, B., Kress, G., & Trew, T. (1979). Language and Control London: Routledge & Kegan Paul. Garfinkel. H. (1967). Studies in Ethnomethodology. Englewood Cliffs, NJ: Prentice-Hall, Inc. (*orgakopoulou, A., & Goutsos, D. (1997). Discourse Analysis. An Introduction. Edinburgh: Edinburgh University Press. Givon, T. (Ed.) (1979). Syntax and Semantics, vol. 12: Discourse and Syntax. New York: Academic Press. Goffman. E. (1959). The Presentation of Selfin Everyday Life. Garden City, N. Y.: Doubleday. Goffman, E. (1961). Encounters: Two Studies in the Sociology of Interaction. Indianapolis: Bobbs-Merrill. Grice. H. (1975). 'Logic and Conversation'. In P. Cole & J. Morgan (Eds.), Syntax and Semantics, vol. 3: Speech Acts (pp. 68-134). New York: Academic Press. Grice. H. P. (1975). 'Logic and Conversation'. In P. Cole andN. L. Morgan (Eds.), Syntax and Semantics, vol. 3: Speech Acts (pp. 41-58). New York: Academic Press. Grimes. J. E. (1975). The Thread ofDiscourse. The Hague: Mouton. Gumnerz- I I f l Q R O o l n.v ^ • • •
xl
EDITOR'S INTRODUCTION
1
Gumperz, J. J. (1982b). Language and Social Identity. Cambridge (England New York: Cambridge University Press. Gumperz, J. J., & Hymes, D. (1972). Directions in Sociolinguistics: The Ethnography of Communication. New York: Holt, Rinehart and Winston. Habermasjj. (1981). Theorie des kommunikativen Handelns. Frankfurt am Main: Suhrkamp. Hall, S., Critcher, C , Jefferson, T., Clarke, J., & Roberts, B. (1978). Policing the Crisis: Mugging, the State and Law and Order. London: Methuen. Halliday, M. A. K., & Hasan, R. (1976). Cohesion in English London: Longman. Herring, S. C. (Ed.). (1996). Computer-Mediated Communication. Linguistic, Social and CrossCultural Perspectives. Amsterdam: Benjamins. Holmes, J. (1992). An Introduction to Sociolinguistics. London New York: Longman. Holmes,}., & Meyerhoff, M. (Eds.). (2003). The Handbook of Language and Gender. Oxford: Blackwell. Hymes, D. (1972). 'Models of the Interaction of Language and Social Life'. InJ.J. Gumperz and D. Hymes (Eds.), Directions in Sociolinguistics: The Ethnography of Communication (pp. 35-71). New York: Holt, Rinehart and Winston. Jaworski, A., & Coupland, N. (Eds.). (1999). The Discourse Reader. London New York: Routledge. Johnson-Laird, P. N. (1983). Mental Models. Cambridge: Cambridge University Press. Johnstone, B. (2002). Discourse Analysis. Oxford: Blackwell. Kamp, H. (1981). 'A Theory of Truth and Semantic Representation'. InJ. A. G. Groenendijk, T. Janssen & M. Stokhof (Eds.), Formal Methods in the Study ofLanguage (pp. 277-322). Amsterdam: Mathematical Centre Tracts. Kintsch, W. (1974). The Representation ofMeaning in Memory. Hillsdale, N. J.: Erlbaum. Kramarae, C. (1980). Women and Men Speaking: Frameworks for Analysis. Rowley, MA: Newbury House Publishers. Kress, G., & van Leeuwen, T. (1990). Reading Images. Victoria: Deakin University Press. Labov, W. (1972a). Language in the Inner City: Studies in the Black English Vernacular. Philadelphia: University of Pennsylvania Press. Labov, W. (1972b). Sociolinguistic Patterns. Philadelphia: University of Pennsylvania Press. Labov, W., & Waletzky, J. (1967). 'Narrative Analysis. Oral Versions of Personal Experience'. InJ. Helm, (Ed.), Essays on the Verbal and Visual Arts (pp. 12-44). Seattle: University of Washington Press, LakofT, G., &Johnson, M. (1980). Metaphors We Live By. Chicago: University of Chicago Press. Lazar, M. M. (Ed.). (2005). Feminist Critical Discourse Analysis. Gender, Power and Ideology in Discourse. Houndmills, Basingstoke, Hampshire New York: Palgrave Macmillan. Leodolter (=Wodak), R. (1975). Das Sprachverhalten sonAngeklagten bei Gericht.-Ansatzezu einer soziolinguistischen Theorie der Verbalisierung. Kronberg/TS.: Scriptor Verlag. Lerner, G. H. (2004). Conversation Analysis. Studies from the First Generation. Amsterdam Philadelphia: John Benjamins Pub. Lutz, B., & Wodak, R. (1987). Information fur Informierte. Linguistische Studien & Verstdndlichkeit und Verstehen von Hbrfunknachrichten (Information for those who ar informed. Linguistic studies of the comprehensibility and comprehension of radio news). Wien: Verlag der Osterreichischen Akademie der Wissenschaften. Mann, W. C , & Thompson, S. A. (1988). 'Rhetorical Structure Theory. Towards a Functional Theory of Text Organization'. Text, 8, 243-281. Martin, J. R. (1992). English Text. System and Structure. Philadelphia: John Benjamins. Martin, J. R., & White, P. R. R. (2005). The Language ofEvaluation. Appraisal in English. New York: Palerave Macmillan.
rrOR'S INTRODUCTION
xli
Mey,J.! 1985). Whose Language?A Study in Linguistic Pragmatics. Amsterdam Philadelphia: j. Benjamins Pub. Co. Mey,j. L. (2001). Pragmatics: An Introduction, 2nd edn. Maiden, MA: Blackwell. Ochs, E.. & Capps, L. (2001). Living Narrative: Creating Lives in Everyday Storytelling. Cambridge, MA: Harvard University Press. Ochs, £.. Schegloff, E. A., & Thompson, S. A. (Eds.) (1996). Interaction and Grammar. Cambridge: Cambridge University Press. Fetfifi.j. S. (1971). Transformationsgrammatiken und eine ko-textuelle Texttheorie. Frankfurt: Athenaum. PDI i.r,J. 11996). Representing Reality: Discourse, Rhetoric and Social Construction. London Thousand Oaks, CA: Sage. Run f. J.. & Wetherell, M. (1987). Discourse and Social Psychology. Beyond Attitudes and 3'haviour. London Newbury Park, Calif.: Sage Publications. Re:ikema-J. (2004). Introduction to Discourse Studies. Amsterdam Philadelphia: John Benjamins Pub. '••« H., Schegloff, E. A., & Jefferson, G. A. (1974). 'A Simplest Systematics for the • 'rganization of Turntaking for Conversation'. Language 50, 696-735. Sai.ngi. S., & Coulthard, M. (Eds.). (2000). Discourse and Social Life. Harlow, Eng. Jew York: Longman. Sav:!le-Troike, M. (1982). The Ethnography of Communication. An Introduction. Oxford: Blackwell. Scliank. R. C . & Abelson, R. P. (1977). Scripts, Plans, Goals, and Understanding: An Inquiry into Human Knowledge Structures. Hillsdale, NJ. Erlbaum. Sch.,ret. & H. Giles (Eds.). (1979). Social Markers in Speech, (pp. 147-210). Cambridge: Cambridge University Press. Schtffrin. D. (1993). Approaches to Discourse. Oxford: Blackwell. 5chufrin. D.. Tannen, D., & Hamilton, H. E. (Eds.). (2001). The Handbook of Discourse Analysis. Maiden, Mass.: Blackwell Publishers. Scarle,J. R. (1969). Speech Acts: An Essay in the Philosophy of Language. London: Cambridge CP. Slo.vne. T. O. (2001). Encyclopedia of Rhetoric. Oxford New York: Oxford University Press. Swalp^.J. (2004). Research Genres. Explorations and Applications. New York: Cambridge University Press. Tfen Have, P. (1999). Doing Conversation Analysis. A Practical Guide. London: Sage Tooian. M. J. (Ed.). (2002). Critical Discourse Analysis. Critical Concepts in Linguistics. New York: Routledge. Tradgill. P. (1995). Sociolinguistics. An Introduction to Language and Society. London, England Mew York. N.Y., USA: Penguin, van Dijk. T. A. (1972). Some Aspects of Text Grammars. A Study in Theoretical Linguistics and Poetics. The Hague: Mouton. van Dijk. T. A. (1977). Text and Context. Explorations in the Semantics and Pragmatics of Discourse. London New York: Longman, van Dijk. T. A. (1980). Macrostructures: An Interdisciplinary Study of Global Structures in Discourse, Interaction, and Cognition. Hillsdale, NJ.: L. Erlbaum Associates, van Dijk. T. A. (1981). Studies in the Pragmatics of Discourse. The Hague: Mouton. van Dijk. T. A. (1985). Handbook of Discourse Analysis. 4 vols. London Orlando: Academic Press. van Dijk. T. A. (1988). News as Discourse. Hillsdale, NJ: L. Erlbaum Associates. van Dijk, T. A. (1993). Elite Discourse and Racism. Newbury Park, Calif.: Sage Publications. van Dijk. T. A. (2007). Context. A Multidisciplinary Theory. In press.
xlii
EDITOR'S INTRODUCTION
van Dijk, T. A. (Ed.). (1997). Discourse Studies. A Multidisciplinary Introduction. 2 vols. London: Sage. van Dijk, T. A. (Ed.). (2006). 'Discourse, Interaction and Cognition'. Special Issue, Discourse Studies, 8(1), 159-177. van Dijk, T. A., & Kintsch, W. (1983). Strategies of Discourse Comprehension. New York London: Academic Press, van Eemeren, F. H., Grootendorst, R., & Henkemans, F. S. (1996). Fundamentals of Argumentation Theory: A Handbook of Historical Backgrounds and Contemporary Developments. Mahwah, N.J.: L. Erlbaum. van Leeuwen, T. (2005). Introducing Social Semiotics. New York: Routledge. Ventola, E., & Mauranen, A. (Eds.). (1996). Academic Writing: Intercultural and Textual Issues. Amsterdam Philadelphia: John Benjamins. \ferschueren, J. (1999). Understanding Pragmatics. London New York New York: Arnold Oxford University Press. Wardhaugh, R. (1998). An Introduction to Sociolinguistics. Cambridge, MA: Blackwell. Wetherell, M., Taylor, S., Yates, S. (2001). Discourse Theory andPractice. A Reader. London: Sage. Wodak, R. (1986). Language Behavior in Therapy Groups. Berkeley, CA: University of California Press. Wodak, R. (1989). Language, Power, and Ideology. Studies in Political Discourse. Amsterda Philadelphia: J. Benjamins Pub. Co. Wodak, R. (Ed.). (1997). Gender and Discourse. London Thousand Oaks, Calif.: Sage Publications. Wodak, R., & Meyer, M. (Eds.). (2001). Methods of Critical Discourse Analysis. London Thousand Oaks Calif.: Sage. Wodak, R., & Schulz, M. (1986). The Language of Love and Guilt. Mother-Daughter Relationshipsfrom a Cross-Cultural Perspective. Amsterdam Philadelphia: J. Benjamins. Wodak, R , & Van Dijk, T. A. (Eds.). (2000). Racism at the Top. Kkgenfiirt: Drava Verlag. Wodak, R , Nowak, P., Pelikan, J., Gruber, H., de Cillia, R , & Mitten, R (1990). "Wir sind alle unschuldige later". Diskurshistorische Studien zum Nachkriegsantisemitismus ("We are all innocent perpetrators" Discourse-historical studies in post war antisemitism). Frankfurt/Main: Suhrkamp. Wunderlich, D. (Ed.). (1972). Linguistische Pragmatik Frankfurt/M.: Athenaum. Wooffitt, R. (2005). Conversation Analysis and Discourse Analysis. A Comparative and Critica Introduction. London Thousand Oaks, Calif.: Sage.
1 L
Interviewing Persons with a Learning Disability: How Setting Lower Standards May Inflate Weil-Being Scores Charles Antaki
H
ow does one ask p e o p l e about their quality of life, a n d what does one do if those p e o p l e h a v e a learning disability? "Few would dispute that a basic aim of everyone w h o formulates questions for use in social research is to have their questions u n d e r s t o o d b y respondents as intended" (Foddy, 1993, p . 183). I n d e e d , b u t "having their questions understood" is a spe cial worry for researchers w h e n the r e s p o n d e n t has b e e n diagnosed as having some cognitive disability. This article is a b o u t w h a t can h a p p e n w h e n questions are asked "sensitively." Interview techniques a n d questionnaire design h a v e b e e n sites for sustained development for at least the past 50 years (Foddy, 1993). T h e debate has thrown up a number of problems with phrasing that seem solvable, and there are a num ber of authoritative sources to which one can turn for such solutions (e.g., Foddy, 1993; Fowler & M a n g i o n e , 1990). But t h e r e is r e a s o n to believe, since the opening-up of the interactional domain b y ethnomethodologists and conversation analysts (Garfinkel, 1967; Sacks 1992), that things are not so simple. Recent de velopments in understanding "institutional talk," for example in the collection it; Drew and Heritage (1992) a n d especially in close analyses of structured tests ' ^arlaire & M a y n a r d , 1990; M a y n a r d & Marlaire, 1992) a n d standardised interviews (Houtkoop-Steenstra, 1995, 1996; S u c h m a n & J o r d a n , 1990), show i t the work that goes o n to actually ask a question a n d receive an answer is much more complicated t h a n the official script might suggest, and this w o r k poses a serious threat to claims that the questions are standardized.
Source: Qualitative Health Research vol. 9, no. 4, 1999, pp. 437-454.
D I S C O U R S E STUDIES
2
T h e p r o b l e m is exacerbated w h e n the respondent is diagnosed to b e in some w a y cognitively disabled a n d u n a b l e to understand complex questions. In such circumstances, the interviewer m a y n o t always b e able to follow the strict letter of the interview schedule a n d is even m o r e likely to fall prey to what Garfinkel (1972) notes h a p p e n s whenever professionals actually m a k e use of idealized in structions, formulas, rules, a n d so on. T h a t is, the interviewer will mobilize her or his set of ordinary, unstated, advisory rules to deal with difficult cases. As M a y n a r d a n d Marlaire (1992) found w h e n examining t h e standardized testing of children on a linguistic task a n d referring to the interviewer's nonofficial talk, "the reportable s c o r e . . . d e p e n d s u p o n this infrastructure, or substrate, or scaf folding, of organised, practical skills" (p. 196). I n what follows, I want to look ai this "interactional substrate." I shall m a k e a close comparison between script and delivery in a set of quality-of-life assessments of people with learning dis abilities (in N o r t h A m e r i c a n terminology, people with mental retardation) to see just w h e r e the interviewers' unstated advisory rules take them. W h a t I shall try to bring out is what the deviations from the script tell us about the interviewer's treatment of the respondent as b e i n g "incompetent," n o t only mentally but also socially, as a person with reduced personal expectations. I n making this claim, I shall b e invoking the sorts of conversational devices identified b y conversation analysis (CA). O n e question that will occur to readers is whether C A , with its base in the talk of people without a learning disability, can cope with the talk of people w h o are diagnosed as being n o t fully competeni cognitively. This is, in part at least, a n empirical matter, a n d so far as it has been tried, the answer seems to b e that it can. Brewer a n d Yearley (1989) find that, leaving aside certain features that require normatively measurable memon abilities (such as large-scale sequential planning), learning difficulties did noi d e b a r people from using the kind of conversational structures a n d devices CA has identified in the talk of people without a learning disability. Separately, Wootton (1989) takes a possibly m o r e difficult case of the talk of a child severely affected b y D o w n ' s s y n d r o m e to demonstrate t h e illumination C A can give. Evidence in the 1990s has b e e n accumulating m o r e quickly, a n d certainly Perkuu (1997), reviewing the work reported in a conference devoted to the analysis ot order a n d disorder in talk, felt able to conclude that the applicability of C A is mm uncontentious. Given all this, w e can at least a p p r o a c h the talk of people with a learning disability with some faith that they will orient to the same communicative regularities in interaction as d o people without such a disability.
The Scripted Q u e s t i o n s T h e standard, psychometric, psychological account is that quality of life is a.-, publicly measurable as any other psychological state a n d b y the same public means. Let m e say that in this article, I take n o position o n whether the construct "quality of life" is a g o o d or b a d o n e n o r o n whether it serves this or that political purpose a n d so on. M y only interest is in comparing the official description ~f i s mpant rn h a n n e n in the interview with exacdy what does happeD
ANTAKI
3
INTERVIEWING PERSONS WITH A LEARNING DISABILITY
(so far as we can tell from what the audio transcript shows us of the participants' own orientations to the action) a n d in saying something about the w a y the inter viewers' unofficial and unstated practical reasoning allows, or prompts, a certain pattern of treatment to emerge a n d a score to b e recorded. Most quality-of-life assessments involve s o m e kind of interview o r ques tionnaire (Hughes, H w a n g , Kim, Eisenman, & Killian, 1995). T h e particular questionnaire whose administration I shall b e looking closely at in this article is the Quality of Life Questionnaire ( Q O L . Q ) (Schalock & Keith, 1993; see also Schalock, Keith. Hoffman, & Karan, 1989,1990, for its development; R a p l e y & Lobley, 1995, for its acceptable reliability for a British population). It is a routinely used, psychometrically validated, unexceptionable addition to the psychologists' battery of tests, and according to the psychologist responsible for its selection in the interviews we shall see below, it was chosen a n d administered without fuss or second thought. 1
What i s t h e "Official" A c c o u n t of t h e I n t e r v i e w ? A word is in order o n what I take to b e the official version of what this interview should be like and against which I shall b e offering a comparison from the inter actional transcript. T h e introductory p a r a g r a p h o n this particular instrument reads as follows: The Q O L . Q , m a y b e administered to persons with mental retardation who have an adequate receptive a n d expressive language. T h e examiner needs to be sensitive to the possibility that the respondent m a y n o t under stand some of the items or the meaning of some of the words. If this happens, it is okay to paraphrase the item to improve understanding. If this happens frequendy, or if the person is k n o w n n o t to h a v e adequate receptive o r expressive skills, it is acceptable to h a v e two persons w h o k n o w the indi vidual well complete the questionnaire. And after a paragraph of instructions to b e read to the respondent, the instructions to the interviewer continue thus: If the respondent consents, t h e examiner proceeds to administer t h e 4 0 items. W h e n reading the items, p a y close attention to the exact word ing. You m a y paraphrase items a n d repeat t h e m as often as necessary to ensure the respondent's understanding of the item content. As with any set of instructions, there is a great deal that remains implicit. Perhaps the most important omission is t h e lack of instruction about h o w to "paraphrase items a n d repeat them." It would s e e m to b e of crucial importance thai this be channeled along some principled guidelines, b u t n o n e are provided. In other words, if interviewers depart from the script (as w e know interviewers do from the work of HoutknnT.-StPf»nctr l o o * i n n c M - r - ^ a
n
4
D I S C O U R S E STUDIES
S u c h m a n &J o r d a n , 1990), then what they will b e doing is n o t adhering to some strict psychometrically correct procedure guided b y mental-heath theory (indeed. GarfinkeFs [1967] argument is that such procedure is a chimera) b u t rather reveal ing some combination of their "training'' and their unofficial folk craft in designing questions for these recipients. I shall try to show that the w a y they design their deviations is in the direction of what might b e called sensitiveness a n d that this can h a v e the effect of raising the respondent's score without official warrant
The Data T h e data are transcripts of 10 administrations of a questionnaire (as noted above, the Schalock & Keith [1993] Q O L . Q ) in a routine Psychological Services "audi!" of people in supported housing in a city in Northern England (for a fuller descrip tion of aspects of these data n o t relevant h e r e , see Antaki & Rapley, 1996a. 1996b; Rapley & Antaki, 1996). Respondents were all former residents of a local long-stay institution for p e o p l e with learning disabilities (in U.S. termino logy, mental retardation) a n d h a d b e e n previously classified as having moder ate or mild learning disabilities. Interviews w e r e conducted i n the participants" n e w h o m e s . All participants w e r e asked for, a n d gave, their consent to the tape recording. I shall concentrate o n one, typical interview between the psychologist and a m a n pseudonymised as M A a n d bring in examples from other interviews as) go along (each interview was conducted b y o n e of two psychologists, and I shall n o t distinguish between them). Perhaps I should m a k e clear that the persua siveness or otherwise of m y case rests o n h o w well or badly I unravel the workings of scripted question delivery. It is n o t primarily a statistical demonstration done b y counting u p categories across cases; it is an argument that this is the kind of distortion that script delivery suffers as the interaction progresses. Nevertheless, there will b e a couple of points in the argument w h e r e some brute statistics will h e l p , and I shall provide t h e m then.
Analysis I shall turn first to an exemplary question-and-answer pair, t h e n go through the range of such pairs as they increasingly deviate from the printed schedule. I will quote the official version of the question a n d the official response alternatives to m a k e points about discrepancies b e t w e e n the official a n d the live versions. The notation used is derived from Gail Jefferson's conventions, which are widely adopted in C A (see Atkinson & Heritage, 1984, p p . ix ff) a n d which are set oui in the appendix. I n the extracts presented h e r e , all personal n a m e s have been changed, as h a v e most place n a m e s a n d any other details that might identify the respondent.
ANTAO
INTERVIEWING PERSONS WITH A LEARNING DISABILITY
5
A n E x a m p l e o f S e n s i t i v e Paraphrase The following is a good example of the sort of paraphrase that reworks the of ficial question into o n e that is easier for the respondent to answer: Qui scion 21: How did you decide to do the job or other daily activities you do now? / rhosc it because of pay, benefits or interests (3), Only thing available I could find (2), Somonr else decidedfor me (1). Actual delivery and response: 40!> 3 sees, including paper rustling) 4 10 Interviewer, how do you (.) decide to i d o what (.) you > d o at 411 the moment< (,)°Tyeh? 412 MA (well) clean Tup 413 int. yeh? (.) d'y (.) d'y do what you do bec'z (...) you 414 ihave to: (.) or because you"re1tbld to (.) or cos you W a n t to do it 415 MA to do it 41fi Im. hah? 417 MA tyeh r 418 Im. icos Tyou want to do things 419 MA yes 42i» Int. °yeh?° 421 (2 sees) 422 Int. ok (sniff) [moves onto next question] Kttorded score: 3 Notice here that the official question is about paid work, b u t the interviewer has previously established (in earlier lines 177-180, n o t r e p r o d u c e d in this article) that NLA is "about 70" years of age (and thus older than retirement age). T h e interviewer has to offer a suitably r e w o r d e d version. This comes out as " h o w do you (.) decide what (.) to > d o at the m o m e n K " y e a h ? " (line 410). T h e previous material does not seem to offer m u c h of an indication as to what such a pro-verb (as Sucks called t h e m ; see 1992, Vol. 1, p p . 342 ff) as do might stand for h e r e (or even its phrasal version do-at-the-moment, delivered in a burst); so this is rather an underspecified paraphrase. M A has an attempt at specifying it with "(well) d e a n up," and the interviewer recycles the question in a further paraphrase, which nevertheless still uses t h e pro-form do what you do. We might w o n d e r whether MA's turns at 415 (an echo of "to d o it" from the interviewer's previous turn , 417 ("yeh"), and 419 ("yes") d o constitute the score 3 a n d the sentiment "1 chose it because of pay, benefits or interest" that it is m e a n t to stand for; b u t for the moment, the thing to notice is the paraphrase. 0
Dijkstra, van der Leen, a n d v a n der Zouwen (1985) report that only about a third of questions in standardized interviews (with respondents without a learning disability) get a literal delivery. Things are rather worse in the interview with MA: Only 3 of the 2 8 question-and-answer pairs come u p to the official presump tion of strict literalness without paraphrase.
6
D I S C O U R S E STUDIES
This is a point at which some further gross statistics might b e helpful. To get a sense of h o w often in the entire 10-interview corpus questions deviated from literalness, I counted the times a question was delivered in a w a y that would count c o m m o n sensically as word for word (a technically rather problematic phrase), nearly word for word, and, finally, not word for word. I hasten to say thai each of these terms would b e a r a great deal of examination, b u t the purpose 1 want to m a k e t h e m serve here is exacdy the sort of rough-and-ready judgmeni that (I imagine) would b e m a d e b y the interviewers themselves while actually delivering the questions and without time for formal reflection and drafting. T h e criteria I used for literalness w e r e that the item stem a n d its response alternatives b e delivered w o r d for word, with allowance m a d e for some small degree of filled pausing ("urns'' a n d "ers") a n d self-repair ("that h a p - happen to you"). M o r e generously, I allowed the substitution of synonyms or synonymous expressions (for example, substituting the w o r d work for employment, or a bit satisfied ion somewhat satisfied). I counted as "nonliteral" anything grosser: This ranged from reducing the n u m b e r of response alternatives to wholesale para phrase of the entire question. T h e first pass is to see what is m a d e of each question the first time it is asked. N o n e of the 10 interviews h a d m o r e than 3 exactly literal first-time deliveries of question-stem and response alternatives, a n d most (6) h a d only o n e such delivery. O n a m o r e liberal - but of course psychometrically discouraged - allowance of n e a r literalness, each interview would yield about 3 m o r e such deliveries. Bui even with this allowance, the most word-perfect interview h a d n o more than 8 literal or nearliteral deliveries out of 3 3 occasions o n which a question was asked for the first time, a n d the m e a n , over all 10 interviews, was about 13°/o. or 1 question in 8. H o w literal w e r e questions w h e n t h e y w e r e r e p e a t e d ? T h e textbook instruction is that in response to unsatisfactory first answers, "the interviewer's j o b is to explain to the respondent that choosing one answer from the list is the way to answer the q u e s t i o n . . . a n d to read the list of responses again" (Fowler & Mangione, 1990, p . 39). But this virtually never h a p p e n e d in this set of interviews. If we took a purely numerical count of all questions in the interviews (on a crude categorization, including first deliveries, repetitions, subsequent versions, corrections, a n d so on), the proportion of literal or near-literal versions of what is o n the official schedule drops to about 2%. I n other words, the conversationanalytic cliche of s o m e p h e n o m e n o n b e i n g "massively" the case works oui here, for nonliteral delivery of a scripted question, to something like "98% ol the time." I n other words, then, a n d speaking only of this set of interviews, an interviewer can b e expected to pose people with learning difficulties literal, scripted, psy chometrically p r o p e r questions only very infrequently; deviation of o n e kind or another is the n o r m . N o w it is worth r e m e m b e r i n g that, as we saw in the instruc tions a b o v e , this particular questionnaire disarms t e x t b o o k injunctions to literalness b y allowing its administrators to paraphrase the items. But it does so without providing any guidance at all about what constitutes a proper or improper
ANTAK)
INTERVIEWING PERSONS WITH A LEARNING DISABILITY
7
departure from the wording o n the page. T h a t is left to the interviewers' dis cretion, so what they actually d o say is a matter of s o m e interest.
How Sensitive Q u e s t i o n s L o w e r t h e Bar to S o c i a l l y Positive A n s w e r s Let us move on n o w to extracts that depart m o r e m a r k e d l y from the official script and in which the lowering of the criteria for a high score is still m o r e exaggerated. Note, however, that I do n o t m e a n to suggest that every deviation from the script can usefully b e characterized as o n e that lowers the b a r to a socially positive answer, only that this is a c o m m o n a n d striking pattern i n the data and one with quite serious consequences, as I shall try to b r i n g out below. I shall also start to bring in m o r e evidence from other interviews in the set just to underline the point that w h a t I a m saying is n o t s o m e eccentricity o n the p a r t of this interviewer or this interviewee. T h e extracts from the interview with M A always come first. In the first case we look at, there are two things w e h a v e already seen (the simplification of the response alternatives a n d the interviewer's paraphrase). The feature I shall m o v e o n to n o w is t h e interviewer's nontake-up of possibly relevant material offered b y the respondent. Nonofficial Material Apparently Ignored S.- lpied Question 24: How much control do you have over things you do every day. like going to bed, eating, and what you do for fun? complete (3), some (2), .:tffr(I).
Actual delivery and response (in part; we shall see its later development after discussing this segment): 52 i 522 523 52.4 525 526 527 528 529 530
Int.-> MA Ini. MA lnt.-> MA Int. MA
how Tmuch:: (.) conitrol (.) d'you have (.) over the things you i d s [°yeh?° [(eh?) how much control do you have (.) over what you Wo what (.) T'e::4re? twell (.) (li-) (.) Tyeji (..) wha- (.) things that-like going out (.) an' (.) when you get up. an' well I (go out) (.) on a T\Vgdnesday::= =yeh up TMerchanUman
Recorded score: 3. On its first delivery (at 521-522), the interviewer's question is hearable as having been concluded at the end of the m a i n clause ("the things y o u do"). So MA's "(eh?)" might b e heard as a pass or a request for m o r e . T h e interviewer recycles the question from the beginning b u t again only reaches the e n d of the m a i n clause now slighdy altered to "what you do"). Again, M A offers something, and this time his "what (.) "e::re?" is a specific candidate for clarification
(understandable as whether the question is about control over things "here* in his current, nonhospital, home). This would b e in keeping with the spirit of the scripted question. But notice that the interviewer's take u p is prefaced with a "well," signaling a disagreement (compare the immediate a n d emphatic "yes. h e r e " h e could h a v e said). H e goes o n to diverge from MA's offer and reorient thr question away from the current h o m e , with the dis-preference markers of "well" a n d two repaired false starts ("(H) yeh (..) wha-") a n d a correction ("things thai like going out"). T h e phrase "an' w h e n you get up a n " that h e adds is a partial m o v e back to the domestic r o u n d that is specified in the official question, but ii is n o t oriented to b y M A . In other words, the interviewer's delivery (and the respondent's uptake) deletes the specific domestic items of the official questionnaire in favor of leisure activity and, for leisure activity, allows the very specific going to (what is recognizable as the p u b to stand for the range of things the scripted question is meant to ask about (the "things you d o every day"). This is in line with the deviation we saw above, where the question was delivered not as in the script. H e r e , a still stronger case can b e m a d e , I think, for the consequentiality of the deviation for the score that the psychologist enters. If w e are to understand (as w e are probably meani to) b y the scripted phrase a general appreciation of all of life's m u n d a n e decisions, it would b e easily argued that this narrowing alone renders the question a poor delivery of the s c r i p t M o r e to the point, it is another example of lowering the bar: T h e respondent can score highly b y controlling one thing in his life (going to the pub), even though the question asks about his life in general, a lowering that would b e curious if offered to a person without a learning disability. But there is m o r e trouble ahead, and it is in the remainder of what M A gives. O r , to put it m o r e finely, it is in the material that M A offers after the interviewer has offered what might b e hearable as the termination of the topic and a pre paration to m o v e to the next. N o t e how, at line 535 below, the interviewer offers "right right." I n the circumstances of the interaction - without a third party mechanism for turn allocation (e.g., a chairperson, as i n the selection inter view discussed in Button, 1992, or the political meeting discussed in Larrue k Trognon, 1993) - it is the interviewer w h o disposes of the old topic as suitablv answered a n d w h o can propose the n e x t one. So his "right right," like the other terminators w e h a v e previously seen, can b e oriented to as the close of a chapter a n d the alert that h e is about to inaugurate a n e w one. But note that MA does n o t allow the interviewer to k e e p the turn: 531 532 533 534 535 536 537 538 539
Int. MA Int. MA Int.-» MA-* Int. MA
right Tso. (.) are you in charge of doin'that? lyes yeah? (..)lyou don"t have to ask anyone if you can Tno no right right (see) on a Thursday (..) I go to't (...) club mhm (thev- thev) complained said I- >not s'posed to (go in) in< club (.) they said (you're-you not the boss) (..)
ANTAKI
540 541 542 543 544 545 546 547 548 549 550
INTERVIEWING PERSONS WITH A LEARNING DISABILITY
ini.
MA 1m. MA Ini. MA int.
9
fHHhhh (..) right so Tvou"re in charge [(rustle) (rustle) (cough) If I (syll) club(syll) on a 4-Thursday I Twould "right nothing to do with them °ok°= =Tis iit? Tno:.4- I think you're Tright (2 sees, including some rustling) Tok(...) [Int. moves onto next question] 0
The interviewer solicits a yes or n o answer to a pointed version of the scripted question at 531 a n d repeats a version of it at 5 3 3 . T h e answers M A gives are consistent, and the interviewer's "right right" at 5 3 5 m i g h t indicate the point at which he makes an entry on the response sheet (a 3, standing for complete [control]). Bui leaving aside the discrepancy we n o t e d above between the range of activities in the script a n d the n a r r o w focus of the delivery, what d o we m a k e of the material M A offers at 536? S o m e aspects of this stretch of talk h a v e b e e n analyzed elsewhere (Antaki & Rapley, 1996a), a n d a case can b e m a d e that here MA is embarking on a troubles telling (Jefferson, 1988), which the interviewer declines to accept as troubles recipient, a n d nothing c o m e s of the story (which seems to hinge o n M A b e i n g treated badly b y people at a club: lines 5 3 8 - 5 3 9 , "they complained said I-not s"posed to [go in] club"). T h e interviewer's specific use of a "so" initiator in his u p t a k e offers w h a t is to b e taken as a formulation of what the respondent has said, b u t the interviewer's formulation is a patently nomroubled account: It puts it that M A is, in fact, "in charge." This effectively rules the trouble out of b o u n d s as a pursuable topic unless M A corrects the interviewer, and we might r e a d 5 4 3 a n d 545 to b e an attempt at correction; b u t the interviewer's assessment of it with a near-idiomatic platitude ("no I think you're right," line 548) is a hearable shutdown (see, as above, D r e w & Holt, 1988, on idioms as termination points of trouble tellings) a n d n o n t a k e - u p . Moreover, the "so" formulation works as a declarative yes-no question, with the projected "yes" answer b e i n g positive; this is a n efficient way of eliciting a "happy" answer from the r e s p o n d e n t (and there is m o r e on this effect of yes-no question design in Houtkoop-Steenstra & Antaki, 1998). The record will show that the interviewer has just r e c o r d e d (or will at s o m e point record) the most positive entry of 3 for this answer, yet it would n o t b e im plausible to hear the trouble tale M A tells (about something like exclusion from a club) as being inconsistent with the complete control over "things y o u d o every day" that a 3 represents. This is the larger point that I w a n t to m a k e : T h e inter viewer is recording a high, positive score in spite of something that is at least reasonably possible to hear as n o t positive. W h a t might b e in operation is, again, the interviewer's u p w a r d recalibration of what would b e , for a person without a learning disability, unsatisfactory.
10
D I S C O U R S E STUDIED
H e r e is another e x a m p l e of h o w answers to this question attract a different lower standard for what counts as a positive answer about the degree of control the r e s p o n d e n t can exercise in his or h e r life: CodeMT/MR/JW >can I just come back to< the:se (.) Anne? Yeah? 605 I (.) how much [control do you have over things you 606 607 AN [(syll syll) do (.) every day like (.) going to bed (.) eating meals 608 I (.) and what you do for Ifun 609 yeah they tdo let me eat meals (.) they make Tinice 610 AN meals for us 611 the staff 612 I the Tstla:ff 613 AN (.) ok (.) who who decides (.) when you go to bed and (.) 614 I what you have to eat 615 (don't) be TsuicJc 616 AN 617 I yeah (.) do you (.) do you decide those things (.) or do you (..) ask the Tstaff [or618 619 AN [(yis) I Task the istaff or do Tthey tell -lyou what to do 620 I dd I Thelp -tem 621 AN-> 622 I yeah? 623 (3.5) be Tgood to iem 624 AN-* 625 I (.)4-good .. 626 AN-» I Tmop fer -Lem yeah 627 I mop the tkitchen iout 628 AN-> 629 I tgood (when they put) swept it Tiu:p 630 AN-* Uright 631 I (.) all the Ubi:ts 632 AN yeah you do those things 633 I (.) (yis) I idfli 634 AN that's Tgood (.) sa lotta Twork Tisn't lit 635 I I've Ttooime iTpi:lls 636 AN (.)°right° [T(when-)} 637 I [I've Thad me 4Tpi:lls 638 AN 639 I good I Tlad them at Itwo o'Uclo-.ck 640 AN right ((clears throat)) 641 I 642 AN me pills (Tzeppelin) and me big iwhite Tone 643 I °hh oh° I idid 644 AN (
0
Recorded score: 2 (some control) Note the familiar pattern: paraphrase a n d reduction of the question, pursi of a n answer in the face of what is treated as an irrelevant response (in line 61i
ANTAKI
INTERVIEWING PERSONS WITH A LEARNING DISABILITY
a
and disarxention to A N ' s concerns about n o t b e i n g sick (616) a n d taking h e r pills (636), which envelope the account of h e r "work." Again, the interviewer has to reduce all this into a single score, and she opts for 2 - some control. We might see this as another example of upgrading. T h e interviewer seems to treat A N ' s menial domestic duties as evidence that she has a middling a m o u n t of control. They might not seem so for a person without a learning disability. Let us see another example of lowering the bar, this time in the w a y a question is "sensitively" edited right at the outset. Question Edited Right at the Outset Question 37: Do you participate actively in those recreational activities? Usually, most ofthe time (3), Frequently, about half ofthe time (2), Seldom or never (1) Actual delivery: CA 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
KK/TH I
JO 1 JO I JO 1
attChrist_mas(.)righ"(.)oktthen.hhher::: (3 sees) °.hh right er::m° (2 sees) >so (Twhen) you're at - when you're< at your iparties? (..) Twhat d'you 4-do (.) -» (>"ve a b"t of a<) (..) have a >bit of a< -Ufa-ink do you? yea [hh: [eh hehh hehh hehh hehh Tdo you ilike your drink (.) John?= =yeah heheh (.) twhat d'you 4-drink I Tager heheheheh (2 sees) [moves on to next question]
We see the same pattern: T h e interviewer m a k e s a colloquial, folk-theoretical guess, thau forJ o h n , what it m e a n s to "participate actively in those recreational activities" is to "have a bit of a drink." N o t only has the interviewer m a d e the question psychometrically u n r e c o g n i z a b l e b y n a r r o w i n g its field, the n o n officialness of the language m i g h t encourage an answer different from w h a t might be elicited b y the formal version. I n a n y case, it is another example of lowering the bar to a positive response in a w a y that would b e o d d , a n d p e r h a p s unacceptable, to a person without a learning disability whose horizons for "par ticipating actively in those recreational activities" would b e wider t h a n "having a bit of a drink." In the last illustration of lowering the bar, let u s take a m o r e lengthy extract to see the interviewer's conflation of several questions.
DISCOURSE STUDIES
12
Questions Conflated Official Questions 36, 37, and 38: Question 36: How often do you attend recreational activities (homes, parties, dances, concerts, plays) in your community? 3-4 per month (3), 1-2per month (2), less then 1 per month (1) Question 37: Do you participate actively in those recreational activities? Usually most of the time (3), Frequently, about half of the time (2), Seldom or never (1) Question 38: What about opportunities for marriage? lam married, or have the oppor tunity to date anyone I choose (3), I have limited opportunities to date or marry (2), I have opportunity to date or marry (1) Actual delivery and response (This is rather a long extract, used to make the point that Questions 37 and 38 are not explicitly delivered): mm_hhm(..) l°right° .hhh td'you urn (...) 775 Int.-> Thow often do you go out and Tdo:: (.) ^things like t g o 776 (ter) (..) do your idancing an" [go to clubs like that 777 [t(well) 778 MA (w'U) I go out on TSaturdays 779 MA yeh? 780 Int. Saturday Tmorning 781 MA every Sat [urday? 782 Int. [(t'ad thi- t'ad thi-) (.) 'e 'ad 783 MA this T(f t irain) I don't (know syll) 784 "right" 785 Int. (1.5 sees) 786 he w's ifive foot Tfour (...) °yeh° (..) 787 MA eh? (..) >I mean you- cos you go to TfaQlball matches 788 Int. as TwelK don't you? 789 t a y e il'm Tthe::ire 790 MA-> yeh? (...) oh you twatch it on the liejly 791 Int.-> tye::s 792 MA what (.) if it's training 793 Int. Too:h (syll syll syll tlousy) (.) [it was 794 MA [eh?= 795 Int. do you Tgo an' watch 'em (..) d'y- (.) Tgo 796 Int. and watch 'em pi [ay 797 [I Twent (.) I went at ex(...) TOld Trafford 798 MA-> yeh(..) 799 Int. watch cricket _match (..) [they got back at T(ever syll 800 MA [right 801 Int. , syll syll (..) Tsyll) (..) 802 MA d'you watch the TTest TTest match on the Ttelly? 803 Int.-> T(y'h) 804 MA yeh? (..) T aven' done very Twell 'ave we? 805 Int. Tno:: = 806 MA =hhhehh [hhhhh 807 Int. [T(d'you play cricket? 808 MA Tmmm (.) a ilitde bit (.-) y [eh? 809 Int. [(d'you Tsyll isyll syll?) 810 MA
ANTAKI
INTERVIEWING PERSONS WITH A LEARNING DISABILITY
811 1 ni. 812 MA 813 int. 814 MA 815 Int. 816 MA 817 int. 818 MA 819 int. 820 821— 822 823 .M\ 824 Int. 825 MA 826 827 Int. 828 MA 829 830 int. 831 MA 832 Int. 833
MA
831 835 836 837 838 839 840 841 842
Int. MA im. MA Int. MA Int. Int.
843 MA
844 Int.—» 845 MA 846 Int.
13
Tno:: I don't play Tfootball (..) I n o no d'you Tknow W h y (.) you tknow Sharon ttold I m e : uhhuh? you say y'r getting too Told (..) tyou're getting too old? no Tyou're tch- she- -l(said) (.) that (t'get) Tyour (syll syll) H H H H H J i (..) w'U I >don't think< I'm too old 4-yet (..) .hh (sniff)= (cough) (loud cough cough sniff) hh .hhh (2 sees) °io:Tk:::° (...) and yer(.) yer Tkeen on yer idancing aren't you? yes= yeh? (...) so y- you Treally get Tinto that ye:h (.) they 'ad a Tparty down the:re an':: 4-like TChristmas car-Ter iparty mm and er't- er (..) the T(syll) comes round to check Te:re (...) (syll syll syll syll syll).hh (loud cough) Twhat's bad for you? (..) Tthis dancing? Tmm inahit'stnot ('es Tright syll syll syll syll syll syll down Tthe:re yeh? (loud cough) come back at nine o'clock at (.) -Inkght. m m (..) [(syll syll) [(syll syll (.) Ttaxi 4-comes an': Tpicks you Tup an" Ttakes[you Tback [ yeh (w'll) it's good exercise though isn"t it? Teh:: y'g t t o u t and meet people? ToohJ-aye yeh? (..) (sniff) (..) °oik° [Int. moves onto question 40] 00
00
e
Recorded score (for each of the three questions): 3 The interviewer's initial question at 775 is a version of t h e official question (Question 36. listed above), which shows the reduction a n d transformation w e noted above: It d r o p s the alternatives, it p a r a p h r a s e s into m o r e colloquial language, and it nominates a singular particular example. T h a t m u c h replicates the manner that has replaced the scripted version as standard as the interview has progressed. Of interest to us here, though, is the way in which the develop ment of the answer (through a series of specifying p r o m p t s and MA's responses) gives the interviewer license to enter positive scores n o t only for this question but also for the related Questions 37 a n d 3 8 , which h e does n o t explicitly ask. The motivation for t h e conflation might b e to simplify t h e questions, p e r h a p s by the avoidance of repetitions that violate c n n v p r s a K n n a i im^u™^ i—^ -
14
DISCOURSE STUDIES
I h a v e b e e n arguing throughout, the way the simplifications are worded has the effect of lowering the b a r to a positive social response. Let us deal first with the cognitive account of the simplification. I n scripted interviewing, the interviewer has little institutional r o o m for m a n e u v e r a n d must deliver the questions in series, even though this might violate conversational implicatures and occasion an interactional difficulty: for example, w h a t m i g h t b e h e a r d as the responded being "puzzled." H e r e is an example from Houtkoop-Steenstra's (1996) transcript of a telephone survey: (The respondent has offered, 10 turns back, that he is "in the military") 37 38 39 40 41
I:
R-» I
.hh Tokay now we'd like to ask about your employment TSTATUS (.) did you do any work for TPAY last week? did I do any work for pay? [yeah [mmhmm (...)
42 43 44 45
I R->
A:nd what was your TjOB last week that is what KIND of work were you doing .hhh uh:: I'm a platoon leader in the::- (0.5) ARMY (...)
We can see that R's response at 40 is prefaced with a check, a n d his response at 4 4 is prefaced with an intake of b r e a t h a n d a n "uh::.," Both of these might bt h e a r d to m a r k s o m e kind of nonsmoothness occasioned b y the questions, and >' would n o t b e implausible to attribute it to the interviewer's violation of a Griceai: m a x i m (perhaps quality) in repeating a question to which the answer has already b e e n offered (that h e is "in the military"). N o w let us see whether this might h a p p e n in our data. T h e official questions ask h o w often the respondent "attends recreational activities," whether they "participate in those recreational activities," a n d whether they have "oppor tunities for marriage." As to the first of these, w e see M A offer at 779 "(w'll) I go out o n Saturdays," which is not obviously a n answer to the question posed. His subsequent turns are n o clearer. N o w (at 788), the interviewer actually nomii! ates a specific activity (going to football matches), which is a concretizadon (and, as we have seen before, a narrowing) of "recreational activities." The answer seems to b e that M A does ("aye, I ' m the::re"), b u t for reasons that are not obvious (but could b e that M A is h e r e pointing to the television set, or television room, as "there"), the interviewer further narrows the activity to the m o r e sedentary "you watch it o n the telly." T h e same lowering of the b a r - in spite of MA's men tion of m o r e actively going somewhere - h a p p e n s for cricket: At 798/800. MA says h e "went e n (...) TOld Trafford [...] watch cricket m a t c h " (Old Trafford is a stadium w h e r e cricket internationals - "Tests" - are staged), b u t the interviewer recalibrates this d o w n w a r d as "d'you watch the Test match o n the telly?" The interviewer uses MA's answers to these unambitious questions - rather than die m o r e active versions that M A himself offers - as license to enter the maximum positive score of 3 .
AHTAKI INTERVIEWING PERSONS WITH A LEARNING DISABILITY
15
As to the question o n participation, it is possible that t h e interviewer h a s come to tin? judgment that such a question would b e inappropriate for s o m e o n e who is (as was established m u c h earlier in the interview) a b o u t 70 years of age. Perhaps it would strike h i m as m e r e l y absurd to ask whether M A is a n active sportsman (at 808, it is M A w h o asks t h e interviewer if h e plays sports). This, of course, is plausible, but we might w o n d e r w h y the m a x i m u m score of 3 is entered. As to Question 3 8 (opportunities for marriage), the answer to this is (apparently) conflated into the same stretch of talk (at least, n o other m o r e obvious material appears anywhere else), perhaps o n the folk grounds that marriage opportunities are well indexed by the degree a n d scope of one's leisure activities (even though, here, these are narrowed b y t h e interviewer to dancing, attending cricket a n d football matches, or watching t h e m o n the television). I t is possible also that the interviewer is. still in a folk way, sensitive to a threat to MA's face a n d is reluctant to ask a question that might b e u n d u l y sensitive. I n any case, there is nothing in the transcript that approximates the official question, unless it is line 844: "y"get out and meet people?" Yet, M A is scored as 3 : lam married, or have the opportunity to date wh') f choose. We might just pick u p this orientation to Question 3 8 as a sensitive topic i n another interview:
MT/KK CL 363
1
364
yeah (.) "iright oTkay°.hhh (..) Her:::m (...) and you've Tgot a Tb_Qy.ifriend Thaven't 4-you
365 CL
yeh
366
2
aheheheh (..)* ? what's his iname again?* ((*high pitch*))
367
CL
Steven
368
I
oh Tthat's 4-right (..) > (so) have you been
369
together for a < 4-long time?
370
CL
yeh(...)
371
1
°oh that's nice" .hhhh (..) bright" (..)
372
so [moves to next question]
Recorded score: 3
|
The interviewer at line 3 6 4 p r e a n n o u n c e s a question about CL's boyfriend. This is intelligible as a version of Question 3 8 ("What about opportunities for marriage?"). Note, again, the familiar paraphrase of the question and its reduction to domestic dimensions, a n d especially n o t e that the interviewer shepherds the answer in the positive direction. By soliciting agreement that the respondent has a boyfriend, she prefigures a "no-problem answer" (Heritage & Sorjonen, 1994). She then asks his n a m e a n d treats it as already k n o w n to her. This, again, might be a matter of the interviewer orienting to t h e sensitivity of "opportunities for marriage" to this respondent: She casts herself as fully appreciating the respond ent's positive interpersonal situation - she knows h e r boyfriend's n a m e . T h e "so* in line 368 to 369 ("so h a v e y o u b e e n together for a long time?") cements the question as having p r o c e e d e d from a n agreed premise (that the respondent does indeed have a boyfriend) because, as Houtkoop-Steenstra (1997) points
Hi
DISCOURSE STTJD1ES
out, so in such contexts treats the presumption as factual. I n doing all of this, however, the official scripted question, a n d its three alternative answers, is ob literated in favor of a question with a lower bar o n which the respondent scores m a x i m u m points.
Discussion W h a t I wanted to d o in this article was to use the apparatus of C A to examine the administration, b y professionally trained interviewers, of a scripted ques tionnaire. These administrations yield quality-of-life scores that are now on the respondents' official records. T h e score is consequential b o t h for the respondents (who might b e resettled in a different e n v i r o n m e n t or r e t u r n e d to the institution they came from) a n d for the social services authority that is responsible for them (whose provision of quality of life might b e c o m p a r e d with other social service authorities). It is, apparently, in everyone's interest that the questionnaire be delivered faithfully to reproduce the item validity of the questions and not 10 i n t r o d u c e t h e sort of e x t r a n e o u s i n f o r m a t i o n t h a t p s y c h o m e t r i c i a n s casi as "noise." W h y , then, d o w e find only a very few question-and-answer exchanges (at best, about 1 in 8; at worst, about 1 in 50) that seem to reproduce the script with high fidelity? A n d w h y d o w e find consistent deviation from the official scrips b y simplifying the response alternatives, or paraphrase, or the offering of a folk version of the question or conflation of several questions? T h e start (but n o t the end) of the answer is, of course, that the script is i fiction, a n d interviewers h a v e to use their practical reasoning to deliver its quev tions in real talk. C A work o n interviews from S u c h m a n a n d J o r d a n (19W through Marlaire a n d M a y n a r d (1990) to Houtkoop-Steenstra (1996) shows thrsi in delivering scripts, interviewers balance two competing motivations: on onh a n d to play square with the psychometric prescription to b e literal and on the other h a n d actually to deliver the questions in an interactionally sensitive wav Some of that interactional sensitivity is foreshadowed a n d licensed by the interview schedule instructions, b y the official diagnosis of the respondents K< learning disabled, and b y the manifest confusion a n d misunderstanding in (sonr. of) their responses. But there is m o r e to it than that. We n o t only see the inter viewers design questions to lower the cognitive difficulty of the task, but the\ design questions to lower the b a r of what will count as a socially positive answer T h e interviewers mobilize w h a t Garfinkel (1967) calls their unstated advisors rules or what M a y n a r d a n d Marlaire (1992) call their unassessable interactional skills and turn the official questions into ones that the respondents can copt with. Thus, a formal question such as "do you participate actively i n . . . recreations, activities?" becomes "so (Twhen) you're at- w h e n you're at your parties (..) why d'you d o (.) f v e a b't of a) (..) h a v e a bit of a i d r i n k d o y o u ? " O r an answer like 1
"I help ihem [...] b e good to t h e m [...] m o p for t h e m [...] m o p the kitchen out]" is taken as evidence of some control over "things y o u d o every day, like going to bed, eating, and what y o u d o for fun." Such questions assume that the respondents cannot answer the " p r o p e r " or "normaF questions. Is lowering the social b a r in this w a y understandable as motivated by generosity or p e r h a p s b y politeness? It is possible; certainly, Houtkoop-Steenstra (1997) suggests that a good n u m b e r of the distortions she finds in survey questions might b e accountable as illustrations of interviewers' attention 10 respondents' face, in Goffman's (1967) terms, a n d w e also find evi dence in this very set of data that y e s / n o questions can get designed so as to promote the most positive alternative o n the official list of options (HoutkoopSteenstni & Antaki, 1998). O n e might say, then, that to lower the bar is generously to head off the threat that the respondents' answers will embarrassingly fail to reach even the lower of the rungs o n the ladder of alternatives. But to assume the worst in that w a y is a curious infection of a technical, 'clinical" exercise that is supposed to assume nothing. Moreover, to assume the worst is arguably against the respondent's interests. If a respondent's "high" score is achieved only because t h e hurdles are set low, t h e n h e or she ends u p with a score that is ostensibly m e a s u r e d o n questions of " n o r m a l " difficulty but actually given in response to easier ones. T h u s , a r e s p o n d e n t might score high on recreational activities just b y watching television or o n opportunities for marriage by agreeing with the question, "y"get out a n d m e e t people?" a n d so on. The questions are revised, a n d the answers edited, to m a k e a j u d g m e n t not about the respondent's quality of life b u t about the respondent's quality of life in the circumstances. But those circumstances disappear w h e n the official score is entered in the records; then it b e c o m e s the score o n the questionnaire as it is printed in black a n d white, with its questions of "normal" difficulty. Maynard and Marlaire (1992) found children scoring b a d l y because of the ways the interactional d e m a n d s of the interview b r o k e through its psychometric floodgates: These interactional d e m a n d s were "good reasons," they observed, following Garfinkel (1967), "for b a d testing performance." T h e interactional de mands at work in these psychological quality-of-life interviews are just as powerful but this time with the converse effect. R a t h e r than work to decrease the subject's score, sensitive questioning tends improperly to increase it: Using unofficial, low, folk criteria is. in other words, a b a d reason for g o o d testing performance. I said at the outset that I would take n o position o n the political debate about quality of-life testing. T h e conclusions of this study are nevertheless suggestive. They are. of course, based o n o n e set of interviewers, with one set of respondents, and using one interview schedule. Nevertheless, the data offer evidence that questions are routinely transformed in such a way as to increase the probabil ity of higher scoring to less ambitious questions. If there is to b e a debate over whether interview testing attends properly to the interests of the respondents, then such a finding might contribute to it.
18
D I S C O U R S E STUDIES
Appendix Transcription notation: The transcription symbols used here were derived from those developed by Gail Jefferson (see Atkinson & Heritage, 1984, pp. ix-xvi). (•) (-.) (-) .hh hh hehh hahh Go(h)d [*t] (guess) culo:ng (syllsyll) run= =on
(.) is a pause of about .2 of a second; (..) is about .5, and (...) is about .8 of a second A dot before an "h" denotes an in breath. The more hs, the longer the in breath. An "h" denotes an out breath. The more hs, the longer the out breath. Laughter syllables with some attempt to capture "color" (h) denotes "laughter" within words dental click Roman text within brackets is the transcriber's guess at an unclear part of the tape. A dash denotes a sharp cutoff of a prior word or sound. Colons show that the speaker has stretched the preceding letter or sound. The more colons the greater the extent of the stretching Unclear speech rendered as approximations to number of syllables Equal signs link material that runs on
Arrows indicate rising or falling intonational shift. Double arrows indicate very marked shifts Upward intonation (not necessarily a question) Underlining indicates emphasis under Capital letters indicate speech noticeably louder than that surround CAPITALS ing it Degree signs indicate speech spoken noticeably more quietly than the "soft" surrounding talk. Double degree signs indicate greater softness >fast<; <slow> "Greater than" and "less than" signs indicate that the talk they en compass was produced noticeably quicker or slower than the sur rounding talk Square brackets between adjacent lines of concurrent speech denote over [lap the start of overlapping talk [lap —> Side arrow indicates point of special interest in the extract, addressed in the text [...] Material left out of the extract
u
Author's N o t e I am grateful to Mark Rapley for making available the tapes analyzed here and for commentary, to Kevin McKenzie for help in preparing the data, and to Hanneke Houtkoop and Doug Maynard for comments on an earlier draft of this article.
Note
ANTAKI
19
INTERVIEWING PERSONS WITH A LEARNING DISABILITY
References Antaki, C , & Rapley, M. (1996a). Questions and answers in psychological assessment schedules: Hidden troubles in "quality of life" interviews. Journal of Intellectual Deficiency Research, 40, 421-437. Antaki, C , & Rapley, M. (1996b). "Quality of life" talk: The liberal paradox of psy chological testing. Discourse and Society, 7, 293-316. Atkinson,J. M., & Heritage,J. (Eds.). (1984). Structures of social action: Studies in conversation analysis. Cambridge, UK: Cambridge University Press. Brewer, J. D., & Yearley, S. (1989). Stigma and conversational competence: A conver sational-analytic study of the mentally handicapped. Human Studies, 72, 97-115. Button, G. (1992). Answers as interactional products: Two sequential practices used in job interviews. In P. Drew &J. Heritage (Eds.), Talk at work: Interaction in institutional settings (pp. 160-171). Cambridge, UK: Cambridge University Press. Dijkstra, W., van der Leen, L., & van der Zouwen, J. (1985). A field experiment on inter viewer-respondent interaction. In M. Brenner, J. Brown, & D. Canter (Eds.), The research interview: Uses and approaches (pp. 56-78). New York: Academic Press. Drew, P., & Heritage, J. (Eds.). (1992). Talk at work: Interaction in institutional settings. Cambridge, UK: Cambridge University Press. Drew, P., & Holt, E.J. (1988). Complainable matters: The use of idiomatic expressions in making complaints. Social Problems, 35, 398-417. Foddy, W. (1993). Constructing questions for interviews and questionnaires. Cambridge, UK: Cambridge University Press. Fowler, F.J.,Jr., & Mangione, T. W. (1990). Standardised survey interviewing: Minimising interviewer related error. Newbury Park, CA: Sage. Garfinkel, H. (1967). Studies in ethnomethodology. Englewood Cliffs, NJ: Prentice Hall. Garfinkel, H. (1972). Remarks on ethnomethodology. In J. J. Gumperz & D. Hymes (Eds.), Directions in sociolinguistics: The ethnography of communication (pp. 94-118). New York: Holt. Goffman, E. (1967). Interaction ritual: Essays onface toface behaviour. New York: Doubleday Anchor. Heritage, J., & Sorjonen, M. L. (1994). Constituting and maintaining activities across sequences: And prefacing as a feature of questioning design. Language in Society, 23, 1-29. Houtkoop-Steenstra, H. (1995). Meeting both ends: Between standardisation and recipient design in telephone survey interviews. In P. ten Have & G. Psathas (Eds.), Situated order: Studies in the social organisation of talk and embodied activities (pp. 91-107). Washington, DC: University Press of America. Houtkoop-Steenstra, H. (1996). Probing behaviour in semi-standardised survey interviews. Quality and Quantity, 30, 205-230. Houtkoop-Steenstra, H. (1997). Being friendly in survey interviews. Journal of'Pragmatics, 28, 591-623. Houtkoop-Steenstra, H., & Antaki, C. (1998). How to create happy people with yes-no questions. Research on Language and Social Interaction, 30,285-314. Hughes, C , Hwang, B., Kim,J-H., Eisenman, L. T., & Killian, D.J. (1995). Quality of life in applied research: A review and analysis of empirical measures. Americanjournal on Mental Retardation, 99, 623-641. Jefferson, G. (1988). On the sequential organisation of troubles talk in ordinary con versation. Social Problems, 35,418-441. Larrue,J., & Trognon, A. (1993). Organization of turn-taking and mechanisms for turntakinff rfinairs i n a rTiairpH m p p f i n r r Tnnr*>/il nfTZr/uvnnaU™ 10 i
n
c
20
D I S C O U R S E STUDIO
Marlaire, C. L., & Maynard, D. W. (1990). Standardised testing as an interactional phenomenon. Sociology of Education, 63,83-101. Maynard, D., & Marlaire, C. (1992). Good reasons for bad testing performance: The interactional substrate of educational testing. Qualitative Sociology, 15, 177-202. Perkins, L. (1997, September). The contribution of CA to analysing disorder in talk. Pape presented at the International Conference on Order and Disorder in Talk, UCL London. Rapley, M., & Antaki, C. (1996). A conversation analysis of the "acquiescence" of people with learning disabilities. Journal of Community and Applied Psychology, 6,371-391. Rapley, M., & LobleyJ. (1995). Factor-analysis of the Schalock and Keith (1993) qualm of-life questionnaire: A replication. Mental Handicap Research, 8, 194-202. Sacks, H. (1992). Lectures on conversation (Vols. 1 & 2) (G.Jefferson, Ed.). Oxford. UK Blackwell. Schalock, R. L., & Keith, K. D. (1993). Quality ofLife Questionnaire. Worthington, OH IDS. Schalock, R. L., Keith, K. D., Hoffman, K., & Karan, O. C. (1989). Quality of life: I; measurement and use in human service programs. Mental Retardation, 27,25-31. Schalock, R. L., Keith, K. D., Hoffman, K., & Karan, O. C. (1990). Quality of Life Question naire: Standardization manual Hastings: Mid-Nebraska Mental Retardation Service Suchman, L. A., & Jordan, B. (1990). Interactional trouble in face-to-face survey inter views. Journal of the American Statistical Association, 85, 232-244. Wootton, A. (1989). Speech to and from a severely retarded young Down's syndrome child. In M. Beveridge, G. Conti-Ramsden, & I. Leudar (Eds.), The language and com munication of mentally handicapped people (pp. 190-212). London: Chapman-Hall.
I
2 Discourse Studies and the Ideology of 'Liberalism' Robert de Beaugrande
The word is the fundamental object of the study of ideologies. (Volosinov, 1973 jorig. 1929]: 15) Look ai the very connotations of the w o r d 'liberal', which outside the political context is taken to m e a n gentle, generous, nice. But p u t the political wrapping a r o u n d i t . . . (data from the Corpus of South African English) particularly in the West, we often use the w o r d 'liberalization'. T h a t is not a bad word, (data from the Bank of English)
A. Science and Ideology £ " ¥ deology' is a deeply problematic term. I n m a n y registers of discourse, it I routinely connotes some fixed, unreasoning d o g m a that foments conflicts, JL;« when Shils (1958) cited Fascism, Nazism, Bolshevism, and McCarthyism as examples. Geertz (1973: 197 ff.) c o m m e n t e d u p o n this conception: Like the politics it supports, it is dualistic, opposing the p u r e 'we' to the evil 'they', proclaiming that h e w h o is n o t with m e is against m e . It is doctrinaire in that it claims complete a n d exclusive possession of political truth and abhors compromise. It is totalistic in that it aims to order the whole of social a n d cultural life in the image of its ideals, futuristic in that it w orks toward a Utopian culmination of history in which such an ordering wiU be realised. rce: Discourse Studies vol. 1, no. 3,1999, pp. 259-295.
22
DISCOURSE STUI
T h e favoured candidate to defend against 'ideology' in this darkly pejorative sense is held to b e 'science' (Geertz, 1973; Zima, 1981; Pecheux, 1982): the essential criteria of an ideology [are its] deviations from scientific ob jectivity [...] T h e p r o b l e m of ideology arises where there is a discrepancy between w h a t is believed a n d what can b e [established as] scientifically correct. (Talcott Parsons) Such a p r o n o u n c e m e n t complacentiy implies that science itself can and should b e free of all ideology, a n d i n d e e d heralds the ' e n d of ideology' (Geertz. K'~x 199) in the final triumph of 'objectivity' a n d 'correctness'. Yet dictionary definitions suggest that 'ideology' is a m o r e neutral, norm.iL a n d even necessary framework: ' a systematic b o d y of concepts esp. about human life or culture' (Webster's Seventh, p . 413); ' a b o d y of doctrine or thought that guides an individual, social m o v e m e n t , institution, or g r o u p ' (Random flomt Webster's, p . 668); or 'a belief or set of beliefs, especially the political beliefs cm which people, parties, or countries base their actions' (Collins COBUILD,p. 7iti Surely these b r o a d definitions might hold for science too. Perhaps the ideology of science is so h a r d to recognize because there we encounter not so m u c h a particular 'body of concepts' (or 'doctrines' or 'thought') as a set of general methods for arriving at 'concepts' - n o t so m u c h a 'theory' as a modality for producing a n d testing theories. Such a n ideology effectively func tions as a meta-ideology - a n ideology of methods as distinct from an ideology of content - a n d cannot b e challenged a n d discredited in the direct and definiDve ways that radical political ideologies like Nazism or McCarthyism can. Refuting a scientific theory like the 'steady-state theory' of the universe did not refute the science of astronomy or its m e t h o d s , b u t rather vindicated its methods (in this case, observing r e d shift in the spectra of light dispersion). But we d o find a popularized ideology which can b e called scientism, holding that only scientific knowledge is true a n d valid, a n d that the progress of science will eventually explain the entire universe and solve all of humanity's problems. This fits Geertz's pejorative sense quoted earlier b y being 'totalistic' in that n aims to 'order the whole' of the universe, a n d 'futuristic in that it works toward a Utopian culmination' of h u m a n knowledge. A n d the scientists m a y tend to be 'doctrinaire' i n 'claiming complete and exclusive possession of the truth and abhorring compromise'. T h e 'doctrinaire' stance can entrain the scientist in a lonely and vertiginous paradox: the history of science demonstrates that all scientific theories so far have b e e n proven wrong, yet the ideology of scientism encourages you to claim your current theory to be, at long last, the sole true one, whereas the scientists who support rival theories must b e mistaken or misled b y personal or institutionil biases (Kuhn, 1970; Gilbert a n d Mulkay, 1984). T h e p a r a d o x can erode the essence of science as an institution that should freely grant a n d defend the legit imacy of alternative explanations, a n d should acknowledge that these may con tribute the most w h e n w e integrate them. Scientism has also h e l p e d to inhibit science from accrediting ideology as >
DE BEAU GRANDE
DISCOURSE STUDIES A N D THE I D E O L O G Y OF 'LIBERALISM'
23
a 'non-evaluative conception of ideology' into his 'sociology of knowledge' were stymied by the prospect that 'nowhere is resistance to claims of objectivity greater dun in the study of ideology', as r e m a r k e d b y Geertz (1973: 195), w h o a d d e d : "men do not care to have beliefs to which they attach great m o r a l significance examined dispassionately, n o matter for h o w p u r e a purpose [...] they m a y find it amply impossible to believe that a disinterested a p p r o a c h to critical matters of social and political conviction can b e other t h a n a scholastic sham'. Yet Geertz also saw a major ' p r o b l e m ' in the w a y 'social science' tries to 'handle ideology as an entity in itself - as an ordered system of cultural symbols rather than in the discrimination of its social a n d psychological contexts' (1973: 195 ff.). Scientism thus subverts the authentic interests of science w h e n the exagger ations and mystifications of the authority a n d p o w e r of science r e n d e r the indi vidual scientist unproductively possessive a n d defensive a b o u t his or h e r o w n 'objective truth'. Moreover, science is prevented from effectively confronting and discrediting anti-scientific ideologies such as racism a n d sexism, a n d from blocking their attempts to expropriate a n d exploit scientism a n d raid sociology, psychology, or genetics for ' p r o o f of the 'natural inferiority' of minorities and women. If, as philosophers of science like K u h n (1970) h a v e emphasized, a scientific theory can be discredited only b y another theory, t h e n w e m a y reasonably as sume that an ideology can b e effectively discredited only b y another ideology and not by some ivory-tower standpoint purporting to b e free of all ideology (discussed later). Science can profitably investigate ideology, and, at the same time, explicidy develop creditable alternative ideologies for itself a n d for the institutions it sustains. Building u p o n Ulric Neisser's (1976: 2) concept of 'eco logical validity', i.e. whether ' a theory has something to say about what p e o p l e do in real, culturally significant situations' a n d says it in ways that ' m a k e sense to the participants', we could p r o m o t e the ideology of ecologism, wherein the theory and practice of science are dialectically reconciled in a transdisciplinary enterprise of sustaining h u m a n e a n d d e m o c r a t i c practices of action, interaction and discourse (de Beaugrande, 1997). Some trends in this direction have already been emerging, witness the impact of the Gaia Atlas of Planet Managqnenl (Myers et al., 1993) a n d of Gell-Mann's (1994) l a n d m a r k expose of The Quark and the Jaguar. I would also see a similar drift in 'critical Hnguistics' a n d 'critical discourse analysis', e v e n if these sometimes d o not seem to provide for humane counter-ideologies.
B. T h e I d e o l o g y of M o d e r n Linguistics One science that has b e e n deeply reluctant to study the nature and functions of ideology is modern linguistics. I n the major treatises I have surveyed in fine detail, dating from the early 20th century u p into the 1970s (de Beaugrande, 1991), the term 'ideology' simply does n o t appear. Adapting Geertz's terms from sociology, its absence might reflect the m a i n s t r e a m p r o g r a m m e of linguistics, believing that the properly scientific m e t h o d would b e to ' h a n d l p l a n o m a i r o qc o-n < m « h , j «
DISCOURSE STUDIES
24
itself - as an ordered system of symbols rather t h a n in the discrimination of its social a n d psychological contexts'. This austere p r o g r a m m e encouraged mainstream linguistics to develop ar, ideology of idealization, holding that language is based u p o n an ideal mode o! order that is n o t readily evident in the ordinary practices of real discourse within society. Such was the source of the m a n y static dichotomies that in effect disconnec ideal language from real language, such as 'langue versus parole' or 'competence versus performance' (de Beaugrande, 1998a). O n e label for this ideology, which is also being heavily camouflaged as a standpoint freed of all ideology, might be the neologistic t e r m linguisticism. 'Linguisticism' sustains the doctrine that 'Ian guage' is an abstract, uniform, a n d stable system whose nature a n d properties can b e determined only b y the ratiocinations of 'linguistic theory' and not from observing and recording discursive practices. D e Saussure (1966 [orig. 1916]: K hinted as m u c h w h e n h e speculated that 'other sciences w o r k with objects thai are given in advance', whereas in 'linguistics', 'it is the viewpoint that creates tht object'. Hjelmslev (1969 [orig. 1943]: 18), w h o aspired to b e a devout successor to d e Saussure, was m o r e dramatic: 'linguistic theory cannot b e verified (con firmed or invalidated) b y reference to any existing texts a n d languages'. T h e same ideology was signalled, albeit less patently, b y the declaration that 'language is a well-defined object in the heterogeneous mass of speech facts', whereas 'speech cannot b e studied', n o r i n d e e d can it b e 'put in any category of h u m a n facts, for we cannot discover its unity'(de Saussure, 1966 [orig. 1916j:14. 9,11). T h e trend was reinforced half a century later, w h e n Chomsky (1965:3 ff.. 201) a n n o u n c e d that 'Unguistic theory is primarily concerned with an ideal speaker-hearer in a completely h o m o g e n e o u s speech-community, who knows its language perfectly', whereas the 'observed use of language' 'surely cannot constitute the subject-matter of linguistics, if this is to b e a serious discipline'; indeed, 'from the point of view of the theory', ' m u c h of the actual speech observed consists of fragments and deviant expressions of a variety of sorts'. T h e key question to challenge linguisticism might b e : what shall b e the sources of evidence for this 'primary concern', if, as C h o m s k y (1977:192) cheerily grani\ the 'ideal speaker does n o t exist in the real world'? T h e d e facto answer would b e : the 'linguistic intuition' of the linguist 'himself as a 'native speaker' (Chomslc, 1965: 20). Yet C h o m s k y (1965: 8) has expressly denied that the 'speaker of i language' 'is aware of the rules of the g r a m m a r or even' 'can b e c o m e aware <:' them'. So either linguists acting as native speakers are also unable to 'report' 'tfc rules of the g r a m m a r ' ; or else they hold s u p e r h u m a n powers for 'becorniiu aware of a n d reporting' what other speakers cannot. Either linguists do represen: the community a n d cannot 'report'; or they can report a n d do not represent the community. Viewed in this light, the ideology of hnguisticism legitimizes a remarkable arrogation of p o w e r b y the theoretical linguists to b e the sole authorized reprc sentatives of a purely hypothetical 'speech-community' a n d therefore immure to contradiction or counter-evidence from real speakers in a real communir This arrogation sharply illuminates the consequences of linguisticism declining , i . — t . . J „ < „ „ < - , . „ ! o „ ^ , V i ' -xnA or-rnnnts fnr the tendency. tVDical of ideology, lo I Q /
DE BEAUGRANDE DISCOURSE STUDIES AND THE IDEOLOGY OF 'LIBERALISM'
25
foment conflicts within t h e discipline. O n c e linguists h a v e cut themselves off from the rational basis for developing a n d validating empirically sound theories of language, t h e fashion arises of fabricating 'linguistic theories' from t h e t o p down and applying t h e m to o d d handfuls of trivial invented sentences; a n d t h e contentious fragmentation of 'theoretical linguistics' w e h a v e actually seen is a natural consequence (de Beaugrande, 1998a). The ideology of linguisticism h a s evidently foundered u p o n its o w n implicit dualism: equating language ('langue', 'system', 'competence', etc.) with perfect order, whilst equating discourse ('parole', 'speech', 'performance', etc.) with massive disorder. The direct corollary would b e that using a language to produce discourse triggers an abrupt and catastrophic transition from stable a n d integrative order over to unstable and disintegrative disorder. Since this corollary is patently absurd, w e must conclude that the linguisticism h a s m a d e a capital mistake b y attributing to 'language' a n idealized mode of order which is fully determined and finalized within the abstract system. T h e ideology of'ecologism' proposes instead a dialectic whereby the real order of language elaborately supports the order of discourse without fully determining or finalizing it, discovering h o w that support actually operates is now a stimulating challenge for a transdisciplinary science of text a n d discourse (de fkaugrande, 1997).
C . T h e I d e o l o g y of C o r p u s L i n g u i s t i c s The early stages of corpus linguistics at present might b e favourable for deliber ating upon what ideology it could develop out of its strong potential to discredit the mainstream ideology of linguisticism. Corpus data immediately deconstruct the vision of 'observed use of language' constituting a mass of disorder a n d 'deviance'. But the order of discourse, n o t surprisingly, is n o t the m o d e of static and abstract order envisioned b y linguisticism. W e can easily recognize h o w the standing constraints that persist o n t h e plane of the system (e.g. the English article going before the noun, n o t after it) continually interact with emergent constraints that are only decided o n t h e plane of t h e discourse (e.g. t h e lexical choices ap propriate to a political debate). A n d t h e local constraints a m o n g sets of selections and combinations interact with the global constraints of register, discourse domain, topic, and so forth. Linguisticism naturally projects disorder onto discourse after taking into account only a n a r r o w a n d arbitrarily defined subset of local standing constraints which get reconstructed i n a ' g r a m m a r ' a n d hugely o v e r b u r d e n e d with the task of sustaining a fully d e t e r m i n e d a n d finalized system which is not and cannot be) 'reflected' in discourse. A s Sinclair (1991: 496) h a s r e marked, 'much of t h e a p p a r e n t disorder is created b y t h e perspective that is initially adopted'. Corpus research also suggests describing t h e order of real language in t e r m s of colligability, s u b s u m i n g t h e 'preferences' of s o m e grammatical options lor appearing with certain others a n d holding t h e ' g r a m m a r ' together; a n d of collocability, subsuming the 'preferences' of some lexical options for appearing .With Certain Others a n d h n l d i n t r thp ' l p v i r n n ' trnn»Hif»r Tn n r l i a l n r t i r a l
narallol
26
DISCOURSE STUDIES
the order of discourse partly realizes and is partly realized b y the order of language in the actually occurring grammatical c o l l i g a t i o n s a n d lexical c o l l o c a t i o n s . T h e c o n c e p t of t h e i n t e g r a t e d l e x i c o g r a m m a r in systemic functional linguistics (e.g. Halliday, 1994) further suggests that the g r a m m a r a n d lexicon h o l d each other together. This factor could explain w h y so m a n y 'grammars' sponsored b y Unguisticism a n d designed to b e i n d e p e n d e n t of the lexicon h a v e r e m a i n e d so fragmentary a n d so r e m o t e from authentic data. T h e lexicon, in its turn b e i n g regarded as another heterogeneous mass of disorder, has received scant attention in mainstream linguistics (Bolinger, 1970). 1
A major principle that corpus linguistics could n o w field against linguisticism would b e that the order of language is dynamic a n d transitory, a n d in principle not describable b y any static or 'synchronic' theory. To a significant degree, a language is always in the process of being created a n d negotiated whilst discourse is in progress; a n d a different generation of linguistic theories will b e n e e d e d to explain how. T h e ideology sustained b y corpus linguistics would accordingly b e a version of dynamism wherein the specification of a theory will b e far m o r e actively data-driven and ' b o t t o m - u p ' than the theories sponsored b y linguisticism. Theories will n o longer originate 'from the top d o w n ' every time some ambitious linguist chooses to fabricate his or h e r personal idealization and illustrate it with a handful of fictional sentences. Instead, data-based theories will evolve b y being adjusted a n d t u n e d through continuing research o n steadily larger corpora of authentic discourse (Sinclair, 1997). This evolution will undoubtedly affect the familiar dichotomies a n d divisions w h e r e b y those older theories aspired to freeze language into a Saussurian 'welldefined object'. If these are to survive the test of authentic data, they n e e d to b e deconstructed a n d re-theorized as dialectical interactions: langue - parole, competence - performance, synchronic - diachronic, syntagmatic - paradigmatic, g r a m m a r - lexicon, language acquisition - language learning, a n d so forth. Alternatively, less familiar terms or concepts can b e introduced wherever the data seem to justify them, such as the parallel dialectics between 'colligabilities' and 'colligations', a n d between 'collocabilities' a n d 'collocations'. I n the present article, I explore the prospect that 'ideology' m a y also find a n e w h o m e as one of the sources of global constraints u p o n the orders of language a n d discourse, e.g. in the b r o a d sense of 'a systematic b o d y of concepts about h u m a n life or culture' (compare Billig, 1991; Christie a n d Martin, 1997; Van L e e u w e n and Wodak, 1998). T h e s e concepts would b e applied during discourse below con scious awareness a n d thus seem to b e fully identical with life or culture; and those w h o h o l d contrasting ideologies seem to b e distorting the 'real world', w h e n c e the well-known tendency of ideologies to foment conflicts. To recognize the value of a contrasting ideology, you might first n e e d to register the contingent a n d partial quality of y o u r own; b u t doing so w o u l d require adopting, at least temporarily, another ideology from whose standpoint your own could b e contem plated a n d contrasted. A vicious cycle impends: h o w to step outside your own ideology w h e n y o u are not even able to see it as anything b u t the reality of life. For similar reasons, criticizing or attempting to change another person's ideolo e v m a v b e perceived as acutely threatening, p e r h a p s like trying to m a k e them
DE BEAUGRANDE DISCOURSE STUDIES AND THE IDEOLOGY OF 'LIBERALISM'
27
switch from reasonable over to unreasonable. People m a y well defend a n d cling to their ideology even w h e n they perceive symptoms of disorientation, fear ing that they might lose whatever orientation they still h a v e . T h e accumulating symptoms would foment increasing alienation, which could readily b e expressed (though n o t alleviated) b y hostility a n d aggression against p e o p l e w h o h o l d a contrasting ideology. H o w t h e n can science investigate ideology if the latter is so deeply anchored at the base of h u m a n awareness? T h e principle of making a n ideology explicit from the standpoint of a contrasting ideology might incur the usual risk of con struing differences as distortions, especially w h e r e the ideology u n d e r investi gation is defiantly positioned against science itself, as is religious fundamentalism. Also, even if w e can design non-threatening techniques for bringing ideology to the conscious attention of the people w h o subscribe to it, w e m a y risk transform ing its functions w e want to describe, p e r h a p s in the w a y that literary techniques deliberately use framing or irony to u n d e r m i n e the illusion of reality in a nar rative, and t h e r e b y disrupt the 'willing suspension of disbelief (Coleridge) that literature solicits. A n alternative m e t h o d would b e to investigate large c o r p o r a of authentic data in terms of h o w ideologies leave traces within the order of discourse, even (or especially) w h e n the discourse reflects or works out the disorders within the society (Wodak, 1996). T h e traces should b e conspicuous wherever the discursive order sponsored b y o n e ideology is experienced as discursive disorder from the stand point of another ideology. Conversely, discursive order m a y b e merely superficial when our data show the same conception b e i n g appropriated b y contrasting ideologies in order to project a deceptive consensus, e.g. onto the m e a n i n g of 'democracy' (de Beaugrande a n d Williams, in press). However, research with large corpora m a y well discover that far less con spicuous selections a n d combinations of discourse options are also sensitive to ideological groundings. Such was the discovery w h e n feminism, t h o u g h using more limited data, m o v e d b e y o n d the usual issues of'sexist language' (e.g. m a l e pronouns for everyone) to investigate the traces of the ideology of patriarchy i n h e r e n t i n t h e m u c h d e e p e r o r g a n i z a t i o n of d o m a i n s like ' g r a m m a r ' (de Beaugrande, 1988b; C a m e r o n , 1992; Wodak, 1997; Kotthoff a n d Wodak, 1998). Certainly, the degrees of detail a n d delicacy within language that h a v e been exposed b y sorting a n d searching large corpora of data extend far b e y o n d anything detected before (Sinclair, 1996). Might n o t a similar exposure b e achieved for the traces of ideology? W h a t then of the ideology of research for the corpus linguists themselves? By the arguments advanced here, might they n o t overlook the traces of their own ideology a n d focus u n d u l y o n contrasting ideologies? O n e answer might be that research could recruit representatives of diverse ideologies, w h e r e b y the potential blind spots of a n y o n e would b e compensated for b y the vigilance of the others. To b e sure, recruiting scientists b y explicit reference to their personal ideologies would b e a sensational tactic after scientism has cultivated for so long the illusion of science standing free of ideoloev. And the* n m K l o m o f
28
DISCOURSE STUDIES
identifying t h e suitable range of ideologies would b e far from trivial, especially during early stages of t h e research. A n alternative answer could b e that corpus research has a general effect of sensitizing researchers toward t h e normally naturalized constraints u p o n dis course, including ideological ones. T h e patterns that emerge from t h e multiple m e a n s to query a database a r e often unpredictable a n d surprising enough that w e b e h o l d an image of o w n intuitions a n d habits of speaking which h a d never entered o u r conscious awareness before. As w e repeatedly discover just which a m o n g the staggering range of potential selections a n d combinations are typically m a d e , w e are p r o m p t e d to examine t h e order of discourses in u n p r e c e d e n t e d detail, a bit like t h e biologists w h o e x a m i n e d cell tissue u n d e r a powerful micro scope for t h e first time. A t the same time, w e can escape t h e q u a n d a r y of linguisticism which replaced real language with ideal language a n d t h e n claimed s u p e r h u m a n access to t h e 'perfect knowledge' of the 'ideal speaker' b y virtue of 'intuition' as distinct from data. Working with large corpus data emphatically refers t h e linguist back into t h e c o m m u n i t y of real speakers w h o p r o d u c e d a n d received the data a n d a m o n g w h o m we ourselves informally belong. H o w typical a n d representative w e might b e a n d h o w useful o u r intuitions might b e a r e questions to b e decided during the research a n d n o t b y t h e glib arrogations of linguisticism (cf. Francis a n d Sinclair, 1994).
D . I d e o l o g y i n Critical D i s c o u r s e Analysis T h e ancestors a n d precursors of what is presently called 'discourse analysis' include t h e data-driven approaches i n linguistics a n d neighbouring disciplines (e.g. anthropology, ethnography, sociology) that were arrayed against the linguis ticism of idealization, a n d were for decades targeted with dismissive polemics promulgating the doctrine that 'observed use of language' 'surely cannot consti tute t h e subject-matter of linguistics' as a 'serious discipline'. W h e n 'discourse analysis' finally emerged as a field with that n a m e in the 1970s, t h e early concerted efforts of western 'critical linguistics' were also getting u n d e r w a y (e.g. M e y , 1979 [orig. 1974]; Fowler et al., 1979). T h e 'critical' work was clearly distinguished b y its resolve to accept 'ideology' (along with 'power', 'domination', etc.) as a legitimate object of investigation, finally i n line with t h e neglected proposals of Volosinov (1973 [orig. 1929]: 9), who envisioned a 'Marxist theory of ideologies' as 'the bases for studies of scientific knowledge' before h e disappeared in Stalin's death camps. Perhaps to eschew the idealizing a n d formalist tendencies of conventional 'linguistics', this work adopted the b r o a d e r heading critical discourse analysis, devoted to 'the analysis of linguistic a n d semiotic aspects of social processes a n d p r o b l e m s ' (Wodak, 1996:15). A key question for t h e n e w discipline is natur ally h o w 'ideology' as such should b e defined. I n most of this research, t h e old pejorative connotations of 'ideology' h a v e persisted b u t i n a m o r e precisely de fined function: 'ideologies are particular ways of representing a n d constructing societv which reDroduce u n e a u a l relations of power, relations of domination
DE BEAUGRANDE DISCOURSE STUDIES AND THE IDEOLOGY OF 'LIBERALISM'
29
and exploitation' (Wodak, 1996: 18; c o m p a r e n o w Wodak, 1997; W o d a k et al., 1998). Fairclough (1992:67) offers a similar definition: 'ideology is significations generated within p o w e r relations as a dimension of the exercise of p o w e r a n d struggle over power'. L e m k e ' s (1995:12 ff., his italics) definition seems t h e most darkly pejorative: 'ideology supports violence a n d is critically shaped b y a n d in a context of violence' a n d b y 'physical pain a n d social dehumanisation'. N o d o u b t such definitions reflect the justified urgency to critically analyse and deconstruct those ideological discourses which most actively legitimize or mystify power, inequality, domination, exploitation a n d violence, such as racism and sexism. Moreover, the traces of such ideologies in discourse should b e t h e most accessible to practical analysis. However, in terms of the present discussion, building these pejorative effects into our basic definition of 'ideology' incurs at least three serious drawbacks. T h e first drawback concerns the ideological com mitment of critical discourse analysis itself, especially w h e n it repudiates the idealizing ideology I h a v e called 'linguisticism'. By the definitions I h a v e just quoted, the ideology of critical discourse analysis might b e charged with sus taining o n e m o r e 'exercise of p o w e r ' a n d 'domination' - a prospect Fairclough (1996) himself has recently aired. H o w could w e prevent our critical engagements with ideologies from b e i n g hit b y the fall-out of a rigorously pejorative concept of ideology? T h e second a n d closely related drawback follows from the plausible pro spect that a n ideology can b e effectively o p p o s e d or deconstructed only from the standpoint of another ideology. A pejorative definition forecloses our efforts to develop ameliorative counter-ideologies, such as 'ecologism', which expressly p r o m o t e equality a n d solidarity, e.g. in alliances with feminism a n d multiculturalism; a n d for defining such a counter-ideology to guide the projects of discourse analysis itself. Fairclough (1996, pers. comm.) h a s expressed concern that if 'ideology' as 'understood in the critical tradition' gets redefined as I pro pose, the concept might get 'appropriated', 'disarmed', or 'compromised', leaving us with the 'relativist conclusion that values are merely different, that all values are as good as all others'. Such relativist tendencies can undeniably b e diagnosed in several trends in 'post-structuralism' a n d 'post-modernism' b u t b y n o m e a n s in 'ecologism', which insists that we cannot evade the choices between promoting either power or solidarity, either inequality or equality, a n d cannot retreat either into relativism or into objectivity. Fairclough would probably concur with ecolog ism that the hallowed 'scientistic' stance of objectivity in pursuit of 'pure truth' can imply an acquiescence to the existing structures of p o w e r a n d domination, as when a hnguisticism dedicated to the 'ideal s p e a k e r - h e a r e r in a completely homogeneous s p e e c h - c o m m u n i t y ' elided t h e significant social consequences of language variations in a real s p e e c h - c o m m u n i t y (de Beaugrande, 1998a). My point is rather that critical discourse analysis should n o t r u n the risk, inherent in m a n y leftist a n d Marxist critiques, of appearing to offer only oppos ition and negation without a coherent alternative p r o g r a m m e . A t this stage of post-modern society, w e behold a one-sided array of entrenched b u t ecologically u n s o u n d ideologies, such as c o n s u m e r i s m , individualism, capitalism, freemarketism- and sn on whilst VPIV fmu affani-nra ; — ^ — :
j
30
side even h a v e established names, let alone organized groups of adherents. Per suading people to exchange one ideology for another is hard enough; persuading t h e m to throw out their ideology in favour of n o n e at all seems frankly unrealistic. T h e third and again closely related drawback concerns the question of whether all discourses are ideological (Hodge a n d Kress, 1988; Gee, 1990; L e m k e , 1995) or only some discourses are ideological a n d others are not (Fairclough, 1995; Wodak, 1996; Van Dijk, 1998a). If (as is often the case) b o t h positions favour a pejorative definition, the first is plainly the m o r e pessimistic t h a n the second in implying that power and domination will always control discursive practices; such a n implication would certainly b e rejected b y emancipatory ideologies such as feminism. T h e second position would imply a n enterprise of 'de-ideologizing' discursive practices: of purging ideologies rather than changing or replacing them. H o w this might b e achieved without fostering some disorienting ideological v a c u u m remains to b e seen. S o m e of Fairclough's (1992:91, m y italics) deliberations might signify differ ences in degree rather t h a n in kind: Is all discourse ideological? I h a v e suggested that discursive practices are ideologically invested if they incorporate significations which con tribute to sustaining or restructuring power relations.[...] But all discourse is not thereby irredeemably ideological. [...] the fact that all types of discourse are open in principle, and no doubt to some extent in fact, in our society to ideological investment does not m e a n that alltypes of discourse are ideologically invested to the same degree. I n a similar vein, Wodak's (1996: 19, m y italics) conception of critical dis course analysis 'does not claim that all discourse is ideological: "it does not follow that because allpractices are in ideology or inscribed by ideology, all practices are nothing but ideology" (Hall, 1985: 103)'. A n d L e m k e (pers. comm.) counsels that ' a dis course formation is not necessarily inherently ideological, but is so only b y virtue of its uses; it is the social function of the discourse which is ideological, and not the discourse itself. T h e question bears o n both theory a n d practice. If we favour the data-driven theories advocated earlier, then w e should n o t p r e - e m p t our conception of 'ideology', which has only r e c e n d y b e g u n to b e theorized in earnest. T h e b r o a d conception I h a v e advocated of 'ideology' b e i n g a source of global constraints u p o n discourse in general would b e m o s t conducive to putting our prospective theories of ideology onto the broadest empirical basis a n d would b e m o r e per ceptive toward the covert ideological groundings which leave less conspicuous traces a m o n g the selections a n d combinations of discourse. A pejorative con ception, in contrast, would tend to attract our vision to the m o r e conspicuous traces we can already link to ideologies of p o w e r a n d domination in intuitive a n d pre-theoretical ways. I n terms of practice, fundamental problems i m p e n d w h e n w e try to distin guish the set of discourses (or the portions within a single discourse) which count as 'ideological' from the set which d o not. Moreover, the pejorative definition
DE BEAUGRANDE DISCOURSE STUDIES AND THE IDEOLOGY OF 'LIBERALISM'
31
implies that o n c e the 'ideological' o n e s h a v e b e e n d i s c o v e r e d , t h e ' n o n ideological' ones can b e left out of our analyses, at least for the time being. But critical analysts, including Fairclough and Wodak, h a v e often noted the significant tendency of ideological discourses of p o w e r to b e c o m e less overt in order to en courage illusions of openness and democracy. If so, the demarcation between the two sets would b e transient a n d unreliable, a n d we would tend to exclude some discourses whose ideological groundings could b e discerned only after detailed critical analysis. This factor could b e especially acute if the findings of critical discourse analysis itself b e c o m e popularized in the news media, a n d the institutions of power a n d domination r e s p o n d b y adopting m o r e devious and covert discursive strategies to escape detection. Some of these problems might b e attenuated b y Fairclough's proposal to re cognize differences in degree. But the p r o b l e m m a y not p r o v e m u c h simpler of distinguishing degrees of ideological investment'(how ardently you support the ideol ogy) from degrees of discursive overtness (how forcefully you express your support). A devious institution could exploit strategic confusion h e r e too, such as coverdy accepting one ideology (e.g. racism) while overdy promoting another (e.g. 'equal opportunity' achieved b y phasing out affirmative action). Fairclough (1992: 88 ff., m y italics) was evidently aware of such p r o b l e m s when h e critiqued the 'textual view of the location of ideology, which one finds in Critical linguistics - ideologies reside in texts': While it is true that the forms a n d content of texts do bear the imprint of (are traces of) ideological processes and structures, it is not possible to 'read off' ideologies from texts, [. ..] because meanings are p r o d u c e d through inter pretations of texts, a n d texts are open to diverse interpretations which m a y differin their ideological import. [...] Claims to discover ideological p r o cesses solely through text analysis r u n into the p r o b l e m n o w familiar in m e d i a sociology that text 'consumers' (readers, viewers) appear some times to b e quite immune to the effects of ideologies which are supposedly 'in' the texts. In m y view, these arguments might lead to a different conclusion. R a t h e r than saying that 'to r e a d off ideologies from texts is just 'not possible', we might say that different interpreters can and often will 'read off different ideologies from the same texts. Even if we leave aside the old disputes over whether the 'sameness of the text' is a meaningful concept - an issue which has hardly b e e n clarified b y the "post-structuralism' with its concepts like 'dissemination', 'bricolage', 'jouissance', 'free-play of the signifier', a n d so o n (de Beaugrande, 1988a) - we m a y incur a replay of the chicken-and-egg p r o b l e m : you 'read off the ideology' of a text under the controls of the ideology you already hold. But, as I h a v e indicated, you are likely to focus o n a contrasting ideology a n d interpret it as a distortion; conversely, y o u are n o t well positioned to 'read off' your o w n ideology, which would appear natural a n d transparent a n d so largely invisible. Moreover, 'immunity' confirms rather then refutes m y view: the control un consciously exerted b y your o w n ideology naturallv e-enprarpc r o « t * o » ^ Q
32
DISCOURSE STUDIES
'effects' of another sustained b y a 'text'. Actual conversions would b e limited chiefly to cases w h e r e you are u n c o m m i t t e d regarding the relevant ideological parameters, or w h e r e you b e c o m e aware of some n e w e r inclinations or sym pathies, e.g. w h e n your social position has switched from 'have-not' over to ' h a v e ' a n d your populism changes into elitism. At all events, whether ideologies can b e 'read off' from texts a n d h o w are empirical questions to b e resolved through practical tests a n d n o t just theoretical pronouncements. A t least, the prospects d o n o t seem unfavourable for gathering a n d c o m p a r i n g a representative spread of 'read-offs' from a t e a m whose ideolo gies mutually contrast. Also, extensive experience with corpus work might render critical analysts keenly attuned to the discursive traces of ideologies a n d hence far better at 'reading ofF than we might h a v e expected. Still, Fairclough is indisputably justified in contending that analyses of the texts b y themselves are far from sufficient. I n his view, ideology is located b o t h in the structures (i.e. orders of discourse) which constitute the outcome of past events and the conditions for current events, a n d in events themselves as t h e y r e p r o d u c e a n d transform the con ditioning structures. It is an accumulated and naturalised orientation which is built into n o r m s a n d conventions, as well as an ongoing work to natural ise and denaturalise such orientations in discursive events. (1992: 89) T h e analyst would seemingly require extensive background data about 'past events' a n d 'current events', as well as about the processes of'accumulating and naturalizing orientations' a n d 'conditioning structures' which h a v e p e r h a p s been 'transformed' anyway. Surely some 'reading off from texts would b e an allowable heuristic strategy for telling u s w h e r e to look? T h e insufficiency of texts in isolation has b e e n m a d e o n e point of contention in a tireless series of attacks b y H . G. Widdowson against critical discourse an alysis. H e diagnoses 'confusion' 'about the nature of discourse (as distinct from text) a n d about analysis (as distinct from interpretation)', a n d has 'suggested that this confusion is b r e d of c o m m i t m e n t ' (Widdowson, 1995: 171). Fairclough (1996: 50) infers that critical discourse analysis is being accused of 'ideological c o m m i t m e n t ' a n d 'prejudice', a n d sees in Widdowson's attacks 'a version of the classical liberal distinction between' 'science a n d impartiality' vs 'ideology, com mitment, prejudice a n d partiality'. But matters can hardly b e so simple after W i d d o w s o n himself has affirmed:
1
all the discourses of theory, including those of Unguistics, are ideologic ally loaded, cultural constructs designed to establish control a n d a sense of security. This is n o t in the least surprising of course, since theories are m a d e out of language. (Widdowson, 1991a: 39)
Notice the implication h e r e t h a t all discourses are ideological. N o r would Widdowson accept the attitude of mainstream linguistics to achieve an ideologyfree scientific status b y idealizing language, after h e has r o u n d l y vowed that
DE BEAUGRANDE DISCOURSE STUDIES AND THE IDEOLOGY OF 'LIBERALISM'
33 I
'there can b e n o idealisation without ideology' (1991a: 39). H i s advocacy is substantially m o r e general: h e would not want to suggest that we should avoid the cultural partiality of dis ciplinary discourses and strive instead to b e neutrally objective, [but rather that] we should guard against being too readily persuaded into believing in the validity of relevance of any particular discourse, n o matter what apparent authority it m i g h t h a v e . (1991a: 40) So Widdowson is rebuking Fairclough a n d others o n the grounds that a good deal of critical discourse analysis talks about the linguistic features of texts as if they inevitably expressed meaning, particularly ideological meaning, w h e t h e r the writer intended t h e m to or not. [...] Such analysis [...] states w h a t the text m e a n s to the particular reader w h o h a p p e n s to be assuming the role of analyst, a n d t h e n claims that this is what the text itself m e a n s . (1991b: 5) There is rarely a suggestion that alternative interpretations are possible. There is usually the implication that the single interpretation offered is uniquely validated b y the textual facts. (1995: 169) W h a t is b e i n g accurately described h e r e is a favoured tactic of b o t h traditional and formalist literary criticism, though neither of the two would want to appeal to 'ideological meanings'. Critical discourse analysis, in contrast, quite plainly asserts that interpreters are m o r e t h a n discourse subjects in particular discourse processes; they are also social subjects with particular accumulated social experiences, a n d with resources variously oriented to the multiple dimen sions of social life, a n d these variables affect the ways they go about interpreting particular texts. [. . .] it is important to take account of the ways in w h i c h interpreters interpret texts if o n e is properly to assess their political a n d ideological effectiveness. (Fairclough, 1992: 136) The 'overwhelming emphasis' in Fairclough's 'recent work' has accordingly been consigned to 'showing h o w shifting discursive practices, manifested in texts which are heterogeneous in forms a n d meaning, can b e analysed as facets of wider processes of social a n d cultural c h a n g e ' (1996: 55). So W i d d o w s o n ' s rebukes are quite simply misplaced and vacuous. The key difficulty I would see - one n o t raised b y W i d d o w s o n - is that Fairclough's m e t h o d consigns 'ideological effectiveness' only to the pursuit of domination, and not to resistance or solidarity. T h e field is thus not well fortified against relativism, as w e encounter it w h e n W i d d o w s o n elsewhere erases the profound ideological differences a m o n g three approaches to language. Labov's (1970:192) contention that 'working-class speakers are m o r e effective narrators,
34
DISCOURSE STUDIES
reasoners, a n d d e b a t e r s t h a n m a n y middle-class speakers w h o t e m p o r i s e , qualify, a n d lose their argument in a mass of irrelevant detail' is c o m m e n t e d in these terms: we m a y wish to acknowledge that Labov's argument is devoted to a good cause and wish to b e associated with the ideology which informs it. But [. . .] its promotion of a particular m o d e of use b y referring to some absolute criterion of precision is, like the campaign of the Plain Language Movement, n o t essentially different from the efforts of [...] custodians to p r o m o t e the m o d e of use of their preference b y referring to some absolute criterion of correctness. (Widdowson, 1988: 346) H e r e , relativism scales stunning heights b y equating egalitarian projects to restore the dignity of stigmatized language varieties and to defend ordinary citizens against manipulation b y powerful corporations and bureaucrats, with elitist projects to perpetuate a n d reinforce that very stigmatization. Moreover, only the 'criteria' of 'correctness' are absolutebecaase they cannot b e rationally measured or justified. T h e criteria of'plainness' a n d 'effectiveness' are not afoo/ate because they can b e rationally m e a s u r e d and justified in terms of mutual respect a n d openness a m o n g socially a n d linguistically diverse groups. T h e sociolinguistics of L a b o v a n d the 'the campaign of the Plain Language M o v e m e n t ' are sustained b y counter-ideologies against a hnguistic elitism which m o d e r n linguistics, in its aspirations to b e an ideology-free science, could not effectively combat as long as it m e r e l y replaced a simplistic, crude idealization with a complex, sophisticated one. Fairclough's writings imply that a counter-ideology would b e 'partial' and 'sustain relations of domination': [Critical discourse analysis] w o u l d argue that we are all [...] writing from within particular discursive practices, entailing particular interests, com mitments, inclusions, exclusions, a n d so forth; [. . .] Aspects of these discursive practices m a y serve to sustain relations of domination and m a y h e n c e b e ideological - n o theory or science is i m m u n e from that possibility. (1996: 53) Some optimism persists. His aspirations w o u l d thus b e that 'the values of [critical discourse analysis] d o not (unlike others) sustain relations of domin ation - they do n o t actually work ideologically' (1996: 49). Moreover, critical discourse analysis is theoretically better-placed to recognize its own 'partiality' t h a n m o s t theories. H e r e I pursue a different tactic than has b e e n advocated in these discussions. Instead of seeking to 'read off' the ideology of a given text (^aceFairclough), and also instead 'offering a single interpretation as uniquely validated b y the textual facts' {pace Widdowson), I examine a large set of contexts for the terms derived from one central t e r m which I think we would all peacefully grant to harbour ideological r e s o n a n c e s , a l t h o u g h (as w e shall see) these are unstable and
DE BEAUGRANDE DISCOURSE STUDIES AND THE IDEOLOGY OF 'LIBERALISM'
35
contested. M y analysis is not intended to b e authoritative, m u c h less 'validated' but rather heuristic in suggesting some topics a n d directions for the deeper an alysis r e c o m m e n d e d b y Fairclough. J u s t as corpus research continually points to questions we can pursue with a still larger or m o r e specialized corpus, so might discourse analysis working with corpora find u n e x p e c t e d leads toward relations between discourse a n d society which only b e c o m e well-defined whilst we exam ine large sets of authentic data.
E . O n t h e M e a n i n g s of 'Liberal' i n T h r e e C o r p o r a I chose the t e r m 'liberal' a n d its derivatives, such as 'liberalism' a n d 'liberal ization', because they have b e e n m u c h used of late b y groups with disparate interests in various meanings, e.g. for freedom from g o v e r n m e n t regulations and for solidarity of white p e o p l e with black people. T h e same tendency m a y well emerge from data-studies of other key terms in the discourses of post-modern society. I compare data drawn from three corpora representing different regions, the United Kingdom, the U n i t e d States, a n d South Africa. Before y o u read on, you might want to test your o w n intuitions about what the terms would m e a n if you used t h e m yourself. W h o are the 'liberals', and would y o u want to b e o n e of them? 2
E.1 United Kingdom Data . The Bank of English was developed at Birmingham University b yJ o h n Sinclair and his team, a n d as of J u n e 1996 h a d reached the unprecedented size of some 323 million words of running text from contemporary spoken and written sources (Sinclair, 1996). T h e U K sources in this corpus returned exactly 1000 lines centr ing on occurrences of the keyword 'liberal' b y itself or in its derivatives. I dis regard the 2 8 8 lines for 'Liberal Democrat(s)' (and 34 for 'Liberal Party') o n the grounds that the n a m e of a political party can b e notoriously arbitrary and op portunistic, witness the vastly different parties bearing this same n a m e in the U K , Russia, and J a p a n . Predictably, a fair n u m b e r of occurrences, especially for the Process-Verb 'liberalize' and its Noun-Form 'liberalization' concerned economic policy. Pro minent collocates included 'economic' a n d ' e c o n o m y ' themselves as well as technological progress', 'capitalism', 'financial policies', 'free market', 'trade', •protection', a n d 'privatization', with the ' I M F ' near at h a n d [1-4]. I underline the contextual clues I find relevant to determining the meanings of the key-words. 3
[1] Driven b y technological progress and economic liberalisation, global capitalism is changing fundamentally. [2] financial policies that are consistent over time; liberalisation a n d privat isation to p r o m o t e free markets. [3] about the economy. T h e I M F wants a n e w government to liberalise the economy a n d privatise public companies.
3fi
DISCOURSE STUDIES
[4] even essential. O n e is trade: to persuade voters that liberal trade is in their interests, less protection at h o m e . I n such contexts, the core m e a n i n g would plausibly b e : activities and policies designed to remove restraints o n the 'free' manipulation of the 'market' of goods, services, a n d labour. I n view of the adverse effects u p o n ordinary wage-earners and consumers, a deal of mystification m a y b e n e e d e d to 'persuade voters that liberal trade is in their interests' [4]. S o m e corpus data gave m o r e precise indications of what a 'liberal' system offers to 'investors', such as 'cheap labour' and 'few environmental restraints' [5]. T h e collocation 'liberalize + prices' (86 occurrences in the corpus) evidently offers a h a n d y e u p h e m i s m for raising prices with n o controls, a n d consumers can b e thankful if a 'small basket of foods' is exempted, though these m a y still b e unaffordable to the ' u n e m p l o y e d ' whose 'benefits' have b e e n 'cut' [6]. [5] in Taiwan finds cheap property, labour,, a n d raw materials, liberal invest m e n t laws and few environmental restraints. [6] the prices of everything b u t a small basket of foods would b e liberalised a n d generous u n e m p l o y m e n t benefits would b e cut. S u c h data m a y suggest w h y p u b l i c discourse m i g h t b e eager to main tain that 'capitalism' and 'liberal democracy' are 'inexorably' allied, e.g. b y assert ing that 'American micro-electronics accelerated the fall of c o m m u n i s m ' [7]. Yet that o p t i m i s m is belied b y d a t a o n societies like Chile, w h e r e ' e c o n o m i c liberalization' a n d 'market reforms' got so easily c o m b i n e d with 'political repres sion' [8], a n d b y data explicitly linking 'liberal' with 'rapaciousness' [9] a n d with opposition to 'populism' [10]. [7] a m o v e m e n t towards ever higher technological capability. This m o v e m e n t has led, h e argues, inexorably towards capitalism a n d liberal d e m o c r a c y . It w a s , for e x a m p l e , A m e r i c a n micro-electronics that threatened to r e n d e r obsolete the entire Soviet arsenal a n d thereby accelerated the fall of c o m m u n i s m . [8] D e m o c r a c y is well established after the traumatic Pinochet interlude. which combined political repression with economic uberalisation. Chile h a d a 10-year start over other Latin A m e r i c a n countries in m a r k e t re forms, a n d b e g a n bringing d o w n its own trade barriers without waiting for others to reciprocate. [9] A tried-and-tested alternative to rapacious western liberalism was at h a n d . India's colonial past still shapes [10] H e will n o t turn populist, b u t instead p u s h ahead with liberal policies a n d go faster with privatisation Particularly intriguing were the attestations r e m i n d i n g us that favouring big capital over labour a n d consumers is also a cherished principle of the traditional "
i
: J „ „ l „ r fi i _ i Hi P n r h a n c a m
wmanKr
DE BEAUGRANDE DISCOURSE STUDIES AND THE IDEOLOGY OF 'LIBERALISM'
37
merger is u n d e r w a y to give us not just occasional alliances on specific issues like opposing ' m o n o p o l y ' [13] - a stance m a n y current 'liberals' h a v e evidendy abandoned anyway - b u t a political species of 'liberal conservatives' [14-15], or at least 'conservative g o v e r n m e n t s ' that find it expedient to call themselves 'liberal' [16]. [11] viewing city d e v e l o p m e n t conservative a n d liberal plans, p l a n s to support the private m a r k e t [12] its A m e r i c a n d e m o c r a t i c i n h e r i t a n c e . T h a t c o n s e r v a t i v e - l i b e r a l philosophy remained, and still remains, powerful [13] I n theory, it [the argument against monopoly] could b e shared b y liberals eager to break the p o w e r of big business a n d conservatives intent on restoring some semblance of perfect competition [14] trace their ancestors to the liberal, parliamentary, elitist a n d m o d e r a t e conservatives served a n y regime which served them. [15] m a n y Conservatives, those w h o can b e called liberal Conservatives, w e r e thoroughly o p p o s e d to the increasing trend towards economic collectivism [16] voting system, it was n o t i m m e d i a t e l y k n o w n w h e t h e r the Liberal (conservative) government would continue to r u n T h e irony is all the richer in view of the habit of the British Conservative Party to [17] b l a m e everything that has gone w r o n g in British society o n the liberal 1960s rather t h a n o n their party which has b e e n in p o w e r since 1979. Toward the opposite pole of the political spectrum, a few data alluded to the prospect of being a 'liberal communist', b u t specifically in eastern Europe, where 'new political m o v e m e n t s ' [18] or 'expulsion from the Party' would b e i m m i n e n t consequences [19]: [18] But it w o u l d also encourage millions of liberal communists to form new political m o v e m e n t s [19] largely communists expelled from the Party for their liberal views. T h e second was the so-called Left Alternative, T h e core m e a n i n g of 'liberal' could still b e : policies designed to r e m o v e re straints o n the 'free' manipulation of the 'market', b u t the implications w o u l d b e starkly different within a strictly state-controlled economic system. I n view of the later deterioration of the Russian economy, the western economists w h o congratulate 'capitalism' o n its victory over 'communism' (e.g. in sample [7]) might ask themselves (or the Russians) if the h o r r e n d o u s effects of n a k e d profiteering and corruption w e r e p e r h a p s just results of the most thorough a n d relentless ap plication of 'capitalism' to a society which h a d n o understanding of h o w to con trol and modulate it. I was surprised to find 'bourgeois liberals' being opposed to 'affluent prole tarians' until I noticed the context concerning h o w 'jokes are b e c o m i n g less
38
i
DISCOURSE STUDIES
[20] Unfortunately, as our h u m a n e , bourgeois liberal culture gives w a y to the affluent proletarian mass culture of Essex a n d Hollywood, our jokes b e c o m e less subtle a n d a n e w element begins to emerge. [21] So w h y is h e so popular? Because thousands love his sexist, racist a n d deformed-people jokes a n d h e has brilliant comic timing. Believes that wishy-washy white liberal do-gooders are u n d e r m i n i n g t h e soul of the British comedy.
H e r e , being 'liberal' would m e a n defending racial minorities, w o m e n , a n d the handicapped against the public ridicule which, we are briskly told, 'thousands love' a n d which constitutes 'the soul of the British comedy'. I have always wanted to ask the users of the dismissive label 'do-gooders' if they admire 'do-badders' a n d aspire to b e a m o n g them. 'Liberals' w h o d o defend racial minorities a n d w o m e n might b e targeted b y the N e w Right conservative backlash as 'radicals', b u t I found only o n e instance in U K data [22], plus o n e in which a concern for 'minority cultural rights' was contrasted with the 'language of tradition' a n d was 'despised' [23]. T h e routine attack u p o n liberalism in the U K data was closer to the opposite: being mindlessly uncommitted a n d vague in one's good intentions [24-26], rather than, say, being 'tough o n crime' [27]. [22] A n d that fits into radical sort of liberal adult education awareness for some encouraging [23] language of tradition b u t instead deploy the despised liberal rhetoric of minority cultural rights. [24] H o w should o n e describe Bresson's cinema? A liberal humanist ap p r o a c h flounders in generalities [25] m e n like Raplin exist shows w h a t y o u get with a liberal agenda; You p u m p p e o p l e full of drugs a n d p i p e d r e a m s [26] W o m e n can't b e sexist as they are t h e oppressed? Bollocks. T h a t is a lily-livered liberal w a n k view a n d holds n o water with m e . I c a n spot massive violent prejudice w h e n I see it [27] his 'tough o n crime' policies u n i m p e d e d b y Tumim's liberal interference. For h o w long, though, is a m o o t point. T h e vulnerability of w o m e n e m e r g e d again w h e n they, rather than m e n , were bluntly called 'sexists' w h o h a r b o u r 'massive violent prejudice' [26], much as the victims of racism a n d their defenders a r e called 'racists' b y t h e New Right (Van Dijk, 1993, 1998b). A vicious twist with a long tradition was to cast aspersions ('lily-livered', 'wank') o n t h e m a n h o o d of 'liberals' w h o champion w o m e n ' s rights. I n sum, the U K data indicated that the m e a n i n g of'liberal' was most precisely determined in contexts of economic activity a n d policy, as a designation for 'privatizing' everything [2-3] a n d allowing t h e free market to r u n its course, unrestricted b y such annoyances as 'environmental restraints' [5] a n d 'unemploy m e n t h f l n p f i t s ' Tfil This stance haoDens to b e dear to the traditional opposite,
DE BEAUGRANDE DISCOURSE STUDIES AND THE IDEOLOGY OF 'LIBERALISM'
39
'conservative' ideology [11-15]. T h e economic m e a n i n g was usually contextualized amelioratively to u p h o l d the credo that [28] the virtues of the e c o n o m i c system that u n d e r p i n s liberal d e m o c r a c y significantly outweigh its vices along with congratulatory invocations of benefits like 'higher technological capability' [7]. For foreign investors, the 'cheap labour' a n d t h e lack of 'envir onmental restraints' in Taiwan [5] simply constitute attractions for shrewd busi nessmen, rather t h a n short-term a n d long-term degradations in the lives of the local inhabitants. Even 'political repression' can get contextualized into a laudable impulse for a ' h e a d start' in ' m a r k e t reforms' to 'bring d o w n the trade barriers' [8], w h e r e b y Pinochet t h a n k e d t h e U S a n d its C I A for overthrowing the d e m o cratically elected socialist g o v e r n m e n t a n d installing h i m as dictator. Intriguingly, the further away the contexts m o v e d from t h e e c o n o m i c m e a n ing, the m o r e pejorative they b e c a m e . O n l y a small part seemed to b e motivated by threadbare N e w Right tactic of 'blaming everything that h a s g o n e w r o n g in British society o n the liberal 1960s' [17]. A m o r e decisive motive s e e m e d to b e the trenchant hostility in public discourse of spokespersons a n d groups against a different type of restrictions, n a m e l y those protecting of the rights of w o m e n , minorities, a n d accused offenders of the law. Such protections w e r e defamed as the concern of'wishy-washy do-gooders' a n d 'lily-livered w a n k e r s ' [26], whilst the British public is being re-educated b y a 'comedy' of 'sexist, racist and deformedpeople jokes' [21]. So the ameliorative a n d pejorative contextualizations did n o t contradict each other after all, however m u c h they m a y contradict t h e spirit of genuine democracy. E.2 United States Data The U S data c a m e from the US-based sources in the Bank of English and totalled 287 lines. Since 'the p r o p o r t i o n of English from identifiable U S sources in t h e Bank of English is maintained at around 2 5 % ' (Sinclair, pers. c o m m . March 1998), the overall frequency of occurrences was n o t significantly lower t h a n for the U K data with 1000 occurrences for a v o l u m e three times as large, if w e set aside the party n a m e of 'Liberal Democrat(s)' with just seven occurrences in the US data vs 2 8 8 in the U K data. Perhaps the Americans, with their fossilized two-party system, take n o k e e n interest in discussing the political parties of other nations. Some data c o n c e r n e d the same economic activities a n d policies w e saw in the U K data, again with 'liberalizing prices' as a h a n d y e u p h e m i s m for freely jacking t h e m u p [29-30], a m o v e sometimes foisted u p o n a nation from the out side [30]. T h e n o t i o n that m a i n t a i n i n g a 'liberal free-market' m a k e s y o u a 'democrat' scaled t h e peaks of irony for the 'changed K h m e r R o u g e ' [31], b u t then who can p r e d i c t just h o w m a n y 'millions' m a y 'die' in the 'killing fields' of the 'free market' as it rationalizes p e o p l e out of their j o b s a n d abolishes welfare and u n e m p l o y m e n t rnmnprtat;™?
DISCOURSE STUDIES
40
[29] H e wants a m a r k e t e c o n o m y instituted rapidly, with liberalized prices. H e wants the state bureaucracies to d r o p the subsidies [30] price reform. J a n Vanous: If Russia liberalizes prices. Ukraine has n o choice b u t to follow [31] s o m e o n e million C a m b o d i a n s died in the infamous K h m e r R o u g e killing fields. Until this m o n t h , the K h m e r R o u g e a p p e a r e d intent o n proving to the world that they've changed a n d are n o w liberal, freem a r k e t democrats. Yet unlike the U K data, the U S data also reflected the contradictory view that 'liberals' are opposed to current economic 'realities' or cannot apply their prin ciples to these [32], a n d are out of touch with the b r o a d population anyway [33]: [32] is a little like Bill Clinton's: h o w to reconcile liberal principles with certain harsh realities [33] T h e vast majority of Americans are not liberals. T h e y should simply listen to the real speeches A l t e r n a t e l y , 'liberalism' w a s identified as t h e i d e o l o g y specific to the 'bourgeoisie' [34] a n d the 'middle class a n d the rich', where the t e r m ' m o n e y liberals' did seem piquantly apt [35]: [34] T h e bourgeoisie is assumed to b e necessarily liberal in politics a n d if a liberal order is not created then ,,, [35] This is the stuff of social equality. But virtually n o n e of these virtues are evident w h e n all the g o v e r n m e n t does is distribute cash benefits - even if, as m o n e y liberals typically r e c o m m e n d , benefits go to the middle class a n d rich as well as the poor. Another contrast to the U K data emerged where 'liberals' were said to actually favour social welfare programs [36-37]. I n b o t h cases, the contexts connoted disapproval, e.g. in suggesting that 'liberals' w e r e s o m e h o w acting unfairly by 'simply imposing taxes o n the rich minority' i n order to the 'fund the public sphere'. i
[36] t h o u g h t t h a t congressional D e m o c r a t s w e r e tax-and-spend liberals, w e d d e d to big government. [37] large tax increases needed to fund the public sphere - liberals can simply impose the increases on the rich minority 4
T h e diversity of social groups a m o n g or close to the 'liberals' was conspicuous in the U S data, e.g. ' h u m a n rights' advocates and environmentalists [38], feminists [39] (whose cause disqualifies m a l e supporters as 'lily-livered w a n k e r s ' [26]?), blacks [40], or w h i t e s p o k e s p e r s o n s for blacks [41] (the c o r e m e a n i n g in South Africa, as I later show):
DE BEAUGRANDE DISCOURSE STUDIES AND THE IDEOLOGY OF 'LIBERALISM'
41
[38] to recoup with a coalition of h u m a n rights-oriented liberals and antiwaste moderates [39] Socialist feminism in the United States r e e m e r g e d in the 1960s, w h e n liberal feminists b e c a m e frustrated with the p a c e of social reform and b e g a n to seek m o r e fundamental sources of w o m e n ' s oppression. By focusing o n w o m e n ' s economic roles [40] says, ' H e ' s got the w r o n g vision. H e ' s n o t liberal a n d most blacks are liberal' [41] It just is another e x a m p l e of benevolent white liberals dissecting a n d analyzing the attitudes of black people Discursive m o v e s to discredit 'liberal' ideology were m o r e imaginative t h a n in the U K data (back in [22-6]). T h e U S data variously portrayed it n o t just as wishy-washy [42-3], b u t also subservient to ethnic groups like Arabs in 'caftans' and African Americans in 'dashikis' [44], 'not progressive' [45], deficient in 'under standing the 'stubborn facts of h u m a n experience' [46], incapable of integrating 'affective life' with 'intellectual interests' [47], 'sanctimonious' [48], 'fearful' of the 'plebs' [49], or c o m p r o m i s e d if n o t i n d e e d criminalized b y the support of 'drug users' w h o hold a 'permissive a n d anti-conventional outlook' [50-1]. [42] all w e ever get to see is the conservative against the liberal, a n d the liberal has n o politics at all, anyhow. [43] the differences a m o n g individuals. This is because, to liberals, there is n o absolute truth, [44] of a caftan or whatever those things are the Yankee liberals are all run ning a r o u n d in now'. 'Dashikis' [45] an idea of a n agency to p u t even these into effect. T h e liberals of today are n o t a part of a progressive m o v e m e n t [46] Lacking a n y true insight into these stubborn facts of h u m a n experience corruption, evil, irrational desire - liberals also fail to understand that evil often lies b e y o n d purely rational treatment [47] T h e b o d y , the unconscious, the pre-rational are all important to s o u n d thought. But because the liberal has sought n o positive discipline for emotion a n d feeling, there is an o p e n b r e a c h b e t w e e n his affective life and his intellectual interests. [48] G O P ticket in 1952. A n d the privileges of sanctimonious liberals chafed against his [Nixon's] prepolitical past: a childhood [49] long e n o u g h to explore. I n a n y event, a m o n g most French liberals, fear of the plebs outran their h o p e s for it. [50] These veterans of the '60s era of sex a n d drugs a n d rock'n'roll support a liberal social agenda. A n d they're m o r e attuned to the special-interest groups linked to the Democratic Party than to the big-business asso ciations of the G O P . [51] a m o n g those w h o h a v e g o n e to college, d r u g users are m o r e likely to have majored in the social sciences, fine arts, a n d humanities than i n
42
DISCOURSE STUDIES
the natural sciences. T h e y are also m o r e likely to favor liberal politics, to b e estranged from religion, a n d to h a v e a generally permissive and anti-conventional o u d o o k T h e oddest linkage of all was m a d e a m o n g groups w h o are n o t convinced that 'the principle of free speech' should protect the hate speech of 'racists': 'bigots' got juxtaposed with 'liberals', and 'feminists' with 'sexists' (compare sample [26] again). [52] n o t w h e n they engage in violence or vandalism. But w h e n they speak or write, racist assholes fall right into this Oliver Wendell H o l m e s defin ition - highly unpopular a m o n g bigots, liberals, radicals, feminists, sexists, " a n d college administrators: 'If there is any principle of the Constitution that m o r e imperatively calls for attachment than a n y other, it is the principle of free speech' Applying 'the principle of free speech' to 'hate speech', w h e r e the 'negative presentation of the outgroup' (Van Dijk, pers comm.) casts aside all restraint, is an alarming demonstration of h o w the principles of democracy can b e used to subvert it. This notion of 'freedom' is difficult to oppose from a n 'ideology' of 'liberalism' that purports to 'outflank or transcend ideology' and puts its 'faith' in 'reason as a faculty that operates i n d e p e n d e n d y of any particular world view' (Fish, 1994: 134 ff.). At all events, 'liberals' in the U S can expect a barrage of 'attacks' [53], and even their 'contented m o d e r a t i o n ' is a n a t h e m a to the 'red-hunters o n the 'Right' [54]:
:
[53] today a n d continued his attacks o n what h e calls the liberal establish ment. His m a i n target was N e w York, a city [54] For the time being, the tones of contented moderation h a d driven out the discourse of anguished zeal. Great Liberal Fear. But r e d hunters on the Right gave liberal intellectuals a terrible shock.
N o t at all surprisingly, U S politicians were far m o r e wary than in the U K of being classed as 'liberal' [55-6], a n d 'measures' to 'repeal liberal rights' are being busily drawn u p even (or especially?) for California [57]. [55] clear from Clinton's reactions over the weekend to the liberal label, him fighting back, that this ticket is n o t liberal [56] M c G o v e r n : Well, I think Bill Clinton is probably m o r e liberal than his present image. I h o p e m y saying that w o n ' t h a r m his campaign [57] the court overturned a portion of the initiative that w o u l d h a v e stripped defendants of any rights that aren't provided b y the U S Constitution. T h a t part of the measure was designed to repeal m o r e liberal rights provisions in the California constitution. ->
DE BEAUGRANDE DISCOURSE STUDIES AND THE IDEOLOGY OF 'LIBERALISM'
43
Scorning 'liberals' for a failure to b e 'tough o n crime' (sample [27]) is an ironic argument in an age where the 'liberal free-market' is rapidly eroding opportunities for honest work. T h e old antipathy b e t w e e n liberalism a n d religious doctrine, which has furnished history with s o m e b e m u s i n g sidelights [58] (but c o m p a r e sample [94] later), is currendy focused u p o n the inflammatory issue of 'abortion' [59-60], where being 'liberal' m e a n s allowing the w o m e n involved to decide for them selves without b e i n g threatened b y criminal prosecution. [58] W h e n in 1889 a canon of Santo Domingo d e Calzado w h o h a d preached against freedom of the press a n d the ' d a m n e d error of liberalism' a n d called it a sin for Catholics to vote for Liberals, was brought to trial for 'having c o n d e m n e d political liberalism from the pulpit' h e was sup ported b y m a n y of the region's clergy. [59] h e argued for better social services a n d m o r e liberal abortion laws. But the sharpest criticism today [60] that voters h a v e overwhelmingly rejected a proposal to liberalize the country's strict abortion laws. Returns from H o w this m e a n i n g might relate to economic ideologies is a d e e p question indeed. Is the fierce opposition to abortion (a) simply a gaudy political foot ball of the N e w Right; (b) the centrepiece of a b r o a d attack u p o n feminism a n d the rights of w o m e n ; (c) a ploy to gain m o r a l high grounds for the ideology of 'Christian fundamentalism' which can accuse the 'liberal' society of slaughtering the innocents; (d) a measure to encourage an oversupply of cheap labour in coming generations; or (e) some queasy mixture of all of these? H e r e is another issue for in-depth research. T h e U S data signalled a curious m i x of disparate visions regarding the history of 'liberal' ideology [61-3], also disagreeing o n whether or n o t it is currendy coming or going [62-8]: [61] rests u p o n a quite different set of premises. Liberalism in this sense was symbolically a child of Voltaire [62] Outside a small circle of theorists, liberalism c o n n o t e d a worldview defined in the 1930s a n d reborn in the early 1960s that n o longer inspired m a n y activists or voters. [63] the N e w Republic is profoundly distressed to discover that liberalism has u n d e r g o n e an eclipse since 1912 w h e n the bulk of [64] argues that the Western idea of economic and political liberalism has triumphed over alternatives such as fascism. [65] own credit and did a little something to vindicate liberalism b y opposing unjustifiable appeals to arms. [66] they will h a v e to p a y m o r e airention to the discrepancies in their record. T h e y will h a v e to either shift liberalism to n e w a n d stronger ground or repair the manifest breaches in their defenses, liberals deserve the eclipse from which liberalism is suffering.
44
DISCOURSE STUDIES
[67] the liberal remains u n p r e p a r e d to face the worst: a n d o n the brink of w h a t m a y turn out another D a r k Ages h e continues to scan the horizon for signs of dawn. T h e record of liberalism during the last d e c a d e has b e e n shameful evasion a n d inept retreat. Liberalism has compromised with despotism because despotism p r o m i s e d economic benefits to the masses [68] Pragmatic liberalism has flatly betrayed ideal liberalism. T h e values that belong to the latter h a v e b e e n c o m p r o m i s e d away, vitiated, ruth lessly cast overboard. T h e p e r m a n e n t heritage of liberalism has b e e n bartered for the essentially ignoble notion of national security, in itself a gross illusion T
Taken all together, the U S data were substantially less focused t h a n the U K data, though important similarities could readily b e detected. Perhaps because there is n o respected 'Liberal Party' in U S politics, the t e r m has not stabilized a n ameliorative economic meaning, b u t has b e e n heavily coloured b y the 'taxand-spend' a n a t h e m a accompanying the long-term attack o n social welfare ironically, a policy elsewhere proudly called 'liberal'. T h e non-economic meaning of 'liberal' is evidendy b e i n g blurred b y the highly vocal N e w Right offensive, enlisting the t e r m to cast scorn u p o n their long a n d diffuse list of enemies all across society, including n o t just minorities a n d w o m e n (especially feminists), but also, intellectuals, environmentalists, a n d advocates of civil rights or of curbs o n the economic p o w e r of the private capital that bankrolls the N e w Right. The blurring is c o n s u m m a t e w h e n those w h o obviously fall u n d e r the traditional economic core m e a n i n g of the term, such as Bill Clinton (contrast [32] with [55]), repudiate the label. i :'.
E.3 South Africa Data T h e Corpus of South African English (CSAE) has recentiy b e e n established at the University of Port Elizabeth u n d e r the supervision of ChrisJeffries a n d Linda Pearce Williams, a n d has grown from o n e million to s o m e three million words. This corpus was originally inaugurated as part of the International Corpus of English (ICE) u n d e r the direction of the late Sidney G r e e n b a u m at University College L o n d o n (Greenbaum, 1996). T h e s e corpora w o u l d provide, for the first time in history, a base of authentic data about local varieties of English that h a v e hitherto led a contradictory existence of being used b y large populations but rarely recognized as valid alternative language systems. Linda Pearce Williams has kindly supplied the total of 247 data lines from the CSAE. This total represents a vastly higher proportional frequency rate than for the U K a n d U S data from the Bank of English, which is roughly 100 times larger. To appreciate why, w e should k e e p in m i n d that in South Africa, the terms 'liberal' a n d 'liberalism' - b u t n o t 'liberalize' a n d 'liberalization', these b e i n g m o r e recent economic terms - h a v e long carried the highly specific mean i n g of mostly white p e o p l e a n d policies that p r o m o t e d the h u m a n rights of black Africans. S u c h a m e a n i n g aroused m u c h discussion a n d controversy in 5
DE BEAUGRANDE
DISCOURSE STUDIES AND THE IDEOLOGY OF 'LIBERALISM'
45
the grim landscape of apartheid especially the founding of the Liberal Party in 1953. It was eventually oudawed in 1968 b y a decree with the amazingly candid name, for a supposedly parliamentary state, of 'Prohibition of Political Inter ference Act'. Thus, the only 'Liberal Party' in m y data to b e actually b a n n e d was also the only o n e whose ideological position was most sharply defined in terms of promoting h u m a n rights. Being 'liberal' in South Africa as the thumbscrews of apartheid were ruthlessly tightened would expose y o u to drastic hostility of the g o v e r n m e n t a n d its allies in institutions a n d the media, in the form of house arrests, 'banning orders' a n d so on, but - presumably to avoid galvanizing white resistance - usually stopping short of the m a r t y r d o m b y detention, torture, a n d killing, visited u p o n black activists like Steve Biko, the key advocate of 'Black Consciousness'. Moreover, some black Africans mistrusted t h e 'liberals' of trying to preserve the system in a milder modality to k e e p from radically transforming it. Biko himself once des cribed 'the liberal establishment, including radical a n d leftist groups', as 'a curious bunch of non-conformists' a n d 'do-gooders' (that slap again!) whose 'artificially integrated circles are a soporific to the blacks while salving the conscience of the guilt-stricken whites' (1978 [orig. 1972]: 63 ff.). In view of this historical background, the high frequency of occurrences in the CSAE is easily understood. M o r e specifically, we can predict that the data will indicate h o w the ideology of 'liberalism' is being forcefully unsettled b y the establishment of a democratic majority g o v e r n m e n t in 1994. T h e uneasiness can be pungently sensed w h e n the Liberal Students Association at Wits University ruefully a n n o u n c e d a lecture on 'Resuscitating Liberalism'. Reflections about the past 'mission' [69] were typically accompanied b y per plexity on having lost it [70-74]. S o m e contextual cues, like 'pivoted a n d sold' [72], 'basking' a n d 'worshipful disenfranchised' [73], or 'jealousy' [74] m a y b e symptoms of a spiteful serve-them-right tone of those w h o opposed the liberals all along. [69] the liberal students of that time h a d a very clear mission in life. O u r mission was to oppose apartheid. T h a t was our purpose [70] Some people say that t h e liberals' day was yesterday, a day w h e n we were secure in the conducting of a holy war [71] As I look at this hall I see some things h a v e n ' t changed. But the cause has changed. I d o n ' t k n o w what liberalism is or m e a n s now. [72] T h e moral high ground that such liberals occupy is because they have been pivoted a n d sold to black people as true freedom fighters. [73] they have lost the formerly serene sense of basking in the gratitude of the worshipful disenfranchised. As the bearer of a famous liberal n a m e told m e n o t l o n g ago: they d o n ' t n e e d u s any m o r e , y o u k n o w \74\ I think there is a jealousy h e r e . T h e r e are liberals, so-called, w h o feel displaced. T h e y ' r e n o longer the black m a n ' s favourite whites A p r o m i n e n t representation of 'university' influence was also indicated 175-6], with hints of elitism and radicalism even as contrasted with other 'liberals'.
46
DISCOURSE STUDIES
e.g. respecting 'one-man-one-vote' suffrage [76], which was a rallying motto central issue b o t h for the Black Consciousness M o v e m e n t and the African National Congress (Biko, 1979: 126 ff.; Mandela, 1994: 2 3 9 ff.). [75] w e took it for granted that being liberal m e a n t being a university person. or ex- or child-of. Also, being white. [76] I t was liberal to stand u p for a better deal for the blacks, even for what was t h e n called ' o n e - m a n - o n e - v o t e ' , w h i c h post-university liberals considered pretty terroristic in itself Elitism was clearly one t h e m e for disparaging white liberals: [77] white 'liberals', however g o o d their intentions, seem oblivious to their racial socialisation. Too many, of'liberal' and sometimes 'left' persuasion, r e m a i n in a comfortable white social cocoon [78] smacks of the typical liberal naivety a n d paternalism b o r n out of white privilege But t h e n only white people w h o were well-situated could h a v e h a d any effective political impact during the apartheid era; a n d those people did after all take a stand against the very system u n d e r which they h a d prospered a n d which could b e expected to sharply curb their 'privileges'. Today, their status as critical voices u n d e r the n e w system is being exploited to represent t h e m having turned against 'blacks' [79], a n d to disqualify t h e m as 'hypocrites' [80] a n d even 'racists' [81]. [79] I n the old days the Afrikaner establishment saw liberalism as a h o b b y , of upper-class English-speakers w h o disliked the Afrikaners m o r e than they disliked the Africans. N o w the n e w liberated establishment sees liberals as upper-class English-speakers taking whacks at blacks, o n every available grounds [80] w e must expose First World hypocrisy b y speeding u p the democratic revolution in South Africa, for that is what your pathetic liberal drivel is really against. [81] We spent all of the afternoon complaining about racist liberals (isn't it refreshing to find n e w ogres?). Despite the sweeping changes, criticism of the n e w social order is subject to m u c h the same derision as was criticism of the old: [82] the p r o p e r liberal role is to m a k e p e o p l e cross with you. You've got to b e gadfly, and if the r u m p of state does n o t get irritated [83] Liberals ought to b e over the m o o n , b u t in fact there is endless carping and griping. O p e n a paper, a n d there is some spokesperson of liberalism running something down, predictable a n d tiresome. [84] the corpus of liberalism, which in the public m i n d b e c o m e s associated
DE BEAUGRANDE DISCOURSE STUDIES AND THE IDEOLOGY OF'LIBERALISM'
47
Small w o n d e r if s o m e s u p p o r t e r s of liberalism feel e x a s p e r a t e d at p u b l i c ingratitude: [85] 'Liberal' is n o w a dirty w o r d a m o n g those w h o w e r e n o t even b o r n w h e n liberals fought for their rights. In contrast to the U K data, the South African data nowhere associated 'liberal' with 'conservative', a term in the n a m e of a political party o n the far right in the apartheid context, witness such sinister data as these: [86] A m n e s t y has also b e e n granted to Conservative Party and Volksfront m e m b e r Saint Michael Schutte for being in possession of seven AK-47 rifles [87] a Conservative Party M P , D P du Plessis, repudiated fellow C P M P Koos Botha for b o m b i n g a school e a r m a r k e d for black students. [88] Convicted H a n i m u r d e r e r [Chris H a n i , secretary-general of the South African C o m m u n i s t Party, shot d o w n in April 1993 to disrupt the tran sition to democracy] a n d former Conservative Party politician Clive Derby-Lewis continued his testimony before the truth commission's amnesty committee But 'liberal' was multiply associated with ' c o m m u n i s t ' [ 8 9 - 9 0 , 9 3 - 9 4 ] , although in a totally different meaning than we saw to b e specific to eastern Europe in samples [18-19]. T h e apartheid regime h a d indiscriminately pasted the label on all their adversaries and persecuted them u n d e r the 'Suppression of C o m m u n ism Act' of 1950. EvenJ o e Slovo, long-time General Secretary of the South African Communist Party a n d imprisoned o n R o b b e n Island together with the leaders of the African National Congress, was suspected of'liberal' tendencies [91] despite being reputed (like Biko) to regard 'liberals' as 'hypocrites' [92]. [89] It was the h o u r of the raised voice of the liberals w h o saw themselves as the conscience of the nation. A n d it was also the time of the communist voices [90] newspapers were seen to b e unpatriotic reckless liberalistic communistic pinko. You k n o w y o u k n o w all those all those labels [91] T h e r e are those w h o think that Slovo r e m a i n e d a true believer to the end, that the m o r e liberal formulations of his last years were merely another tactical feint [92] For Slovo, of course, all liberals are hypocrites and h e is happily illiberal. In m o r e academic registers, the two ideologies were idiosyncratically linked in terms of their world-views [93-4], along with the 'semi-socialist ideas of the Christian culture' [94] (but c o m p a r e sample [58]): [93] the vacuity of the communist vision of h u m a n nature is equally discern ible in the liberal anthropology of rationalist, autonomous individuality:
48
DISCOURSE STUDIES
it cannot h o p e to escape the fate which befell its c o m p a n i o n ideology. I n this sense liberalism a n d c o m m u n i s m are b u t the two sides of a single Modernist coin [94] This vociferous form feeds o n the radical liberal perspectives a n d some times semi-socialist ideas of the Christian culture of the educated Nelson M a n d e l a (1994: 285) himself certified in 1962 that 'the Liberal Party and the C o m m u n i s t Party were arch-enemies', b u t h e too noticed h o w the terms were getting blurred together in o r d e r to discredit t h e m both. T h e academic data that collocated 'liberal' with 'egalitarian' a n d 'equality' were sternly pejorative: [95] the appeal to basic rights can settle only a modest n u m b e r of contentious issues, a n d the tendency of egalitarian liberal theory to m a k e all con tentious issues a question of right, [and to] blur the separation of the public a n d private spheres has even m o r e disastrous consequences [96] the d o m i n a n t liberal obsession with equality leads to incoherence in theory a n d increasing conflict a n d antagonism in practice After reviewing all these pejorative data we m a y feel a littie wistful to h e a r recent calls to action like this one: [97] J o i n the Democratic Party if y o u share its c o m m i t m e n t to the values of liberalism a n d democracy a n d its vision of fighting for the freedom of every person to shape his or h e r o w n future. T h e contexts were quite different a n d far less n u m e r o u s for 'liberalization' in the m e a n i n g w e saw in the U K data. Perhaps to avoid confusion with 'liberal' in the traditional South African meaning, 'neo-liberal' is currently used for the economic meaning. I found the same usage 2 2 4 times in the U K data but, inter estingly enough, not even once in the U S data, where we noticed some confusion about the history of the ideology ([61-8]). T h e data displayed the typical collocatable terms like 'invest' [98], 'global isation' [99], 'competition' [100], a n d 'privatisation' [100]. T h e 'suspicion' a n d 'hostility' of labour unions [101] provided clues about the implied meanings: 'invest offshore' [98] -> evade paying the national income taxes vitally n e e d e d for better social services; 'join the international e c o n o m y ' [99] -» b o w to the will of multinational corporations a n d b a n k s ; 'fiscal discipline' [100] -> n o upgrading social welfare p r o g r a m m e s ; 'flexible labour markets' [100] -» p o w e r of owners and m a n a g e m e n t to fix wages; ' i m p r o v e d competition' + 'export drive' [100] -» undersell other countries b y u n d e r p a y i n g the productive workforce at h o m e (cf. d e Beaugrande a n d Williams, in press). If [102] was a bit m o r e frank, [103] was extremely so in calling these economic policies b y some well-earned names. [98] foreign exchange controls will b e liberalised b y allowing individuals to invest offshore for the first time
DEBEAUGRANDE DISCOURSE STUDIES AND THE IDEOLOGY OF 'LIBERALISM'
49
[99] the whole business of globalisation and liberalisation. We are n o w u n d e r e n o r m o u s pressure as a country to join the international e c o n o m y [100] the foundation proposes a 'multi-pillared' strategy of legal reform, financial liberalisation, fiscal discipline, flexible labour markets, im p r o v e d competition a n d an export drive. [101] union m o v e m e n t s , which n o w view privatisation with suspicion a n d hostility and which h a v e expressed misgivings about trade liberalisation [102] A n y finance obtained from the World B a n k would b e o n policy terms dictated b y it. T h e usual neo-liberal conditions - privatisation, liberal isation of trade, m o n e t a r y restraint or high interest rates, deregulation, cutting state subsidies, a n d so on. [103] the global c o m m u n i t y w e are supposed to b e p a r t of. T h e shift to the left is primarily a resounding rejection of neo-liberal society, in its un caring, alienated politics, as well as its free m a r k e t e c o n o m i c s of u n a b a s h e d profiteering We m i g h t well speculate that the future well-being of t h e d e m o c r a t i c South Africa will d e p e n d chiefly o n whether the old 'liberalism' of egalitarian h u m a n rights can withstand the ' e n o r m o u s pressure' from the 'neo-liberalism' of 'uncaring, alienated politics' a n d 'unabashed profiteering', whilst the h u g e majority of black Africans continue to live in poverty a n d hopelessness.
F. Further Issues for Analysis T h e data presented h e r e are but a small sample of 103 passages out of the total of 1534 data samples examined - not even o n e tenth, although still far m o r e real data than y o u will find in m a n y publications b y linguists, especially b y ones who idealize language. To work as a discourse analyst confronting plentiful corpus data, you are compelled to b e selective a n d to rely in early stages on your o w n expectations a n d intuitions about which data might b e m o r e relevant or inter esting. Later, the parameters you find will send y o u back to the larger sets in search of m o r e specific or subtie indicators. Obviously, the m e t h o d s for doing a 'critical discourse analysis' of corpus data are far from established yet. Even w h e n we have e x a m i n e d a fairly large set of attestations, w e cannot b e certain whether our own interpretations of key items and collocations are genuinely representative of the large populations who produced the data. But we can b e fairly confident of accessing a range of interpretative issues that is b o t h wider a n d m o r e precise than we could access by relying on our o w n personal usages and intuitions. Moreover, when we observe our own ideological position in contest with others, w e are less likely to overlook it or take it for granted. Nor can I point to any consensus about which a m o n g the m a n y analytic techniques of previous linguistics w e might apply. D o we focus o n morphology, syntax, semantics, or pragmatics, or some combination of these (de Beaugrande, 1996)? O r d o we focus o n less familiar factors like 'colligability' and 'collocahilitv'
50
DISCOURSE STUDIES
which cut across all of those in ways we are just beginning to grasp (de Beaugrande, 1998b)? O r , again, d o w e focus o n the 'discursive structures, strategies, and moves' w h e r e b y ideological discourse enacts, 'at all levels of the text, the positive presentation of the ingroup a n d the negative presentation of the out-group' b y m e a n s of 'polarisation', 'metaphor', 'hyperbolic emphasis', 'negative lexicalisation', 'rhetorical mitigation', 'euphemisation', a n d so on (Van Dijk, 1998b: 18)? Moreover, what are the implications of using key words as the units of search and analysis when our major interest is in contexts and in the positions of speakers (Lemke, pers. comm.)? We plainly n e e d some m e a n s for searching and analysing relations a m o n g meanings or belief-systems as well, but the necessary software would d e m a n d a highly innovative design. To b e consistent with the principles described i n section C, w e cannot setde these questions from the top down, b u t only from our ongoing b o t t o m - u p work with plentiful data. N o t merely do corpus data in n o w a y manifest the disorder projected b y 'linguisticism' since d e Saussure, they manifest an amazingly delicate order despite their e n o r m o u s quantity and variety. To deal with all that, we shall n e e d to pursue multiple inquiries guided a n d co-ordinated b y some evolving dialectical agenda, w h e r e a 'critical' ideology could strategically guide our focus toward interconnected data that are indicative of p o w e r a n d domination but also others of solidarity a n d equality. Consider in this connection 'the principal observation of corpus linguistics in the last decade': 'meaning affects the structure profoundly' (Sinclair, 1991:496). O n e significant reflex of this is that the collocability a n d colligability of any one item, such as a N o u n or a Verb, are unequally distributed a m o n g its alternative forms, such as Singular and Plural (Sinclair, 1991: 494 ff.). I in turn found that the Adjective 'liberal' is far m o r e variable a n d contested in its collocations than the N o u n 'liberalism', whilst the Verb 'liberalise' or (in U S data) 'liberalize' is by far the most restricted. T h e (invariably T h i r d Person) Subjects of the Verb were nearly always governments, institutions, or whole countries, all representing e n o u g h p o w e r to m a k e resistance seem futile. T h e Objects w e r e nearly always abstractions, coming mainly from a single semantic d o m a i n with the headword 'economy', such as 'prices', 'trade', 'markets', 'financial system', a n d 'tariffs'; 'laws' appeared at a m u c h lower frequency. A few data signalled ongoing evolutions in usage, such as using the Verb in Intransitive colligations, e.g.: 'two regimes which, having liberalised u n d e r pressures'; a n d 'steps were taken to explore whether Tanzania should liberalise politically as well as economically'. Clear innovations included the 'closet liberalisers' in the British Parliament, a n d the way Poland a n d H u n g a r y w e r e said to ' r u n i m p o r t regimes that out-liberalise those of m a n y O E C D countries'. I can see n o rational w a y to account for all these emergent constraints except b y reference to the ideological status of the key words. Whilst engaging with such data, you m a y detect a dual evolution in your interpretative work: b o t h toward a convergence of meanings for a given key word a m o n g sets of data; a n d toward a divergence between sets so loosely associated as to m a k e a c o m m o n core meaning quite difficult to formulate. W h a t might b e the c o m m o n core shared b y the set of economic contexts (these from the U K data)
DE BEAUGRANDE DISCOURSE STUDIES AND THE IDEOLOGY OF 'LIBERALISM'
51
such as 'open and liberal economic policies' or 'the money-capital concept linked to the classical liberal ideology', a n d b y the set of ironic goodwill contexts such as 'liberal Christian g o d ' or 'dressing u p their doctrine with a liberal sprinkling of post-Daisy A g e h a p p y harmonies', or again b y the set of plenitude contexts such as ' a liberal supply of live food including small insects' or ' a liberal appli cation of face p o w d e r o n top of foundation'? O r might the expectation that a given expression has a stable and determinate core meaning b e u n d u l y proximate to the 'linguisticist' vision of language being a stable a n d determinate system? Corpus data m a y b e telling us that the m e a n i n g of an expression typically requires a b a n d of fluctuation a n d indeterminacy in order to adapt so effortlessly to emergent constraints during discourse. Contest ations over meanings are o n e implication of this b a n d , as w h e n the connotations of'liberal' with generosity a n d freedom can b e handily exploited for policies of selfishness and compulsion. Still, the fluctuation could plausibly b e constrained b y the attractions of ana logies among the meanings. We might consider if whether the m e a n i n g of 'liberal' as 'wishy-washy' [26] m a y stand in some analogy to the officially laissez-faire policies of the core economic meaning, or is merely a n e m p t y projection of 'negative lexicalization' (in the sense of Van Dijk, 1998b) to dismiss everyone to the left of the N e w Right, whatever their ideologies. If the analogy is implied, then it is itselfjust manipulative as the 'free' in 'free market'. T h e massive evidence marshalled b y Martin a n d S c h u m a n n (1996) a m o n g others demonstrates that economic 'liberalism' in its current form is far from laissez-faire; multinational corporations, the World Bank, a n d the International M o n e t a r y Fund, are literally forcing their 'free market' ideology onto whole populations a n d governments with the conviction that continual wage reductions, longer work-hours, cuts in social welfare, even the complete a b a n d o n m e n t of social welfare in the U S , are supposed to make the nations 'fit' for global competition. (Martin and Schumann, 1996: 312; m y translation) The two financial journalists are m o v e d to ask: 'how m u c h free m a r k e t can democracy stand?' Now w h a t the founders of the post-war welfare states h a d learned from bitter experience is becoming ever m o r e clearly visible: free-market econ omy and d e m o c r a c y are b y n o m e a n s inseparable blood-brothers, w h o harmoniously nourish prosperity for everyone. Instead, t h e two central guiding models of the old industrial states of the West stand in constant contradiction to each other. (1996: 311; m y translation) We could r u n extensive data queries to detect the m e a n s w h e r e b y public discourses in countries like the U K a n d the U S carry a h e a v y 'ideological invest ment' (Fairclough) in p a p e r i n g over the 'contradiction', e.g. b y thematizing the 'virtues of the economic system that underpins liberal democracy' [281.
52
DISCOURSE STUDIES
We could also ran multiple data queries to observe a n d describe the emer gence of a set of evasive terms within the ideological proximity of economic neoliberalism to designate a n d yet obscure the complex a n d devious tactics for profiteering at the expense of ordinary workers: 'rationalizing', 'privatizing', 'restructuring', 're-engineering', 'downsizing', 'outsourcing', 'multiskilling', 'increasing productivity', a n d so on, all m e a n : fewer j o b s with h a r d e r work for lower wages a n d n o benefits at all, whilst the gains for owners and shareholders soar. T h e New York Times calculated from the statistics of the U S D e p a r t m e n t of Labor that, between 1979 a n d 1 9 9 5 , 4 3 million workers lost their j o b s ; a n d two thirds of those w h o found other j o b s were forced to accept m u c h poorer working conditions (reported in the International Herald Tribune, 6 M a r c h 1996). T h e n irony is consummated w h e n neo-liberalism adopts the strategy of claim ing to b e n o ideology at all. T h e E u r o p e a n U n i o n Commissioner for Competi tiveness, Karl v a n Miert, recently defended privatization b y declaring: 'the decision to liberalize certain branches which offer public services is by no means ideological, but the expression of a natural adaptation to economic a n d technical developments' (LeMonde diplomatique,]aaxxaxy 1996, m y translation and italics). His 'choice of words', Martin a n d S c h u m a n n (1997: 190 ff.) c o m m e n t , 'already b e t r a y s t h e u n e x a m i n e d i d e o l o g y t h a t is always r e c o g n i s a b l e w h e n pol iticians invoke "nature" whilst h a n d i n g out state properties, tax revenues, and economic privileges'. T h e echoes in Fairclough's concept of 'naturalization' are highly serendip itous here. We can return to our data to look for m o r e subde choices of language intended to m a k e 'liberalism' s e e m natural if n o t virtually inevitable. I would highlight what is underlined in these data: [la] D r i v e n b y technological progress [2a] financial policies that are consistent over time [7a] a m o v e m e n t towards ever higher technological capability [...] has led [...] inexorably towards capitalism a n d liberal d e m o c r a c y [10a] H e will [. . .] p u s h a h e a d with liberal policies a n d go faster with privatisation [29a] H e wants a m a r k e t e c o n o m y instituted rapidly [30a] If Russia liberalizes prices, U k r a i n e has n o choice b u t to follow [12a] T h a t conservative-liberal p h i l o s o p h y r e m a i n e d , a n d still remains, powerful [34a] T h e bourgeoisie is assumed to b e necessarily liberal [64a] the Western idea of economic a n d political liberalism has triumphed over alternatives W h e n efforts are m a d e to 'persuade voters that liberal trade is in their interests' [4], the contradiction is to b e actually reproduced in the awareness of the population. T h e y are to see themselves only as shoppers wanting cheaper goods, a n d to forget being also workers w h o will get lower wages to shop with w h e n 'protection at h o m e ' has b e e n abolished.
D E B E A U G R A N D E DISCOURSE STUDIES AND THE IDEOLOGY OF 'LIBERALISM'
53
However, w e also n e e d to consider the prospect that ideological terms m a y BO/appear in a corpus with a frequency corresponding to their influence, because they h a v e b e e n quietly i n c o r p o r a t e d into t h e state policies. T h e radically 'conservative' R e a g a n - B u s h governments a n d their supporters in the business and financial communities did n o t h a v e to confuse the voters b y expressly ad vocating 'liberal' policies, leaving their N e w Right allies free to hurl the term 'liberals' at the advocates of the rights of w o m e n a n d minorities and of social welfare p r o g r a m m e s . Yet o n an international plane, 'liberal' remains in vogue as an agreeable label for of economic policies that could b e m o r e honestly called 'unabashed profiteering'. H e r e too, instability in the m e a n i n g is a n ideolo gical advantage. Still another advantage of the instability is to h i n d e r the consolidation of a n effective counter-ideology. W h a t can we rationally call t h e opposite of 'liberal' policies w h e n the usual converse 'conservative' is n o t m e r e l y unattractive for critical thinkers a n d ecologically oriented intellectuals, b u t is also b e c o m ing hardly distinguishable from 'liberal' anyway, as our data indicated? We certainly would n o t declare o u r s e l v e s ' illiberal' which the Random House Webster's variously defines as ' n a r r o w - m i n d e d , b i g o t e d , miserly, lacking culture or refinement' (p. 670). Now, if ideological terms are so strategic for the a d v a n c e m e n t of the ideology itself, a lack of accredited terms for counter-ideologies is a serious p r o b l e m . I have suggested t h e p r o b l e m to b e a general o n e , also holding for consumerism, individualism, capitalism, a n d free-marketism, n o w that the terms 'socialism' and ' c o m m u n i s m ' are losing favour. T h e p r o b l e m is i n d e e d o n e major motive for advocating a 'critical' enterprise that formulates a n d p r o m o t e s a n explicit counter-ideology such as 'ecologism', whose n a m e can profit from the serendip itous echoes with the 'ecology m o v e m e n t ' (e.g. Myers et al., 1993) a n d also with Neisser's (1976: 2) concept of 'ecological validity'. I certainly could not d e n y that the sincere ambition of critical discourse ana lysts to transcend ideology is admirable, since their goal is to overcome practices of power and domination, as w e have seen i n section D . But I feel uneasy recalling the ominous precedents of various groups, including scientists and economists, whose self-serving denials to h o l d any ideology constituted one major motive for critical discourse analysis in the first place. O r , if we agree to call the 'ideology' (in m y sense) of critical discourse analysis b y s o m e other n a m e - say, 'position', 'agenda', or ' p r o g r a m m e ' - w e risk being scolded for terminological squeamishness and also risk obscuring the oppositional status of our p r o g r a m m e vis-a-vis the ideologies of domination. Besides, critical discourse analysis currently does have power, as manifested in being a favoured d o m a i n for books, journals, a n d even academic j o b descriptions - a n d also a conspicuous target for continual attacks such as Widdowson's. At all events, I h o p e to have m a d e a reasonable case for broadening our base to explore t h e ideological aspects of discourse m a d e available b y large corpus data before we can decide such questions as whether or not all discourse implies ideology. T h e various debaters of this question would p r e s u m a b l y agree that the ideological constraints on a n d in discourse are substantially m o r e pervasive
DISCOURSE STUDIES
54
and subtle than most people, including most linguists, remotely suspect. If so, then surely we should look at extensive discourse data with an o p e n m i n d re specting its potentially ideological character. O u r m e t h o d s will of course d e m a n d extensive refinement for such tasks as sorting out corpus data according to ideologically relevant factors, such as political settings and historical periods, along with sources, registers, genres, audiences, a n d p e r h a p s even individual speakers or writers. For example, w e might well find significant distinctions between public speeches or editorials vs casual con versations in h o m e s , pubs, or grandstands, even for the same groups of speakers. If an ideology encourages consistency on some issues, it m a y encourage inconsist ency o n others in order to avoid drawing distasteful conclusions a n d making painful admissions (Potter a n d Wetherell, 1987). Psychotherapy a n d psycho analysis labour in a veritable minefield of such avoidances. If, as so m u c h work in critical discourse analysis implies, ideology is essentially a construct for p o w e r a n d closure, a p a r a d o x might b e conjectured for the enter prise of developing a n ideology of solidarity a n d equality wherein concurrent alternatives are accepted with openness a n d respect. I would again draw an analogy to a scientific theory, which in classical science wages w a r against all others until a newer theory discredits it. I would submit that this adversarial and confrontational notion of science should itself b e regarded as a discredited theory or falsified paradigm. T h e significant problems of the post-modern, multicultural society can b e solved o n l y b y 'post-classical science' that integrates alternatives (de Beaugrande, 1997). So too should 'ecologism' as a n integrative ideology for b o t h society and science transcend the combative proclivities of the ideologies of earlier ages with their horrifying rosters of atrocities. O n l y this integration can provide the consensus and strength to effectively deconstruct such anachron istic a n d recidivist ideologies as an 'economic neo-liberalism' propelling global society backwards into the primitive stages of robber-baron capitalism, nonunionized labour, subsistence wages, chronic u n e m p l o y m e n t , a n d endemic poverty. T h e alternatives are as clear as our imperatives: w e cannot choose not to choose between p o w e r a n d solidarity, between inequality a n d equality.
Acknowledgements I am much indebted to Teun van Dijk, Norman Fairclough, Jay Lemke, Jim Martin and Ruth Wodak for reading and commenting upon a previous draft of this paper.
Notes 1. On the origins of these terms, compare Firth (1968); Greenbaum (1974). 2. All three data sets were kindly extracted by Linda Pearce Williams from the Bank of English in Birmingham during her visit early in 1998, and from the Corpus of South African English at her home University in Port Elizabeth. I also wish to thank Stephen Bullon at COBUILD for completing some data lines for me.
DE BEAUGRANDE DISCOURSE STUDIES AND THE IDEOLOGY OF 'LIBERALISM'
55
3. 'Process-Verb' in the sense of functional grammar (Halliday, 1994; de Beaugrande, 1997), i.e. not an Auxiliary or a Modal Verb. 4. This line appeared in a UK source, but quoting a US source. 5. My sketch of South African politics is purely for the orientation of unfamiliar readers, and has drawn upon Plaatje (1982 [1916]); La Guma (1972); Biko (1978); Paton (1981); Kuzwayo (1985); Mandela (1994); and on the recent South African press (e.g. Mail and Guardian) as well as the corpus itself. Whether these representations are currendy accredited and by whom is a question far beyond the scope of the paper; the continuing flood of discoveries, admissions, confessions and so on has plunged South Africa's sense of its own history into a ferment which may not setde down for many years.
References Beaugrande, R. de (1988a) Critical Discourse: A Survey of Contemporary Literary Theorists. Norwood, NJ: Ablex. Beaugrande, R. de (1988b) Achieving Feminist Discourse: The Difficult Case of Luce Irigaray', College English 50/3: 1-20. Beaugrande, R. de (1991) Linguistic Theory: The Discourse ofFundamental Works. London: Longman. Beaugrande, R. de (1996) 'The "Pragmatics" of Doing Language Science: The "Warrant" for Large-Corpus lAngaistics', Journal ofPragmatics 25: 503-35. Beaugrande, R. de (1997) New Foundations for a Science of Text and Discourse. Greenwich, CT: Ablex. Beaugrande, R. de (1998a) 'Performative Speech Acts in Linguistic Theory: The Ration ality of Noam Chomsky',/oarna/o/iVflgmato29: 765-803. Beaugrande, R. de (1998b) 'Corpus and Controversy: Ideal Language versus Real Language', Internationaljournal of Corpus Linguistics. Beaugrande R. de and Williams, L. (in press) 'Discourse Analysis and 'Democracy': Some Evidence from the Corpus of South African English', Discursoy Sociedad 1. Biko, S. (1978 [1972]) I Write What I Like. London: Heinemann. Billig, M. (1991) Ideology and Opinions. London: Sage. Bolinger, D.L. (1970) 'Getting the Words in', American Speech 45: 78-84. Cameron, D. (1992) Feminism and Linguistic Theory; 2nd rev. edn. London: Macmillan. Chomsky, N. (1965) Aspects of the Theory ofSyntax. Cambridge, MA: MIT Press. Chomsky, N. (1977) Language and Responsibility. New York: Pantheon. Christie, F. and Martin, J.R., eds. (1997) Genre and Institutions: Social Processes in the Workplace and School. London: Cassell. Fairclough, N. (1992) Discourse and Social Change. Cambridge: Polity. Fairclough, N. (1995) Critical Discourse Analysis. London: Longman. Fairclough, N. (1996) A Reply to H. Widdowson's "Discourse Analysis - A Critical View"', Language and Literature 5(1): 49-56. Firth, J.R. (1968) Selected Papers of J. R. Firth 1952-1959, ed. by F. R. Palmer. London: Longman. Fish, S.E. (1994) There is No Such Thing as Free Speech. Oxford: Oxford University Press. Fowler, R., Hodge, R., Kress, G. and Trew, T. (1979) Language and Control. London: Roudedge & Kegan Paul. Francis, G. and Sinclair, J. McH. (1994) 'I Bet he Drinks Carling Black Label: A Riposte to Owen on Corpus Grammar', Applied Linguistics 15: 190-200. Gee, J. (1990) Social Linguistics and Literacies. London: Falmer.
56
DISCOURSE STUDIES
Geertz, C. (1973) The Interpretation of Cultures. New York: Basic Books. Gell-Mann, M. (1994) The Quark and TheJaguar: Adventures in the Simple and the Comple London: Little, Brown. Gilbert, N. and Mulkay, M. (1984) Opening Pandora's Box; A SociologicalAnalysis of Scientis Discourse. Cambridge: Cambridge University Press. Greenbaum, S. (1974) 'Some Verb-Intensifier Collocations in American and British English', American Speech 49: 79-89. Greenbaum, S., ed. (1996) Comparing English Worldwide: The International Corpus of Engli Oxford: Clarendon. Hall, S. (1985) 'Signification, Representation Ideology: Althusser and the Post-Structuralist Debate', Critical Studies in Mass Communication 2: 91-114. Halliday, M.A.K. (1994) An Introduction to Functional Grammar, 2nd rev. edn. London: Arnold. Hjelmslev, L. (1969 [1943]) Prolegomena to a Theory ofLanguage. Madison: University o Wisconsin Press. Hodge, R. and Kress, G. (1988) Social Semiotics. Ithaca, NY: Cornell University Press. Kotthoff, A. and Wodak, R. eds (1998) Communicating Gender in Context. Amsterdam: Benjamins. Kuhn, T. (1970) The Structure of Scientific Revolutions. Chicago, IL: Chicago University Press. Kuzwayo, E. (1985) Call Me Woman. Randburg: Ravan. Labov, W. (1970) 'The Logic of Non-Standard English', i n j . Alatis (ed.) 20th Georgetown University Round Table on Languages and Linguistics 1969, pp. 1-43. Washington, DC Center for Applied Linguistics. La Guma, A. (1972) In the Fog of the Season's End. London: Heinemann. Lemke.J. (1995) Textual Politics. London: Longman. Mandela, N.R. (1994) Long Walk to Freedom. London: Litde, Brown. Mannheim, K. (1936) Ideology and Utopia. New York: Harcourt Brace. Martin, H.P. and Schumann, H. (1996) Die Ghbaliserungsfalle: Der Angriff aufDemokratie und Wohlstand. Reinbeck: Rowohlt. Mey,J. (1979 [1974]) 'Zur kritischen Sprachtheorie', inJakob Mey (ed.) Pragmalinguistics: Theory and Practice, pp. 412-34. The Hague: Mouton. Myers, Norman et al. (1993) Gaia: An Atlas of Planet Management. New York: Doubleday. Neisser, U. (1976) Cognition and Reality. San Francisco: Freeman. Paton, Alan (1981) Ah, But Your Land Is Beautiful. Harmondsworth: Penguin Pecheux, M. (1982) Language, Semantics, and Ideology. London: Macmillan. Plaatje, S. (1982 [1916]) Native Life in South Africa. Randburg: Ravan. Potter, J. and Wetherell, M. (1987) Discourse and Social Psychology. London: Sage. Saussure, F. de (1966 [1916]) Course in General Linguistics, trans. W. Baskin. New York: McGraw-Hill. Shils, E. (1958) 'Ideology and Civility: On the Politics of the Intellectual', Sewanee Review 66: 450-80. Sinclair, J. McH. (1991) 'Shared Knowledge', inj. Alatis (ed.) Georgetown University Roun Table on Languages and Linguistics 1991, pp. 489-500. Washington, DC: Georgetown University Press. Sinclair, J. McH. (1996) 'What do We Know about Language, How do We Get to Know it, and What has all That Got to do with Language Teaching?' Paper presented ai the International Conference on Analysis and Description: Applications to Language T o a A i n r at Lienan College and at the Hong Kong University of Science and
DE BEAUGRANDE DISCOURSE STUDIES AND THE IDEOLOGY OF 'LIBERALISM'
57
Sinclair,J. McH. (1997) 'Corpus Linguistics at the Millennium', i n j . Kohn, B. Ruschoff and D. Wolff (eds) New Hori&ms in CALL: Proceedings ofEUROCALL 96 in Szombathety, Hungary, pp. 1-10. Van Dijk, T.A. (1993) Elite Discourse and Racism. London: Sage. Van Dijk, T.A. (1998a) Ideology: A Multidisciplinary Study. London: Sage. Van Dijk, T.A. (1998b) 'Critical Discourse Analysis', unpublished manuscript. Van Leeuwen, T. and Wodak, R. (1998) 'Legitimizing Immigration Control: A Discourse Historical Study', Discourse Studies 1: (1) 83-118. Volosinov, V.N. (1973 [1929]) Marxism and the Philosophy of Language. New York: Seminar. Widdowson H.G. (1988) 'Language spread in Modes of Use', in P.H. Lowenberg (ed.) Georgetown University Round Table on Languages and Linguistics 1987: Language Spread and Language Policy, pp. 342-60. Washington, DC: Georgetown University Press. Widdowson, H.G. (1991a) 'Discourses of Enquiry and Conditions of Relevance', in J. Alans (ed.) Georgetown University Round Table on Languages and Linguistics 1990. Washington, DC: Georgetown University Press. Widdowson, H.G. (1991b) 'Types of Equivalence', Paper presented at the Triangle Seminar, Paris, January 1991. Widdowson, H.G. (1995) 'Discourse Analysis - A Critical View', Language and Literature 4: 157-72. Wodak, R. (1996) Disorders ofDiscourse. London: Longman. Wodak, R., ed. (1997) Gender and Discourse. London: Sage. Wodak, R. et al. (1998) Diskursive Konstruktion nationaler Identitat. Frankfurt: Suhrkamp. Zima, P. (1981) 'Les mecanismes discursifs de I'ideologie', Revue deflnslitut deSociologies.
3 The Discourse Structure of News Stories Allan Bell
1 O n Analysis
T
his chapter offers a n d exemplifies a framework for analysing the discourse structure of news. W h y would one want to undertake such a n enterprise? W h y analyse news story structure? W h a t is the pay-off for the labourintensive analysis of the text of news stories? Stories are central to h u m a n nature. T h e stories people tell are a core part of their social identity, a n d the construction of a life story is crucial to our selfidentity. T h e idea of the story is also central to news media. Journalists do not write articles, they write stories - with structure, order, viewpoint a n d values. So the daily happenings of our societies are expressed in the stories we are told in the media. I n addition, the m e d i a are important social institutions. T h e y are crucial presenters of culture, politics a n d social life, shaping as well as reflecting how these are formed a n d expressed. M e d i a 'discourse' is important b o t h for what it reveals about a society a n d because it also itself contributes to the character of society. Linguistic research o n the media has always emphasized this last concern, focusing w h e r e issues of ideology a n d p o w e r are closest to the surface. But pre requisite to all such questions is a sound discourse analysis, a n d this is something that students of m e d i a - in their eagerness to get to the 'real m e a t ' of ideological detective w o r k - h a v e sometimes skimmed over at their peril. To the linguist, a first answer to a question about w h y undertake such a study is 'because it is there'. News is a major register of language. U n d e r s t a n d i n g how it works is important to understanding the functioning of language in society. Such a study also enables us to c o m p a r e news with other kinds of stories, such as the ones people tell in face-to-face conversation. A n d we can c o m p a r e the Source: Allan Bell and Peter Garrett (eds), Approaches to Media Discourse, 1996, pp. 64-104.
BELL THE DISCOURSE STRUCTURE OF NEWS STORIES
59
discourse structure of news with other m e d i a genres such as editorials, a n d one type of news m e d i u m with another, for instance tabloid a n d b r o a d s h e e t press, or newscasts o n different television stations. T h e text is central to news. News content is n o t i n d e p e n d e n t of its expression, and we can only h o p e to have a clear understanding of the nature of news content by close analysis of the news text. A close, Unguistically proficient analysis of the text n e e d s to b e the foundation for all attempts to u n p a c k t h e ideologies underlying t h e news. Such a n analysis shows that even simple-looking news stories are often rather complex, a n d the events they describe rather less distinct than we supposed. It also illuminates h o w stories are m a d e . This is i m p o r t a n t for news audiences - to k n o w something of h o w the news products they consume are manufactured. T h e very idea that the news is a 'product' m a y itself c o m e as something of a surprise, a n d analysis can show something of the make-up of the product. My a p p r o a c h to news discourse focuses o n the question 'what does this story actually say h a p p e n e d ? ' . It begins b y taking the news media's concentration o n factuality o n its o w n terms, together with the ordinariness of our acceptance, as readers, of such stories a n d w h a t they tell us (cf. Scannell). It is n o t a question at least initially - of w h e t h e r these reports represent w h a t 'really' h a p p e n e d . What we are told h a p p e n e d is important in the news just as it is in courts of law, which devote m u c h of their time to scrutiny of accounts of events. This analytical framework seeks chiefly to deduce an 'event structure' for a story - that is, to reconstruct, from the often fragmented information presented, what the story says actually h a p p e n e d . It thus concentrates o n the basics of storytelling, which in the n e w s are encapsulated b y the journalist's 'five W s a n d an H': who, what, when, where, why, how. O n l y after we are clear what the story says will we b e in a position to see what it does not say. We will find, as shown below, that news stories are regularly n o t saying w h a t we think they say o n first reception. T h e y are not telling a simple, clear tale, but are replete with ambiguity, unclarity, discrepancy a n d cavity. Such close analysis is prerequisite to a m o r e sophisticated a p p r o a c h to ques tions of ideology in news texts. It makes us aware of the complexity a n d ambiguity of news. It enables us to e x a m i n e whether a headline fairly represents t h e story it accompanies. It shows w h o are t h e sources of information in the news - a n d which information has n o explicit source at all. It leads us to consider w h y these particular events h a v e b e e n r e p o r t e d at all, a n d w h y they h a v e b e e n gathered together into a single published news story. This chapter is organized thus: first I outline a framework for analysing the discourse structure of news, then exemplify it briefly through examining a n u m ber of single-sentence newspaper stories. I t h e n present detailed guidelines for analysis of discourse structure, a n d p r o c e e d to close analysis of one news story. Lasdy, I e x a m i n e one of the m o s t important a n d interesting aspects of news story structure - time - again with a n e x a m p l e analysis of press stories.
60
DISCOURSE STUDIES
2 T h e Framework T h e approach to m e d i a discourse analysis used in this chapter was developed in Bell (1991). It draws o n elements from general frameworks of story analysis, especially Labov's analysis of narratives of personal experience told in con versation (Labov a n d Waletzky, 1967; Labov, 1972), as well as from v a n Dijk's framework for analysing news discourse (1988b). T h e analysis of time in news stories derives from Bell (1995b, 1996). T h e frameworks used to analyse the structure of different kinds of stories h a v e a lot of their elements in c o m m o n (e.g. Labov, 1972; Rumelhart, 1975; v a n Dijk, 1988b). Labov's is one of the most familiar a n d contains six elements: abstract, orientation, action, evaluation, reso lution a n d c o d a (see Bell, 1991 for application of this framework to news stories). Figure 1 shows the elements n e e d e d to describe the discourse structure of news stories. A story normally consists of attribution, an abstract, a n d the story proper. Attribution of w h e r e the story c a m e from is not always m a d e explicit. It can include credit to a news agency a n d / o r a journalist's byline, a n d m a y also state place a n d time ('dateline'). T h e abstract consists of the lead sentence or 'intro' of the news story and - for press news - also a headline. T h e lead covers the central event of the story, a n d possibly one or m o r e secondary events. This necessarily entails giving in the lead itself some information o n actors and setting involved in the event. T h e b o d y of the story itself consists of one or m o r e episodes, which in turn consist of one or m o r e events. Events must describe actors and action, usually express setting of time a n d place, a n d m a y h a v e explicit attribu tion to a n information source. Episodes are clusters of events which share a c o m m o n location or set of news actors (and need only b e specified when a single story contains two or m o r e clearly distinct sets of events). As well as the above elements which present the central occurrences, there are three additional categories of material in a news story: background, com m e n t a r y a n d follow-up. T h e s e represent the past, the (non-action) present, and the future of the events described in the m a i n action of the story. T h e category of background covers any events prior to the current action story past time. T h e s e are frequently previous events which p r o b a b l y figured as news stories in their own right at a n earlier stage of the situation. If the background goes b a c k b e y o n d the n e a r past, it is classed as 'history'. E x a m p l e 3 of the onesentence stories below contains just such background. Commentary provides the journalist's or news actor's present-time observations o n the action, assessing a n d c o m m e n t i n g o n events as they h a p p e n (rather than the actual narration of the events themselves, or other parties' verbal reaction to them). It m a y provide context to assist understanding of w h a t is happen ing, or evaluative c o m m e n t o n the action, or expectations of h o w the situation will d e v e l o p . T h e story in figure 4 analysed b e l o w contains b o t h context (S9: S = Sentence) a n d evaluation (SI, S3). Follow-up covers story future time - any action subsequent to the m a i n action of a n event. It can include verbal reaction b y other parties or non-verbal conse quences. Because it covers action which occurred after what a story has treated
BELL
DISCOURSE STRUCTURE OF NEWS STORIES
62
DISCOURSE STUDIES
as the m a i n action, follow-up is a p r i m e source of subsequent updating stories which are themselves called 'follow-ups' b y journalists. T h e story in figure 6 below contains follow-up material in S2.
3 A n a l y s i n g t h e S i n g l e - S e n t e n c e Story This chapter deals with the structure of press news only, a n d a specific kind of press news: the ' h a r d ' or 'spot' news which w e recognize as the staple diet of daily m e d i a - stories of fires, wars, accidents, disasters, dangers a n d all m a n n e r of the m a y h e m that befalls h u m a n life. T h e m i n i m a l well-formed h a r d news story is just one sentence long. M a n y newspapers publish one-sentence stories, either to fill o d d corners, or gathered in a column of news briefs. I n a lot of broadcast news, m a n y stories m a y consist only of a single sentence. T h e singlesentence story is also an appropriate p r o x y for examining the structure of longer stories in general. T h e news story is always focused in its first sentence - its lead or intro (Bell, 1991). We can see this most clearly w h e n the story itself is r e d u c e d to just that sentence, b u t the lead is itself a microstory even w h e n a full story follows. It compresses the 'news values' that h a v e got this story through to publication. T h e single-sentence story is therefore a good place to start as an introduction to analysis. H e r e are the texts of five one-sentence spot news stories published in British daily newspapers in February 1994. T h e y are all from the international news agencies - Associated Press, Reuters a n d Agence France Presse - which act as m o d e l s of Western journalistic style. (1) Clashes kill eight At least eight people have died in tribalfightingin the Bimbila region of northern Ghana. - AFP (2) Fumed out Tokyo: Two sake brewers were seriously ill after being overcome by fumes when one fell into a half-full vat and the other was trapped trying to rescue him. Reuter (3) Deportation setback Storms over Iceland delayed the deportation from Norway yesterday of 12 American anti-abortion activists who had allegedly planned to stage demon strations during the Winter Olympics and were detained by police when they arrived at Oslo's airport. - AP (4) Icicle h o r r o r A woman was fighting for life last night after a giant icicle fell 30 storeys from a New York skyscraper and speared her. (5) Awaiting the e n d Communist North Korea is building an underground mausoleum and water proof glass coffin to await the death of its "Great Leader", President Kim Il-sung, a South Korean press report said. - AFP
BELL THE DISCOURSE STRUCTURE OF NEWS STORIES
63
Example 1 covers a single event a n d can b e d i a g r a m m e d as in figure 2 (in this and subsequent figures, some nodes of the structure are omitted to simplify the tree diagrams). It has a headline - 'Clashes kill eight' - as h a v e all these singlesentence stories, a n d an attribution to the source agency, Agence France Presse. It consists of a single event, which specifies: • actors - 'at least eight p e o p l e ' • action - 'have died in tribal fighting' • setting, specifically place - 'in the Bimbila region of n o r t h e r n G h a n a ' . The structure of this story is as minimal as it can be. Yet, of course, even h e r e it is quite likely that w e are n o t dealing with just a single action - the eight victims may well h a v e perished in widely separated incidents. E x a m p l e 2 compresses a chain of five actions into one sentence: 1 2 3 4 5
first sake b r e w e r falls into vat second b r e w e r goes to save h i m second b r e w e r also gets trapped b o t h brewers overcome with fumes both n o w seriously ill.
The most notable thing about example 2 is the order in which it tells these. M y list above is in chronological order, that is the time sequence in which things actually h a p p e n e d , beginning at (1) a n d ending at (5). But the news typically begins with the most recent m a i n h a p p e n i n g (as also in examples 3 a n d 4). So the narrated order of the actions in story 2 is 5-4-1-3-2. It starts with the most NEWS TEXT
ATTRIBUTION
ABSTRACT
AGENCY
HEADLINE
AFP
Clashes kill eight
STORY
EVENT
ACTORS
ACTION
SETTING PLACE
At least eight people
have died in tribal fighting
in the Bimbila region of northern Ghana
Figure 2: Discourse structure of the one-sentence story in example 1 (some nodes of the structure are omitted for simplification)
64
DISCOURSE STUDIES
recent happening, then the next most recent, then goes back to the first a n d causal action, then fills in the remaining steps (again in reverse chronological order). T h e story is thus told in a radically nonsequential fashion - a n issue we will return to in detail later in the chapter. Most of these single-sentence stories overtly narrate m o r e t h a n o n e event. I n terms of the m o d e l story structure indicated in figure 1, their structure is therefore complex. Example 3 contains a sequence of four events, with a complex e m b e d d i n g of background involving multiple time points a n d locations. T h e four events are (in chronological order): 1 2 3 4
Anti-abortion activists h a d allegedly p l a n n e d to stage demonstrations during the Winter Olympics detained b y police w h e n they arrived at Oslo airport h a d b e e n scheduled for deportation from N o r w a y yesterday deportation delayed b y storms over Iceland.
T h r e e of these events have their o w n different place specification. Each has a different time specification: event 1 includes a future ('planned to stage demon strations') which did n o t in fact h a p p e n , and event 3 is as yet a 'non-event' which h a s still n o t occurred - deportation. Figure 3 diagrams the discourse structure of this example (again with some of the potential n o d e s of the tree diagram omitted to reduce complication). I t shows some of the complexity of action, place a n d time t h a t c a n b e covered b y e v e n a single-sentence story. Such a storv offers a considerable challenge to the reader's comprehension, as 1 shall discuss farther below, particularly because of its multiple e m b e d d i n g of disparate event, some of which h a v e n o t in fact (yet) happened. T h e r e is also obviously a lot more to b e told about what has h a p p e n e d 'behind the scenes' of this story, the factual^ of which will doubdess b e in dispute b e t w e e n the different parties. The sti r\ also has one notable gap that readers h a v e to fill from their world knowledge w h y storms over Iceland should delay a deportation from Norway. We are prob ably to u n d e r s t a n d that the storms delayed or diverted a n incoming aircraft OD which the activists w e r e m e a n t to leave. Looking across these five example stories, w e can see patterns in their dis course structures. Firstly, all of t h e m specify m a i n event, n e w s actors and place the journalist's what, who a n d where. E x a m p l e 1 contains only those element!. T i m e is expressed in most of them. News agency attribution is present in il except (4), and (5) also credits a local agency. But n o n e attribute thekinformatia to any other source such as bystanders, spokespeople or officials. Secondly, these stories proclaim their news value. As single-sentence stories, they probably originated as the lead paragraphs of longer stories from theintanational agencies. Their role is therefore to concentrate the news value of whole story. News values are those factors which take a story into the news, include attributes such as negativity, i m m e d i a c y (or recency), proximity, lade ambiguity, novelty, personalization a n d eliteness of the news actors (Bell. I A story's n e w s value focuses in the lexicon of newsworilnuess i n its lead sent
65
BELL THE DISCOURSE STRUCTURE OF NEWS STORIES
NEWS TEXT
"
STORY
ATTRIBUTION ABSTRACT
I
EVENT 4
HEADLINE
I AP
Deportation setback
ACTION
SETTING
BACKGROUND
PLACE storms delayed
BACKGROUND
EVENT 2
over Iceland
ACTORS
ACTION
Activists
(scheduled) deportation
ACTORS
Activists
EVENT3
SETTING
PLACE
TIME
from Norway
yesterday
BACKGROUND
detained by police
when they arrived
at Oslo airport
ACTORS
EVENT 1
ACTION
SETTING
I TIME Activists
planned during the demonstration Winter Olympics
Figure 3: Discourse structure of the one-sentence story in example 3
fl()
DISCOURSE STUDIES
I n (4), for instance, almost every w o r d makes a claim to news value: 'fightingfor life', 'giant icicle', 'fell 30 storeys', 'New York skyscraper', 'speared'. Most of the stories contain death, violence, or the i m m i n e n t threat of these. T h e facticity of the stories is stressed through the detail of person, place a n d time already men tioned, and through the use of figures. Thirdly, all but one story (3) begin with who - the m a i n news actor. T h e stress o n personalization a n d elite news actors guides the order of constituents within the sentence, even if this will result in a passive-voice verb. News-writing myth ology holds that verbs should b e active, b u t passivization is quite common as the only m e a n s of getting the m a i n news actors to the start of the sentence. We thus h a v e a grammar of news value as well as a lexicon. Syntactic rules such as passivization are applied to serve the values of n e w s discourse, as is linkage through temporal conjunctions such as after a n d as (2, 4). I n most cases one event is the sequel to or result of another, as also in examples 3 a n d 4. Fourthly, time is sometimes expressed directly in these stories as 'yesterday' or 'last night'. But often there is n o direct reference to current calendar time. We interpret these stories as 'reported within the past day'. But this immediacy cri terion m a y n o t always b e the case, and some of these stories could well b e daysold news (e.g. 1, 2) and we readers would b e n o n e the wiser. Fifth, since these are all international agency stories, they often carry explicit place of origin in the dateline (2). Otherwise place is sometimes expressed in the canonical form of a prepositional phrase, such as 'in the Bimbila region of n o r t h e r n G h a n a ' (1). Lastly, information content, brevity a n d clarity can b e seen tussling with each other. Sometimes, as i n (3), clarity is the loser a n d the result is a sentence or it story in w h i c h the degree of layering a n d the dispersion of the events reported jeopardizes comprehension. This accords with van Dijk's finding (1988b: 77 that lead sentences i n news stories can often b e syntactically a n d reformationally complex.
4 W h a t Does t h e Story Actually Say? I n o w turn to consider in m o r e depth o n e longer n e w s story, subjecting it v close analysis of its discourse structure. T h e G u i d e to analysis below present* step-by-step procedure for establishing what a story says h a p p e n e d - that is. th event structure of the story. T h e fruit of analysing the discourse structure of stor: in this way is an understanding of what the story says actually happened. At leaat this stage, I a m concerned only with w h a t the story says, n o t with whether h o w closely that corresponds to 'the facts' of w h a t h a p p e n e d . T h e event analysis takes into account n o t just the overt specifying of eveiv themselves, b u t what the story says about news actors, locations and times occurrence in order to tease out the structure of the story. M o v i n g through eac of these aspects, w e will c o m m o n l y h a v e to modify our conception of wh ;
BELL THE DISCOURSE STRUCTURE OF NEWS STORIES
()7
happened as further specification of persons, time or place m a k e s it clear that our earlier models w e r e inadequate. Figure 4 contains the text of a very routine news story from the (UK) Daily Mirror. I chose such a story because it is c o m m o n rather t h a n exceptional. T h e first group of analyses seek to u n c o v e r w h a t the story is telling us actually happened - what events occurred, w h e r e they occurred a n d w h e n , a n d who was involved (that is, four of the journalistic five W s and the H , with the exception of'why' and 'how': cf. Manoff and Schudson, 1987). This analysis yields structures for these four aspects: events, times, places a n d news actors. I n theory, these four should all mesh into a unified picture of what the story says happened. I n practice, they often do n o t quite fit together, a n d it is precisely these differences which may cast the most interesting light o n the story. T h e category labels I use below are from Bell (1991), r e p r o d u c e d earlier as figure 1, the m o d e l tree diagram of news story structure. Working through this example, I follow the sequence of steps indicated in the Guide, although n o t all the points listed there are useful for analysing this - or any individual - story. T h e r e is a logic to the order in which I have listed the steps, b u t not all steps are necessarily going to contrib ute equally to a n understanding of every story. Neither are they necessarily always best followed in this particular order or independently of each other - it may often m a k e sense to carry out some later steps before those I h a v e listed here as earlier.
Guide How to analyse and interpret the discourse structure of a news story. First number the sentences of the story. T h e first steps listed u n d e r each n u m b e r e d point beiow (in r o m a n type) are basic analytical moves. T h e later steps u n d e r each point - indented a n d in italics - are the m o r e interpretive procedures.
What 1 Headline AVhat events take place in the headline? Summarize a n d n u m b e r each event. 2 Lead What events take place in the lead or intro? Summarize and number. 3 Events What events take place in the story? Summarize a n d n u m b e r , t h e n enter numbers alongside each sentence of the story (as in figure 6). Re-categorize events in headline a n d lead as necessary to correspond with the fuller picture vou n o w h a v e from the story as a whole.
It
DISCOURSE STUDIES
W h a t is the central event of the whole story? (Usually t h e m a i n 'hard' news event in the l e a d ) 4 H e a d l i n e , l e a d a n d story W h a t is the relationship of the headline to the lead? What events in the lead are included/excluded in the headline? What news values lie behind these inclusions and exclusions? Is the headline a valid representation of the lead?
W h a t is the relationship of the lead to the story as a whole? What events in the story are included/excluded
in the lead?
What news values lie behind these inclusions and exclusions? Is the lead a valid representation of the whole story? Is there any information that is given in the lead but not returned to in the rest of the story? How does the lead begin telling the story as well as act as an abstract for iff
Who 5
Story attribution Is the story as a whole attributed? To w h o m (agency, journalist)?
6 Sources attribution Is there any attribution within the story? W h o is attributed? (list) Beside each sentence, note d o w n w h o m it is attributed to (if anyone). Precisely what is attributed and to w h o m ? W h a t speech v e r b is used in the attributions? (list) What claims do the attributed sources have to authority? Who is quoted directly?Indirectly? Why have the particular speech verbs been used? What parts of the story are not attributed? Why? Where is attribution unclear or ambiguous? Does this have any repercussions?
7 N e w s actors What news actors are mentioned? (list: people, organizations, nations, etc.) H o w are they labelled or referred to? (list) What kinds of people or entities are mentioned in the stoty? Why are they in the news? Are the news actors elite? Is the news story personalized? Are there patterns in the way the story refers to them or labels them?
Does specifying w h o the news actors are modify the event structure you developed earlier?
BELL THE DISCOURSE STRUCTURE OF NEWS STORIES
69
Where 8
Places W h a t place expressions are used? (list) W h e r e d o they occur in the story? (sentence n u m b e r )
9 Place structure What locations does the story take place in? (list) Does the story stay in one location or move from place to place and back? Why? Is it clear what is happening in which location? What sort ofplaces are the events happening in ? Is there a pattern to this? Does specifying locations in this way modify t h e e v e n t structure you developed earlier?
When JO T i m e s What time expressions are used in t h e story? (list) Where d o they occur in t h e story? (sentence n u m b e r ) 11 T i m e s t r u c t u r e What is the time structure of the story? Take the time of the central event as T i m e 0. Label earlier events as T i m e - 1 , - 2 e t c , a n d later events as Time + 1 , +2 etc. in the chronological sequence in which they actually occurred. Beside each sentence, note d o w n t h e n u m b e r of the time or times at which the actions m e n t i o n e d there occurred. How does the order in which the story is told relate to the chronological order of events? What has the story been written in this order? What values lie behind the order? Does the order help or hinder a reader in understanding what is going on in the story? Does specifying times of occurrence in this way modify the event structure developed earlier? 12
Background Is any background given (events prior to the central action - either recent previous events or m o r e historical events)? Does any of the background indicate any particular ideologicalframe behind the story?
DISCOURSE STUDIES
70
13
Commentary Is there any commentary o n events? - evaluation of events (editorializing)? Context, for what has h a p p e n e d ? Expectations of h o w the situation will develop? Does any of the commentary (especially evaluation) indicate any particular ideology behind the story?
14
Follow-up Is there any follow-up to the central action of each event (subsequent events, either reaction (verbal) or consequences (non-verbal))? Does any of the follow-up indicate any particular ideology behind the story?
Event and Discourse
Structure
15 E v e n t structure Collate your successive re-categorizations of what h a p p e n e d in the story, drawing o n news actors, place and time as well as the actions themselves. List in chronological order the events a n d their associated actors, times and places a n d the sentence n u m b e r s in which they occur (as in table 1 example). This represents the event structure as you finally assess it to be. Note a n y alternatives, which represent discrepancies or unclarities in the story itself. Is the story told in instalments ? That is, do the eventsfollow one after another or are they interspersed with each other? 16 Discourse structure At this stage you can draw a. tree diagram of the discourse structure (e.g. figures 2, 3, 5). Note that the apparent order which these example figures display is only arrived at after a succession of r o u g h drafts! Such a structure is also idealized in the sense that it masks ambiguities and unclarities. 17
Cohesion W h a t linkages are expressed or implied between the sentences or events in the story? H o w ? (list) W h a t linkages are omitted? (list) What do the linkages (or their absence) between sentences or events mean for understanding the story? Js a cause-and-effect relationship between different events implied by the way they are ordered? Does the story flow smoothly, or does it jump about? Why?
18
Confusion H a s it now b e e n possible to say precisely what h a p p e n e d in the story? O r are there still ambiguities, gaps or confusions, as exemplified in the possibility of alternative event or discourse structures? (list, explain)
BELL THE DISCOURSE STRUCTURE OF NEWS STORIES
EVENT STRUCTURE
'
71
TIME STRUCTURE
3 1 1(2?)
HL
2 held as IRA blitz is foiled
SI
1(2?)
S2
I? (2?)
S3
3
S4
POLICE and MI5 agents have swooped in an undercover operation which netted bomb-making equipment - and may have foiled a major IRA blitz on the British mainland. They recovered enough Semtex explosive to devastate a large office block. And security services believe they have intercepted a highly-placed IRA quartermaster who supplies equipment to terror cells. Last night two men were being held in London under the Prevention of Terrorism Act.
+2 0 0 (+1?)
0 (+1?) 0? (+1?)
+2
Pistol 1
S5
I
S6
i
S7
1
S8
i
S9
One, in his 30s, was arrested at Accrington, Lanes, with between 10 lb and 15 lb of high explosives, bomb-making equipment, a pistol and ammunition in his car. Plain-clothes police wearing flak jackets and carrying semi-automatic weapons seized the suspect in a car park next to Charlie Brown's Auto Centre. Within two hours another man, in his 50s, was held at Wembley in North London. A terraced house in Accrington was under police guard last night. The suspects can be held for up to seven days before being charged or released.
0
0
+1 +2 +3
Figure 4: Example story with Event and Time Structures (Daily Mirror, London, 23 February 1994; reproduced by permission of Mirror Group Newspapers) 1 Headline The headline appears to narrate two events - (1) the holding of two suspects, and (2) the foiling of an I R A blitz. T h e two events are clearly closely linked, or may even in effect b e the same action - the blitz m a y h a v e b e e n foiled precisely through the capture of t h e two people. T h e journalistic 'as' of the headline would thus not b e a n accident; it would specify precisely the t e m p o r a l a n d causal relationship between the two events. Alternatively, there m a y h a v e b e e n action which foiled the blitz, a n d then led to the capture of two suspects. We cannot tell from the headline alone. W e m u s t r e a d on.
DISCOURSE STUDIES
72
2 Lead T h e lead appears to tell of one event: a security service 'swoop' which discovered b o m b - m a k i n g equipment. This action is clearly related to, but n o t quite identical with, the two events in the headline. T h e 'foiled' statement in the headline is now s h o w n to b e a n evaluation of the significance of the raid rather than a separate action of any kind. We thus find that the lead sentence clarifies the ambiguity of the headline, but it of course remains possible that the operation consisted of raids o n m o r e t h a n just o n e location. 3 Events We can n o w m o v e to tease out the event structure of the story as a whole. The lead a n d the following three sentences (i.e. S l - 4 ) invite the interpretation that there has b e e n a single raid. But the second half of the story tells of two suspects being arrested in widely separated locations. It takes until S7 to b e c o m e clear that w e are dealing with quite separate events, a n d it remains unclear whether they were in some sense part of a single security operation, or whether they coincided b y chance. T h e central action of b o t h events is clearly the respective raids (assuming that Wembley involved a raid - w e are told only that the m a n was 'held', in a strangely active use of a stative verb, akin to 'arrested'). To this point, it seems we can describe the story as being about a total of three events, assuming we treat the continuing detention of the two suspects as a separate event: 1 Accrington arrest 2 Wembley arrest 3 detention of two suspects. T h e Accrington arrest seems to b e the central event of the story a n d most of the story apparently concentrates o n it (see figure 4 for Event Structure). 4 Headline, Lead and Story We h a v e already seen that the content of the lead in this story disambiguates the headline. But a closer look at the headline shows that the '2 held' statement with which it begins does n o t derive from the lead at all b u t rather from S4. This is an infrequent pattern, since a majority of headlines are derived solely from lead sentences a n d n o t from information further d o w n a story (Bell, 1991). T h e r e is also a contrast in the modality of the verbs used in the headline and the lead. T h e lead hedges its evaluation as ' m a y h a v e foiled', but the headline presents this as unvarnished fact - 'is'. T h e central event of the lead, the raid, is represented in the headline only b y the gloss of 'foiled' - that is, the tentative evaluation expressed in the lead has b e c o m e the u n h e d g e d description of action in the headline. This is a classic news 'over-assertion' of a kind I found commonly in a study of editing changes m a d e to news c o p y (Bell, 1983,1984). T h e origina)
BELL THE DISCOURSE STRUCTURE OF NEWS STORIES
73
copy-editor. T h e shift is driven b y the attempt to m a k e a story as definite as pos sible ('unambiguity' is o n e of the classic news values: Galtung a n d R u g e , 1965; Bell, 1991). So it appears that the headline is n o t a valid representation of the story as a whole to the extent that through omitting the m o d a l verb, the headline overstates the certainty of the evaluation contained in the lead. Modality is probably rarely expressed in headlines. Further, w e are n o t told w h o s e opinion it is that a major I R A blitz m a y h a v e been foiled. T h e r e is n o sourcing of the evaluation. It is n o t m e n t i o n e d again in the b o d y of the story. We are probably to assume the source is the 'security services' quoted in S 3 , b u t it m a y in fact b e the p r o d u c t of the journalist w h o wrote the story. A m o r e likely possibility is that this statement is a gloss inserted by the copy-editor in order to raise the news value of the story - to 'beat it u p ' . Evidence for this is the fact that it appears in SI after a dash - an avoided piece of punctuation that betrays late a n d hasty addition of the second half of S I . 5 Story Attribution This story attributes n o journalist as author n o r agency as provider. W e must assume it is written b y a staff reporter w h o was n o t 'bylined'. 6 Sources Attribution More importantly, the story does not directly specify the source of its information. The nearest w e get to sourcing is in S3 w h e r e we are told what 'security services believe' in relation to the results of their actions. We can presume they were also the source of s o m e of the information about the event itself, as well as possibly of the evaluation in S I . T h e use of 'believe' rather than 'say' in S3 is p r o b a b l y significant, indicating either that the security services were n o t p r e p a r e d to pro vide information 'on the record' for the media, or that this was a chance c o m m e n t that the journalist has built into m o r e t h a n the source would h a v e wanted. We can note that the S9 background about detention powers p r o b a b l y comes from the journalist's o w n knowledge rather than any source. 7 News Actors The news actors specified in,the story are: security services (S3) police (SI, S6, S8) MI5 agents (SI)
1
two IRA suspects (S4) one arrested in Accrington (S5) second arrested in Wembley (S7). There is also another reference to a n I R A m e m b e r - ' a highly-placed I R A
DISCOURSE STUDIES
74
character is n o t the same as the p e r s o n / s involved in the events of S l - 2 . Use of co-ordinator ' a n d ' to introduce the sentence, plus indefinite article 'a . . . quartermaster', implies a different person. If it was the same person we would rather expect a phrasing like 'security services believe the suspect is a highlyplaced I R A quartermaster . . . ' . O n the other h a n d , given the detail of w e a p o n r y listed with the Accrington suspect in S5, S3 seems m o r e likely to also refer to him. But since that would entail S l - 2 referring to the Wembley suspect, this interpretation is n o t entirely satisfactory either. T h e remaining possibility is that this refers to a third person n o t otherwise m e n t i o n e d , b u t that seems least likely of all. Notice h o w w e are gaining m o r e information as our analysis moves into the story, and that n e w information modifies our understanding of earlier infor mation. But it does n o t always clarify that understanding, a n d at this point we h a v e to say that the identity of the quartermaster is unclear. Such questions of reference identity are n o t u n c o m m o n in news. I n studies of editing practice (Bell, 1984), I h a v e found cases w h e r e editors were obviously u n a b l e to decode whether expressions in different sentences w e r e m e a n t to refer to the same or different places or people. Their attempts to clarify reference sometimes took the w r o n g interpretation, a n d converted a second reference to a city already m e n t i o n e d in a story into a first reference to another city. 8 Places T h e story uses expressions of place in the following: blitz on the British mainland (Si) i • held in London (S4) . arrested at Accrington, Lanes (S5) with . . . ammunition in his car (S5) seized the suspect in a car park next to Charlie Brown's Auto Centre (S6) held at Wembley in North London (S7) a terraced house in Accrington (S8). W h a t is n o t said is m o r e striking than what is said: there is n o specification of w h e r e the raids a n d arrests took place until S5 a n d below. This contributes to our unclarity about what has taken place. We m a y also w o n d e r w h y the story details 'Charlie Brown's Auto Centre', since this is unlikely to b e k n o w n to anyone b e y o n d the immediate AccringtoD area. For locals, however, the description might signify that the raid was in a central, busy area of the town, for example. This m a y b e a clue that the origin of this part of the story was in a local Accrington reporter's account which has not b e e n appropriately edited for national readership. I n contrast, Wembley - which n o Briton needs a n introduction to - is glossed as being in N o r t h London, pre sumably for international readership. So within two sentences we have conflicting indications of the p r e s u m e d readership, local versus international, through the kind of shared knowledge that is a n d is n o t presupposed.
BELL THE DISCOURSE STRUCTURE OF NEWS STORIES
9 Place Structure
75
1
It seems then that the story has (at least) three locations - raids in Accrington, Lancashire, a n d in Wembley, N o r t h L o n d o n , plus the site or sites in L o n d o n where the two suspects are being held prisoner. So we can confirm our specifi cation of probably three events in the story. If we h a d m o r e detail, w e might wish lo regard other actions - such as the guarding of a terraced house in Accrington S8) - as further separate events, since it seems likely that the house was not located at the car p a r k which was the scene of the Accrington suspect's arrest. What is most striking, however, is that the further we get into the analysis, ihe less clear it b e c o m e s exactiy what we are b e i n g told h a p p e n e d . W h a t o n the surface a p p e a r e d a simple little story turns out to b e rather complicated a n d opaque. In particular, it is b y n o m e a n s obvious which of the two sets of news actors, locations a n d actions detailed in S 5 - 7 are actually being referred to in S l - 3 . because those first three sentences contain n o unambiguous references to persons, and n o reference to place at all. For instance, w e cannot b e sure which place the Semtex explosive was found in (S2), or which (if either) suspect is the quartermaster of S3. 10 Times The specification of time is often surprisingly sparse i n a n e w s story (aside from tense and aspect m a r k i n g o n verbs). It is frequendy left to the inference that this must b e recent because it is news'. T h e time expressions in this story are: last night (S4) within two hours (S7) last night (S8) for up to seven days (S9). No time is expressed until the fourth sentence of the story, with the presumption of recency governing our reading u p to that point. T h e r e is n o specification of the time of day w h e n the two arrests were m a d e , although we know there was n o more than two h o u r s b e t w e e n t h e m . We p r e s u m e the arrests took place during the day before the story was published - certainly before 'last night'. But in fact it is possible that the n e w s is older t h a n this, dating from the night before last or even earlier. It is not u n c o m m o n for stories to r e m a i n silent o n the time of their events. Occasionally it b e c o m e s obvious that the presumption of recency is wrong - and sometimes this silence o n timing appears to serve the misleading implication of immediacy. 77 Time Structure Closely related to teasing out the event structure of a story is specifying its time structure: what h a p p e n e d in what order, what is the story's chronology? T h e
DISCOURSE STUDIES
76
time structure of this story is rather straightforward except for the lack of clarity over absolute timing, a n d what actions are part of what event. Time structure is shown o n the right of figure 4. We take the time of the main news action as TO - the operation described in S I , which I will assume to refer to the Accrington raid (though it m a y cover both). T h u s SI, 2 , 5 and 6 all describe aspects of the raid. I will interpret S3 as also referring to Accrington. All these are therefore TO events. S7 occurs at the time of the W e m b l e y raid 'within two hours', therefore at T + 1: the n e x t time point after TO. S4 a n d 8 both refer to 'last night', subsequent to b o t h arrests, that is T + 2 . S9 can b e considered subse quent again, therefore T + 3 . This is a comparatively simple time structure, as even a short news story can easily involve ten or m o r e actions in non-chronological order, as w e will see in a later example. This story is also unusual in n o t leading with the m o s t recent event, p e r h a p s because of the staticness of 'held' that I n o t e d earlier. But we can also note that the headline does in fact major o n the most recent event - the detention of the two suspects. T h e r e has thus b e e n a difference of opinion over what constitutes the central news event in this story. T h e journalist w h o wrote the story, including the original lead, thought it was the raid. T h e copy-editor w h o edited it a n d wrote the headline thought it was the detention of two suspects. Such disagreements are a c o m m o n locus of n e w s r o o m anguish. Because time is such a n important element in news stories a n d merits more discussion a n d illustration, I will return to it below with examples w h e r e it plays a m o r e important role. 72 Background T h e category of background (cf. figure 3) covers events prior to the central action of the story, a n d m a y include either recent past events or m o r e remote history. This story contains n o background. 1
13 Commentary i ••
C o m m e n t a r y concerns context, evaluation or expectations for events that are happening. I n this story we h a v e already noted the evaluative c o m m e n t s con cerning foiling a n I R A blitz (S1, headline), a n d the assessment of the significance of the security services' discoveries (S3). I n addition, there is context given about the Prevention of Terrorism Act (S4), which spells out (S9) the provisions under which the suspects can continue to b e detained. 14 Follow-up If w e take the raids to b e the central event of the story, t h e n the detention of the two suspects counts as follow-up to that event. Follow-ups commonly become the lead in subsequent stories, a n d we h a v e noted already that the headline (written later t h a n the story, of course) tends in this direction.
BELL THE DISCOURSE STRUCTURE OF NEWS STORIES
77
15 Event Structure We have noted the ambiguity a n d lack of clarity in the who a n d the where of this story. T h e left side of figure 4 shows the event structure as best I can j u d g e it. Usually it is time that is the problematic factor in understanding what h a p p e n e d in a story. In this case it is the place structure which is the m o r e confusing, a n d to a lesser extent the identity of the news actors, because it is b y n o m e a n s clear which place is referred to in different parts of the description - Accrington or Wembley. Table 1 schematizes this event structure, linking action, actors, place and time for each event. Table 1: Event structure of IRA story (figure 4) Event
Sentence
Actors
Place
Time
1
Raid, arrest, seizure Arrest
Headline SI, 2, 3, 5, 6,8 S7 (SI? 2? 3?)
Accrington
0 +2
Wembley
Detention ('held')
Headline S4,9
Security services, IRA Security services, IRA 2 IRA members
+1 Within 2 hours +2,-1-3 Last night
2
3
London
16 Discourse Structure Figure 5 diagrams the discourse structure of the story in accordance with the interpretation I h a v e offered, again omitting some of the detail so that an already complex diagram d o e s n o t get m o r e complicated still. T h e neatness of the even tual tree should n o t disguise the fact that it p r e s u p p o s e s a series of (albeit increasingly refined) rough drafts. Event 3 is placed as a follow-up to both events ) and 2. T h e constituents of each event are the straightforward ones of actors, action and setting, plus commentaries a n d follow-ups. If we take other interpretations of the ambiguous information in the story, it will mean shifting some constituents to other points of the diagram. So, if we understand the 'blitz foiled' c o m m e n t as applying to b o t h the Accrington a n d Wembley raids, we will attach that Evaluation to both events 1 a n d 2 (or possibly to the Story as a whole). If we decide the quartermaster was the Wembley suspect rather than the Accrington one, we will re-attach that Evaluation to event 2. 17 Cohesion News stories are standardly written as a series of one-sentence paragraphs, a n d commonly express little linkage between the sentences. W i t h each news sentence usually also its o w n paragraph, there is n o larger unit of text organization. T h e r e js routinely n o flow of time sequence from one sentence to the next, a n d a lack
NEWS TEXT
ABSTRACT HEADLINE LEAD
r^i
EVENT 1
EVENT 2
EVENT 3 EVENT 1
ACTORS
ACTION SETTING
COMMENTARY
FOLLOW-UP
FOLLOW-UP
ACTORS
ACTION
SETTING
PLACE EVALUATION EVALUATION Security IRA services
raid, Accrington arrest, seizure
blitz foiled
quartermaster intercepted
house watched
Security IRA services
arrest
TIME
PLACE
within 2 hours
Wembley
EVENT 3
ACTORS
2 IRA members g u r e 5: D i s c o u r s e structure o f I R A s t o r y i n figure 4
ACTION
held
SETTING
COMMENTARY
TIME
PLACE
CONTEXT
last night
London
Prevention of Terrorism Act
BELL THE DISCOURSE STRUCTURE OF NEWS STORIES
79
of devices such as adverbs expressing linkages between sentences. It is com mon for cohesion b e t w e e n sentences to b e unclear or non-existent, a n d w e m a y be genuinely doubtful what actions within the story b e l o n g together, at what point location actually shifted, or w h a t material is attributed to w h o m . It is the lack of such signposting, together with the m i x of locations, that con tributes to the ambiguities of this story. I n the absence of any indications to the contrary, we would normally take S l - 3 to b e describing a single set of h a p p e n ings, although there are n o explicit linkages to m a r k this. Things b e c o m e clearer after S4 where links are expressed: ' o n e ' at the start of S5 links the news actor back into S4, a n d 'the suspect' in S6 continues the thread of reference. T h e signposting in S7 equally clearly indicates that this is a different event - 'another man' in another place at another time. However, the j u m p back to Accrington again in S8 is rather disconcerting to a reader. 18 Confusion I have noted a n u m b e r of points where it is unclear, ambiguous or confused what actually h a p p e n e d in this story, to w h o m , a n d where. To list these as ques tions (such as a journalist doing a follow-up story might ask a source): Were the two arrests part of a single operation or coincidental? (SI) Which arrest did security forces consider represented a 'foiled IRA blitz'? (SI) Was Semtex recovered at Wembley as well as Accrington? (S2) Was the Accrington or the Wembley suspect the quartermaster? (S3) The grounds for confusion in this story are possible because of the lack of cohe sion and specific reference across the story. This is c o m m o n l y a source of p r o b lems for accuracy of b o t h reporting a n d understanding. I n a study of reporting of climate change (Bell, 1994), I found that the practice of not expressing cohesive links between sentences of a story enabled one journalist to refer alternately to the greenhouse effect a n d ozone depletion between successive sentences of a story in a way that seriously confused these two basically distinct atmos pheric p h e n o m e n a . The packaging of the two raids in one story implies that they and the suspects were closely linked. However, we are told so little about the Wembley arrest that we cannot b e sure of this. I n fact, w h e n we examine the role of the security services as news actors in the story, 'police' are twice specified in relation to Accrington (S6, S8), b u t M I 5 are m e n t i o n e d only generally in the lead sentence. It seems possible that Accrington involved only the police, Wembley involved only M I 5 , a n d the timing was coincidence rather than a single operation. This unclarity of reference in fact serves to enhance news value through co-option of a rninor story into a bigger story (Bell, 1991). O n e reading of the story is that most of the serious action is linked to Accrington, a n d very little to Wembley. T h e ambiguity of reference therefore enables the Wembley event to b e co-opted to the Accrington raid, building it into m o r e t h a n it would b e just o n its o w n when it might n o t h a v e b e e n reported at all.
DISCOURSE STUDIES
80
T h e detailed analysis I h a v e r u n through in this section has s h o w n us a g o o d deal about the character of this story - indications of h o w it was m a d e , evidence of inconsistencies a n d gaps, a n d manifestations of the news values b e h i n d it. We can see the claims it makes to b e newsworthy mainly in the negativity of its content - in the presence of conflict, discoveries of w e a p o n s a n d explosives, use of force - together with involvement of elite news actors (the security services), co-option of possibly separate stories to reinforce each other, reduction of ambi guity over the significance of the arrests, a n d possibly e n h a n c e d immediacy. T h e journalist's o w n ultimate abstract of a story is the one-word 'slugline' b y which each story is identified in the news processing system - n o w in practice also presumably the filename u n d e r which the story is saved i n the computer. T h e slugline is often a g o o d guide to news value, a n d it is instructive to speculate what slugline a particular story might have carried. T h e likely candidates on this example story seem to m e to b e : 'swoop', 'blitz' or 'foiled'.
5 News Time T i m e , as we h a v e seen in the analyses so far, is an important dimension of news stories a n d merits m o r e discussion in its o w n right. T i m e is expressed in stories at different levels of language - in the morphology a n d syntax of the verb phrase, in time adverbials w h e t h e r lexical or phrasal, a n d in the discourse structure of stories above the sentence. O n e characteristic of all kinds of stories appears to b e so shared across genres a n d cultures that it has b e e n called 'canonical' events tend to b e told in the chronological order in which they occurred. T i m e in story has interested widely different groups of scholars (Toolan, 1988) - literary theorists, a n d cognitive psychologists researching discourse com prehension, as well as text a n d discourse linguists. Literary theory tends to treat chronological order as the u n m a r k e d or basic form underlying fictional narrative, and identifies the time structure of a novel as interesting just where it departs from chronology. Genette's general theory of narrative discourse (1980) developed in relation to Proust's A la recherche du temps perdu is primarily devoted to time. T i m e order is his dominant category, analysed in terms of 'anachronies' - depart ures from chronological sequence. Cognitive psychologists m a k e the distinc tion between event structure a n d discourse structure (Brewer, 1985) - b e t w e e n the order in which events actually h a p p e n e d a n d the order i n which they are told in a story (my o w n framework above allows for an event structure in discourse which is non-chronological). T h e r e is only one real-world event structure but m a n y possible discourse structures. T h e chronological is apparenuy the 'natural' order because it matches its discourse structure to the event structure. Fairy stories are typical - even archetypal - of the canonical time ordering of stories. So, in ' T h e T h r e e Little Pigs': The first little pig meets a man carrying straw, buys the straw, and builds a house out of i t The big bad wolf comes, huffs, puffs, blows the house of straw down and eats up the pig.
BELL THE DISCOURSE STRUCTURE OF NEWS STORIES
81
The second little pig meets a man carrying sticks, buys the sticks, builds a house, and ends up getting eaten too by the wolf. The third little pig meets a man carrying bricks,... etc. But the wolf fails to blow the brick house down, the pig puts a pot on the lire to boil, the wolf climbs down the chimney to get at the pig, but instead gets boiled for dinner. End of story. Analysis of a Story's Time Structure News is b y n a t u r e time-bound - it is a perishable commodity. T i m e is a defining characteristic of the n a t u r e of news, a major compulsion in news-gathering procedures, a n d a n influence o n the structure of news discourse. Figure 6 shows the text of a typical agency story, as published in Wellington's Evening Post newspaper. T h e time structure of the story's events is listed d o w n the right-hand side. Time zero is the story present, the time of the lead event in the lead sentence. Times prior to this are labelled T i m e - 1 for the event immediately preceding, moving back u p to T i m e - 6 in this story, the earliest occurrence in the reported background. T h e story also reports o n events subsequent to T i m e 0, labelled Time + 1 (the diagnosis), a n d so on. H e r e I report only briefly for illustrative purposes o n a m o r e detailed analysis drawn from Bell (1995b). T h e lead sentence covers three temporally a n d causally sequential events, but in reverse chronological order. T h e result (Higuchi's departure to hospital) precedes cause (hunger strike), which itself precedes prior cause (disqualifica tion from elections). T h e y are linked b y temporal expressions, usually after for a sequence as in examples 2 a n d 4 of the single-sentence stories analysed earlier, or Of for simultaneous events (as in the headline in figure 4). T h e temporal ex pressions c o m m o n l y imply a causal link, even though this m a y n o t always seem warranted. I n figure 6, T i m e Zero is explicitly identified in the lead sentence - 'today' with other earlier or later time points signposted in the later sentences. S o m e of the time references situate events in calendar time (since August in S6), others in relation to each other [just 24 hours after... in S I , earlier in S3), still others are deictic with the present as reference point. Labov found that a defining characteristic of personal narrative as a form is the temporal sequence of its 'clauses'. T h a t is, the action is invariably told in the order in which it h a p p e n e d , what I h a v e labelled canonical order - 'matching a verbal sequence of clauses to the sequence of events' (Labov, 1972: 360). News stories, b y contrast, are seldom if ever told in chronological order. T h e time struc ture of the story in figure 6 is v e r y complex, with nine points in time identified in the analysis. T h e story as a whole divides into three sections: S l - 2 , S 3 - 5 , S6. The lead sentence alone covers three events a n d times, as w e h a v e seen. E a c h of the three sections represents a cycle through events taking us further back in time, presenting in reverse order of actual occurrence (plus a couple of excursions into story future time) the chain of events which h a v e culminated in the lead
DISCOURSE STUDIES
82
TIME STRUCTURE Protest cut short 51 Lima, Jan 18. - The estranged wife of Peru's President Alberto Fujimori was taken to hospital today just 24 hours after she began a hunger strike to protest at her party's elimination from congressional elections.
0 :
-2 -3
52 Doctors said she was suffering from tachycardia, or an accelerated heartbeat.
+1
53 Earlier, [deposed first lady Susanna] Higuchi, sitting under an umbrella in a scorching summer sun outside the National Electoral Board's headquarters, had pledged to press on with her protest.
-1
54 The electoral board said on Monday Higuchi's Armonia-Frempol party had not qualified for the April Congressional vote because it failed to present a full list of candidates for the 120-member legislature.
-3
55 Board member Manuel Catacora said today that since Higuchi had presented her party's congressional slate just 10 minutes before the filing deadline, a provision allowing parties five days to correct any error did not apply.
i
i r ' 1
+2 -4 0 -4 0
l
56 Higuchi, a 44-year-old civil engineer, has been estranged from Fujimori since August when she protested an election law that banned her from running from [sic] public office. - Reuter
, -5 -6
Figure 6: International news agency story as published in the Evening Post, Wellington. 20 January 1995 (reproduced by permission of Reuters)
event of this story. T h e earliest events are reported last of all, with the final sen tence describing events from some six m o n t h s previously which are the antece dents to the present occurrences. I n the b o d y of the story, perceived news value overturns t e m p o r a l sequence a n d imposes an order completely at odds with linear narrative. It m o v e s back wards and forwards in time, picking out different points o n each cycle or giving m o r e detail o n previously m e n t i o n e d matters (S4 a n d S5). This violation of our expectations that narratives usually proceed in temporal succession is a distinctive feature of news stories - the inverted p y r a m i d structure. Van Dijk has called this the instalment m e t h o d (1988b: 43), b y which an event is introduced then returned to in m o r e detail two or m o r e times. T h e radical discontinuity of time between sentences imparts a general lack of syntactic, semantic a n d discourse cohesion
BELL THE DISCOURSE STRUCTURE OF NEWS STORIES
83
within the news story, as we have seen in the figure 4 I R A story. T h e reversechronological structure has a direct pay-off (and in part doubtless a n origin) in journalistic practices - stories are routinely cut from the b o t t o m u p to fit into available space. T h e i m p o r t a n t information must therefore c o m e as early as possible, a n d the story should b e capable of ending at any sentence. So, in the IRA story, at least the last two sentences could b e cut without affecting the earlier part of the story (for m o r e discussion of this question, see Bell, 1991). Earlier Forms ofNews Telling As news consumers we are so accustomed to this a p p r o a c h that w e forget how deviant it is c o m p a r e d b o t h with other narrative genres a n d with earlier norms of news reporting. Figure 7 shows a story from the New Zealand Herald of 11 J u n e 1886. It reports the eruption of M o u n t Tarawera, a volcano located some 200 k m to the south-east of Auckland, w h e r e the Herald is published. T h e eruption caused significant loss of life a n d reshaped a n extensive area of the New Zealand landscape. It is a typical disaster story which would r u n in any m o d e r n newspaper, but here it is narrated in absolute chronological order. Events are told from the viewpoint of the readers in Auckland - first hearing the sounds of the distant eruption a n d speculating o n their cause, then eventually receiving the n e w s of the real cause. W h e r e the historical story begins at the beginning, the m o d e r n lead sentence would r u n something like this: 'Mount Tarawera erupted last night killing at least 30 people and sending hundreds more fleeing . ..' Research o n the d e v e l o p m e n t of m o d e r n news narrative style indicates that the inverted p y r a m i d structure developed in A m e r i c a n journalism at the e n d of the nineteenth century. Schudson (1982,1989) found that until the 1880s, stories covering presidential State of the U n i o n addresses did n o t summarize the key points at the beginning. H o w e v e r , b y 1910 the lead as s u m m a r y was standard. AT an early hour on Thursday morning a noise as of the firing of cannon was heard by many Auckland residents. From the continuousness of the firing, the loudness of the reports, and the apparent occasional sound resembling salvoes of artillery, many people both here and at Onehunga were under the impression that a man-of-war, probably the Russian Vestnik, had run ashore on the Manukau bar and that these were her signals of distress. Vivid flashes, as from the firing of guns, were also seen both at Onehunga and also from the cupola of the Herald Office, which served to almost confirm the impression that there had been a marine disaster. At about 8.30 a.m., however, it began to be circulated about town that a
DISCOURSE STUDIES
catastrophe, far surpassing in horror even the most terrible of shipwrecks, had taken place. The first news was, through the courtesy of Mr Furby, the officer in charge of the telegraph department, issued by us in an extra and consisted of the following message, sent from Rotorua, by Mr Dansey, the telegraphist there, who manfully and bravely "stuck to his instrument" in the face of the most dreadful danger: "We have all passed a fearful night here. At 2.10 a.m. there was a heavy quake, then a fearful roar, which made everyone run out of their houses, and a grand, yet terrible, sight for those so near as we were, presented itself. Mount Tarawera suddenly became active, the volcano belching out fire and lava to a great height. A dense mass of ashes came pouring down here at 4 a.m., accompanied by a suffocating smell from the lower regions.
1
'
, , ,,
,
"Several families left their homes in their nightdresses with whatever they could seize in the hurry, and made for Tauranga. Others more lucky, got horses and left for Oxford." Figure 7: News as chronological story: New Zealand Herald, 11 June 1886 j
News Time and News Comprehension T h e discontinuous, non-chronological nature of the time structure of contem p o r a r y news stories invites the question of h o w well readers a n d hearers under stand them. Introducing their study of narrative comprehension, O h t s u k a and Brewer (1992: 319) state: 'If the r e a d e r is to c o m p r e h e n d a narrative text, the reader m u s t b e able to derive the underlying event sequence from the given text sequence.' If this is strictly so, t h e n it has dire consequences for the comprehensibility of news. T h e example analyses in this chapter show it takes consider able effort for the academic analyst - let alone the casual audience - to unravel the time structure. Psychological research o n the comprehension of narratives h a s illuminated b o t h the time structures of different kinds of stories, a n d the effects which different orderings h a v e o n audiences' comprehension. Ohtsuka a n d Brewer (1992) found that readers understood the canonical/chronological version of a story most easily. T h e r e was a significant d r o p in comprehension level for a second version of the same story which was presented in direcdy reverse chronological order, a n d for a third version which told some of the events using flashbacks. A final version contained 'flashforwards' which could n o t b e immediately related to what h a d already b e e n narrated, a n d h e r e c o m p r e h e n s i o n was little better than chance. Relating these principles a n d findings to news stories, w e can see that while the canonical/chronological pattern is n o t used in m o d e r n news, nor is the backwards pattern used in its p u r e form of telling a n entire story from last evem 1
BELL THE DISCOURSE STRUCTURE OF NEWS STORIES
85
Flashback appears to b e the staple time structure of news stories, and this is a b u n d a n d y evident in the stories examined in this chapter. T h e findings to date o n comprehension of news stories are equivocal o n whether the non-linear time sequence of standard news writing is in fact h a r d e r to c o m p r e h e n d , or whether chronological order is easier. Lay readers clearly have some ability to reproduce news formats. Lutz a n d W o d a k (1987) tested the form in which subjects retold n e w s stories, asking t h e m to write a lead sentence themselves. T h e y found that most middle-class students r e p r o d u c e d a standard, reverse-chronology news format But while some lower-class informants included much of the desired information, their retellings did n o t follow news style. I n another study, Duszak (1991) h a d readers reconstruct the r a n d o m i z e d sentences of a n e w s story. Most identified the lead sentence correcdy, b u t there were two relevant tendencies in h o w they ordered sentences after the lead. Infor mants a p p e a r e d to b e avoiding the radical discontinuousness of news formats by grouping apparently cohesive sentences together. A n d they tended to reas semble events in a m o r e chronological order t h a n the original h a d , for instance putting background material earlier in the story rather than at the end. It seems that readers h a v e a default strategy b y which - in the absence of cues to the contrary - they re-impose chronological order o n events in the discourse. Duszak rightly interprets this as a 'powerful drive' to retell events chronologically, clash ing with readers' knowledge of the ordering p r o m o t e d b y news schemata. More research is n e e d e d o n the effects of time structure o n news compre hension. T h e hypothesis that the non-canonical news format does adversely affect understanding is a reasonable o n e o n the basis of comprehension research into other narrative genres, but the degree to which familiarity with news models may mitigate these p r o b l e m s is unclear. Brewer hypothesizes that certain discourse genres are particularly intended for comprehension. I n these, 'the order of events in the discourse will m a p the order of the underlying events' (Brewer 1985:187) - that is, they will b e written in chronological order. A m o n g the genres supposedly designed for compre hension h e explicitly a n d rightly includes news stories - with n o c o m m e n t o n the fact that news stories virtually never follow the easily comprehensible can onical order. T h e y consist of series of flashbacks, flashforwards a n d reverse tellings which, w h e n they are used in films or novels, challenge the audience's powers of decoding. So why d o journalists write in a n order which we k n o w to b e less easily com prehended, w h e n one of their declared goals is reader comprehension? O h t s u k a and Brewer (1992: 331) ask the same question of non-chronological orderings in any kind of narrative: 'why d o n ' t authors always write texts in canonical form?' Their answer is that authors h a v e purposes - and narratives h a v e func tions - other than just comprehension. T h e y serve to entertain a n d persuade, for example. W i t h n e w s stories, however, the question is especially sharp, be cause comprehension b y the audience is one of the m a i n explicit aims of news >vriting. T h e answer lies in a marriage ofjournalistic values, journalistic practices and technological development, which is strong enough to overturn the drive to comprehensibility.
8fi
DISCOURSE STUDIES
T h e values of m e d i a control the way in which news is presented. We can account for the w a y news stories are structured only with reference to the values b y which one 'fact' is j u d g e d m o r e newsworthy than another - a n d so more worth r e m e m b e r i n g a n d understanding. I n particular the value of immediacy, as we h a v e seen, is a powerful force in news discourse. T h e reason for this is the dominating role that time plays in the practice of news-work (Bell, 1995b). T i m e rules news-work in a way it does few other professions: the product must b e finished b y a set time or it is useless. Schlesinger's study of B B C newsroom practices rightly characterizes news-work as ' a stop-watch culture' (1987: 83). Newsrooms operate against the clock, measuring daily achievement b y the ability to p r o d u c e a required n u m b e r of stories for the deadline. T h e drive to get the news first is e m b e d d e d d e e p in the news ethos, a n d radically affects the structure of the news text - to the extent of motivating attempts to disguise the age of news which is less than fresh. T h e s e work practices b o t h drive and are reinforced by technological development. F r o m the invention of the telegraph, which gave rise to m o d e r n news coverage as w e k n o w it, to the presentation of live news as it h a p p e n s through the satellite a n d the cable, innovation in technology has b e e n aimed at reducing the time between the occurrence of a n e w s event and its telling.
6 Conclusion: Application and Applicability T h e detailed analysis of news stories presented in this chapter has told us a lot about the character of the stories - indications of h o w they w e r e m a d e , evidence of discrepancies a n d gaps, a n d manifestations of the news values b e h i n d them. T h e r e is considerable pay-off from a comparatively straightforward discourse structural analysis of this kind. It can o p e n u p a subtext of the news, for instance of implied linkages which m a y n o t stand u p to scrutiny. It can lead us closer to 'what the story actually says h a p p e n e d ' . I n the I R A story, for example, we h a v e b e e n able to infer a good deal aboui its production a n d potential reception as well as about the text itself. We have seen the joins between the different input texts that h a v e g o n e to m a k e this one story; the divergent local a n d international readerships at which these inputs w e r e targeted; a n d the conflicting views held b y journalist a n d copy-editor over what was really the story in these events. We h a v e n o t e d the unspecificness of m u c h of what we h a v e read, reflected in alternative possible discourse structures for the story, which m a k e it impossible from this text alone to determine exacdy what w e r e the persons, places a n d events involved. We h a v e seen how the routinely j u m b l e d time structure a n d lack of cohesion between sentences has served to m a k e s o m e things unclear, a n d l o o k e d at s o m e of the repercus sions that m a y h a v e for comprehension of news stories. We h a v e b e e n able to assess the validity of the headline, a n d the reliability of evaluations put forward in the story.
BELL THE DISCOURSE STRUCTURE OF NEWS STORIES
87
It is particularly salutary h o w our specification of events, persons a n d places changed as the analysis m o v e d further d o w n the story. T h e interpretation based only on the headline a n d lead was rather misleading. If this h a d b e e n a singlesentence story of the kind analysed earlier in the chapter, the level of unclarity and ambiguity would h a v e r e m a i n e d very high. Projecting this experience back on to the earlier examples, it seems likely that the interpretations I offered of their event a n d discourse structures would b e p r o v e d quite mistaken if we h a d access to the full text of the stories from which they h a v e b e e n cut. T h e system of analysis I presented in t h e Guide earlier covers most of what we might want to k n o w about a story's discourse structure, b u t there are other questions we m i g h t ask in relation to it, such as: • W h a t vocabulary is used in the lead to claim news value for the story? W h a t is the order of sentence constituents within the lead? W h a t news values lie b e h i n d that order? Is there a conflict b e t w e e n the journalist's goals of brevity a n d clarity in the lead? • W h a t figures or similar ' h a r d facts' does the story contain? H o w does the story use these to buttress its news value? • H o w m u c h of the story consists of talk? W h a t kind of talk - statement, accusation, reaction, a n n o u n c e m e n t ? • W h a t are the news values which h a v e m a d e this a published news story? The analytical framework concentrated o n the journalist's five W s a n d an H . But it is significant that two of those did n o t figure explicitiy - ' h o w ' a n d par ticularly 'why'. ' H o w ' , of course, is in m a n y ways represented as the detail of 'what', but 'why' is n o t overfly addressed in the I R A story - or in m a n y others, it is no accident that C a r e y characterizes ' h o w ' a n d 'why' as the 'dark continent' of journalism (Carey, 1987; cf. Bell, 1996). Secondly, while there is g o o d insight to be gained from a n understanding of a profession's own understanding of its work (Manoff a n d Schudson, 1987), it must n o t b e forgotten that such categories -.hould themselves b e subjected to analysis a n d deconstruction. The analytical work involved in this framework is n o t small. T h e d e m a n d is w fact a true reflection of the real complexity of texts such as n e w s stories, which sometimes a p p e a r deceptively simple. I n doing such work, it is essential •o move systematically, picking off the analysis point b y point. Yet while we n e e d to b e selective in our analyses a n d k n o w what is an appro priate point to cut off at, it is evident to m e that m a n y of the insights yielded tome only near the e n d of detailed work following a pattern such as that outiined in the analysis G u i d e . Because t h e analyses are time-consuming if u n d e r taken at all comprehensively (cf. van Dijk, 1991: 10), we n e e d to c o m p l e m e n t full-scale work o n a few texts with m o r e piecemeal, specific analyses o n larger samples. But the r e w a r d of such work is a better knowledge of what the stories of our time are saying, and an understanding of the w a y they m a y b e p r o d u c e d and received.
DISCOURSE STUDIES
88
Notes This chapter is in large part the fruit of a fellowship I held in 1994 at the Centre for Language and Communication Research, University of Wales Cardiff. So also was the idea for the Cardiff Round Table as a whole and this publication of a range of approaches to media discourse. I am grateful to the Centre - and particularly to its Director, Nikolas Coupland - for support and warm hospitality on that and later occasions. During the fellowship, I taught a course on news language and discourse, and much of the present chapter grew from that material and experience. I work normally as a researcher and journalist rather than a teacher, and it was salutary how hard the students found it to understand and apply these analytical approaches. This was the main motivation behind the Round Table meeting and this collection. I am greatly indebted to co-editor Peter Garrett for the encouragement and insight he brought to my chapter and to the book as a whole. 1. In this section, I report at second hand the findings of other researchers on story 'comprehension' which are relevant to my concern with temporal order. This begs methodological and theoretical questions that are beyond my scope to address in this chapter - such as the relationship between recall and comprehension, and between written and oral channels, and the dynamic way in which audiences interpret meanings in everyday life. Compare Richardson's and Allan's chapters in Approaches to Media Discourse (1996).
References Bell, Allan, 1991. The Language of News Media. Oxford: Basil Blackwell. Brewer, William F, 1985. 'The story schema: universal and culture-specific properties.' In David R Olson, Nancy Torrance & Angela Hildyard (eds), Literacy, Language, and Learning: the Nature and Consequences of Reading and Writing. Cambridge: Cambridge University Press. 167-94. Cary, James W, 1987. 'Why and how: the dark continent of American journalism.' In Robert Karl Manoff & Michael Schudson (eds), Reading the News. New York: Pantheon. 146-96. Duszak, Anna, 1991. 'Schematic and topical categories in news story reconstruction.' Text 11/4: 503-22. Gaining, Johan & Mari Holmboe Ruge, 1965. 'The structure of foreign news.' Journal of Peace Research 2/1: 64-91. Genette, Gerard, 1980. Narrative Discourse: an Essay in Method. Ithaca, NY: Cornell University Press. Hjarvard, Stig, 1994. 'TV news: from discrete items to continuous narrative? The social meaning of changing temporal structures.' Cultural Studies 8/2: 306-20. Labov, William, 1972. Language in the Inner City: Studies in the Black English 'Vernacular. Philadelphia: University of Pennsylvania Press. Labov, William & Joshua Waletzky, 1967. 'Narrative analysis: oral versions of personal experience.' In June Helm (ed.), Essays on the Verbal and Visual Arts (Proceedings of the 1966 Annual Spring Meeting of the American Ethnological Society). Seattle: University of Washington Press. 12-44.
BELL THE DISCOURSE STRUCTURE OF NEWS STORIES
89
Lutz. Benedikt & Ruth Wodak, 1987. Information fur Informierte: Linguistische Studien zu Verstandlichkeit und Verstehen von Horfunknachrichten [Information for the Informed: linguistic Studies on the Comprehensibility and Comprehension of Radio News]. Vienna: Verlag der Oesterreichischen Akademie der Wissenschaften. Manoff, Robert Karl & Michael Schudson (eds), 1987. Reading the News. New York: Pantheon. Ohtsuka, Keisuke & William F Brewer, 1992. 'Discourse organization in the compre hension of temporal order in narrative texts.' Discourse Processes 15/3:317-36. Pollard-Gott, Lucy, Michael McCloskey & Amy K Todres, 1979. 'Subjective story structure.' Discourse Processes 2/4: 251-81. Rumelhart, David E, 1975. 'Notes on a schema for stories.' In Daniel G Bobrow & Allan Collins (eds), Representation and Understanding. New York: Academic Press. 211-36. Schlesinger, Philip, 1980. 'Between sociology and journalism.' In Harry Christian (ed.), The Sociology of Journalism and the Press (Sociological Review Monograph 29). Keele: University of Keele. 341-69. Schlesinger, Philip, 1987. Putting 'Reality' Together: BBC News (2nd edition). London: Methuen. Schudson, Michael, 1987. 'When? Deadlines, datelines and history.' In Robert Karl Manoff & Michael Schudson (eds), Reading the News. New York: Pantheon. 79-108. Schudson, Michael, 1989. 'The sociology of news production.' Media, Culture & Society 11/3:263-82. rhorndyke, Perry W, 1979. 'Knowledge acquisition from newspaper stories.' Discourse Processes 2/2: 95-112. Toolan, Michael J, 1988. Narrative: a Critical Linguistic Introduction. London/New York: Routledge. van Dijk, Teun A, 1988. News as Discourse. Hillsdale, NJ: Lawrence Erlbaum Associates, van Dijk, Teun A, 1991. Racism and the Press. London/New York: Routledge.
Crossing Genders, Mixing Languages: The Linguistic Construction of Transgenderism in Tonga Niko Besnier
1 Introduction
T
his c h a p t e r takes as p o i n t of d e p a r t u r e t h r e e s e e m i n g l y unrelated developments in social a n d cultural anthropology. T h e first concerns recent rethinking of anthropological approaches to gender as a social and cultural category. H e r a l d e d b y feminist anthropologists in the last decades of the twentieth century, this shift is spurred o n b y the insistence that gender (and, b y implication, all other social categories) is always e m b e d d e d in a complex m a z e of other social divisions that criss-cross all social groups: social class, race a n d ethnicity, religious identity, age, sexuality, citizenship in its various manifest ations, position in structures of production a n d consumption, a n d so on. On b o t h large-scale dimensions a n d in microscopic fashion, all aspects of social identity a n d dimensions of social difference can potentially inform or even deter mine the m e a n i n g of gender, dislocating sameness where it is least expected, and potentially establishing connections b e t w e e n surprisingly distinct categories, persons, a n d entities. A corollary to the recognition of the inherently embedded nature of gender is the assertion that "all forms of patterned inequality merit analysis" (di L e o n a r d o 1991: 31), a n d that such analysis is the sine qua non of an anthropological coming-to-terms with the m e a n i n g of gender. T h e second d e v e l o p m e n t I a m concerned with arose with the increasing malaise a m o n g anthropologists, also characteristic of the 1980s a n d 1990s, with the tacit equation of culture with place, a n d the continued assumption that social
Source: Janet Holmes and Miriam Meyerhoff (eds), Handbook of Language and Gender, 2003, pp. 279-301.
BESNIER
CROSSING GENDERS, MIXING LANGUAGES
91
groups could simply b e defined in terms of geographic co-presence. A p p a d u r a i (1996), a m o n g others, demonstrates that locality is a problematic category for an ever-increasing n u m b e r of people, for various possible reasons: place (of origin, affective ties, residence, etc.) m a y n o t b e a singular, well-defined entity, as is often the case of the migrant. Place of origin m a y b e a site of violence and horror, which is best erased from m e m o r i e s a n d daily lives, as in the case of refugees from civil wars a n d genocidal situations (e.g. D a n i e l 1996; Malkki 1995). Alter natively, place can h a v e shifting, context-bound characteristics that vary with persons and contexts (Gupta and Ferguson 1997; Lovell 1998). Consequendy, as Marcus (1995) argues, the age-old pattern of anthropological fieldwork that objectified "the O t h e r " in distant lands is giving w a y to a m o r e dynamic, "multisited" pattern of research, in which t h e ethnography "follows" persons, objects, or metaphors as they travel across geographies a n d histories. T h e third anthropological preoccupation I invoke is the effort to c o m e to grips with the various forms a n d meanings of modernity. Modernity, the condi tion of experience associated with capitalism, industrialism, consumption, a n d other characteristics of life in "the West," has long occupied a privileged if backstaged place in anthropology and the social sciences. At its inception, anthropology was defined as the study of what m o d e r n i t y was not; even recently, m u c h work in anthropology continued to tacitly assume a n unproblematic contrast between modernity a n d traditionalism (Spencer 1996: 378-9). H o w e v e r , recent thinking has unsettled the facile dichotomy b e t w e e n tradition a n d modernity, d e m o n strating, for instance, that the two categories are mutually constitutive, a n d that forms of tradition a n d forms of modernity are commensurable in m a n y contexts. Furthermore, neither tradition n o r m o d e r n i t y is a unitary condition: there m a n y forms of modernity (as illustrated b y the "alternative modernity" ofJ a p a n , for example) and, as Comaroff a n d Comaroff point out, "[n]or should this surprise us. With hindsight, it is clear that the cultures of industrial capitalism h a v e never existed in the singular, either in E u r o p e or in the myriad transformations across the surface of the earth" (1993: xi). In this chapter, I explore h o w these various strands of thinking can b e tied together, a n d inform concerns of language a n d gender. I explore the role of lan guage use in constructing gender in the context of a n investigation of h o w other social and cultural categories define gender. For example, m e n a n d w o m e n in many societies h a v e different interests (in the various senses of the term) in "tradition" and "modernity," in the maintenance of the status quo or the emergence of new social arrangements, a n d language behavior a n d ideologies are often constitutive of these differing investments. I n this project, I take gender n o t as a given, but as potentially emerging out of conflict and negotiation between members of a society, conflict a n d negotiation in which language plays a n important role. The empirical basis of m y discussion is a n ethnographic examination of the lives of transgendered males in Tongan society. Like all larger societies of the Polynesian region (Besnier 1994), Tongan society counts in its ranks a substantial number of m e n w h o "act like w o m e n , " a category that Tongans refer to variously asfakaleiti, leiti, or fakafefine. T h e first t e r m is the most c o m m o n l y h e a r d at this moment in history; it is a lexical c o m p o u n d m a d e u p of the ubiquitous polysemic
92
DISCOURSE STUDIES
prefix faka-, which in this context m e a n s "in the m a n n e r of"; leiti is borrowed from the English word "lady," which is only used to refer to transgendered per sons (i.e. never to female "ladies"). Transgendered Tongans prefer the unprefixed version of the t e r m to refer to themselves, arguing somewhat tongue-in-cheek that they are n o t like ladies b u t they are ladies (I explore additional reasons for the preference of the shorter word in Besnier 1997:19-20). T h e last term, fakafefine, literally "in the m a n n e r of a w o m a n , " is slighdy old-fashioned, but it is readily understood because its m e a n i n g is transparent from the s u m of its parts. I n the discussion that follows, I first introduce Tongan society as a diaspora scattered widely in the Pacific R i m , whose center of gravity is a n independent nation-state coterminous with a group of islands in the Southwestern Pacific, the K i n g d o m of Tonga. I briefly describe the sociocultural m e a n i n g of the two principal languages spoken b y m e m b e r s of this diaspora, Tongan and English, a m e a n i n g which is undergoing rapid change as expatriate Tongans in N e w Zealand, Australia, and the U n i t e d States increase in n u m b e r a n d p r o m i n e n c e . I then turn to the position offakaleiti in Tongan society, which I show to b e varied and full of inherent contradictions. I demonstrate that English has b e c o m e a trade m a r k of fakaleiti identity in the islands, as it encodes a cosmopolitanism and m o d e r n i t y which m a n y leiti find useful to foreground in their daily lives. However, this trademark has a price, in that m a n y leiti are n o t fluent in English a n d most d o n o t have access to the material m e a n s of backing claims of cosmo politanism with tangible tokens of it. I n addition, mainstream society can utilize the claims associated with the use of English to dislocate leiti from the local context a n d further marginalize t h e m . i•
2 T o n g a n S o c i e t y as a D i a s p o r a T h e fieldwork o n which this chapter is b a s e d was conducted principally in the capital of Tonga, Nuku'alofa. However, the Tongan diaspora figures prominently in all aspects of the economic, social, a n d cultural life of the island society, and its importance continues to increase, despite efforts from some quarters to contain a n d minimize it. A s a nation-state a n d a n island-based society, Tonga therefore cannot b e considered independentiy of overseas Tongan communities. Altogether, about 150,000 persons claim Tongan descent, of w h o m about 97,500 reside in the archipelago, a loose clustering of 150 islands, 36 of which are perman ently inhabited. Overseas Tongans live principally in Auckland, Sydney, the San Francisco Bay Area, u r b a n Southern California, a n d Salt Lake City, but there are small groups of Tongans or single individuals just about everywhere in the world. T h e size, diversity, a n d importance of the diaspora is particularly striking in light of the fact that significant emigration only b e g a n in the 1970s. Tongans are Polynesians, a n d their society has b e e n one of the most stratified a n d politically centralized of the region since its early-nineteenth-century unifica tion u n d e r t h e rule of a sacralized king. A British protectorate b e t w e e n 1900 a n d 1970, Tonga is today a n independent state. State a n d society are b o t h founded o n a marriage b e t w e e n selected aspects of a p u r p o r t e d tradition a n d selected
BESNIER CROSSING GENDERS, MIXING LANGUAGES
93
aspects of a version of modernity (Philips 2000: 2 3 5 - 6 ) . For example, the State is "the only remaining Polynesian k i n g d o m " a n d a n u p h o l d e r of Christianity, features that Tongans consider to b e illustrative of timeless tradition, while also emphasizing the fact that Tonga is a n economically forwardlooking entity, a symbol of modernity. However, different elements of society a n d the State m a y differ on key points as to which aspects of tradition and which aspects of modernity should b e m a d e relevant to Tonga: for example, parliamentary representation and the scope of the nobility's political p o w e r are topics of acrid debate, parti cularly since the emergence of a Tongan Pro-Democracy M o v e m e n t in the 1980s (Campbell 1992: 218-22). T h o s e in p o w e r view dissenting voices as signs of a n undesirable modernity, often associating t h e m with the diaspora. Despite rapid increases in the Tongan populations of cities such as Auckland, the most important u r b a n centre for Tongan society continues to b e Nuku'alofa, the capital of the nation-state, inhabited b y about 25,000 people, m a n y of w h o m have m o v e d there from rural areas of the country in the last few decades. Nuku'alofa is the p r i m e destination of overseas Tongans' visits to the island kingdom, in part because its international airport is the most important point of entry into the country from Hawaii, N e w Zealand, Australia, a n d Fiji. It is the venue of most national celebrations, including ceremonies relating to kingship, government, and n a t i o n h o o d (e.g. coronations, important funerals, a n d yearly festivals of culture). Nuku'alofa is the focal point of b o t h the rest of the nation and the diasporic dispersion. It serves as the point of convergence for most of the intensive flows of goods, m o n e y , a n d people that k e e p the diaspora together. In the context of the rapidly increasing transnationalism of their society, many Tongans see the maintenance of a quality of "Tonganness," as well as the very definition of this quality, as areas of concern (Morton 1996; Small 1997). Tongans refer to this quality as angafaka-Tonga, "behavior in the fashion of Tonga," or. when speaking English, "the Tongan way," echoing comparable phrases used in neighboring societies. T h e quality is concretized most forcefully in high culture, including the performing arts, the manufacture a n d exchange of koloa "valuables" (tapa-cloth a n d mats), ceremonies affirming hierarchy a n d kinship, and of course language. H o w e v e r , Tongans often invoke anga faka-Tonga w h e n referring to culture in the b r o a d e r anthropological sense, particularly w h e n the context calls for a contrast between locality and extralocality. For example, overseas Tongans and locally based but cosmopolitan Tongans lay claims o n "Tonganness" that other Tongans sometimes challenge (Morton 1998). "Tonganness" is deeply tied to place, b u t in potentially conflicting ways.
3 T o n g a n a n d English The tensions associated with the definition a n d maintenance of local identity and related dynamics are p e r h a p s m o s t clearly enacted in the competition between the two principal languages utilized in Tongan society, Tongan a n d English. Just about everyone in Tonga knows at least rudiments of English, which is a prominent language in schooling a n d even, in the case of a few schools, the
DISCOURSE STUDIES
94
only language of instruction. However, Tongans vary widely in terms of their fluency in English a n d the degree to which they feel comfortable speaking and writing English. Both fluency and readiness to speak English (which are not necessarily coterminous, as I will illustrate presently) d e p e n d o n an aggregate of factors closely linked to the structuring of social inequality in Tonga. First, English is a prestige language, as elsewhere in the Pacific where it is the m a i n postcolonial cosmopolitan language: linked to a colonial past, it dominates contexts of e m p l o y m e n t , education, m o d e r n i t y , transnationalism, contacts with the external world, a n d n e w forms of socio-economic h e g e m o n y such as entrepreneurship. Elite Tongans of either r a n k or wealth are m o r e likely than non-elite Tongans to h a v e resided in English-speaking countries u n d e r favorable circum stances (pursuing their education or visiting, for example), and therefore generally have h a d m o r e opportunities to b e c o m e fluent speakers of English. T h e y are also intimate with the privilege and cosmopolitanism that English indexes. I n contrast, most non-privileged Tongans are often reluctant to speak English, ostensibly, according to explanations offered, because they fear making linguistic errors. I n practice, their reluctance is n o t so m u c h a matter of defective gram matical competence, but of n o t having the social self-assurance to assert oneself credibly as a privileged, m o d e r n , and cosmopolitan person without fearing shame (ma) a n d exposing oneself to ridicule. W h i l e m a n y non-elite Tongans have resided overseas, they h a v e invariably b e e n employed in menial j o b contexts, i n which communication with native speakers of English is confined to jobrelated topics (e.g. understanding directives). I n Tongan communities in cities such as Auckland a n d the San Francisco Bay Area, the life of m a n y less-thanprivileged first-generation m i g r a n t Tongans continues to b e predominantly Tongan-centered a n d Tongan-speaking. As is the case of m a n y migrant com munities, it is only the overseas-born generation that acquires fluency in the d o m i n a n t language. T h e association of English with privilege is n o t unmitigated, for at least two reasons. First, m o s t Tongans exhibit a high degree of allegiance to their own language. It is n o t u n c o m m o n to h e a r Tongan being used as an everyday tool of resistance to the h e g e m o n y of English. For example, it is used widely in the workplace, h o w e v e r steeped this workplace m a y b e in the English language a n d associated symbols. I n Nuku'alofa streets, youngsters d o n o t fail to crack loud jokes in Tongan at the expense of any foreigner (Palangi) t h e y pass, w h o m they assume not to understand the language. But the prestige of Tongan is also asserted in contexts w h e r e English is not a competing code, as in oratory, ceremonialism, a n d song-and-dance concerts, a n d thus it is n o t solely associated with resistance. Second, there are contexts in which people use English widely without access to the material resources to "justify" their code choice, a n d without a n y obvious fear of s h a m e either. O n e example is the v e r y popular Nuku'alofa flea market, w h e r e English is a c o m m o n m e d i u m a m o n g sellers and often also customers. W h a t is interesting, though, is that the flea m a r k e t is also o n e of the most visible local sites of modernity and transnationalism, for several reasons. Most simplistically, the goods sold (principally second-hand clothing) are from overseas, a n d thus the market is a place where people go to b u y the product of transnational 1
2
BESNIER CROSSING GENDERS, MIXING LANGUAGES
95
links. I n addition, socially marginal groups a n d "local Others," that is, persons who are already marginalized because of their non-mainstream religious affili ation or lifestyle (e.g. Mormons, Charismatic Christians, entrepreneurs), are overrepresented a m o n g the sellers. Furthermore, the act of selling, particularly second hand objects, flies in the face of the "traditional'' order: in the "Tongan way," selling used items makes others suspect that the sellers are so p o o r that they are forced to sell their possessions, a state of substantial ma "shame." However, sellers w h o m I interviewed described with pride h o w they h a d overcome the strictures of traditionalism a n d b e c o m e m o d e r n persons, a process that some at tributed to their religious affiliations. T h e p r o m i n e n c e of English a n d the modernity that suffuse the flea market are thus not coincidental, and they indicate that Tongan a n d English are embroiled in potentially complex structures of competing prestige, along with the categories with which each language is associated, a t h e m e which will figure p r o m i n e n d y in the analysis that follows. 3
4 Leiti i n T o n g a n S o c i e t y It is impossible to come u p with a precise definition of w h o ufakaleiS is in Tongan society, for the same reasons that defining " m a n " or " w o m a n " in any social con text is neither feasible not fruitful. A s for all social categories, one cannot isolate a set of necessary a n d sufficient conditions to determine w h o is a fakaleiS a n d who is not. Nevertheless, stereotypes abound, as they d o wherever a marginalized minority is concerned in all social groupings. O n e can therefore utilize these stereotypes to provide a working definition of the category, bearing in m i n d at all times that they are stereotypes, a n d h e n c e that they are p r o n e to distortions, underlain with covert moral judgments, a n d subject to socio-political manipulation. Mainstream Tongans stereotypically associate a fakaleitVs presentation of self with a "feminine" c o m p o r t m e n t (e.g. emotional way of talking, a n animated face, "swishy" walk). I n domestic or rural settings, leiti d o " w o m e n ' s " work (e.g. laundry washing, cooking, flower gardening, child-minding, caring for elderly parents) and d o n ' t d o or d o n ' t like to d o physically d e m a n d i n g work associated with men (e.g. subsistence gardening, w o o d chopping, construction). I n u r b a n contexts, they h o l d occupations that h a v e feminine associations (e.g. seamstress, hairdresser, cook, "house-girl"), because they either cater to w o m e n or are com monly performed b y w o m e n . Fakaleiti are c o m m o n l y characterized as wearing women's clothes a n d make-up, although in practice m o s t leiti wear either m e n ' s or gender-neutral clothes. Their leisures a n d interests are concerned with beauty, creativity, and femininity (e.g. talking a n d doing fashion, hairstyles, a n d decor). They play netball a n d definitely n o t r u g b y (but m a n y , like m e n and in contrast to women, do get drunk, a n d often). Finally, "because" they are like w o m e n , as the local logic goes, fakaleiti have sexual relations with "straight" m e n , that is, with men w h o are not identified as fakaleiti. Most "straight" m e n engage t h e m in frequent banter over their "true" gender identity a n d the possibility of sexual relations, often portraying fakaleiti as the sexual aggressor, a strategy designed in part to emphasize the out-of-control nature of fakaleitVs sexuality (a t h e m e
DISCOURSE STUDIES
96
familiar to m a n y sexually defined minorities a r o u n d the world), a n d in part to invalidate their claim that they are "real w o m e n , " since sexual aggression is a male trait. W h a t these stereotypes do n o t capture is that leiti identity is highly variable, considerably m o r e complex, a n d criss-crossed b y dynamics that reach far beyond the confines of n a r r o w characterizations of gender a n d sexuality. A n important t h e m e that will n o t often arise u n d e r elicitation is the notable w a y in which leiti orient their lives toward aspects of modernity to a n extent a n d in ways that other Tongans d o not. While mainstream Tongans tacitly recognize, in their rapports with a n d attitudes toward leiti, that this orientation is part a n d parcel of who they are, they do not explicidy point to it as a characteristic marker of the identity. I will argue h e r e that it is as central to understanding the m e a n i n g of the category as its gendering.
5 Leiti a n d E n g l i s h It is h e r e that language a n d language use begin to offer a particularly rich entree into the intricacies of the problem. First of all, verbal behavior is one of the most consciously foregrounded features of leiS identity, yet also o n e of the vaguest. W h e n asked, " H o w do y o u k n o w w h e n s o m e o n e is a fakaleiS?,'" mainstream and leiS Tongans often r e p l y , ' Oku te 'ilo 'i H he le'o "You know b y the voice," where le'o "voice" also m e a n s , m o r e generally, "way of speaking, speech mannerism." W h e n pressed further, informants typically suggest that leiS speak with a highpitched voice a n d at a fast t e m p o , a n d engage in dramatic emotional displays. However, attempts to determine this distinctiveness m o r e precisely r u n into the same conceptual and analytic difficulties as characterizations of the linguistic characteristics of gender or sexual minorities elsewhere in the world (cf. Hall and O ' D o n o v a n 1996; Gaudio 1997; O g a w a a n d Smith 1997; a n d m a n y others). W h a t is particularly striking but often left u n m e n t i o n e d b y informants is the salience of English in leiti's linguistic repertoire. T h e most immediate piece of evidence of this salience is the n a m e of the category itself: a borrowing from English used exclusively to refer to transgendered males, the w o r d 'fakaleiti in and of itself indexes the English language, its contexts of use, a n d its symbolic associations with modernity a n d cosmopolitanism, an indexicality that probably operates largely at a subconscious level. This indexicality m a y b e further re inforced b y two factors: the original m e a n i n g a n d connotation of the English w o r d "lady" (evoking sophistication, class, g o o d breeding); a n d leiti's own preference for the unsuffixed version of the term, which "denativizes" the term even further b y stripping it of the Polynesian m o r p h e m e faka-. (Going one step further, leiti sometimes pronounce the t e r m as if it were an English word, voicing the dental stop, diphthongizing the vowel cluster, a n d shifting the stress from the word-final long vowel onto the diphthong.) T h e orientation to English that is p a r t and parcel of leiti identity goes further. N o matter h o w fluent or elementary their English proficiency m a y be, leiti pepper their conversations with one another and others with English. Leiti's code-switching v
4
BESNIER
CROSSING GENDERS, MIXING LANGUAGES
97
can occur in a n y context, a n d can target a wide variety of linguistic units, from single words to large discourse chunks. T h e most frequent examples in my corpus, not surprisingly, are to be found in face-to-face interviews with me, since leiti see me primarily as a speaker of English, even though m y Tongan is perfectly adequate, a n d p e r h a p s m o r e importantly as s o m e o n e with w h o m they wish to establish a r a p p o r t for which the appropriate language is English. T h e following excerpt from a typical one-to-one interview illustrates the ubiquitous nature of borrowings a n d code-switched strings (I = interviewee, N = Niko [myself]): 3
I: Ka koe'uhf, 'e ki'i- te nau feel secure. N: Hm. I: Pea mo e anga ko e fie nofo faka-Tonga, you know, how our culture, 'oku- 'oku tight up pe'a e respect N: Hm. I: ki he matu'a mo e sisters mo e brothers mo e me'a. N: Hm. I: Ka ko e taimi ko e 'oku nau- nau mavahe ai ko e 'o nofo faka'apitanga, pehe N: 'Io. I: 'a e camp. N: Hm. I: Ko e fo'i- fai tahataha pe 'oku tu'u 'i he 'ulu, that's all. N: Hm. I: They don't really care, pe ma'u ha me'akai pe 'ikai. N: Hm. (Transcript 1993: 3, p. 6) Translation I: And because, just- they will feel secure. N: Hm. I: And they have a desire to live in the Tongan way, you know, how our culture, the respect is quite tight up N: Hm. I: for the parents and the sisters and the brothers and so on. N: Hm. I: But when they move out and start living together as roommates in a house, it [becomes] like N: Yes. I: a camp [i.e. an encampment, where norms of respectability are ignored]. N: Hm. I: Every- each does whatever goes through his head, that's all [i.e. and nothing more]. N: Hm. I: They don't really care whether they even get food or not. N: Hm. This excerpt, taken from an interview with a leiti w h o is relatively fluent in English, presents several interesting features. First, m a n y words a n d phrases that the interviewee utters in English could equally h a v e b e e n uttered in Tongan, and in a couple of instances the Tongan equivalent m a y have been m o r e felicitous. Second, some of the terms that m y interviewee utters in English in fact refer to concepts that are highly specific to Tonean soriprv ar>A c.._i-
98
of "respect," a word that in Tongan English has the locally specific meaning of "avoidance behavior between cross-sex siblings and some inter-generational relations," which is much more succincdy denoted in Tongan b y the widely used term faka'apa'apa. Such is also the case of the English kinship terms "brothers" and "sisters," which do not do a good job of capturing the kinship categories relevant to "respect," best understood in terms of crosssiblings (tuonga'ane "[woman's] brother", tuofefine "[man's] sister"). What is particularly interesting is that even leiti who do not have grammatical fluency in standard English nevertheless engage in code-switching with a frequency and poise that would rarely be witnessed among mainstream Tongans of com parable linguistic abilities. The following is an excerpt from an interview with a leiti who is much less fluent in English than the interviewee in the prior excerpt, despite years spent working in Australia. Nevertheless, English words and sentences abound in the interview: I: Ne- 'Aositelelia, sai'aupito 'a 'Aositelelia ia ki ke kau leiti. He ko e- mostly ko e sio ki he- ki he-, have you heard about the Mardi Gras, N: 'Io.'Io. I: 'Oku topu 'a 'Aositelelia he, N: Hm. I: () he nofo pehe. N: Hm. I: E? Lesbian. N: Hm. I: And also the- ladies and the gay. N: Hm. I: () understand? N: Na'akefa'akaukiai? I: / only joins but I- na'e kai- na'e 'ikai ke u 'alu au ki he ngaahi fale pehe. N: 'Io. I: I just went inside and watch them, N: Hm, hm. I: E? But I never do this one. (Transcript 2000: 2, p. 6) Translation I: It was- Australia, Australia is very good to its transgenderedpeople. Because it's- mostly if you look at the- at the- have you heard about the Mardi Gras, N: Yes, yes. I: Australia is top [topu, a recent borrowing from English] on that font, N: Hm. I: () living like that. N: Hm. I: Right? Lesbian. N: Hm. ' " " , I: And also the- ladies and the gay. N: Hm. h {) understand? N: Did you ofkn partake in it? I: I only joins but 1-1 didn't-1 didn't go to that kind of houses \ptesumab\y, gay bar»V
BESNIER CROSSING GENDERS, MIXING LANGUAGES
B9
N: Yes. I: I just went inside and watch them, N: Hm, km. I: Hm? But I never do this one. I n short, grammatical competence, concerns for efficiency of expression or the untranslatability of certain terms, and the fear of shaming are of litde relevance to m y interviewees' code choices. Rather, what is foregrounded in their code choices in interviews with m e , as well as in face-to-face interaction with every one else, is the indexical m e a n i n g of English and possibly the indexical m e a n i n g of the very act of code-switching (cf. Stroud 1992).
6 T h e P u b l i c Construction of Leiti
Identity
With Kulick (1999: 615), I consider a n analysis based o n talk p r o d u c e d in the context of ethnographic interviews b o t h limited a n d limiting (although n o t completely devoid of value, as long as the ethnographer places his or her o w n position u n d e r ethnographic scrutiny). W h a t is of interest in the Tongan material is that the patterns of c o d e choice I elicited during ethnographic interviews with my informants echo strikingly patterns of language use in other contexts, a n d thus are representative of patterns of wide social scope." Take, for example, public talk in the context of the annual beauty pageants that leiti have staged, with increasing a p l o m b since the early 1990s, in some of the most p r o m i n e n t v e n u e s in t h e country. T h e s e events are particularly interesting because, for m a n y Tongans, they represent a context in which fakaleiti identity is most clearly elaborated. Leiti themselves and their non-leiti champions (principally m e m b e r s of a cadre of influential professional w o m e n d'un certain age) see the pageant as a prime opportunity to present themselves in the best light and to seek control of their public image, and thus as a subtle b u t efficacious context for political affirmation. T h e Miss Galaxy beauty pageant is the most salient of these events, although it is only o n e of several comparable events held throughout the year. Like other important events in Tonga, the pageant has a high-ranking or otherwise prominent patron, w h o in recent years has b e e n recruited from within the ranks of the royal family. Half of the j u r y of six or seven is com posed of non-transgendered Tongan dignitaries (e.g. high-ranking a r m y officers, intellectuals, a n d the winner of the mainstream Miss Heilala pageant for "real" women, which precedes the transgendered pageant), while the other half are "distinguished" Expatriates (i.e. t e m p o r a r y foreign residents of Tonga, such as businessmen, spouses of diplomats, a n d the occasional visiting anthropologist). Sponsored b y various businesses a n d organizations (e.g. hotels, hairdressing salons, rugby teams), contestants appear on stage in various costumes, ranging a gamut familiar from South Pacific pageants in general, which includes evening dress, pule taha "island wear" (ankle-length skirt and matching short-sleeved top, worn with a tasseled fiber belt), a n d "their own creations" (see Photograph 1). Each appearance is ostensibly designed to allow contestants to present themselves
DISCOURSE STUDIES
100
as attractive a n d feminine persons, following familiar patterns of beauty pageants around the world. T h e core of the pageant consists of several j u d g e d events, including a n individual talent display, a brief interview (of the what-would-youdo-to-save-the-world? type), and catwalk parades. Interspersed are entertainment routines, which m a y include a h u l a performance b y the emcee, a rock-and-roll standard sung b y a local talent, a dance routine performed b y all contestants to a popular Tahitian or disco tune, and a short classical a n d torch-song concert b y non-transgendered performers. W h a t I designate "extra-locality'' pervades the entire atmosphere of the Miss Galaxy pageant. It is a feature of the pageant that organizers a n d contestants take great pains to elaborate, and that the audience expects of the show, although these expectations are always mitigated b y the view that this extra-locality is fraudulent. T h e most immediate a n d spectacular manifestation of extra-locality is the very n a m e of the event. Both funny a n d poignant, "Miss Galaxy" lays a claim o n as ambitiously cosmopolitan an image as can b e imagined, a n d plays o n hyperbole in the same fashion as some of the c a m p aspects of the pageant (e.g. the m o r e extravagant costumes a n d performances), creating h u m o r while attempting to retain control of this h u m o r . But extra-locality also saturates other aspects of the pageant. For e x a m p l e , o n e of the events requires contestants to a p p e a r in "national" costume as representative of foreign "countries" (e.g. Miss Rarotonga, Miss Switzerland, Miss South America). Similarly, at the organizing stage, can didates provide their age, vital statistics, occupation, a n d personal aspirations, which one of the organizers enters o n bio-data sheets. Clearly, what participants in the pageant aim for in this emulation of international pageant practices is the 7
BESNIER CROSSING GENDERS, MIXING LANGUAGES
101
appearance of a glamor whose reference reaches b e y o n d the confines of the local context. T h e extent to which participants are aware of the inspiration for these practices depends o n their relative worldliness. While some leiti involved in the p r o g r a m m i n g of the show h a v e h a d the opportunity to watch televised international pageants, others m u s t rely o n second-hand reports of such events, what they can infer about t h e m from watching the mainstream Miss Heilala pageant, a n d their imagination. I n addition to bearing the n a m e s of the countries they represent, Miss Galaxy contestants go b y female-sounding stage names of their own choosing, and which they often use in everyday contexts. These stage n a m e s are often coinages that bear linguistic similarity to the person's original Tongan n a m e (e.g. "Suzie" from Sosefo), a n d are either English n a m e s (e.g. Priscilla Pressland) or exotic-sounding names with n o c o n n o t a t i o n other t h a n their generic foreignness (e.g. Aisa De Lorenzo, Aodushi Kiroshoto), b u t never Tongan names. T h e extra-local flavor also pervades the stage decorations (in 1997, flower arrangements a n d rather unfortunate bouquets of phallic-shaped multicolored balloons), the background music (for the opening, a m e d l e y of triumphalist classical themes such as the William Tell Overture), a n d the singing a n d dancing. W h e n events are explicitly designed to add local color (e.g. a tau 'olunga performance, a popular Tongan tune sung b y one of the organizers), they are bracketed entertainment routines designed to fill the time while contestants are getting changed back-stage, and often look like strained token gestures. W h e n a contestant does decide to perform a Tongan dance for a j u d g e d event, it is generally a spoof. Perhaps the most powerful index of extra-locality is language use. Throughout the pageant, the d o m i n a n t language is English. W h e n contestants first present themselves, for instance, they d o so in English: 8
Aisa: {(walks up to the mike)) Good evening ladies and gentlemen. My name is Aisa De Lorenzo, I'm eighteen years of age, and I represent, ((pauses, raises arms triumphantly)) BLUE PACIFIC TAXIS! ((walks down catwalk)) (1997: Sony: 2 1:07:36-1:08:20) Each contestant will have memorized and rehearsed her lines prior to the pageant, and will take utmost care to p r o n o u n c e t h e m correcdy and loudly. This does not prevent occasional slip-ups, which the audience will immediately ridicule boisterously. T h e important point is that, for most contestants, speaking English before a large a n d distinguished audience of elite Tongans (many of w h o m are bilingual) and foreigners represents a serious challenge: m a n y leiS, particularly pageant contestants, speak minimal English, as poverty and marginality h a v e barred t h e m from opportunities to learn the language. A significant n u m b e r have not traveled overseas, a n d those w h o h a v e resided in industrial countries have not d o n e so u n d e r privileged conditions. By centralizing the English language a n d its associations, leiti position themselves o n the side of prestige a n d worldliness, a n d in opposition to the use of Tongan and its localized connotations. But their sociolinguistic behavior, b o t h in and out of the pageant, adds further complexity. I n d e e d , despite the obvious
102
DISCOURSE STUDIES
difficulties that leiti experience in speaking English during the pageant, m a n y Tongans expect t h e m to speak English m o r e readily o n a day-to-day basis than non-transgendered Tongan m e n , for a n u m b e r of reasons. First, Tongans generally see fakaleiti as self-assured a n d brash creatures that k n o w n o shame (ta'ema). While in actuality a significant percentage of leiti are self-effacing, the d e m e a n o r of other leiti underscores this stereotype. O n e illustration of this shamelessness is their very participation in a pageant that constitutes the p r i m e locus of the formation a n d reinforcement of popular stereotypes of fakaleiti: contestants' behavior in the pageant can b e moderately outrageous a n d is certainly viewed as exhibitionistic. 9
Second, stereotypes of leiti view t h e m as oriented toward modernity, the West, transnationalism, a n d social change. O n c e again, the extent to which this stereotype reflects reality varies across individuals, b u t h e r e as well it is certainly founded o n undeniable (if partial) evidence. T h e uncompromisingly extra-local design of the pageant falls right in line with this expectation, b o t h establishing a n d confirming the stereotypes held b y audience m e m b e r s . Viewed in this light, the prominence of English in both public a n d private contexts is hardly surprising, since English is the language of extra-locality. Finally, Tongans tend to view the use of English as having feminine under tones: as in m a n y other societies in which a language of modernity competes with a code of traditionalism (e.g. Gal 1979), the former is associated with women's aspirations for u p w a r d m o b i l i t y a n d e m a n c i p a t i o n from t h e strictures of traditionalism (compare Meyerhoff). W h e n questioned o n the matter, most Tongan m e n a n d w o m e n will state that w o m e n speak better English overall than m e n , and that this is due to the fact that girls study harder in school and that w o m e n are talkative "by nature." These famihar-sounaing assertions b e a r witness to the fact that the gendering of language use is tacit a n d e m b o d i e d in practice, rather t h a n explicit and g r o u n d e d in overt consciousness. As a result of this gendering, m e n w h o speak "too m u c h " English do so ai the risk of compromising their masculinity in the eyes of society at large. This concerns fakaleiti, w h o willingly go to great lengths to dislocate themselves from their masculine attributes. Interestingly, it also concerns overseas-born Tongans: their awkwardness in performing Tongan maleness, including speaking Tongan as a preferred language, frequently brands t h e m as fakaleiti-Hke, regardless o! whether they present any identifiable sign of effeminacy in their comportment T h e use of English thus has m a n y associations in addition to extra-locality: i; potentially indexes deficient Tonganness, deficient masculinity, femininity, and transgendered identity, traits which m a y or m a y not overlap b u t which are all readily equated to one another. T h u s failure to perform Tonganness can easily b e c o m e a sign of imperfect masculinity a n d vice versa, unless it is mollified by convincing mitigating factors, such as elite status or wealth. 10
Patterns of language use in the Miss Galaxy pageant, as well as the overall non-local ambience to which they contribute, are n o t without irony. As discussed earlier, most contestants live in relative poverty. I n tune with their under-privileged status, m a n y leiti speak English poorly. Sustaining the level of extra-locality expected of t h e m is therefore difficult for m a n y contestants, w h o switch to Tongan
BESNIER CROSSING GENDERS, MIXING LANGUAGES
103
once they h a v e delivered simple m e m o r i z e d lines. But English still r e m a i n s dominant in the pageant: it is the language that the e m c e e uses to address the audience and, w h e n h e addresses the contestants, h e does so first in English a n d then provides a Tongan translation, usually sotto voce. These communicative prac tices maintain English in the foreground, at the expense of Tongan. T h e difficulty contestants h a v e in maintaining English as their working language during the pageant places t h e m in an awkward position. For example, in the interview event, contestants are given the choice of answering in English or Tongan, and most choose the latter. I n 1997, o n e contestant chose English, a n d the audience initially reacted with a loud m u r m u r of temporary admiration for her courage. H o w e v e r , it took little time for h e r to stumble, as she searched for an English w o r d while waving h e r h a n d campily, while the audience, satisfied with the expected proof of the fraudulence of h e r claim to cosmopolitanism, began hooting a n d ridiculing, forcing her to abort h e r brave attempt: 11
Emcee:
Masha:
What would you say about being a hairstylist, or- being- a working- whatwhat does it mean, like, to be working at Joy's Hair Styles? {(sotto voce, summaries the question in Tongan)) Koehaeme'a 'oku he fai 'i he hair salon? ({takes cordless mike)) Well thank you very much, ((audience laughs, then shouts with admiration and encouragement)) If you want your hair to be curled, ((beckons with her hand)) come over, ((audience explodes in laughter and whooping, Masha laughs and then becomes serious and requests silence with the hand)) Uh, I like it very much, and uh-I enjoy working there, with uhmm- ((pauses, word-searches, waves her hand, audience explodes in laughter, drowning the remainder of the answer)) blowers, ((unable to finish, mouths)) (thank you). ((hands mike back and returns to her position)) (1997: Sony: 4 0:02:45-0:03:55)
Contestants thus are caught b e t w e e n a rock a n d a h a r d place: if they answer in English a n d m a k e mistakes, they will b e laughed at, and if they answer i n
Photograph 2: Masha Entura searches for the English word she needs to answer her infeniew question.
DISCOURSE STUDIES'
104
Tongan, this v e r y fact will b e ridiculed as evidence that they are unable to carry through the artifice of extra-locality to its logical end. T h e ridicule that greets the choice of Tongan is congruent with m a n y other aspects of m a i n s t r e a m Tongans' attitudes t o w a r d fakaleiti, b o t h at the p a g e a n t a n d in day-to-day interactions. Mainstream Tongans indeed consider fakaleiti identity as essentially bogus: h e r e are these m e n pretending to b e w o m e n , a n d n o t just any w o m e n b u t cosmopolitan sophisticates, a n d yet they cannot even maintain their end of a simple conversation in English. At the pageant, it is n o t u n c o m m o n for drunken m e n or w o m e n to try to rip contestants' outfits a n d expose t h e m as what they "really" are, n a m e l y persons with m a l e physiologies. N o t h i n g generates greater hilarity than contestants losing their b r a in the middle of a performance. I n dayto-day interactions between fakaleiti a n d mainstream Tongans, the latter often express m o c k a n n o y a n c e at the "fraudulence" of leiti self-presentation and identity, while leiti argue back with "proofs" that they are "real w o m e n . " However, like all ideological linkages that disadvantage some a n d benefit others, the linkages I h a v e described are n o t i m m u n e to contestation o n the part of those w h o m they marginalize. This was powerfully illustrated b y a minor h u m o r o u s incident in the 1997 pageant, w h e n one of the contestants, the quick witted 'Amini or L a d y A m y l a n d , sponsored b y j o e y ' s U n i s e x Hair Salon, turned the tables o n the audience during the interview event (and, p e r h a p s , o n society at large, even if only for a fleeting moment). Before she has a chance to answer the emcee's question, L a d y A m y l a n d is heckled b y a d r u n k e n leiti in the audience, w h o urges h e r to answer h e r interview question in English {faka-Palangi). The heckling draws some laughter, since everyone knows that L a d y Amyland's English is p o o r a n d that she would m a k e a fool of herself if she tried. But 'Amini's repartee wins the prize: Emcee:
Miss Joey's Unisex Hair Salon! What do you have to say to promote Joey's Unisex Hair Salon? ((lowers voice, translating into Tongan)) Koehae me'a 'oku kefai ke promote ai 'a e- ((rolls eyes, searches for Tongan word)) fakalakalaka ai 'aJoey's Unisex Hair Salon. 'Ahi: ((heckling from audience)) Faka-Palangi, 'Amini! Audience: ((laughter)) 'Amini: Sorry excuse me, I'm a Tongan () {(rest of answer drowned by deafening laughter, vigorous applause, cat-calls) (1997: Sony: 4 0:05:42-0:06:26, see Photograph 3) 'Amini answers the heckler b y reaffirming h e r Tongan identity a n d therefore her duty a n d privilege to answer the question in Tongan, an u n e x p e c t e d move which the audience (and a n y Tongan viewer of the video recording) found extremely h u m o r o u s , because the claim is e m b e d d e d in a context in which everything is d o n e to foreground non-locality. W h a t L a d y A m y l a n d is doing h e r e is part of a wider tacit project o n the part of at least s o m e contestants to take greater charge of the pageant a n d its effect o n the audience. This projecr consists in stripping the audience (and society at large) of its privilege to ridicule contestants, a n d to take control of the b o u n d a r y between h u m o r and seriousness 12
BESNIER CROSSING GENDERS, MIXING LANGUAGES
105
Photograph 3 : Lady Amyland savors the effect of her quick-minded repartee to a heckler. Sut the project goes further, and its m e a n i n g b e c o m e s clear w h e n viewed in light of the previous analysis. N o t e that L a d y A m y l a n d asserts h e r claim to Tongan identity n o t in Tongan, but in English; the covert message is that o n e can assert one's Tonganness while controlling the tools with which o n e does so, and while using tools that are not part of the sanctioned repertoire. I n addition, the preface of her repartee ("Sorry excuse me") is a n inside j o k e which non-leiti audience m e m b e r s are unlikely to m a k e sense of, a reference to another kit»'s awkward attempt, a few years earlier, to speak English to a prospective Palangi date. T h e overall effect of L a d y A m y l a n d ' s r e p a r t e e contests the p o w e r of dominant forces to dictate what counts as markers of locality a n d what does not; asserts that the claim to b e part of the "galaxy" does n o t necessarily d e n y one's local identity; a n d proclaims that b e i n g a leiti does m e a n giving u p one's place in Tongan society. 13
7 C o n c l u s i o n : T h e Linguistic Constructions of T o n g a n T r a n s g e n d e r i s m This chapter has investigated the linguistic behavior a n d ideologies oifakaleiU and mainstream Tongans, and the relationship of these various behaviors and ideologies in the constructions of identities, stereotypes, a n d life trajectories. I argued that, in a society that remains essentially monolingual, the p r e s e n c e of English is strongly felt, being associated with contexts w h e r e cosmopolitanism, modernity, and capitalism are foregrounded, elements of increasing importance to the very nature of Tongan society. A m o n g the subgroups of Tongan society who are enthusiastic users of English, fakaleiti figure prominently, even though
106
DISCOURSE STUDIES
most d o n o t h a v e access to the kind of resources which might justify, in the eyes of greater Tongan society, the implicit claim to prestige status that the choice of English entails: wealth, status in the traditional hierarchy, cosmopolitanism, and grammatical fluency in English. Fakaleiti code-switch for complex and diverse reasons, a n d in this respect they do not differ from code-switchers in all other societies of the world. However, one of the m o s t salient, although largely unarticulated, motivations for codeswitching that this chapter has explored is the fact that the use of English represents for m a n y fakaleiti a symbolic escape hatch out of social marginality (compare Meyerhoff, o n w o m e n o n M a l o , Vanuatu). T h e claims e m b e d d e d in their use of English a n d their code-switching serve as a n idiom of resistance against the symbolic a n d material oppression that they experience as b o t h transg e n d e r e d persons a n d p o o r Tongans. However, this strategy is not without risk, l i k e all resistant action, these claims can b e turned a r o u n d a n d used against t h e m to further marginalize them. Leitfs language choices place t h e m at risk of b e i n g perceived b y non-transgendered Tongans as alienating themselves from a local context that offers b o t h unpleasant b u t also potentially rewarding symbols a n d resources for everyone. Being generally poor, leiti are not in a good position to define for the rest of society what counts as "local," a n d the perception that they are alienating themselves from a pre-defined localness over which they h a v e little control is potentially disadvantageous. This chapter has attempted to explore the intersection of gender, modernity, and locality b y focusing o n the differences a n d conflicts in the subjectivities of m e m b e r s of one society. Reading d o m i n a n t characterizations of modernity from sociology a n d cultural studies (e.g. Featherstone, Lash, a n d R o b e r t s o n 1995; J a m e s o n a n d Miyoshi 1998), we are led to expect that Tongans would experience tokens of modernity a n d globalization, for example, i n a k i n d of Durkheimian (solidarity-enhancing) unison. W h a t l h a v e shown h e r e is that they n o t only differ from o n e another in the w a y they experience these tokens a n d i n w h a t they do with t h e m , b u t they also actively challenge each other's experiences of these tokens. Furthermore, t h e y enlist these experiences to argue over the meaning of seemingly highly localized categories a n d dynamics, including gender. m t m s c ^ a p t e r , m l i n e w i t h a s u b s t a n t i a l b o d y of recent research, I have explored the potentially heterogeneous nature of gender as a social category, and have sought to unravel this heterogeneity in terms of the varied positions that members of the "same" gender can take vis-a-vis modernity and localness. I have also soughi to distinguish b e t w e e n different m e a n i n g s of m o d e r n i t y , from material to ideational manifestations of it. Finally, I h a v e investigated the complex interplay of modernity with locality. T h e chapter has explored the role of language in creating and indexing these social and cultural dynamics. T h e discourse- a n d ethnography based analysis I h a v e developed h e r e illustrates the complex role that categories other t h a n g e n d e r play in defining gender. It also shows that the meaning and valuation of such categories as gender, modernity, a n d localness are objects of mnflict a n d contradiction, b o t h across subgroups of society a n d across contex; ~
BESNIER CROSSING GENDERS, MIXING LANGUAGES
107
Acknowledgments I conducted fieldwork in 1994,1995,1997, and 1999-2000, grounding myself on a general understanding of Tongan society acquired during residence in a Tongan village in 1978-80. I thank many Tongan informants and friends for their help and devotion, and Janet Holmes and Miriam Meyerhoff for useful comments on an earlier version of this chapter. Financial support for fieldwork from the following sources is gratefully acknowledged: Marsden Fund of the Royal Society of New Zealand, Wenner Gren Foundation for Anthropological Research, Yale University Social Science Faculty Research Fund, and Victoria University of Wellington Faculty of Humanities and Social Sciences Research Fund. Portions of this chapter are reprinted from Besnier (2002) by permission of the American Anthropological Association, which is gratefully acknowledged.
Notes 1. The fact that, in Tonga as in many other parts of the Pacific, English is the language of choice when one is drunk lends further support to this analysis (compare Harvey 1991 on the role of Spanish in Quechua drunken conversation). 2. Tongan has a notable system of honorifics ("speech levels"), centered principally on the lexicon: certain words are used solely when addressing or speaking of members of the nobility or the royal family other than the sovereign, and others when addressing or speaking of the sovereign or God. These register variations are the subject of ideo logical elaboration, but in practice they concern a very restricted range of linguistic structures and their use is very flexible (Philips 1991). They are of no significant relevance to the materials presented here. 3. One interviewee, who belongs to a small Charismatic Christian sect, expliciuy linked her "liberated" stance to the fact that she had accepted Jesus into her heart, which enables her to ignore tradition-based gossip and shaming. Because they reject the (often oppressive) structuring of mainstream Christian denominations, Charismatic Christians place themselves on the margin of a society where church-mediated and church-directed exchange is so determinative of social life. This is also true, to a lesser extent, of Mormons (Gordon 1990) and other people who have somehow extricated themselves from the duties of reciprocity and exchange, often at a cost to their social standing. '<•- There is a substantial and evergrowing corpus of borrowings from English in the contemporary Tongan lexicon, many of which have been phonologically nativized (Schiitz 1970). Some words were borrowed early in the history of contact (e.g. taimi "time", siasi "church"), and have lost all connotations of foreignness. More recent borrowings, while highly integrated in everyday linguistic usage, continue to subtly index the connotations of English as a medium of communication, as evidenced, for example, by cases where both a borrowing and a word of Polynesian origin have roughly the same meaning (e.g. kiti and leka "kid"). The borrowing of "leiS"prob ably dates back to the early decades of the twentieth century (Futa Helu, personal communication). In the orthography in general use for Tongan, an apostrophe represents a glottal stop, a macron superscripted to a vowel represents gemination, and an acute accent above a word-final vowel indicates that stress shifts from the penultimate to the accented vowel to denote the definiteness and specificity of the noun phrase ending with the
t
DISCOURSE STUDIES
108
6. Don Kulick extends his criticism to analyses that focus primarily on talk produced in other "on-stage" circumstances, for example, for media dissemination, or during performances of various kinds. The point is well taken, and falls in line with a long tradition in linguistic anthropology of emphasizing the importance of seeking an understanding of social dynamics by focusing on day-to-day interaction. However, one should also not forget that "public" discourse may also act as an important medium through which identities are created and negotiated, representations constructed and challenged. 7. Some of the information provided is fake or unrealistic, while other details are designed to be humorous. For example, contestants regularly claim "high-status" feminine occupations such as "nurse" and "public relations" [sic) to add glamor to their profile, as well as "future plans" to be "computer operator," "flying attendant" (sic), and "to be a good wife." The same practice of emulating international beauty contests is found in the pageants that transgendered persons stage in Jolo, Southern Philippines (Johnson 1997) and in urban South Africa (Reid 1999), both of which exhibit fascinating similarities to the Tongan material. 8. In the following discussion, I have not attempted to hide the identity of those con cerned since my analysis is based on a public event. Extracts are identified by year of recording and video reference number. 9. A Tongan businessman told me that he had employed a.fakaleitt~ to sell his products door-to-door precisely because fakaleiti worry littie about shame, in addition to being gregarious and talkative. These traits are thus not necessarily seen as negative assets. 10. Many of the symbolic associations I describe here of course echo patterns found in many other societies. One is reminded of Willis's (1977) celebrated analysis of workingclass masculinity among adolescents in English schools, Bourdieu's (1985) analysis of social class and "refinement" in France, particularly as it relates to gender, and Ortner's (1991) study of social class and gender in New Jersey, among many other relevant examples. 11. English, as with other tokens of modernity and cosmopolitanism, also occupies a prominent role in many other public events in Tonga, including the Miss Heilala beauty pageant for "real" women. However, in other events, these tokens are com monly on a par with Tongan and tokens of "Tonganness." In the Miss Heilala pageant, for example, the contestants' ability to perform tokens of Tonganness, including their linguistic skills, are scrutinized very closely. This scrutiny frequendy places overseas-born contestants at a disadvantage, as discussed in Teilhet-Fisk (1996) and Besnier (2002). 12. The humor already began with the heckle itself, which is uttered in Tongan, despite the fact it urges the contestant to speak English, and which refers to the contestant by his everyday name, rather than her transgendered name. 13. I do not wish to imply that Lady Amyland's act of resistance was the result of a care' fully engineered strategy on her part. For one thing, she was probably drunk, as many contestants are. However, we know from Scott (1985, 1990) that everyday acts of resistance need not be the outcome of calculated designs.
References •*
I
Appadurai, Arjun 1996: Modernity at Large: Cultural Dimensions ofGlobalizfditm; Minneapolis: T Tniuorsitv o f Minnesota Press.
BESNIER CROSSING GENDERS, MIXING LANGUAGES
109
Besnier, Niko 1994: Polynesian gender liminality through time and space. In Gilbert Herdt (ed.) Third Sex, Third Gender: Beyond Sexual Dimorphism in Culture and History. New York: Zone, pp. 285-328, 554-66. .Besnier, Niko 1997: Sluts and superwomen: The politics of gender liminality in urban Tonga. Ethnos 62: 5-31. Besnier, Niko 2002: Transgenderism, locality, and the Miss Galaxy beauty pageant in Tonga. American Ethnologist 29: 534-67. Bourdieu, Pierre 1985: Distinction: A Social Critique ofthejudgement ofTaste. Translated by Richard Nice. Cambridge, MA: Harvard University Press. Campbell, I. C. 1992: Island Kingdom: Tonga Ancient and Modem. Christchurch: Canterbury University Press. Comaroff, Jean and Comaroff, John 1993: Introduction. In Jean Comaroff and John Comaroff (eds) Modernity and its Malcontents: Ritual and Power in Postcolonial Africa. Chicago: University of Chicago Press, pp. xi-xxxvii. Daniel, E. Valentine 1996: Charred Lullabies: Chapters in an Ethnography of Violence. Princeton Studies in Culture/Power/History. Princeton, NJ: Princeton University Press, di Leonardo, Micaela 1991: Introduction: Gender, culture, and political economy: Feminist anthropology in historical perspective. In Micaela di Leonardo (ed.) Gender at the Crossroads of Knowledge: Feminist Anthropology in the Postmodern Era. Berkeley, CA: University of California Press, pp. 1-48. Featherstone, Mike, Lash, Scott, and Robertson, Roland (eds) 1995: Global Modernities. London: Sage. Gal, Susan 1979: Language Shift: Social Determinants of Linguistic Change in Bilingual Austria. New York: Academic Press. Gaudio, Rudolph P. 1997: Not talking straight in Hausa. In Anna Livia and Kira Hall (eds) Queerly Phrased: Language, Gender, and Sexuality. New York: Oxford University Press, pp. 416-29. Gordon, Tamar 1990: Inventing the Mormon Tongan family. Injohn Barker (ed.) Christianity in Oceania: Ethnographic Perspectives. Association for Social Anthropology in Oceania Monographs, 12. Lanham, MD: University Press of America, pp. 197-219. Gupta, Akhil and Ferguson, James (eds) 1997: Anthropological Locations: Boundaries and Grounds ofa Field Science. Berkeley, CA: University of California Press. Hall, Kira and O'Donovan, Veronica 1996: Shifting gender positions among Hindispeaking Hijras. In Victoria Bergvall, Janet M. Bing, and Alice F. Freed (eds) Rethinking Language and Gender Research: Theory and Practice. New York: Longman, pp. 228-66. Harvey, Penelope 1991: Drunken speech and the construction of meaning: Bilingual competence in the Southern Peruvian Andes. Language in Society 20: 1-36. Jameson, Fredric and Miyoshi, Masao (eds) 1998: The Culture of Globalization. Durham, NC: Duke University Press. Johnson, Mark 1997: Beauty and Power: Transgendering and Cultural Transformation in the Southern Philippines. Oxford: Berg. Kulick, Don 1999: Transgender and language: A review of the literature and suggestions for the future. GLQ5: 605-22. Lovell, Nadia (ed.) 1998: Locality and Belonging. European Association of Social Anthropologists Series. London: Routledge. Malkki, Iiisa 1995: Purity andExile: Violence, Memory, and National Cosmology amongHutu Refugees in Tanzania. Chicago: University of Chicago Press. Marcus, George E. 1995: Ethnography in/of the world system: The emergence of multisited ethnography. Annual Review of Anthropology 24: 95-117
no
DISCOURSE STUDIES
Morton, Helen 1996: Becoming Tongan: An Ethnography of'Childhood. Honolulu: University of Hawaii Press. Morton, Helen 1998: Creating their own culture: Diasporic Tongans. The Contemporary Pacific 10: 1-30. Ogawa, Naoko and Shibamoto Smith, Janet 1997: The gendering of the gay male sex class in Japan: A case study based on Rosen No SobyS. In Anna Livia and Kira Hall (eds) Queerly Phrased: Language, Gender, and Sexuality. New York: Oxford University Press, pp. 402-15. Ortner, Sherry B. 1991: Reading America: Preliminary notes on class and culture. In Richard G. Fox (ed.) Recapturing Anthropology: Writing in the Present. Santa Fe, NM: School of American Research Press, pp. 163-89. Philips, Susan U. 1991: Tongan speech levels: Practice and talk about practice in the cultural construction of social hierarchy. In Robert Blust (ed.) Currents in Pacific Linguistics: Papers on Austronesian Languages in Honour ofGeorge Grace. Canberra: Pacific Linguistics C-117, pp. 369-82. Philips, Susan U. 2000: Constructing a Tongan nation-state through language ideology in the courtroom. In Paul V. Kroskrity (ed.) Regimes of Language: Ideologies, Polities, and Identities. School of American Research advanced seminar series. Santa Fe, NM: School of American Research Press, pp. 229-57. Reid, Graeme 1999: Above the Skyline: Integrating African, Christian and Gay or Lesbian Identities in a South African Church Community. MA thesis, Department of Anthro pology, University of the Witswatersrand. Schtttz, Albert J. 1970: Phonological patterning of English loan words in Tongan. In S. A. Wurm and D. C. Laycock (eds) Pacific Linguistic Studies in Honour of Arthur Capell. Canberra: Pacific Linguistics C-13, pp. 409-28. Scott,James C. 1985: Weapons of the Weak: Everyday Forms of Peasant Resistance. New Haven, CT: Yale University Press. Scott, James C. 1990: Domination and the Art of Resistance: Hidden Transcripts. New Haven, CT: Yale University Press. Small, Cathy A. 1997: Voyages: From Tongan Village to American Suburbs. Ithaca, NY: Cornell University Press. Spencer, Jonathan 1996: Modernism, modernity and modernization. In Alan Barnard and Jonathan Spencer (eds) Encyclopedia ofSocial and Cultural Anthropology. London: Roufledge, pp. 376-9. Stroud, Christopher 1992: The problem of intention and meaning in code-switching. Text 12: 127-55. Teilhet-Fiskjehanne 1996: The Miss Heilala beauty pageant: Where beauty is more than skin deep. In Colleen B. Cohen, Richard Wilk, and Beverly Stoeltje (eds) Beauty Queens on the Global Stage: Gender, Contests, and Power. London: Routiedge, pp. 185-202. Willis, Paul 1977: Learning to Labour: How Working Class Kids Get Working Class Jobs. Westmead, England: Saxon House.
I
The tower and Politics of Genre Vijay K. Bhatia
Introduction
A
ll discourse forms, especially those used in institutionalized contexts, are socially negotiated. At the very h e a r t of most frameworks for the analysis of discourse, especially as genre, is the belief that there is nothing like a universal form of discourse for structuring knowledge. T h e r e can only b e a 'consensus or a n agreement' (Bruffee, 1986: 777) a m o n g the m e m b e r s of specific disciplinary communities to express their concerns in specific discursive forms. Most institutionalized forms of discourse, therefore, are socially constructed, inter preted a n d used. Goodrich (1987) explains this institutionalization of discoursal practices in terms of 'social authorship' as against the m o r e familiar subjective authorship. The right to a discourse is organized a n d restricted b y a wide variety of means, to particular roles, statuses, professions a n d so on. Similarly the institutionalisation of discourse is limited in t e r m s of its legitimate appropriation, a n d the restrictive situations of its reception - church, court, school, hustings a n d so on. Foucault (1981) also sees social authorship of discourse in terms of two interrelated aspects. T h e first one, according to him, is: W h o is speaking? W h o m a m o n g the totality of speaking individuals, is accorded the right to use this sort of language? W h o is qualified to do so? Who derives from it his own special quality, his prestige, and from w h o m , in return, does h e receive if n o t the assurance, at least the presumption
Source: World Englishes \o\. 16, no. 3, 1997, pp. 359-371.
DISCOURSE STUDIES
112
that w h a t h e says is true? W h a t is the status of the individual w h o alone - h a v e the right, sanctioned b y law or tradition, juridically defined or spontaneously accepted, to proffer such a discourse? T h e second, h e explains in terms of institutional sites from which the authorized speaker m a k e s his discourse a n d from which the discourse derives its 'legitimate source a n d point of application.' Pecheux (1982) also stresses the importance of institutional forms in discourse w h e n h e says w h a t can b e a n d should b e said - in the form of a speech, a sermon, a pamphlet, a report, a p r o g r a m m e , etc. Like other forms of discourse, genres are also socially constructed a n d are even m o r e intimately controlled b y social practices. Genres are the media through which m e m b e r s of professional or academic communities communicate with each other. T h e y are, as Berkenkotter a n d H u c k i n (1995) point out, . . . intimately linked to a discipline's methodology, a n d they package information in ways that conform to a discipline's n o r m s , values a n d ideology. Myers (1995: 5) also points out, Disciplines are like cultures in that their m e m b e r s h a v e shared, taken for granted beliefs; these beliefs can b e mutually incomprehensible b e t w e e n cultures; these beliefs are encoded in a language; they are e m b o d i e d in practices; n e w m e m b e r s are b r o u g h t into culture through rituals. T h e consensus is arrived at a n d negotiated through professional conversations a n d practices amongst the informed a n d practising m e m b e r s of a professional community. Interactions a n d conversations enable consensus, o n the o n e hand, a n d h a v e a regulatory or limiting effect o n the other, as to what should or should not b e admitted into a community's b o d y of knowledge. Genres, in other words, are socially authorized through conventions, which, in turn, are e m b e d d e d in the discursive practices of m e m b e r s of specific disciplinary cultures. These discursive practices, to a large extent, reflect n o t only conventions used b y specific disci plinary communities, but also social conventions, including social changes, social institutions a n d social knowledge, all of which, in a way, could b e seen as significantiy contributing to what in genre theory is regarded as 'genre knowledge.' Genres, in whatever m a n n e r o n e m a y identify them, whether as a typification of rhetorical action (as in Miller (1984), a n d m o r e recently, Berkenkotter a n d Huckin (1995)), as a staged, goal oriented social process (as in Martin (1993)), o r as shared communicative purposes (as in Swales (1990) a n d Bhatia (1993)), are products of an understanding or a prior knowledge of generic conventions. These generic con ventions are responsible for regulating generic constructs, giving t h e m internal
BHATTA THE POWER AND POLITICS OF GENRE
U3
ordering, which in discourse literature, for various reasons, has b e e n given differ ent n a m e s , s o m e of which include discourse structuring, staging of discourse a n d generic structure potential (GSP). Complexity of Generic Forms Although generic forms are products of conventional knowledge e m b e d d e d in disciplinary cultures, they are d y n a m i c constructs. Typical realizations of these institutionalized forms are often characterized b y their generic integrity, o n the one hand, a n d their propensity for innovation, o n the other (see Bhatia, 1993, 1995). T h e s e two aspects of g e n r e m a y a p p e a r to b e somewhat contradictory at first, but as w e shall soon discover, these two characteristics are c o m p l e m e n t a r y to each other. I n fact, it will n o t b e inaccurate to suggest that o n e is, in a way, a n essential prerequisite to the other. Generic integrity is the product of the conven tional features of a specific generic construct. Although these conventions are e m b e d d e d in the rhetorical context, they often constrain the use of linguistic resources (lexico-grammatical as well as discoursal), and are frequentiy invoked to arrive at a reasonable interpretation of the genre or even determine the choice of the genre to suit a particular context. Within generic boundaries, experienced users of genre often manage to exercise considerable freedom to manipulate generic conventions to respond to novel situations, to m i x what Bhatia (1993) calls 'private intentions' with socially recognized communicative purposes, and even to produce new forms of discourse. Therefore the tension between conformity a n d creativity, so often m a d e a n issue of in applied discourse studies, is n o t necessarily real. A s Dubrow (1982:39) points out, 'a concern for generic traditions, far from preclud ing originality, often helps to produce it.' Similarly, Fowler (1982: 31) points out: Far from inhibiting the author, genres are a positive support. T h e y offer r o o m , as o n e might say, for h i m to write in - a habitation of mediated definiteness, a proportioned mental space; a literary matrix b y to order his experience during composition . . . T h e writer is invited to match experience and form in a specific yet u n d e t e r m i n e d way. Accepting the invitation does n o t solve his problems of expression . . . But it gives h i m access to formal ideas as to h o w a variety of constituents might suitably b e c o m b i n e d . G e n r e also offers a challenge b y provoking a free spirit to transcend the limitations of previous examples. In fact, a subde exploitation of a certain aspect of generic construct is always seen as tactically superior a n d effective. It is almost like the advertiser's exploit ation of the cliche the shape of things to come in the following opening headline of an advertisement for a car. T h e shape of things to c o m e : Mitsubishi C o r d i a Or, the use of the famous statement about the British colonial e m p i r e i n t h e Lufthansa advertisement, The sun never sets on Lufthansa tenitam ™.j~u-~*-" • clnrrQn
DISCOURSE STUDIES
114
idea of waste of energy is lost unless it is associated with 'Don't b e foolish.' T h e whole point about such associations is that they communicate best in the context of what is already familiar. In such contexts, words o n their own carry n o meanings; it is the experience which gives t h e m the desired effect. Therefore, if one is not familiar with the original, the value of the novel expression is u n d e r m i n e d . Just as the advertiser m a k e s use of the well-known a n d the familiar in existing know ledge, a clever genre writer makes use of what is conventionally available to a discourse c o m m u n i t y to further his or h e r o w n subtle ends. T h e innovation, the creativity or the exploitation becomes effective only in the context of the already available a n d familiar. T h e m a i n focus of this p a p e r is on these two interrelated aspects of genre theory, i.e., the constraints o n generic construction, a preknowledge of which gives p o w e r to insiders in specific discourse communities, a n d the exploitation of this p o w e r b y experienced a n d expert m e m b e r s of such disciplinary cultures to a c h i e v e t h e i r ' p r i v a t e i n t e n t i o n s ' w i t h i n 'socially recognized communicative purposes.' i Organisational Preferences and Generic Controls T h e other interesting area of generic variation, although within a restricted range, one finds in organizational preferences. I n the case of academic publications, we often come across what we c o m m o n l y refer to as housestyles. Although every single journal claims to have its own style sheet, most of t h e m can b e characterized m o r e b y their overlap rather than variation. Similarly, in the case of newspaper genres, especially the n e w s reports and the editorials, w e find an unmistakable 'generic identity' (Bhatia, 1993) in almost all of the exploits of these genres from various newspapers, although all of them have their own preferences in terms of style, stance a n d substance. Some m a y be m o r e objective, while others m o r e interpretative; some m o r e socially responsible, while others more sensational. I n spite of all these differences, most of them display c o m m o n characteristics in terms of their use of generic resources, in terms of their structure, interpretation a n d communication of intentions. These somewhat different orientations to the events of the day do n o t m a k e their stories very dif ferent in terms of their generic form. Even in the case of business communities, w e often find different organizations displaying their unique identities through their organizational preferences in the matters of their choice of generic forms, but the b r o a d range of genres they tend to exploit to further their organizational objectives show remarkable similar ities rather than differences. All these areas of generic use indicate that although their preferred generic forms show a subtle degree of variation for w h a t could b e seen as 'tactical advantage,' they never disregard some of the basic features of individual generic constructs, which give these genres their essential identities.
T h e P o w e r of G e n r e T h e r e is n o better illustration of the saying 'knowledge is p o w e r ' t h a n the one in the case of generic power. Power to use, interpret, exploit a n d innovate novel
BHATIA THE POWER AND POLITICS OF GENRE
115
generic forms is the function of generic knowledge which is accessible only to the m e m b e r s of disciplinary communities. H o w d o these disciplinary communities maintain what we h a v e called generic integrity in their discursive practices? Let us look at this p h e n o m e n o n b y looking at the academic community. Maintaining Generic Integrity: Editorial Intervention In some forms of academic discourse, especially the research articles, o n e can see generally two kinds of m e c h a n i s m in place to ensure generic integrity: the peer review process, and editorial intervention. Both these mechanisms, though operating at different levels, are actively invoked to ensure that all accounts of new knowledge conform to the standards of institutionalized behaviour that is expected b y a c o m m u n i t y of established peers in a specific discipline. Although individual j u d g e m e n t s can vary within the m e m b e r s h i p of specific disciplinary communities, a high degree of consensus is often ensured b y selecting like-minded scholars from within well-defined disciplinary boundaries. For example, if w e survey a few journals which regularly publish articles o n discourse analysis, we will find that although all of t h e m publish articles on various aspects of discourse, they have a very different set of reviewers to certify accounts of knowledge claims for inclusion in the respective journals. If o n e encounters n a m e s like Cazden, Geertz, Goffman, Gumperz, H y m e s , Milroy, Saville-Troike, Scollon, Tannen, and Z i m m e r m a n o n the editorial committee of a journal, one could safely guess that they will b e unlikely to accept articles outside a sociolinguistic orientation o n discourse. Articles o n other aspects of discourse are m o r e likely to b e discouraged and even rejected. If, o n the other hand, one finds names like Ackerman, Bazerman, Berkenkotter, Comprone, Doheny-Farina, Huckin, Linda Flower, Miller, or Odell, one would expect t h e m to welcome papers with a strong rhetorical orientation. Similarly, if one finds names like Carter, Christie, Halliday, Hasan, Kress, Martin and Rothery, one will come to the inevitable conclusion that the journal will favour a m o r e systemic orientation to discourse. After p e e r review, the second m o s t important intervention comes from the editors, w h o enjoy all the p o w e r o n e can imagine to maintain the identity a n d integrity of the research article genre. Berkenkotter a n d Huckin (1995) document an in-depth a n d fascinating study of this kind of editorial control to maintain generic integrity. T h e y point out that for the construction a n d dissemination of knowledge 'textual activity' is as important as the 'scientific activity.' Generic Conventions as Authority: The Case of Citations and References To us academics, the p o w e r of genre is n o w h e r e better illustrated than in the publication of research articles. Swales in his research report Aspects of Research Article Introductions (1981) was the first one to point out the importance of the description of previous research o n the rhetorical activity of knowledge dis semination as distinct from knowledge creation. I n order to b e c o m e acceptable to
DISCOURSE STUDIES
116
the specialist community of fellow researchers, one must relate his or her know ledge claims to the accumulated knowledge of the discipline, without which his or h e r claims in the field are unlikely to find recognition through publication. I n this context it is hardly surprising that literature review occupies a n import ant place in the researcher's repertoire of skills in most academic disciplines. R e f e r r i n g to t h e i m p o r t a n c e of c i t a t i o n i n scientific r e s e a r c h activity, A m s e r d a m s k a a n d Leydesdorff (1989: 451) point out, I n a scientific article 'the n e w encounters the old' for the first time. This encounter has a double significance since articles not only justify the n e w b y showing that the result is warranted b y experiment or observation or previous theory, but also place a n d integrate innovations into the context of'old' a n d accepted k n o w l e d g e . . . References which a p p e a r in the text are the most explicit m a n n e r in which the arguments presented in the article are portrayed as linked to other texts, a n d thus also to particular b o d y of knowledge. Power to Innovate (Mixing and Embedding) Although this pressure for the 'democratisation' (Fairclough, 1992) of discourse is b e c o m i n g increasingly intense in some countries, especially in the U S A , it is unlikely to make a significant dent in the so-called integrity of professional genres, at least n o t in the foreseeable future. However, one can see an increasing 'frag mentation of discursive n o r m s a n d conventions' (Fairclough, 1992: 221), often leading to genre-mixing a n d e m b e d d i n g in institutionalized orders of discourse (see Bhatia, 1994, for a detailed discussion of this), o n the o n e h a n d , a n d creation of new genres, on the other. To a large extent, these changes in discursive practices are m a k i n g professional genres increasingly dynamic a n d complex. T h e dynamic complexity of academic a n d professional communication is further increased b y the role of multimedia, the explosion of information tech nology, the multi-disciplinary contexts of the world of work, the increasingly competitive professional environment, a n d a b o v e all, the overwhelmingly com pulsive nature of promotional a n d advertising activities, so m u c h so that our present-day world of work is being increasingly identified as a 'consumer culture' (Featherstone, 1991). T h e inevitable result of this is that m a n y of the institutionalized genres, whether they are social, professional or academic, are seen as incorporating elements of promotion. Fairclough (1992: 207) righdy associates some of these changes with what h e calls 'commodification' of institutional orders of discourse. Referring to such changes in discourse practices, h e (1993: 141) points o u t . . . there is a n extensive restructuring of boundaries between orders of discourse a n d b e t w e e n discursive practices, for example, the genre of consumer advertising has b e e n colonising professional a n d public service orders of discourse o n a massive scale, generating m a n y n e w hybrid ™^tmnol rrpnrps . . . 11
BHATIA THE POWER AND POLITICS OF GENRE
117
As an instance of such a h y b r i d genre, Fairclough (1993) discusses the case of c o n t e m p o r a r y university prospectuses, w h e r e h e highlights an increasing tendency towards marketization of the discursive practices of British universities. Bhatia (1995), in his discussion of genre-mixing in professional discourse, gives examples from several settings, where genre-mixing a n d e m b e d d i n g is becoming increasingly c o m m o n . H e also mentions several instances where o n e m a y find an increasing use of promotional strategies in genres which are traditionally considered non-promotional in intent, especially academic introductions, including book introductions, forewords, prefaces of various kinds, which are b e c o m i n g increasingly difficult to distinguish from publishers' blurbs. Shared Knowledge - Privileged Access/Insider Information If generic conventions, o n the o n e h a n d , give suitable expression to the com municative intentions of genre writers (who are m e m b e r s of a particular discourse community), o n the other h a n d , they also match their intentions against their intended reader's expectations. This is possible only w h e n all the participants share, not only the code, b u t also the knowledge of the genre, which includes the knowledge of its construction, interpretation a n d use. A necessary implication of this shared genre knowledge is that it is not routinely available to the outsiders, which creates a kind of social distance between the legitimate m e m b e r s of a discourse c o m m u n i t y a n d those w h o are considered outsiders. Although this creates conditions of homogeneity between the insiders, at the same time it also increases social distance between t h e m and the outsiders, sometimes resulting in disastrous consequences for the o n e w h o does n o t h a v e access to such shared knowledge. This shared knowledge could b e in the form of hnguistic resources used to construct a generic form, or it could b e in the awareness of the rules of language use, some of which are socially learnt, as the ones associated with classroom discourse a n d academic genres, while other scan b e legally enforced, such as the ones associated with courtroom procedures. Allen a n d G u y (1989), based on a personal communication from Worthington (1984: personal c o m m u nication), report a n excellent e x a m p l e of the lack of shared knowledge from the account of the courtroom interaction. An off-duty policeman in a store had shot and killed an intruder. Investigation had shown a set of burglar tools at the back of the store. The prosecutor was trying to show that there was no ground for presuming criminal intent, and that this was cold blooded murder. The victim's wife was testifying for the prosecution. Here she is being cross-examined by the defense.
J
Defense Lawyer: Could you tell the court and the jury what your husband's occupation was? Wife: He was a burglar. This supported the defense's contention of criminal intent, and secured ioquitfal for the policeman.
DISCOURSE STUDIES
116
the specialist community of fellow researchers, one must relate his or her know ledge claims to the accumulated knowledge of the discipline, without which his or h e r claims in the field are unlikely to find recognition through publication. I n this context it is hardly surprising that literature review occupies a n import ant place in the researcher's repertoire of skills in most academic disciplines. R e f e r r i n g t o t h e i m p o r t a n c e of c i t a t i o n i n scientific r e s e a r c h activity, A m s e r d a m s k a and Leydesdorff (1989: 451) point out, I n a scientific article 'the n e w encounters the old' for the first time. This encounter has a double significance since articles n o t only justify the new b y showing that the result is warranted b y experiment or observation or previous theory, b u t also place a n d integrate innovations into the context of 'old' a n d accepted k n o w l e d g e . . . References which appear in the text are the most explicit m a n n e r in which the arguments presented in the article are portrayed as linked to other texts, a n d thus also to particular b o d y of knowledge. Power to Innovate (Mixing and Embedding) Although this pressure for the 'democratisation' (Fairclough, 1992) of discourse is b e c o m i n g increasingly intense in some countries, especially in the U S A , it is unlikely to m a k e a significant dent in the so-called integrity of professional genres, at least n o t in the foreseeable future. H o w e v e r , one can see an increasing 'frag mentation of discursive n o r m s a n d conventions' (Fairclough, 1992: 221), often leading to genre-mixing a n d e m b e d d i n g in institutionalized orders of discourse (see Bhatia, 1994, for a detailed discussion of this), o n the o n e h a n d , a n d creation of new genres, on the other. To a large extent, these changes in discursive practices are m a k i n g professional genres increasingly dynamic a n d complex. T h e dynamic complexity of academic a n d professional communication is further increased b y the role of multimedia, the explosion of information tech nology, the multi-disciplinary contexts of the world of work, the increasingly competitive professional environment, a n d above all, the overwhelmingly com pulsive nature of promotional a n d advertising activities, so m u c h so that our present-day world of work is being increasingly identified as a 'consumer culture' (Featherstone, 1991). T h e inevitable result of this is that m a n y of the institutionalized genres, whether they are social, professional or academic, are seen as incorporating elements of promotion. Fairclough (1992: 207) rightly associates s o m e of these changes with what h e calls 'commodification' of institutional orders of discourse. Referring to such changes in discourse practices, h e (1993: 141) points out, . . . there is a n extensive restructuring of boundaries b e t w e e n orders of discourse a n d b e t w e e n discursive practices, for example, the genre of consumer advertising has b e e n colonising professional a n d public service orders of discourse o n a massive scale, generating m a n y n e w hybrid partly promotional genres . . .
BHATIA THE POWER AND POLITICS OF GENRE
117
A s an instance of such a hybrid genre, Fairclough (1993) discusses the case of contemporary university prospectuses, w h e r e h e highlights an increasing tendency towards marketization of the discursive practices of British universities. Bhatia (1995), in his discussion of genre-mixing in professional discourse, gives examples from several settings, where genre-mixing and e m b e d d i n g is becoming increasingly c o m m o n . H e also mentions several instances where one m a y find an increasing use of promotional strategies in genres which are traditionally considered non-promotional in intent, especially academic introductions, including book introductions, forewords, prefaces of various kinds, which are b e c o m i n g increasingly difficult to distinguish from publishers' blurbs. Shared Knowledge - Privileged Access/Insider Information If generic conventions, o n the o n e h a n d , give suitable expression to the com municative intentions of genre writers (who are m e m b e r s of a particular discourse community), o n the other h a n d , they also match their intentions against their intended reader's expectations. This is possible only w h e n all the participants share, n o t only the code, b u t also the knowledge of the genre, which includes the knowledge of its construction, interpretation a n d use. A necessary implication of this shared genre knowledge is that it is not routinely available to the outsiders, which creates a kind of social distance between the legitimate m e m b e r s of a discourse c o m m u n i t y a n d those w h o are considered outsiders. Although this creates conditions of homogeneity b e t w e e n the insiders, at the same time it also increases social distance between t h e m a n d the outsiders, sometimes resulting in disastrous consequences for the o n e w h o does n o t h a v e access to such shared knowledge. This shared knowledge could b e in the form of linguistic resources used to construct a generic form, or it could b e in the awareness of the rules of language use, some of which are socially learnt, as the ones associated with classroom discourse a n d academic genres, while other scan b e legally enforced, such as the ones associated with courtroom procedures. Allen a n d G u y (1989), based on a personal communication from Worthington (1984: personal commu nication), report a n excellent e x a m p l e of the lack of shared knowledge from the account of the courtroom interaction. An off-duty policeman in a store had shot and killed an intruder. Investigation had shown a set of burglar tools at the back of the store. The prosecutor was trying to show that there was no ground for presuming criminal intent, and that this was cold blooded murder. The victim's wife was testifying for the prosecution. Here she is being cross-examined by the defense.
.
Defense Lawyer: Could you tell the court and the jury what your husband's occupation was? Wife: He was a burglar. This supported the defense's contention of criminal intent, and lecured acquittal for the policeman.
If only the wife h a d b e e n slightly m o r e familiar with the conventions of the courtroom examination, the task of the defence lawyer would n o t h a v e b e c o m e that easy. Another example of the use of insider information to get access to information can b e illustrated b y the following headline from a n advertisement for 'The Schroder Singapore Trust,' which reads, The Schroder Singapore Trust Has Grown Over 60% In 3 Years T h e information being given h e r e can b e extremely misleading, except to those w h o are well aware of the discursive practices of the professional c o m m u n i t y of financial managers. A n y b o d y trying to m a k e sense of this statement should k n o w that this 60 p e r c e n t growth in three years o n its face value could be misleading, to say the least. Although it carries the usual statutory disclaimer in the form of a note in small print saying, 'Past performance is n o t necessarily a guide to future performance, the price of units m a y fall as well as rise a n d cannot b e guaranteed,' a lay person might still b e led to think that his investment will probably get h i m close to a 60 percent return. T h e fact, o n the other hand, could b e that the unit value might h a v e declined b y 100 percent in the last one year or so, a n d m a y still b e showing the d o w n w a r d trend at the time of the ad vertisement. T h e r e could b e several other possible scenarios w h i c h will be accessible only to those with the inside knowledge of the w a y these genres function rather t h a n to outsiders. Maintaining Solidarity within a Professional Community O n e of the m o s t noticeable characteristics of any professional or academic discourse c o m m u n i t y is the availability a n d typical use of a range of appropriate genres, which their m e m b e r s think serve the goals of their community. The recurrent use of such discoursal forms create solidarity within its membership giving t h e m their most powerful w e a p o n to keep the outsiders at a safe distance. H u d s o n (1979: 1) righdy claims, If o n e wished to kill a profession, to r e m o v e its cohesion a n d its strength, the most effective w a y would b e to forbid the use of its characteristic language. I n this context, it is hardly surprising that most of the attempts b y the powerful reformist lobbies in m a n y Western democracies to introduce plain English in legislative contexts are seen as imposition from outside and h a v e b e e n firmly rejected b y the professional legal community. T h e m a i n purpose of legislation, as Bhatia (1993) points out, is to govern the behaviour of individuals a n d institutions in society through the use of rules and regulations. I n order to keep control firmly in the h a n d s of the legislature rather than the judiciary in a parliamentary democracy, statutory acts are written noi only clearly, precisely a n d unambiguously but all-inclusively too. This rigour and
TIA THE POWER AND POLITICS OF GENRE
119
adequate specification of scope in legislation helps the legislature to control a totally subjective and idiosyncratic interpretation of the statute book. All attempts to reform legislative language, including those b y the plain English campaign (see T h o m a s , 1985; Eagleson, 1988; Kelly, 1988), h a v e to a large extent m e t with very limited success, for the simple reason that they are seen as a transgression of the generic integrity of the whole tradition in the legis lative process. Although the plain English m o v e m e n t has b e e n quite effective in influencing the redrafting of general commercial and administrative documents, including insurance policies, residential leases, tax return forms, social benefit claim forms and other papers for better accessibility and usability b y a larger section of society, w h e n it comes to legislative provisions, it has not b e e n able to soften the attitude of t h e p a r l i a m e n t a r y draftsmen significantly i n m a n y of the C o m m o n w e a l t h countries. T h e other argument for the preservation of these generic characteristics of legislative discourse is that the real legislative p o w e r in all parliamentary democracies must rest with the legislature a n d n o t with the judiciary. This is o n e of the most important reasons w h y clarity, precision, unambiguity a n d all-inclusiveness are so highly prized in British legislative dis course, which gives a relatively high degree of transparency to legislative intentions. Power and Control in Legislative Context Writing legislative discourse in terms of simple principles without adequate speci fication of the required scope, o n the other h a n d , m e a n s giving wider powers to the judges a n d the courts to interpret the intentions of the legislature, which is not considered highly desirable in parliamentary democracies. We find a n excellent illustration of this point in the Basic Law drafted b y the People's Republic of China. It is m e a n t to b e a mini-constitution for post-1997 Hong K o n g based o n the Sino-British J o i n t Declaration of 1978. It is written in the form of very simple principles a n d guidelines, a generic form, which can b e seen as almost exactiy the opposite of what we find in a typical legislative discourse from any of the countries of the Commonwealth. It is written in the form of basic principles o n which there can hardly b e any disagreement, lacking very sig nificantly in details of all kinds. O n e of the important issues raised there is that of status of the laws previously in force in H o n g Kong. Article 8 of the Basic Law states, T h e laws previously in force in H o n g Kong, that is, the c o m m o n law, rules of equity, ordinances, subordinate legislation a n d customary law shall b e maintained, except for any that contravene this law, a n d subject to any a m e n d m e n t b y the legislature of the H o n g K o n g special Admin istrative Region. As one m a y see, like the other Articles of the Basic Law, this one too is expressed in terms of somewhat universal principles which are applicable to everything me could think of in the context of pre-existing legislative machinery. O n e m a y
120
DISCOURSE STUDIES
b e tempted to point out that there should b e n o serious p r o b l e m in expressing legislation in terms of general principles. T h e r e are several legal systems which adopt such a strategy. T h e French legislative system is a g o o d e x a m p l e of this. However, p r o b l e m s h a v e occurred in this context, primarily because t h e transi tion of p o w e r has still s o m e w a y to go. A n u m b e r of issues a r e still unresolved and to m a k e it worse, all this is h a p p e n i n g in the context where two very different systems are in operation, t h e most elaborate a n d exhaustive legislative style used in t h e U K a n d their extremely diluted plain English versions captured in the Basic Law. Every time a n e w ordinance is considered or promulgated in Hong Kong, it becomes a matter of fresh negotiation between the two governments. Obviously, this generic form of writing gives m a x i m u m p o w e r to t h e one w h o h a s t h e authority to interpret it. Since the Basic L a w is m e a n t to take effect only after J u l y 1, 1997, after the territory is h a n d e d over to China, t h e eventual control over its interpretation will b e concentrated i n the h a n d s of the future S A R g o v e r n m e n t of the P R C . However, i n the intervening period leading u p to the hand-over, the interpretations of m a n y of its sections h a v e b e e n disputed by b o t h sides, for t h e simple reason that t h e genre i n which it is written allows m a x i m u m p o w e r to interpret it to those w h o h a v e t h e p o w e r to d o it, which for the time b e i n g is shared b y t h e two parties. T h e p o w e r a n d politics of genre a r e t h e two sides of the same coin. I n one context, it c a n b e seen as a legitimate force often used to maintain solidarity within a disciplinary community, whereas o n t h e other h a n d , it is used to keep outsiders at a respectable distance. O n t h e one h a n d , it empowers some people, the insiders, while at the same time, it tends to silence others, especially the out siders. T h a t is what I m e a n t b y t h e p o w e r a n d politics of genre. W e h a v e seen the p o w e r of genre, let's turn to its politics now.
T h e Politics of G e n r e Exploitation, Innovation and Manipulation of Generic Conventions I h a v e tried to present genre as a dynamic a n d clarificatory rather than a static or classificatory construct. I h a v e also tried to maintain that it h a s a propensity for innovation, exploitation a n d manipulation. I would n o w like to take this argu m e n t further to discuss t h e nature of this exploitation or manipulation and con straints o n such exploitation. Genres are dynamic constructs, even though they a r e essentially seen as e m b e d d e d in conventions associated with typical instances of language use in social, academic or professional settings. A n understanding or a prior knowledge of conventions is considered essential for its identification, construction, interpret ation, use a n d ultimate exploitation b y m e m b e r s of specific professional commu nities to achieve socially recognized goals with some degree of pragmatic success. T h e nature of genre manipulation is invariably realized within the broad limits of specific genres a n d is often very subtle. This can only b e handled witliin the concept of genre because such liberties, innovations, creativities, expioit— — '•ail f V i o m a r e invariablv realized wiihin 1
BHATIA THE POWER AND POLITICS OF GENRE
121
rather t h a n outside the generic boundaries, whichever w a y o n e m a y draw them, in terms of recurrence of rhetorical situations (Miller, 1984), consistency of communicative purposes (Swales, 1990), existence and arrangement of obligatory structured elements (Halliday a n d Hasan, 1985) or a combination of these (Bhatia, 1993). T h e m o m e n t it b e c o m e s a free-for-all k i n d of activity, communication itself will b e c o m e m o r e of a problem. T h e reason is that the flouting of generic conventions leads to the opting out of the genre altogether a n d is noticed b y the members of the concerned disciplinary c o m m u n i t y as being o d d . A n y attempt, therefore, to overlook, ignore or u n d e r m i n e the p o w e r of conventions at this stage can result in disastrous consequences. Although a g o o d understanding of genre knowledge is a prerequisite to a n y manipulation of generic resources, it is b y n o m e a n s sufficient to get such innov ations and exploitations accepted i n a disciplinary community. Kress (1987:42) mentions two significant ways in which generic innovations are accepted, either they are backed b y a stable social occasion or b y authority. Unless . . . there is a change in the social structures - a n d in the kinds of social occasions in which texts are p r o d u c e d - the n e w generic forms are unlikely to succeed. T h a t is w h y childish innovations fail; n o t because they do n o t constitute perfectly plausible solutions to particular textual/ cognitive p r o b l e m s , but because they are supported neither b y a stable social occasion, n o r b y 'authority.' This latter is of course the case w h e r e a writer of 'authority' creates a n e w generic form, which, seemingly be cause of the writer's authority alone, succeeds in establishing a n e w generic convention. Gate-keeping Function ofDiscourse Communities Berkenkotter and Huckin (1995:102) in their study of gate-keeping at an academic convention, i.e., the study of abstracts submitted for the Conference o n College Composition a n d C o m m u n i c a t i o n ( C C C C ) , discuss an interesting instance of the power of generic control in well-defined contexts. O n the basis of their analysis of the process of selection of papers for the C C C C convention, they claim that, 1. T h e high-rated abstracts all addressed topics of current interest to active, experienced members of the rhetoric and composition community, the low-rated abstracts often did not. 2. Almost all of the high-rated abstracts clearly defined a problem; the lowrated abstracts often did not. 3. The high-rated abstracts all discussed this problem in a way that would be seen by experienced insiders as novel and therefore interesting, whereas virtually none of the low-rated abstracts did. 4. The high-rated abstracts usually projected more of an insider ethos through the use of terminology, special topoi, a n d / o r explicit or implicit refer* ences to the scholarly literature of the field than did the low-rated abstracts. I tEmohasis adrlprl 1
L
DISCOURSE STUDIES
122
T h e y also point out that often 'the genre was shaped significandy b y the interests of the p r o g r a m chair.' This is generally d o n e through the t h e m e statement issued w h e n p a p e r s are invited for the convention. D e p e n d i n g u p o n the interests of the p r o g r a m chair or of the discourse community he/she represents, the emphasis can shift from one year to the other. Based o n their study of C C C C abstracts for four years b e t w e e n 1 9 8 8 - 9 2 , they find two m a i n levels of gate-keeping (1995: 115): (a) the external reviewers and (b) the p r o g r a m chair. We h a v e observed m a n y cases w h e r e the reviewers rated a n abstract Excellent a n d yet it was n o t included in the p r o g r a m . Presumably, the chair disagreed with the reviewers' j u d g e m e n t s I n short, each convention bears the stamp of its principal gatekeeper. T h e y further point out that (1995: 115), I n one particularly unfortunate case, a very interesting abstract was sub mitted to the Technical C o m m u n i c a t i o n area o n e year, w h e r e it received a n Excellent rating from a reviewer a n d the p r o g r a m chair b u t was not included in the p r o g r a m (presumably because of a b a d 'fit'). It was revised slightly a n d resubmitted the following year to the Discourse Analysis area. Again it received an Excellent rating, b u t again it was n o t included in the p r o g r a m . T h e author of this abstract p r o b a b l y n e v e r k n e w that she h a d written an outstanding abstract. All she w o u l d h a v e b e e n told was that h e r paper h a d b e e n rejected for the program. Another interesting case of such a gate-keeping encounter, though of a shghtly different n a t u r e , b e t w e e n two different discourse groups was recently referred to b y William Bright (1996). Giving his view of 3 0 years of A m e r i c a n linguistics, h e referred to the following extract from a letter written b y C h o m s k y , . . . t h e level of rumour-mongering a n d of personal h o s t i l i t y . . . outright falsification so scandalous that they raise serious questions about the integrity of the f i e l d . . . I do not want to b e associated with a j o u r n a l . . . . which publishes flat lies . . . couched in rhetoric of a sort that might be appropriate to some criminal, but that one is surprised to find in a scholarly journal. N o w o n d e r Chomsky never published in the journal, not because h e was kept om. but h e decided to keep himself out. Gate-keeping can obviously work both ways. Hegemony and World Englishes - Generic Variation and Control Another important aspect of generic control raises the issue of hegemonic attitude to maintain generic standards, which in m u c h of c o n t e m p o r a r y discourse and g e n r e studies are d o m i n a t e d a n d e v e n d e t e r m i n e d b y essentially Western
BHATIA THE POWER AND POLITICS OF GENRE
123
used global language for academic as well as professional purposes, it is n o longer the sole property of a n y o n e c o m m u n i t y of people, b e they English, American, Australian or a n y other. Like cricket, English has also b e c o m e m o r e universal not only in usage b u t also in its character. True to the reality of present-day vari ation in English, o n e needs to think in terms of world Englishes, rather t h a n English as a single monolithic variety of English (for a detailed discussion a n d references see Kachru, 1986,1994 a n d 1996). This variation in the use of English across the globe is getting increasing recognition in the sociohnguistic literature in the last decade or so; however, in some of the genres, especially used in the academia, the p o w e r to control a n d maintain generic standards can b e , a n d often is, interpreted in terms of the d o m i n a n t community, which undoubtedly happens to b e a Western community. Anything which appears to b e different from the n o r m s set b y the d o m i n a n t c o m m u n i t y is viewed as deficient a n d i n need of correction. In some areas, genre writers h a v e b e c o m e increasingly sensitive to local knowledge a n d h a v e started constructing, interpreting a n d using genres in forms which display such sensitivities, especially in the case of advertising a n d s o m e other business genres, w h e r e it has b e c o m e an established practice n o w to de velop local teams to act alongside the expatriates in most of the multinational advertising companies. T h e reason for such sensitivities is also n o t difficult to understand. I n the case of academic genres, especially in research publications, the politics is still controlled b y those w h o have the power. M u c h of the academic discourse still fails to acknowledge the sources of variations, especially those of marginality a n d exclusion, giving the impression that there is, or should b e , n o variation in the way genres are constructed, interpreted and used.
Implications for L a n g u a g e T e a c h i n g a n d L e a r n i n g What are the implications of all this for language teaching? Applied genre analy sis, unlike m a n y other analytical frameworks, is neither static n o r prescriptive. Potentially, it is dynamic a n d explanatory. It is for the language teacher to use it the way one would like to use it, for innovative exploitation of generic resources or for a limited exposure to standardized generic contexts. Although it is essential for the learner to b e familiar with specific generic conventions associated with a particular professional setting, it is neither necessary n o r desirable to restrict the experience of linguistic behaviour to just the conventionalized and standardized aspects of genre construction a n d use. How can one bring in creativity in genre-based language teaching a n d learnine? Since genre analysis gives a g r o u n d e d description of linguistic behaviour in professional settings, it is possible to bring in a fair a m o u n t of creativity in language teaching b y adjusting communicative purposes, the nature of participatk a in a particular communicative setting, the social a n d professional relationship between the participants taking p a r t in a particular genre-construction exercise, and above all, b y bringing in variability in the use of generic strategies to achieve the same communicative purposes.
124
DISCOURSE STUDIES
T h e r e are two schools of thought, I should say. T h o s e w h o believe in the explicit teaching of genres, especially the regularities of textual form and typifications, a n d others w h o see this as too constraining a n d advocate free expression. T h e truth, however, rests somewhere in the middle. All genres, primary as well as secondary, involve regularities and, h e n c e , these regularities must b e learnt b y a n y o n e w h o has even the slightest ambition of being part of any specialist dis ciplinary community. As Bakhtin (1986: 80) points out 'genres must b e fully mastered to b e used creatively.' However, in order to m a k e this h a p p e n , the first prerequisite is to h a v e an awareness of the conventional k n o w l e d g e that is situated within a specific disciplinary genre or a 'system of genres.' Bazerman (1993: viii) attempts to resolve this tension between institutionalized expression and individual expression w h e n h e points out, . . . the individual learns to express the self against the compulsive society . . . We are n o t ourselves because we set ourselves apart from each other. We b e c o m e ourselves as we realize ourselves in relation to each other. T h e social is everything we do with each other a n d w h a t we b e c o m e as we d o it. W e individuate b y identifying ourselves o n a social landscape, a landscape c o m e to k n o w as w e interact with it. We discover a n d create ourselves a n d others b y w h a t w e d o with each other. T h e r e are at least three things which stand out clearly from the foregoing discussion. First, language learners n e e d to b e c o m e aware of the conversations of the disciplinary c o m m u n i t y to which they aspire to b e m e m b e r s of, which could b e d o n e through 'centripetal participation in the learning curriculum of the ambient c o m m u n i t y ' (Lave a n d Wenger, 1991: 100). Second, acquisition of g e n r e k n o w l e d g e , w h i c h leads to a n u n d e r s t a n d i n g of generic integrity, is necessary b u t not sufficient for any subsequent exploitation or manipulation of generic conventions. A n d , finally, genre knowledge should b e best viewed as a resource to exploit generic conventions to respond to recurrent a n d n o t so recur rent rhetorical situations, rather than a blueprint for replication.
References Allen, Donald E. and Guy, Rebecca F. (1989) Non-routine conversation in operational crisis. In Working with Language: A Multidiseiplinary Consideration ofLanguage Use in Work Contexts. Edited by Hywel Coleman. Berlin, New York: Mouton de Gruyter. Amsterdamska, O. and Leydesdorff, L. (1989) Citations: indicators of significance? Scientometrics, 15,449-471. Bakhtin, M. (1986) The problem of speech genres. In Speech Genres and Other Late Essays. Edited by C. Emerson M. Holquist. Trans. V W. McGhee. Austin: University of Texas Press, pp. 60-102. Bazerman, Charles (1993) Foreword to Professional Communication: The Social Perspective Edited bv Nancy Roundy Blyler and Charlotte Thralls. Sage. pp. VH-VIII.
BHATIA THE POWER AND POLITICS OF GENRE
125
Bazerman, Charles (1994) Systems of genres and the enactment of social intentions. In Genre and the New Rhetoric. Edited by Aviva Freedman and Peter Medway. London: Taylor & Francis, pp. 79-101. Berkenkotter, C. and Huckin, Thomas N. (1995) Genre Knowledge in Disciplinary Communication - Cognition/Culture/Power. NewJersey: Lawrence Erlbaum Associates. Bhatia, V. K. (1993) Analysing Genre - Language Use in Professional Settings. London: Longman. Bhatia, V. K. (1994) Generic integrity in professional discourse. In Text and Talk in Professional Contexts. Edited by Britt-Louise Gunnarsson, Per Linell and Bengt Nordberg. ASIA: sskriftsrie 6, Uppsala, Sweden. Bhatia, V. K. (1995) Genre-mixing and in professional communication: the case of'private intentions' v. 'socially recognized purposes.' In Explorations in EnglishfirProfessional Communication. Edited by Paul Bruthiaux, T. Boswood and B. Bertha. Hong Kong: Department of English, City University of Hong Kong. Bright, William (1996) The view from the editor's desk: 30 years of American Linguistics. A talk given at the City University of Hong Kong. Bruffee, K. A. (1986) Social construction, language and the authority of knowledge: a bibliographical essay. College Composition, 48, December, 730-790. Dubrow, Heather (1982) Genre. London: Methuen & Co. Ltd. Eagleson, R. D. (1988) Efficiency in legal drafting. In Essays on Legislative Drafting: In Honour of J QEwens, CMG, CBE, QC. Edited by D. Kelly. Adelaide: University of Adelaide, The Adelaide Law Review Association Law School, pp. 13-27. Fairclough, N. (1992) Discourse and Social Change. London: Polity. Fairclough, N. (1993) Critical discourse analysis and the marketization of public discourse: the universities. Discourse & Society, 4(2), 1993, 133-168. Featherstone, M. (1991) Consumer Culture and Postmodernism. London: Sage. Foucault, M. (1981) The Archaeology ofKnowledge. New York: Pantheon Books. Fowler, Alastair (1982) Kinds ofLiterature. Oxford: Oxford University Press. Freedman, Aviva, and Medway, Peter (eds.) (1994) Genre and the New Rhetoric. London: Taylor & Francis. Goodrich, P. (1987) Legal Discourse. London: Macmillan. Halliday, M. A. K. and Hasan, R. (1985) Language, Context, and Text: Aspects of Language in a Social-Semiotic Perspective. Victoria: Deakin University Press. Hudson, Kenneth (1979) TheJargon ofthe Professions. London: The Macmillan Press Ltd. Kachru, Braj B. (1986) The power and politics of English. World Englishes, 5(2-3), 121-140. Kachra, Braj B. (1994) World Englishes: approaches, issues and resources. In Readings on Second Language Acquisition. Edited by D. Brown and S. Gonzo. New York: PrenticeHall. (An earlier version in Language leaching: The International AbstractingJournal of Language Teachers and Applied Linguistics. January, 1992. Cambridge: Cambridge University Press, pp. 1-14. Kachru, Braj B. (1996) The paradigms of marginality. World Englishes, 15(3), 241-255. Kelly, D. L. (ed.) (1988) Essays on Legislative Drafting: In Honour of/ QEwens, CMG, CBE, QC. University of Adelaide, The Adelaide Law Review Association, Law School. Kress, Gunther (1987) Genre in a social theory of language: a reply to John Dixon. In The Place of Genre in Learning: Current Debates. Edited by I. Reid. Geelong, Australia: Deakin University Press. Lave, J. and Wagner, E. (1991) Situated Learning: Legitimate Peripheral Participation. Cambridge, MA: Cambridge University Press. Martin, J R. (1993) A Contextual Theory of Language. In The Powers of Literacy -A Genre Ap proach to Teaching Writing. Pittsburg, PA: University of Pittsburgh Press, pp. 116-136.
126
DISCOURSE STUDIES
Miller, C. R. (1984) Genre as social action. QuarterlyJournal ofSpeech, 70, 151-167. Myers, Greg (1995) Disciplines, departments, and differences. In Writing in Academic Contexts. Edited by Britt-Louise Gunnarsson and Ingegerdv Backlund. Uppsala Universitet. pp. 3-11. Pecheux, M. (1982) Language, Semantics and Ideology. London: Macmillan. Swales,John M. (1981) Aspects ofArticle Introductions. LSU Research Report. University of Aston in Birmingham. Swalesjohn M. (1990) Genre Analysis-English in Academic andResearch Settings. Cambridge: Cambridge University Press. Thomas, R. (1985) Plain English and the law. Statute Law Review, 9(3), 144.
6 On the Complexity of Discourse Complexity: A Multidimensional Analysis Douglas Biber
1. Introduction " ^ h e r e h a s b e e n considerable interest within linguistics, psychology, a n d e d u c a t i o n c o n c e r n i n g t h e linguistic c h a r a c t e r i z a t i o n of d i s c o u r s e ^ J. complexity. Although most of this research takes a comprehension per spective, addressing the question of w h a t linguistic characteristics of a text m a k e it more or less understandable or readable, there are also two major production perspectives o n this issue: complexity differences in the language p r o d u c e d b y different social groups, and complexity differences in the language p r o d u c e d in different situations. T h e present article focuses o n the analysis of linguistic complexity across situations of use. Much research o n text comprehensibility has b e e n motivated b y an applied interest in the linguistic factors influencing the readability of a text (see e.g., Klare, 1984, for o n overview), whereas other processing research has b e e n theor etically motivated, analyzing the influence of more-or-less derivationally complex linguistic structures o n performance. These two perspectives are c o m b i n e d in the recent collection of p a p e r s edited b y Davison a n d G r e e n (1988). Readability research has long recognized that lexical a n d structural factors m a k e separate contributions to the overall complexity of a text; the recent p a p e r s b y Frazier (1988) and Smith (1988) futher claim that there are different kinds of structural complexity involved in the syntactic processing of sentences. Variation in discourse complexity across social groups is often assoctaied with the proposals of Bernstein (1962, 1970), w h o posited the existence of two competing "codes" - "restricted" and "elaborated" - a n d attempted to characterize the discourse of different social groups with respect to these codes (cf. the critique Source: Discourse Processes vol. 15.1992. nn laa-ifia
DISCOURSE STUDIES
128
b y Dittmar, 1976). Studies analyzing the linguistic complexity of texts produced b y different social groups h a v e considered b o t h structural and lexical features. For example, Poole (1973, p . 98) considers 19 "indices of code elaboration," including "indices of structural complexity" (such as the frequencies of sub ordinate clauses a n d relative clauses), "indices of language elaboration and modification" (such as the frequencies of total adjectives a n d adverbs), and "indices of v e r b complexity" (such as the frequency of passive verbs a n d prev e r b length). Van d e n Broeck (1977) focuses o n five measures of complexity: m e a n length of T-unit (i.e., a m a i n clause a n d all associated d e p e n d e n t clauses), a subordination index, a multiple e m b e d d i n g index, a passivization index, and the ratio of agentless passives to iy-passives. Findings h a v e not b e e n consistent in this research area, a n d further research is required to investigate the kinds and degree of complexity differences across social groups. T h e present article, however, approaches discourse complexity from the perspective of differences across situational varieties, or "registers," rather than taking a social or psycholinguistic perspective. This perspective adopts a product (rather than process) approach, focusing o n the linguistic characteristics of texts p r o d u c e d u n d e r differing situational constraints a n d purposes. For example, m a n y studies h a v e claimed that written texts are m o r e structurally complex than spoken texts because authors h a v e m o r e time for production in typical writing situations than speakers in typical speaking situations. Similarly, studies h a v e claimed that speakers with informational purposes will p r o d u c e more complex texts than speakers with personal purposes. Although most studies focusing o n speech and writing h a v e reached the overall conclusions that writing is m o r e complex, elaborated, a n d explicit than speech (e.g., DeVito, 1967; O'Donnell, 1974; Olson, 1977), s o m e studies reach quite different conclusions; for example, Poole a n d Field (1976) a n d Halliday (1979) claim that speech is m o r e structurally complex t h a n writing. I n part these contradictory findings can b e attributed to the failure to control for the kind of text being analyzed (cf. the critiques b y Tannen, 1982, and Biber, 1986, 1988). However, they also reflect our incomplete understanding of t h e linguistic characteristics of discourse complexity. D e Beaugrande a n d Dressier (1981, p . 45) m a k e the general c o m m e n t that it would b e "wrong to imply that the production a n d reception of texts are well explored at this time. . . . T h e true complexity of the operations involved doubtless exceeds b y several orders of magnitude the most c o m p l e x models developed so far." I n a series of recent studies (Biber 1986,1988; Finegan & Biber, 1986), I ana lyze s e v e r a l linguistic correlates of discourse c o m p l e x i t y within a multi dimensional m o d e l of linguistic variation a m o n g English registers. Biber (19861 identifies three major "dimensions" of variation a m o n g spoken and written registers and finds that complexity features are associated with all three. Finegan and Biber (1986) further analyze the complexity features included in Biber (1986). showing that discourse complexity is at least a two-dimensional construct. Biber (1988) uses an exploratory factor analysis to identify six basic dimen sions of variation a m o n g registers in English. T h e study analyzes t h e distribution of 67 linguistic features across 481 texts taken from 2 3 spoken and written register 1
BIBER DISCOURSE COMPLEXITY
129
(e.g., a c a d e m i c prose, press reportage, conversation, radio broadcasts). T h e dimensions h a v e b o t h linguistic and functional content. T h e linguistic content comprises a group of linguistic features (such as p r o n o u n s , h e d g e s , present tense verbs) that co-occur with a m a r k e d l y high frequency in texts. T h e s e co-occurrence pattens are identified quantitatively b y the factor analysis. T h e functional content of a dimension is b a s e d on the assumption that linguistic co-occurrence reflects shared function; thus, e a c h d i m e n s i o n is interpreted to characterize the situ ational, social, and cognitive functions m o s t w i d e l y shared b y the co-occurring linguistic features. A l t h o u g h these dimensions represent a w i d e range of linguistic features a n d functions, t h e y include m a n y features associated w i t h discourse c o m p l e x i t y , and they confirm the finding that discourse c o m p l e x i t y is a multidimensional construct that interacts in intricate w a y s w i t h other functions. For e x a m p l e , dif ferent types of d e p e n d e n t clauses, w h i c h h a v e traditionally b e e n r e g a r d e d as being a m o n g the m o s t important indicators of discourse c o m p l e x i t y , are found on all six of the dimensions. O n the first dimension, causative subordination, sentence relatives, a n d W H clauses c o - o c c u r w i t h interactive and i n v o l v e d features (as opposed to features m a r k i n g informational integration); o n the second dimension, present participial clauses co-occur w i t h narrative features; on the third dimension, W H relative clauses function to m a r k an explicit and elaborated style of reference (as o p p o s e d to situated reference); o n the fourth dimension, infinitives a n d conditional adverbial clauses co-occur with m o d a l s and suasive verbs marking an overtly persuasive and argumentative style; o n the fifth dimen sion, past participial postnominal clauses and past participial adverbial clauses co-occur with passive constructions a n d conjuncts as markers of a technical, impersonal style; and o n the sixth dimension, that c o m p l e m e n t clauses to v e r b s , that c o m p l e m e n t clauses to adjectives, and /AaJrelative clauses o n object position co-occur with demonstratives, final prepositions, a n d existential there, reflecting an on-line informational focus. (These interpretations are b a s e d on the shared communicative functions o f the co-occurring features o n e a c h dimension and analysis of the distribution of features across genres; a full discussion o f the inter pretations is g i v e n in B i b e r , 1988.) T h e a p p r o a c h u s e d in B i b e r (1988) m i g h t b e l a b e l e d a macroscopic, explora tory approach. It is m a c r o s c o p i c i n that it analyzes a b r o a d r a n g e of linguistic features representing m a n y grammatical classes and discourse functions, and it analyzes a b r o a d r a n g e of the registers i n English. A major advantage of a m a c r o scopic approach is that it enables analysis of a discourse system relative to a variety of other interacting systems. T h u s , Biber (1988) s h o w s that the system of discourse c o m p l e x i t y interacts in important w a y s with a n u m b e r o f other dis course systems, i n c l u d i n g interactiveness a n d p e r s o n a l i n v o l v e m e n t , infor mational focus, narrative focus, m a r k i n g of reference, persuasive stance, and impersonal or technical style. T h e u n d e r l y i n g assumption of this a p p r o a c h is thai discourse systems are n o t autonomous; b u t rather that they represent a series of interlocking form/function sets that c a n o n l y b e fully understood b y considering their relation to the w h o l e r a n g e o f v a r y i n g linguistic features.
DISCOURSE STUDIES
130
A macroscopic a p p r o a c h can b e usefully c o m p l e m e n t e d , though, b y m o r e detailed analyses of particular systems, because the functional interpretation of macroscopic co-occurrence patterns depends o n prior analysis of the form/function associations within each system. It is for this reason that the present study focuses exclusively o n the system of discourse complexity. T h e a p p r o a c h used in Biber (1988) is exploratory in that it does n o t test par ticular hypotheses concerning the dimensions; rather, the p r i m a r y statistical procedure (exploratory factor analysis) identifies a set of co-occurrence patterns that fit the data, a n d the researcher then must interpret those patterns post-hoc. T h e r e are two potential problems with a n exploratory approach: (a) the analysis can capitalize o n chance co-occurrence patterns, and (b) certain results can be difficult to interpret because they h a v e little basis in prior research studies. Thus for some analyses a theory-based a p p r o a c h is preferable to a n exploratory one, because it is m o r e constrained, less likely to capitalize o n chance variation, and direcdy interpretable. I n a theory-based approach, multidimensional models are p r o p o s e d o n theoretical grounds, a n d statistical procedures (such as confir matory factor analysis) are used to test the adequacy of each m o d e l a n d compare the relative strengths of competing models. In this case, only those models that h a v e a solid theoretical grounding are tested, a n d the functional interpretation of the dimensions in a m o d e l is carried out before the statistical analyses. T h e present study thus differs from the analysis in Biber (1988) in two respects: it focuses exclusively on linguistic features associated with discourse complexity, and it adopts a theory-based rather than exploratory approach. It builds direcdy o n that previous study, however, b y testing several of the associations among complexity features indicated there. 2
3
2. O v e r v i e w of t h e Study T h e present study analyzes the distribution of 33 linguistic markers of complexity across 2 3 spoken and written registers. If discourse complexity is a single, unified construct, and all kinds of texts are complex (or not) in the same way, then all 33 of these linguistic features should b e distributed o n a single dimension of vari ation. However, if these features instead comprise distinct co-occurence patterns, t h e n a multidimensional account of discourse complexity is required. The exploratory analyses in Biber (1986,1988) and Finegan and Biber (1986) indicate that this latter possibility is the case; that discourse complexity is a multidimen sional construct. I n the present analysis I statistically c o m p a r e aunidimensional m o d e l with several m u l t i d i m e n s i o n a l m o d e l s h y p o t h e s i z e d o n theoretical grounds, using confirmatory factor analysis to determine the relative strengths of each competing m o d e l . T h e analysis confirms the earlier findings concerning the multidimensional nature of discourse complexity a n d identifies a particular five-dimensional m o d e l as the most adequate representation of the parameters of complexity in these data.
131
2.1 Texts Used in the Analysis T h e texts used in the study are summarized in Table 1. T h e s e texts represent variation along several situational parameters relating to discourse complexity, including: 1. Opportunity for careful production and revision versus the d e m a n d s of online production (e.g., an academic p a p e r vs. academic lecture); 2. Topic a n d purpose (e.g., a n academic p a p e r with a n informational p u r p o s e vs. a face-to-face conversation where the focus is o n interpersonal concerns); 3. O p p o r t u n i t y for interaction a n d clarification; a n d 4. T h e possibility of direct reference to the physical a n d temporal situation. Table 1: Distribution of texts across 23 registers Register
No. of texts
Written - Registers 1-15 Prom the LOB Corpus 1. Press reportage 2. Editorials 3. Press reviews 4. Religion 5. Skills and hobbies 6. Popular lore 7. Biographies 8. Official documents 9. Academic prose 10. General fiction 11. Mystery fiction 12. Science fiction 13. Adventure fiction 14. Romantic fiction '•• ' i 15. Humor ' 16. Personal letters , > 17. Professional letters
44 27 17 17 14 14 14 14 80 29 13 6 13 13 9 6 10
Spoken - From the London-Lund Corpus 18. Face-to-face conversation 19. Telephone conversation 2fl. Public conversations, debates, and interviews il. Broadcast J.2. Spontaneous speeches J3. Planned speeches ,
44 27 22 18 16 4
1
Total Approximate Number of Words
,
1
481 960,000
For example, regarding No. 3 a n d N o . 4, referents must b e clearly specified in written exposition, because reader and writer d o not typically share space
DISCOURSE STUDIES
132
and time, a n d t h e y c a n n o t easily interact; in contrast, referents can b e less fully specified in face-to-face conversation, because both speaker a n d h e a r e r can rely o n a direcdy shared space a n d time, a n d hearers can direcuy request clarification.) By analyzing the linguistic forms typically found in a n u m b e r of different registers, associated with different production opportunities, communicative pur poses, a n d interactional possibilities, it is possible to analyze the various types of discourse complexity from a textual perspective. T h e texts included in the present study include m a n y combinations of these situational parameters, as illustrated b y Table 2. Table 2: Summary of some situational differences among selected registers used in the study Interaction
Production
Purpose
Academic prose, official docs., press editorials Fiction Speeches, interviews Letters
Careful production and revision
Informational exposition
Minimal
Careful On-line Moderately careful
Minimal Moderate Moderate
Conversations
On-line
Aesthetic Informational Interpersonal and affective Interpersonal and affective
Extensive
2.2 Linguistic Markers of Discourse Complexity T h e present study focuses on 3 3 linguistic features that h a v e b e e n associated with discourse complexity. T h e o c c u r r e n c e s of these features are counted in each text, and these counts are normalized to frequency p e r 1,000 words. Table 3 lists the features in groups according to their p r i m a r y discourse function a n d grammatical category. T h e first six features are actually associated with a reduced complexity; these features fall into three groups: structural reduction, less specified reference, a n d fragmented structure, representing reductions in structural elaboration, specificity, a n d integration. At the same time, these features reflect m o r e difficult or complex tasks, because they are typically produced in demanding, on-line situations. T h e remaining features in Table 3 directly represent increased discourse complexity. These fall into four major groups: integrated structure (features that enable a large a m o u n t of informational content to b e packed into relatively few words), lexical specificity (features that reflect a careful word choice and words having relatively specific semantic content), passive constructions, a n d dependent clauses. T h e category of d e p e n d e n t clauses, which is the feature most frequently associated with structural complexity, actually comprises five subgroups: posln o m i n a l modifying clauses, attitudinal clauses, c o m p l e m e n t clauses, adverbial clauses, a n d participial clauses. As n o t e d in Table 3, these are not exclusive cate gories (e.g., certain participial clauses can also b e considered passive construe 4
BIBER DISCOURSE COMPLEXITY
Table 3: list of 33 surface features associated with linguistic complexity I. Features Associated With Reduced Complexity A. Structural reduction 1. That deletions 2. Contractions B. Less specified reference 3. Pro-verb Do 4. Pronoun It 5. Demonstrative pronouns C. Fragmented structure 6. Clause coordination II. Features Associated With Increased Complexity D. Integrated structure 7. Nouns 8. Prepositions 9. Attributive adjectives 10. Nominalizations 11. Phrasal coordination E. Lexical specificity 12. Word length 13. Type/token ratio F. Passive constructions (cf. past participial constructions) 14. Agendess passives 15. .fly-passives G. Dependent clauses Gl. Structural elaboration of reference - postnominal modifleM (cf. participial postnominal clauses) 16. WH relative clauses on subject position 17. WH relative clauses on object position 18. "Pied piping" relative clauses 19. That relative clauses on object position 20. That relative clauses on subject position G2. Attitudinal clauses 21. Sentence relatives G3. Complement clauses 22. WH clauses 23. That complement clauses to verbs 24. That complement clauses to adjectives 25. Infinitives G4. Adverbial clauses (cf. participial adverbial clauses) 26. Conditional adverbial subordination 27. Causative adverbial subordination , 28. Concessive adverbial subordination 29. Other adverbial subordination G5. Participial clauses , 30. Present participial postnominal clauses JjL 31. Past participial postnominal clauses 32. Present participial adverbial clauses
133
DISCOURSE STUDIES
134
All of the features in Table 3 are surface structure markers, and as such they are n o t fully representative of the linguistic correlates of discourse complexity. A complete m o d e l would include features representing cohesion, informational packaging (e.g., given vs. n e w information, topic, theme), a n d rhetorical organ ization. T h e present article is restricted to a comparison of models of discourse complexity as m a r k e d in the surface structure; b u t it shows that, even in this domain, discourse complexity is an extremely rich construct, comprising at least five separate dimensions. 2.3 Overview of Confirmatory Factor Analysis T h e present investigation of discourse complexity is based on statistical com parisons a m o n g a series of confirmatory factor analyses. I n a confirmatory analysis, models are p r o p o s e d o n theoretical grounds, a n d competing models are evaluated in terms of their relative goodness of fit (see, e.g., Long, 1983). This approach can b e best described b y contrasting it with a traditional ex ploratory factor analysis (such as in Biber, 1988). T h e goal of b o t h techniques is to uncover underlying dimensions of variation, where each dimension, or "factor," is represented b y a group of features that co-occur frequentiy in texts. A n explora tory factor analysis identifies factors quantitatively in a way that accounts for the m a x i m u m a m o u n t of "shared variance" a m o n g features, w h e r e shared vari ance corresponds to a frequent co-occurrence of the features in texts. T h e result ing factors are interpreted o n the assumption that features co-occur in texts because they share discourse functions. T h e theoretical analysis in this case is post hoc. T h a t is, although theoretical considerations are important to an explora tory factor analysis in that they guide preliminary decisions concerning the texts a n d linguistic features to b e i n c l u d e d , t h e statistical analysis itself is not constrained b y theory, a n d thus it can build o n chance correlations among features as well as theoretically significant correlations. I n contrast, a confirmatory factor analysis can b e motivated at all stages by theoretical considerations. Similar to exploratory factor analysis, these include, first of all, decisions concerning the texts and linguistic features to b e analyzed. H o w e v e r , w h e n using confirmatory techniques, the statistical analysis itself can b e constrained b y theoretical considerations to determine the number of factors, the particular features associated with each factor, and whether the factors are intercorrelated or not. T h e strength of this a p p r o a c h is that a number of competing models, each motivated o n theoretical grounds, can b e statistically c o m p a r e d to determine which best fits the data. T h u s , w h e n there is adequate background research to enable sound theoretical predictions, a confirmatory approach is preferable to an exploratory one. T h e confirmatory factor analyses were c o m p u t e d using the L I S R E L VI stat istical package (Joreskog & Sorbom, 1984). T h e analyses are based on a cor relation matrix of all linguistic features. For each particular m o d e l , the researchei specifies various theory-based restrictions, such as the n u m b e r of factors and the particular features included on each factor; the L I S R E L statistical procedure t h e n estimates the statistical p a r a m e t e r s of that m o d e l , including the actual 5
BIBER DISCOURSE COMPLEXITY
135
"loadings" of the features o n the factors and the intercorrelations a m o n g the factors. Factor loadings, which can range from - 1 . 0 to +1.0, m e a s u r e the strength of t h e association between a factor a n d a linguistic feature. If theoretical con siderations predict that a feature is n o t associated with a given factor, then the researcher specifies the loading of that feature o n that factor to b e 0.0. Otherwise, L I S R E L estimates the m a g n i t u d e of the loading. I n a later step, the estimated p a r a m e t e r s are used to generate a matrix of predicted correlations a m o n g the linguistic features. T h e adequacy of each m o d e l can b e evaluated b y c o m p a r i n g t h e original correlation matrix to the predicted correlation matrix; if the two matrices are highly similar, t h e n the p r o p o s e d model is j u d g e d to b e statistically adequate. This approach can also b e used to c o m p a r e t h e relative strengths of competing m o d e l s - the m o d e l that produces the best predicted fit to the original correlation matrix is considered the strongest. A m o r e detailed discussion of the statistics used to evaluate a n d compare statistical models is given in the A p p e n d i x , b u t the discussion in the following sections does not assume familiarity with those details. 2.4 Theoretical Models Indicated by the Exploratory Analysis in Biber
(1988)
As noted in Section 1, m a n y features associated with discourse complexity were included in the exploratory analysis of Biber (1988). T h e four models of discourse complexity tested in t h e present study are based o n structural a n d functional criteria identified in this earlier analysis. (In addition, I test a one-dimensional model here because of its widespread adoption in previous research.) T h e first m o d e l reflects a basic three-way distinction in t h e 1988 analysis, among complexity features m a r k i n g reduced structure, elaborated structure (i.e., the various kinds of d e p e n d e n t clauses), and integrated structure. Although this model receives strong support from structural considerations, because it groups all dependent clauses o n a single dimension, it is n o t well-supported functionally, because d e p e n d e n t clauses are distributed o n all six speech/writing dimensions in the 1988 analysis. T h e earlier analysis does indicate, however, a fundamental distinction between referential elaboration (Dimension N o . 3 in the 1988 analysis) and other types of elaboration. T h e second complexity m o d e l tested h e r e thus distinguishes between referential elaboration and other kinds of "framing" elabor ation (in addition to the dimensions of reduced structure and integrated structure). The other two models are five-dimensional. T h e first of these separates passive constructions (both m a i n clause and dependent clause) from the other four dimen sions, based o n t h e patterning of passive constructions as a separate dimension in the 1988 analysis (Dimension N o . 5). Finally, a second five-dimensional m o d e l is tested that divides "framing" elaboration into two groups on structural grounds: adverbial clauses versus c o m p l e m e n t clauses. This division is indicated in part by the 1988 analysis (e.g., that clauses as v e r b complements a n d as adjective complements are two of t h e m a i n features in D i m e n s i o n N o . 6), b u t it is based primarily on structural considerations. It is possible to p r o p o s e o t h e r m o d e l s o n structural g r o u n d s (e.g., disnnguishing participial clauses from other types of d e p e n d e n t clauses^. H n w » w «
DISCOURSE STUDIES
136
these additional m o d e l s are n o t indicated in the exploratory 1988 analysis, a n d there is n o other evidence that these groupings are functionally as well as structurally coherent; thus, they d o n o t h a v e sufficient theoretical basis for con sideration h e r e . Table 4 presents s u m m a r y statistics for all five of the models analyzed in the present study. I n the following sections, the relative strengths of each m o d e l will b e discussed in turn. 6
Table 4: Goodness-of-fit statistics for five models of discourse complexity Model M l . One dimension: Reduced versus elaborated and integrated structure Chi-square Ml
3435.0
Delta GFI*
(df= 495)
.594 .632
Chi-square difference
MO - Ml = 5 0 3 6 (df= 33, p< .001)
Model M3. Three dimensions: Reduced structure/specificity, Elaborated structure, and Integrated structure Chi-square M3 Delta GFI
3163.4
(<2/=492)
' '
.626 .670
Chi-square difference Ml - M3 = 272 (df= 3,p< .001) Improvement in Delta M3 - Ml =.032 (3.2%) Model M4. Four dimensions: Reduced structure/specificity, Structural elaboration of reference, Integrated structure, and "Framing" structural elaboration Chi-square M4 Delta GFI
3018.8
(#"=489)
.644 .691
Chi-square difference M3 - M4 = 144 (df= 3, p < .001) Improvement in Delta M4 - M3 =.018 (1.8%) Model M5a. Five dimensions: (a) Reduced structure/specificity, (b) Structural elaboration of reference, (c) Integrated structure, (d) Adverbial structural elaboration (causative, conditional, concessive, and "other" adverbial clauses', present and past participial adverbial clauses; sentence relative clauses), and (e) Complement clauses (WH clauses, THAT clauses as complements to verbs and adjectives, infinitives) Chi-square M5a 2989.4 Delta GFI
(#=485)
.647 .694
Chi-square difference M4 - M5a = 29 (df= 4, p< .001) Improvement in Delta M5a - M4 =.003 (0.3%) (Table 4 continued}
BIBER DISCOURSE COMPLEXITY
137
[Table 4 continued) Model M5b. Five dimensions: Reduced structure/specificity, Structural elaboration of reference, Integrated structure, "Framing" structural elaboration, and Passive constructions Chi-square M5b Delta GFI
2600.3
(df= 485)
.693 .730
Chi-square difference M4 - M5b = 419 {df= 4, p < .001) Improvement in Delta M5b - M4 = .049 (4.9%) Chi-square difference M5a - M5b = 389 (df= \,p< Improvement in Delta M5b - M5a = .046 (4.6%)
.001)
Chi-square difference Ml - M5b = 835 (df= 10, p < .001) Improvement in Delta M5b - M l = .099 (9.9%)
*, '
1
'" •
Correlations Among Factors: i Reduced Reduced Ref. Elab. - — Integrated Framing Elab. Passives
1.00 -0.41 -0.83 0.94 -0.69
Ref. Elab. 1.00 0.51 -0.22 0.39
Integrated
Framing Elab.
Passives
-
-
1.00 -0.58
1.00
1.00 -0.67 0.79
3 . M o d e l s of Surface-Structure D i s c o u r s e C o m p l e x i t y 3.1 A One-Dimensional Model: Reduced versus Elaborated Structure and Specificity The first model to be considered is a one-dimensional (1-D) model of discourse complexity (model M l ) . The exploratory analysis in Biber (1988) does not pro vide theoretical support for this model, but because it is the model often assumed in research studies, it is considered here. According to this model, discourse complexity is a unitary construct, and thus all linguistic features associated with discourse complexity should load onto a single factor. A summary of the goodness-of-fit of the 1-D model is presented in Table 4. The 1-D model is statistically significant in that it represents a significant im provement over the "null model," which is a model that specifies that there are no relationships among these linguistic features. (The chi-square difference score of 5036 shows this to be a significant model.) The model is also a theoretically meaningful one in that it accounts for a large portion of the observed correlations. This is indicated b y the Delta coefficient value of .594, which shows that 59.4% of the relationships among these linguistic features is accounted for by the 1-D model. The 1-D model is thus a significant and theoretically meaningful one: the following sections show, however, that it is less adermato A m
138
DISCOURSE STUDIES
3.2 A Three-Dimensional Model: Reduced Structure and Specificity, Elaborated Structure, and Integrated Structure T h e first alternative m o d e l to the 1-D m o d e l tests w h e t h e r the features marking structural elaboration a n d those m a r k i n g informational integration should be considered as separate constructs. T h e 1-D m o d e l is based o n the claim that these two groups of features represent a single underlying construct, discourse complexity, which is direcdy opposed to features m a r k i n g reductions in struc ture, a n d specificity. T h e m o d e l tested i n this section is based o n the claim that integrated structure, elaborated structure, a n d r e d u c e d structure represent three separate constructs or dimensions. Features such as that deletions, contractions, a n d the pro-forms do a n d it are specified as representing the r e d u c e d structure dimension. Features such as nouns, prepositions, attributive adjectives, a n d nominalizations reflect a h i g h informational focus a n d a relatively dense integration of information in a text; long words a n d diversified vocabulary (i.e., high type/ token ratio) reflect a careful, precise w o r d choice. Together, these features rep resent a dimension m a r k i n g integrated structure. Finally, all forms of struc tural dependency, including relative clauses, c o m p l e m e n t clauses, a n d adverbial s u b o r d i n a t e clauses, are specified as r e p r e s e n t i n g t h e e l a b o r a t e d structure dimension. I n one form of this m o d e l , we might hypothesize that these three dimensions are completely i n d e p e n d e n t (or "orthogonaP); according to this hypothesis, structural reduction, integration, a n d elaboration are unrelated i n their dis tributions a n d functions. It is possible to test this hypothesis b y specifying the intercorrelations a m o n g the factors to b e zero. T h e c o m p u t e d goodness-of-fit statistics for this model, however, show that it is less adequate than the 1-D model; the 3-D m o d e l with uncorrelated factors has a chi-square value of 4032.1, result ing in a Delta coefficient of .524 (52.4%), which represents a p o o r e r fit than the 1-D m o d e l (which h a d a Delta coefficient of 59.4%). A m o r e p l a u s i b l e m o d e l is t h e 3-D m o d e l a l l o w i n g intercorrelations a m o n g the factors, which represents the claim that integrated structure, elabor ated structure, a n d reduced structure are separate b u t interrelated dimensions. T h e statistical s u m m a r y of this m o d e l is given in Table 4 (Model M3). The chi-square value of 3163.4 (df = 492) represents a significant i m p r o v e m e n t over the 1-D m o d e l , a n d the Delta value of 62.6% represents a 3.2% increase over the 1-D m o d e l . T h e 3-D m o d e l with correlated factors thus provides a significandy better representation of the relationships a m o n g these linguistic features than the 1-D model.
1
3.3 A Four-Dimensional Model: Reduced Structure and Specificity, Structural Elaboration of Reference, Integrated Structure, and "Framing" Structural Elaboration
T h e dependent clauses included in this study can b e divided into several different subclasses. T h e exploratory analysis i n Biber (1988) indicated t h a t one of the basic divisions a m o n g d e p e n d e n t clauses is b e t w e e n postnominal modifying
BIBER DISCOURSE COMPLEXITY
139
clauses, which specify a n d elaborate the identity of referents, a n d w h a t I call here "framing* d e p e n d e n t clauses, which provide discourse frames of various types. Postnominal clauses include W H relative clauses, that relative clauses, and participial postnominal modifiers (clauses with "WHIZ-deletion"). All of these modify a n o u n h e a d , specifying t h e reference of that n o u n or providing elaborating details a b o u t that n o u n . T h e framing clauses are less closely related structurally; they include sentence relatives, W H clauses, that c o m p l e m e n t clauses, infinitives, a n d various types of adverbial subordination. T h e s e can b e considered "framing" in the sense that they either provide a discourse frame for a portion of text (as in the case of m a n y types of adverbial subordination; see, e.g., Ford & T h o m p s o n , 1986; T h o m p s o n , 1983, 1985), or they provide a n overt assessment of the speaker/writer's stance (in the case of sentence relatives, that c o m p l e m e n t c l a u s e s , a n d W H c l a u s e s ; see B e a m a n , 1 9 8 4 ; Q u i r k , Greenbaum, Leech, & Svartvik, 1985; Winter, 1982). Model M 4 in Table 4 presents the summary statistics for this four-dimensional model in which referential structural elaboration is separated from "framing" structural elaboration. T h e chi-square value of 3018.8 represents a statistically significant i m p r o v e m e n t over the 3-D m o d e l , a n d the Delta value of 64.4% represents a 1.8% i m p r o v e m e n t . Consideration of the loadings for t h e features in this m o d e l (not presented in Table 4) shows that the Framing Factor is relatively diverse. Four features h a v e strong loadings over .40: sentence relatives, W H clauses, conditional adverbial subordination, a n d causative adverbial subordination. H o w e v e r , t h e o t h e r features h a v e relatively low loadings, s o m e n e a r zero (that complements, infin itives, and concessive adverbial subordination), indicating that this dimension is less well-defined t h a n the others. 3.4 Five-Dimensional Models It was noted previously that there are a n u m b e r of ways that t h e d e p e n d e n t clauses included in the study can b e subgrouped, a n d given that t h e "framing" elaboration dimension is relatively diverse, alternative models with theoretical support are worth further investigation. O n e of these other groupings, separating out complement clauses as a separate factor, is indicated in the 1988 explora tory analysis and is thus tested h e r e . T h e s u m m a r y of M o d e l M 5 a in Table 4 shows, however, that this m o d e l represents a chi-square difference of only 29 over M o d e l M 4 ; this is a statistically significant i m p r o v e m e n t , b u t it cor responds to an i m p r o v e m e n t in Delta of only 0.3%. This m o d e l thus contributes little towards further understanding of d e p e n d e n t clauses. A second five-dimensional m o d e l s e p a r a t e s o u t passive c o n s t r u c t i o n s including agendess passives, Sy-passives, past participial postnorninal clauses, and past participial adverbial clauses) as a separate factor. This m o d e l receives strong support from the 1988 analysis, in which passive constructions constituted trie major features on o n e of the speech/writing dimensions. T h e s u m m a r y of Model M5b in Table 4 shows that it provides the best fit of all models considered here. It represents a significant i m p r o v e m e n t o v e r all nth**1
relatively strong improvement in Delta over other models - nearly a 5% improve m e n t over the 4-D m o d e l , a n d nearly a 10% i m p r o v e m e n t over the 1-D model. Table 4 also shows the intercorrelations a m o n g the factors in M o d e l M 5 b . These intercorrelations are moderately high. For the Passive factor, they range from .79 (62.4%) to .39 (15.2%), whereas those for the Reference factor range from .51 (26%) to - . 2 2 (only 4.8%). T h e Framing Elaboration factor shows an extremely high correlation of .94 (88%) with the R e d u c e d factor, indicating an extremely strong positive relationship between structural reduction and a type of structural elaboration. T h a t is, framing structural elaboration is strongly associated with these forms of reduced structure a n d specificity, indicating that these two groups of features are c o m m o n l y used in the same communicative situations. T h e factor loadings for this last five-dimensional m o d e l (Model M 5 b ) , given in Table 5, indicate that the factors are relatively well-defined. O n the Reduced Structure factor, all loadings are over .50; o n the Integrated Structure factor, all loadings except o n e are over .50; a n d o n the Passive factor, three of the four features h a v e l o a d i n g s o v e r .70. I n contrast, t h e Referential Elaboration factor shows some lower loadings, with that relative clauses having a loading near 0.0, a n d the Framing Elaboration factor is the least coherent in the analysis, with five features having loadings below .20. Overall, M o d e l M 5 b accounts for 6 9 . 3 % of the relationships a m o n g these linguistic markers of complexity (or 7 3 % according to the G F I estimate). This leaves approximately 30% of the variance unaccounted for. I n part, this remaining variance is due to the particular functions of individual features apart from their shared functions; a n d in part, this remaining variance could represent additional parameters of complexity n o t isolated here. I n particular, the Framing Elabor ation factor needs further investigation, to identify the salient subfactors rep resented there and additional linguistic features associated with those factors. M o d e l M 5 b , though, accounts for a large a m o u n t of the shared variance among these linguistic features a n d thus provides a good first m o d e l of the dimensions of discourse complexity with respect to these surface structure features in English. I n the following section, I consider the linguistic characteristics of different spoken a n d written registers with respect to these dimensions. 7
8
Table 5: Factor loadings for Model M5b, with five dimensions: reduced structure specificity, structural elaboration of reference, integrated structure, "framing" structural elaboration, and passive constructions
1. 2. 3. 4. 5.
That deletions Contractions Pro-verb Do Pronoun It Dem. pronouns
Reduced
Reference Elab.
Integration
Framing Elab.
.83 .92 .79 .75 .75
.0 .0 .0 .0 .0
.0 .0 .0 .0 .0
.0 .0 .0 .0 .0
Passive .0 .0 .0 .0
(Table 5 continutfi
OTHER TJISCOURSE COMPLEXITY (Table
141
5 continued) Reference Elab.
Reduced 6. 7. 8. 9. 10. 11. 12. 13. 14. IS. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 2«.
Clause coord. Nouns Prepositions Attributive adjs. Nominalizations Phrasal coord. Word length Type/token ratio Agentless passives .By-passives WH rels. - subj. pos. WH rels. - obj. pos. "Pied piping" rels That rels. - obj. pos. Thatxeh. - subj. pos.. Sentence relatives WH clauses That clauses - vbs. That clauses - adjs. Infinitives Cond. adv. subord. Caus. adv. subord. Concess. adv. sub. Other adv. subord. i0. Pres. part, postnom. tt. Past part, postnom. t2. Pres. part. adv. els. i3. Past. part. adv. els.
.56 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0
'.
.0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .45 .85 .87 .20 .04 .0 .0 .0 .0 .0 .0 .0 .0 .0 .30 .0 .0 .0
Integration
1
.0 .66 .84 .82 .76 .51 .96 .36 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0
Framing Elab.
Passives
.0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .54 .41 .17 -.03 .06 .42 .65 -.08 -.19 .0 .0 -.26 .0
.0 .0 .0 .0 .0 .0 .0 , .0 .76 .71 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .0 .77 .0 .42
4. The D i s c o u r s e C o m p l e x i t y o f S p o k e n a n d Written Registers Two conclusions resulting from the m o d e l testing carried out in Section 3 are: 1. Discourse complexity is n o t a single, unified construct; rather, the results here indicate five separate b u t interrelated dimensions of complexity based on surface structure features, a n d consideration of linguistic features asso ciated with information packaging, cohesion, a n d rhetorical organization will likely reveal additional dimensions. 2. Given this multidimensional nature of discourse complexity, a complete description of a text (or register) cannot focus exclusively o n the degree of i complexity; rather, the complexity of a text is described h e r e within a fiveI dimensional space, so that texts are complex in different ways as well as to I differing extents.
DISCOURSE STUDIES
142
I n the present section, I use this five-dimensional m o d e l to provide a pre liminary description of the discourse complexity of spoken a n d written registers. To d o this, a dimension score for each complexity dimension was c o m p u t e d for each text, b y s u m m i n g the frequencies of the linguistic features on that dimension. For e x a m p l e , the dimension score for the R e d u c e d Structure factor is computed b y s u m m i n g the frequencies of $a/-deletions, contractions, pro-verb do, p r o n o u n it, demonstrative p r o n o u n s , a n d clause coordination for each text. Before the dimension scores are computed, the individual frequency counts are standardized to a m e a n score of 0.0 a n d a standard deviation of 1.0, so that all counts have comparable scales. T h e dimension scores are used for relative comparisons a m o n g texts (rather than for absolute characterizations). Individual texts can be c o m p a r e d directly in terms of their dimension scores, a n d registers can b e com p a r e d in terms of their m e a n dimension scores. Table 6 presents statistics from a General Linear Models p r o c e d u r e for each of the five dimension scores as predictors of the differences a m o n g 14 spoken a n d written registers. T h e F scores show that there are significant differences a m o n g the m e a n scores of these registers with respect to all five dimensions, and the r scores s h o w that all five dimensions are strong predictors of the register differences. T h r e e of the dimensions - R e d u c e d Structure, Integrated Structure, a n d Passive Constructions - are extremely strong predictors, having f values ranging from 59.5% to 83.8%. T h e other two dimensions - Elaboration of Refer ence a n d F r a m i n g Elaboration - are weaker b u t still relatively strong predictors, with r values of 36.6% a n d 32.1%. Figures 1 to 5 present plots of the dimension scores of these 14 spoken and written registers. T h e figures can contribute to the interpretation of the factor analysis b y identifying t h e situations associated with m a r k e d l y high or low occurrence of each type of discourse complexity. T h a t is, although the fac tors can b e interpreted directly b a s e d o n previous research, those interpret ations can b e extended b y analysis of the distribution of each complexity type across texts. F r o m a quick overview of Figures 1 to 5, it is a p p a r e n t that each of these five types of complexity is distributed differentiy a m o n g registers, confirming the claim that each dimension is responsive to different functional considerations. Figure 1 shows that the R e d u c e d Structure D i m e n s i o n is strongly associated 9
2
2
Table 6: Statistical significance and importance among the mean scores of 14 spoken and written registers with respect to the five complexity dimensions Dimension Reduced structure Elaboration of reference Integrated structure Framing elaboration Passive constructions
77(13, 337) 128.2 14.3 71.2 11.8 36.4
P< .0001 .0001 .0001 .0001 .0001
r
3
83.8*. 36.6'K74.2* 32.1* 59.5%
BIBER DISCOURSE COMPLEXITY
143
Official docs. Academic prose Editorials Prof, letters Press reportage Biographies General fiction Romance fiction Personal letters Prep, speeches Spon. speeches Broadcasts Interviews Conversations -r
- 4 - 2 0 Dimension Score Figure 1: Inverse reduced structure dimension scores for 14 registers
with the m o d e distinction between speech and writing - all spoken registers have negative scores o n this dimension, showing a greater than average frequency of reduced forms, whereas nearly all written registers have positive scores, reflecting a less-than-average frequency of reduced forms. (The poles of this dimension have been reversed, so that negative scores represent high frequencies of reduced features a n d positive scores represent a m a r k e d absence of r e d u c e d features.) However, it would n o t b e accurate to characterize this dimension as a simple distinction between speech a n d writing. First of all, two written registers h a v e negative scores: a v e r y small negative score in the case of r o m a n c e fiction (cf. the small positive score for general fiction), and a large negative score in the case of personal letters. I n addition, there is a considerable range of values among the negative scores of the spoken registers: from nearly - 1 0 for faceto-face conversations, d o w n to approximately - 1 for p r e p a r e d speeches a n d broadcasts. In fact, the distribution of scores along this dimension m i g h t b e better explained in terms of the interaction between purpose of communication a n d opportunity for careful production. Conversations are the least formal and least informational register included here, and thus they permit the most reduction in form. I n addition, they are p r o d u c e d on-line without prior planning, m a k i n g it relatively difficult to use fuller forms. Text Sample 1 illustrates r e d u c e d forms in a conversational text; these forms (marked in bold) include n u m e r o u s con tractions, as well as deletion of that in c o m p l e m e n t clauses (e.g., Iknow
one can't. . .), a n d frequent use of pronominal it and that (e.g., that is a great deprivation). (This text portion also illustrates the use of Framing Elaboration features, discussed later.)
DISCOURSE STUDIES
144
Text Sample 1: Conversation (Reduced Structure features are in boldface, Framing Elaboration features are capital ized and italicized, and < 0 > indicates deleted that in complement clauses. A: you're rushing off # — B: I was thinking # B: THAT IF I go*. B: I might get into Salisbury's BEFORE they close #. B: our housekeeper's left # A: how absolutely awful B: so . I mean < 0 > that is a — a great deprivation # A: well it is # A: i s n ' t i t # - B: she's got a good office job # B: good money # B: but. I know < 0 > one can't suppose # A: just saying to mine this morning # A: I don't know WHAT I'd do IF she left # A: she will of course # A: BECAUSE she said WJSiVshe came # A: THAT WHEN the children are old enough # A: she would get a driving job # . Interviews and spontaneous speeches are similarly influenced by on-line production, but they have a much more informational purpose and thus show a less frequent use of reduced forms. Planned speeches are less influenced by the production circumstances due to the effect of prior planning, and so they show relatively few reduced forms. (Because there is considerable ellipsis of whole words and phrases in broadcasts, they are actually more reduced in form than their score o f - 1 indicates.) The written registers are fairly uniform in their infrequent use of reduced forms, in line with their informational purposes and extensive opportunity for production. However, their scores do range along a continuum: official docu ments and academic prose with the largest positive scores, corresponding to their extreme informational focus and formality; biographies, press reportage, professional letters, and editorials with relatively large positive scores, cor responding to their relatively informational and formal purposes; general and romance fiction with scores near zero, reflecting their aesthetic purpose and their mixing of narrative with fictional dialogue; and finally, personal letters with a large negative score, reflecting their interpersonal purpose and a relative lack of planning and revision during production. In contrast, the distribution of scores with respect to the Structural Elaboration of Reference Dimension, shown in Figure 2, appears to be completely unrelated to m o d e differences (between speech and writing) or to differences in production circumstances. Rather, this type of complexity seems to be stricdy related to purpose: informational versus noninformational. Informational purposes re quire a frequent use of this type of elaboration, whether in speech (interviews,!
BIBER
DISCOURSE COMPLEXITY
145
Official docs. Academic prose Editorials Prof, letters Press reportage Biographies General fiction Romance fiction Personal letters Prep, speeches Spon. speeches Broadcasts Interviews Conversations -2
-1
0
1 2 3 4 Dimension Score
5
6
7
8
Figure 2: Referential elaboration dimension scores for 14 registers spontaneous speeches, p r e p a r e d speeches) or in writing (official documents, academic prose, editorials, professional letters, press reportage, biographies). Noninformational purposes, o n the other h a n d , show a m a r k e d l y infrequent use of these forms, again, whether in speech (conversations and broadcasts) or in writing (fiction a n d personal letters). T h e frequent use of these features in informational registers is illustrated in Text Samples 2 a n d 3 , taken from a (written) official document a n d a (spoken) interview. Text Sample 2: Official Document (Referential Elaboration Dimension features are in boldface, Integrated Structure Dimension features are italicized, and Passive Constructions are capitalized.) Questions about marriage and children WERE again INCLUDED, as they had BEEN at the 1911 and 1951 censuses.... The 756/ census questions related to all women who WERE or had BEEN MARRIED, and so repeated the enquiry MADEfifteenyears earlier NY the 1946family census CONDUCTED on behalf ofHie royal commission on population. The questions about children... extended to all women who WERE or had BEEN MARRIED. Text Sample 3: Interview (Referential Elaboration Dimension features are in boldface, and Framing Elaboration features are capitalized and italicized.) Speaker A: for instance WHEN we're looking # at the legislative framework within which marketing. operates # it will be fairly important # to - - give # . 1
•
DISCOURSE STUDIES
our - students # a full understanding # . of the - different legal systems # . operating in different. countries # - which provide the environment for # exporting # — i or importing indeed # i SINCE a number of our students who comefromabroad # may be engaged # in importing . activities # into their - . - native countries # — again # WHENone # is looking at the way in which one the techniques # . by which one - analyzes # . markets # behaviours # and social situations # - we would - hope # THATOUT students would have a full understanding # of the cultural differences # which - will help t h e m . . . .
'
i
•: • i
'
Text Sample 2 illustrates t h e c o m m o n use of relative clause constructions for referential specificity i n official documents (e.g., all women who were or had been married). Relative constructions are used even m o r e frequently in the portion of an interview given in Sample 3. I n this case, these constructions are used both to specify t h e identity of referents a n d to provide elaborating details regarding a referent (illustrated in turn b y the following two relative clauses: the - different legal systems operating in different. countries # — which provide the environment for# - exporting). T h e distribution of scores with respect to the Integrated Structure Dimension, given in Figure 3 , d e p e n d s o n a n interaction b e t w e e n communicative purpose and m o d e , distinguishing informational written registers from all other registers. T h u s the informational written registers all show a frequent use of integrative features; these registers include the expository registers (press, official documents, and academic prose) as well as informational narrative (biographies) and infor mational written interaction (professional letters). T h e extremely dense use of these features i n written informational registers is illustrated b y Text Sample 2 (from a n official document). T h e s e features include n o u n s , often in noun-noun (-noun) sequences (e.g., census questions, family census), attributive adjectives, and prepositional phrases (e.g., on behalf of the royal commission on population). All other registers show a n infrequent use of this type of complexity; thest include all spoken registers, as well as fictional narrative a n d personal written interaction (personal letters). With regard to t h e spoken informational register(e.g., speeches and interviews), production constraints apparentiy limit the extern —i-.-_u . „ f „ , - „ h carefully integrated; a n d with respect to the m
a
t
l
n
r
n
n
f i
BIBER
147
DISCOURSE COMPLEXITY
Official docs. Academic prose Editorials Prof, letters Press reportage Biographies General fiction Romance fiction Personal letters Prep.speeches Spon. speeches Broadcasts Interviews Conversations -t—
1
'
'I
=
-2 0 2 Dimension Score Figure 3: Integrated structure dimension scores for 14 registers
noninformational written registers (e.g., fiction a n d personal letters), the p u r p o s e does not require a dense integration of information. T h e "Framing" Structural Elaboration D i m e n s i o n (Figure 4) seems to b e related primarily to communicative purpose, with little relation to m o d e or pro duction differences. T h e s e features are relatively frequent in those registers that promote the expression of personal attitudes, justifications, a n d feelings: inter views, s p o n t a n e o u s speeches, conversations, professional letters, p r e p a r e d speeches, a n d personal letters. T h e y are m u c h less frequent in those registers having informational, impersonal, "factual" purposes, such as press reportage, biographies, a n d academic prose; a n d they are m a r k e d l y infrequent in official documents, which are typically direct statements of "fact" with n o acknow ledged author. T h e s e features are also m a r k e d l y absent in broadcasts, which are radio reportage of events in progress (sports events, parades, etc.). T h e s e texts were recorded in t h e 1960s, preceding the current practice of interactive, per sonal reportage of events; these broadcasts are thus similar to official documents in being direct factual reportage with n o acknowledged speaker and little per sonal commentary. Text Sample 1 illustrates the use of framing elaboration in conversation, •Sample 3 illustrates these features in an interview, a n d Sample 4 illustrates these features in a professional letter. These samples differ in m o d e and in informational purpose (conversations b e i n g spoken and noninformational, interviews b e i n g spoken but informational, professional letters b e i n g written a n d informational), but they share a focus o n the expression of personal attitudes and opinions with justification for these positions. Several of the framing functions of these elabor ation features are illustrated in these text samples. For e x a m p l e , W H clauses and conditional clauses are used to contrast various n n s s i h l o *nHr.™ ~ r
DISCOURSE STUDIES
148
Official docs. Academic prose J Editorials Prof, letters Press reportage Biographies General fiction Romance fiction Personal letters Prep, speeches Spon. speeches Broadcasts Interviews Conversations T 4 6 Dimension Score Figure 4: Framing elaboration dimension scores for 14 registers -1
12
view (e.g., I was thinking THAT IF I go I might get into Sainsbury's before they close and I don't know what I'd do IF she left, from the conversation sample; for instance WHEN we're looking at the - legislative framework, from the interview). Causative adverbial clauses are used to justify certain attitudes or opinions, as in the fol lowing example from the interview sample, w h e r e the speaker justifies a selfcorrection: a full understanding of... legal systems <provides> the environment for exporting --or importing indeed - SINCE a number ofour students... may be engaged in importing activities. Similarly in the professional letter sample, causative clauses are used to explain the opinion that: any such drastic change would ultimately require the action of the board of directors BECAUSE if would invoke a change in the constitution... and BECAUSE any such change would in turn require a vote. That complement clauses are used to overtly frame a n attitude or position relative to a "stance" verb or adjective (e.g., we would hope THAT our students would have a full understanding, from the interview sample; and I understand THAT certain XYZ populations are... impacted, from the professional letter). T h u s , although these elaboration features vary in their particular characteristic functions, they share general "framing" uses c o m m o n in m o r e personal, attitudinal registers. 10
11
Text Sample 4: Professional Letter (Framing Elaboration features are capitalized and italicized.) . . . Any such drastic change would ultimately require the action of the board of directors BECAUSEit would involve a change in the constitution... and BECAUSE any such change would in turn require a vote . . . /Fit is not possible TO add your concern this year, it would certainly be possible TO add it next year . . . Please understand THAT WHILE I am sympathetic to what you are trying to achieve, and THAT WHILE I understand THAT certain XYZ populations are . . . impacted..., T a m nnt at n r p s p n r p n t i r e l v i n s v m D a t h y . . .
BIBER DISCOURSE COMPLEXITY
149
Finally, the Passive Constructions Dimension, plotted in Figure 5, seems to reflect quite n a r r o w functional considerations, b e i n g frequent only in written informational exposition (official documents, academic prose, a n d to a lesser extent press reportage and editorials). T h e s e features are m a r k e d l y infrequent in all spoken registers, all narrative registers (fiction a n d biography), and all written interactional registers (personal a n d professional letters). Passive features differ from the Integrated features in that they are found only in written exposition rather t h a n in all written informational registers (i.e., including biographies a n d professional letters). This dimension thus seems to b e related specifically to a n expository organization of information. Text Sample 2 illustrates the frequent use of passive constructions in official documents. Thus, each of these complexity dimensions comprises a distinct set of lin guistic features having particular grammatical a n d discourse functions; each dimension defines a u n i q u e set of relations a m o n g spoken a n d written registers; and each dimension is responsive to different aspects of the communicative situation, including m o d e , production circumstances, and purpose. As previously noted, these dimensions define a five-dimensional space of discourse complexity, and the complexity of particular registers can b e analyzed within this space with respect to b o t h the types a n d extent of complexity. Figures 6 a n d 7 plot the five-dimensional complexity profile of 11 registers: six written registers in Figure 6 and five spoken registers in Figure 7. These figures show that registers differ m a r k e d l y in their complexity characterizations, a n d that n o register is consistendy c o m p l e x with respect to all dimensions. Official documents a n d professional letters (on Figure 6) can b e considered the two most consistendy complex registers, but official documents are markedly noncomplex with respect to the Framing Elaboration Dimension, a n d professional letters are markedly n o n c o m p l e x with respect to the Passives Dimension.
Official docs. Academic prose Editorials Prof, letters Press reportage Biographies General fiction Romance fiction Personal letters Prep, speeches Spon. speeches Broadcasts Interviews Conversations i
1
i
0 1 2 3 Dimension Score Figure 5: Passive constructions dimension scores for 14 registers
150
DISCOURSE STUDIES
10
Not Reduced RefElab
Integrated
Framing
Passives
—•— Official Documents —a— Prof. Letters —•— General Fiction — P e r s o n a l Latters —Biographies —O— Press Reportage Figure 6: Complexity dimension scores for six written registers 15
1
-10? , Not Reduced RefElab
! Integrated
! Framing
1
1
Passives
—•—Conversations —*—Interviews —•—Prepared Speeches —Q— Spon. Speeches —6— Broadcasts Figure 7: Complexity dimension scores for five spoken registers I n addition, comparison of Figures 6 a n d 7 indicates a fundamental distinction between the discourse complexities of written a n d spoken registers: Written registers differ widely a m o n g themselves in both the extent a n d kinds of discourse complexity, whereas spoken registers follow a single pattern with respect to their kinds of complexity, differing only with respect to extent. T h u s , Figure 6 shows that written registers use m u c h of the available complexity space defined b y these five dimensions. Four basic profiles can b e distinguished among the written registers: 1. N o n c o m p l e x with respect to the dimensions of R e d u c e d Structure, Refer ential Elaboration, Integrated Structure, and Passives, b u t markedly complex
BIBER
DISCOURSE COMPLEXITY
151
with respect to Framing Elaboration - this profile characterizes personal letters a n d general fiction, with the only difference between t h e m b e i n g that fiction is u n m a r k e d (rather t h a n m a r k e d l y noncomplex) with respect to R e d u c e d Structure a n d Referential Elaboration; 2. Moderately c o m p l e x with respect to the first four dimensions (i.e., r e d u c e d structure, referential elaboration, integrated structure, a n d framing elabor ation) and n o n c o m p l e x with respect to the use of passive constructions this profile characterizes press reportage and biographies, which are very similar to each other in their kinds a n d extent of complexity; 3. Moderately c o m p l e x in avoiding reduced structures, highly complex with respect to referential elaboration, integrated structure, a n d framing elabor ation, a n d n o n c o m p l e x with respect to the use of passive constructions this profile characterizes professional letters; a n d 4. Highly complex with respect to r e d u c e d structures (i.e., avoiding them), referential elaboration, integrated structure, a n d passive constructions, a n d markedly n o n c o m p l e x with respect to the use of framing elaboration - this profile characterizes official documents. In contrast, spoken registers follow a single complexity profile o n Figure 7: noncomplex in permitting the use of r e d u c e d structures, moderately complex in the use of referential elaboration, n o n c o m p l e x in the use of integrated struc tures, highly complex with respect to the use of framing elaboration, a n d noncomplex with respect to the use of passive constructions. (The only departure from this profile is shown b y the broadcasts register, which is u n m a r k e d o n all dimensions.) These two figures thus indicate that there is a fundamental difference in discourse complexity associated with the production differences between speech and writing. Written p r o d u c t i o n gives the addressor m a x i m u m freedom to manipulate the linguistic characteristics of a text in accordance with a n u m b e r of situational parameters, including purpose, interactiveness, and degrees of per sonal involvement. I n spoken production, though, the on-line production circum stances apparently restrict the kinds of discourse complexity possible. T h u s , Figure 7 shows differences in the extent of complexity associated with situational parameters such as purpose a n d involvement, b u t only a single basic profile of the kinds of complexity associated with the spoken m o d e . Overall, the complexity characterizations of registers differ in three respects: (a) the particular dimensions m a r k e d as complex or n o t for each register, (b) the extent to which a register is m a r k e d as complex on each dimension, a n d (c) the number of dimensions o n which a register is m a r k e d as complex. Both the ex tent of complexity a n d the particular type of complexities for a given register are determined b y the situational characteristics of that register. However, as previously shown, the exploitation of complexity features is apparently restricted by mode, so that written registers differ greatly a m o n g themselves in b o t h the kinds and extents of discourse complexity, whereas spoken registers are mostly restricted to differences of extent.
152
DISCOURSE STUDIES
5. S u m m a r y a n d C o n c l u s i o n O n the basis of a n analysis of m a n y kinds of surface complexity features across m a n y kinds of spoken a n d written texts, the present study has identified a fivedimensional m o d e l as the m o s t adequate for these data, a n d it has described the complexity characteristics of spoken a n d written registers with respect to that m o d e l . T h e five dimensions reflect the influence of a n u m b e r of situational par ameters, including production circumstances, purpose, a n d involvement. N o n e of the dimensions is associated with an absolute distinction between spoken and written texts; rather, m o d e interacts with other situational characteristics (such as a n informational purpose) to determine the relations a m o n g registers with respect to these dimensions. H o w e v e r , w h e n all five d i m e n s i o n s are considered together, they identify a fundamental distinction between written a n d spoken registers: Spoken registers are apparently limited i n the kinds of complexity they can exploit, whereas written registers show m u c h greater dif ferences a m o n g themselves with respect to b o t h their kinds a n d extents of discourse complexity. T h e study raises several issues requiring further investigation. First of all, the study is based only o n surface structure features of English. Inclusion of features relating to information packaging, cohesion, a n d rhetorical organization would enable fuller m o d e l s (cf. Biber, in press). Second, some of the dimensions require further analysis even with respect to surface structure features. I n par ticular, the "Framing" Elaboration Dimension should b e analyzed further, to better u n d e r s t a n d the discourse functions underlying this dimension, and to identify a n y additional dimensions inadvertently collapsed with this one. Third, the situational interpretations in Section 4 are preliminary: m o r e detailed analyses of these features in particular texts will enable a richer interpretation of the situ ational a n d functional correlates of each dimension. Finally, the analysis here is based only o n the distribution of linguistic features in English texts; further research is required to u n c o v e r the dimensions of complexity in other Western a n d non-Western languages. T h e present analysis provides a n e w framework for the analysis of discourse complexity a n d identifies five important dimensions of complexity in English; additional research is n e e d e d to e x p a n d this analysis to other discourse systems a n d languages.
H
Author's N o t e
I Would like to thank E. Finegan, J. B. Carroll and an anonymous reviewer for Discount Processes for their helpful criticisms on earlier drafts of this article.
Notes 1. Some previous register-based studies have proposed multiple parameters of variation that include complexity features on different parameters. Chafe (1982) proposes two
BIBER
DISCOURSE COMPLEXITY
153
parameters (fragmentation versus integration, and involvement versus detachment) on functional grounds: Carroll (1960) and Marckworth and Baker (1974) use multivariate statistical techniques to analyze the parameters of variation among written registers. 2. For example, see the analysis of the "stance" System (including evidentiality and affect) in Biber and Finegan (1989). 3. As shown in Biber (1990), however, the dimensions resulting from the exploratory factor analysis of Biber (1988) do seem to be quite stable and reliable (in that they are replicable in various subsets of the text corpus). 4. The linguistic feature counts were carried out by computer programs, described in Biber (1988, especially Appendix II). The analyses in Biber (1990) indicate that the feature counts are highly reliable across text samples from within texts, and across texts within registers. Apart from extremely common features, such as nouns, verbs, and prepositional phrases, the features have positively skewed distributions. These distributions reflect the fact that these are functional/stylistic markers that occur relatively rarely (or not at all) in the majority of texts but with relatively high fre quencies in particular registers. A visual inspection of scatterplots indicated that these features tend to have linear (rather than curvilinear) relationships with one another. 5. These analyses are undertaken within the general framework of covariance structure analysis, which provides a set of techniques for theory testing using correlational data (see, e.g., Hayduk, 1987). 6. The models tested here are thus based on theoretical considerations: a one-dimensional model commonly assumed in other studies, and four multidimensional models indi cated by Biber (1988). Modification indices were not used in the analysis here. For each factor in each model, the functional/structural grouping of features indicated by previous research was specified as loading on that factor with "free" loadings (i.e., loadings that were estimated by LISREL), and all other features were fixed as having a loading of 0.0. 7. Correlations over .9 can be taken to indicate colinearity (Hayduk, 1987, p. 176), which is especially problematic in regression analyses (when two predictors are highly intercorrelated). Thus, the high correlation between the Reduced Structure Dimension and the Framing Elaboration Dimension indicates that these should not be used as separate predictors in a multiple-regression analysis; rather, a higher-order factor should be constructed out of the two. In the present study, though, I keep these two factors separate, because they are not used together as predictors in any analysis, and because they are clearly distinct in their structural (i.e., linguistic) characteristics. 8. This finding replicates the different distributions of WH relatives and that rela tives found in Biber (1988). Ed Finegan (personal communication, 1990) has pointed out two functional differences between these relative clause types. First, whereas WH relative clauses can be either restrictive or nonrestrictive, that relative clauses are consistently restrictive. Second, there is an apparent tendency for which to be used in preference to that for plural (versus singular) antecedents. In addition, the use of that entails some loss of information relative to the more specific WH relative pronouns. The alternation between these forms in present-day English requires further empir ical investigation. !* For example, nouns have a mean frequency of 180.5 occurrences per 1,000 words (in the entire corpus) and nominalizations have a mean frequency of only 19.9 occurrences per 1,000 words. If these raw frequencies were used in computing the dimension score for the Integrated Structure factor, the frequency of nouns would carry much greater weight than nominalizations. Standardized scores put all frequencies on a single scale.
t
154
DISCOURSE STUDIES
so that all features on a factor cany the same weigth in the dimension score. For example, if a text had 216 nouns and 34 nominalizations, it would have a standardized score of approximately 1.0 for both features, because the standard deviation for nouns is 35.6, and for nominalizations it is 14.4. That is, for nouns: (216 - 180.5) / 35.6 = .997 and for nominalizations: (34 - 19.9) / 14.4 = .979. Thus, this text is similarly marked for both nouns and nominalizations, having one standard deviation over the mean for both features, and the standardized scores re flect this fact. 10. Stance marking in English, including evidentiality and affect, is further discussed in Biber and Finegan (1989). 11. Given this interpretation, it is surprising that editorials, which are personal and opinionated, do not have a higher score on this dimension. Their intermediate char acterization might be due to their relatively "protected" status. That is, because most readers of editorials cannot respond, there is less risk that editorial writers will lose face due to a direct challenge and, thus, the writers have a lesser need to overtly frame and justify opinions (in contrast to the direcdy interactional, and potentially confrontational, nature of letters, conversations interviews, and even speeches).
i ,
References
.
•
Beaman, K. (1984). Coordination and subordination revisited: Syntactic complexity in spoken and written narrative discourse. In D. Tannen (Ed.), Coherence in spoken and written discourse. Norwood, NJ: Ablex. Bender, P. M., & Bonett, D. G. (1980). Significance tests and goodness-of-fitin the analysis of covariance structures. Psychological Bulletin, 88, 588-606. Bernstein, B. (1962). Social class, linguistic code and grammatical elements. Language and Speech, 5,221-240. Bernstein, B. (1970). Class, codes and control, Vol. 1. Theoretical studies towards a sociology of language. London: Routledge & Kegan Paul. Biber, D. (1986). Spoken and written textual dimensions in English: Resolving the con tradictory findings. Language, 62, 384-414. Biber, D. (1988). Variation across speech and writing. Cambridge: Cambridge University Press. Biber, D. (1990). Methodological issues regarding corpus-based analyses of linguistic variation. Literary and Linguistic Computing, 5,257-269. Biber, D. (in press). Using computer-based text corpora to analyze the referential strategies of spoken and written texts, inj. Svartvik (Ed.), Directions in corpus linguistics: Proceedings of Nobel Symposium 82, Stockholm, 4-8 August 1991. Mouton. Biber, D. & Finegan, E. (1989). Styles of stance in English: Lexical and grammatical marking of evidentiality and affect. Text, 9, 93-124. Carroll, J. B. (1960). Vectors of prose style. In T. A. Sebeok (Ed.), Style in language. Cambridge, MA: MIT Press. Chafe, W. (1982). Integration and involvement in speaking, writing, and oral literature. In D. Tannen (Ed.), Spoken and written language: Exploring orality and literacy. Norwood, NJ: Ablex. , Davison, A., & Green, G. M. (Eds.), (1988). Linguistic complexity and text comprehension^ Readability issues reconsidered Hillsdale, NJ: Erlbaum. J
BIBER DISCOURSE COMPLEXITY
155
de Beaugrande, R., & Dressier, W. (1981), Introduction to text linguistics. London: Longman. De Vito, J. A. (1967). Levels of abstraction in spoken and written language. Journal of Communication, 17, 354-361. Ditmar, N. (1976). A critical survey of sociolinguistics: Theory and application. New York: St. Martin's Press. Finegan, E., & Biber, D. (1986). Two dimensions of linguistic complexity in English. In J. Connor-JJnton, C.J. Hall, & M. McGinnis (Eds.), Social and cognitive perspectives on language (Southern California Occasional Papers in Linguistics II), Los Angeles: University of Southern California, Department of Linguistics. Ford, C. E., & Thompson, S. (1986). Conditionals in discourse: A text-based study from English. In E. C. Traugott, A. T. Meulen, J. S. Reilly, & C. A. Ferguson (Eds.), On conditionals. Cambridge: Cambridge University Press. Frazier, I. (1988). The study of linguistic complexity. In A. Davison & G. M. Green (Eds.), Linguistic complexity and text comprehension: Readability issues reconsidered. Hillsdale, NJ: Erlbaum. Halliday, M. A. K. (1979). Differences between spoken and written language: Some implications for literacy teaching. In G. Page, J. Elkins, & B. O'Connor (Eds.), Com munication through reading: Proceedings of the 4th Australian Reading Conference (Vol. 2). Adelaide: Australian Reading Association. Hayduk, I. A. (1987). Structural equation modeling with USREL: Essentials and advances. Baltimore, MD: Johns Hopkins University Press. Joreskog, K., & Sorbom, D. (1984). LISREL, VI: Analysis of linear structural Relationships by maximum likelihood, instrumental variables, and least squares methods, Mooresville, IN: Scientific Software. Klare, G. R. (1984) Readability. In D. Pearson (Ed.), Handbook of reading research New York: Longman. Long, J. S. (1983). Confirmatory factor analysis: A preface to LISREL. Beverly Hills, CA: Sage. Marckworth, M. L., & Baker, W.J. (1974). A discriminant function analysis of co-variation of a number of syntactic devices in five prose genres. Americanjournal of Computational Linguistics. (Microfiche 11). O'Donnell, R. C. (1974). Syntactic differences between speech and writing. American Speech, 49, 102-110. Olson, D. R. (1977). From utterance to text: The bias of language in speech and writing. Harvard Educational Review, 47, 257-281. Poole, M. E. (1973). A comparison of the factorial structure of written coding patterns for a middle-class and a working class group. Language and Speech, 16, 93-109. Poole, M. E., & Field, T. W. (1976). A comparison of oral and written code elaboration. Language and Speech, 19, 305-311. Quirk, R., Greenbaum, S., Leech, G., & Svartvik, J. (1985). A comprehensive grammar of the English language. London: Longman. Smith, C. S. (1988). Factors of linguistic complexity and performance. In A. Davison & G. M. Green (Eds.), Linguistic complexity and text comprehension: Readability issues re considered. Hillsdale, NJ: Erlbaum. lannen, D. (1982). Oral and literate strategies in spoken and written narratives. Language, 58 1-21. Thompson, S. (1983). Grammar and discourse: The English detached participial clause. In F. Klein-Andreu (Ed.), Discourse perspectives on syntax. New York: Academic, fhompson, S. (1985). Grammar and written discourse: Initial versus final purpose clauses in English. Text, 5, 55-84.
DISCOURSE STUDIES
156
van den Broeck, J. (1977). Class differences in syntactic complexity in the Flemish town of Maaseik. Language in Society, 6, 149-181. Winter, E. (1982). Towards a contextual grammar ofEnglish: The clause and its place in the definition ofsentence. London: George Allen & Unwin.
Appendix The observed and predicted correlation matrices are compared using a chi-square goodness-of-fit statistic (generated by LISREL). A small chi-square value indicates that the predicted correlation matrix is not statistically different from the observed matrix, and that the model thus provides a very good fit to the data. In practice, however, the chi-square goodness-of-fit test does not by itself provide an adequate evaluation of a model, since the chi-square statistic is greatly influenced by sample size. Thus, practically any model based on a large sample will be rejected as inadequate (% < .05, indicating that the predicted correlation matrix is significantly different from the observed matrix), although it is possible to find numerous statistically acceptable models based on very small samples. For this reason, Bentler and Bonett (1980) developed a technique for comparing the relative adequacy of various models, rather than focusing on outright rejection or acceptance of a model. The Bentler and Bonett technique uses a "null model" (referred to here as M 0 ) as a baseline for evaluating the adequacy of particular models and the degree of improve ment between models. In a factor analysis, the null model specifies that the observed variables are mutually independent; thus, the model has no common factors, the unique variances are equated with the variances of the observed variables, and all common factor loadings are set to zero. In the present analysis, the null model (M0) has a chisquare value of 8471.3 for 528 degrees of freedom. Theoretically based models are evaluated positively to the extent that their chi-square goodness-of-fit statistic is sig nificantly smaller than this chi-square value for the null model. In particular, if the difference in chi-square values between the null model and a theory-based model is significant, then that model represents a statistically significant improvement over the null model. (This is because the difference score between chi-square values also has a chi-square distribution.) For example, Table 4 shows that the factor analysis for a onedimensional model of complexity (Ml) represents a statistically significant improvement over the null model: 2
i 2
2
8471 (the MO x ) - 3435 (the Ml 3C) = 5036, where the degrees of freedom is 528 - 495 = 33. (This is significant a t ^ < .001.) The null model can also be used as a baseline to asses the theoretical importance of a given model, by determining the percentage of the observed correlations accounted for by the model. This is indicated by the Delta coefficient (Bender & Bonett, 1980, p. 599), which represents the difference in chi-square values between the null model and a theory-based model as a proportion of the total chi-square value for the null model. For example, for the one-dimensional model of complexity (Table 4), A <= (8471 - 3435) / 8471 = .594.
I
IBER
DISCOURSE COMPLEXITY
157
The Delta coefficient of .594 shows that 59.4% of the relationships among these linguistic features is accounted for by the 1-D model. The Goodness-of-Fit Index is a slightly more liberal measure of strength produced by LISREL (.632 in this case). Both of these coefficients indicate that the 1-D model is relatively strong and represents an important theoretical improvement over the null model.
7 The Dialogic Unconscious: Psychoanalysis, Discursive Psychology and the Nature of Repression Michael Billig
T
his p a p e r will discuss the possibility of m a k i n g links between discursive psychology a n d psychoanalytic theory. At first sight, these two ways of understanding psychological p h e n o m e n a seem utterly incompatible. Dis cursive psychology, which is being developed as a challenge to o r t h o d o x social psychology, discourages speculation about 'inner' processes (e.g. Antaki, 1994; Billig, 1997a; Edwards, 1991,1994; Edwards & Potter, 1 9 9 2 , 1 9 9 3 ; H a r r e , 1995; H a r r e & Gillett, 1994; Parker, 1992; Potter, 1996; Potter & Wetherell, 1987,1995). Discursive psychology takes inspiration from the philosophical tradition of Wittgenstein's later philosophy and from the development of ethnomethodology and conversation analysis. These traditions of analysis stress the n e e d to examine in detail the outward accomplishment of social life, showing h o w social order is r e p r o d u c e d through discursive interaction. Discursive psychology applies this project to psychological p h e n o m e n a . It argues that p h e n o m e n a , which traditional psychological theories h a v e treated as 'inner processes', are, in fact, constituted through social, discursive activity. Accordingly, discursive psychologists argue that psychology should b e based o n the study of this outward activity rather than u p o n hypothetical, a n d essentially unobservable, inner states. I n this respect, discursive psychology seems inimical with psychoanalytic theory, which presumes that h i d d e n , unconscious motive-forces lie b e h i n d the surface of social life. Psychoanalytic theorists often treat outward social activity as a cipher for unobser vable, inner motivational processes. T h e apparent incompatibility between the n e w discursive psychology a n d psychoanalysis might b e expressed crudely, discursive psychologists turn the person inside out, converting inner mental life
Source: BritishJournal of Social PsycAoiogy vol. 36,1997, pp. 139-159.
BILLIG THE DIALOGIC UNCONSCIOUS
150
into outward social activity, while psychoanalysts m o v e in the opposite theoretical direction b y turning social life outside in. It will b e suggested that matters are not quite so simple, especially if one takes seriously the notion that conversational interaction is not only a m o d e of expression, but is also m o d e of repression. T h e r e is a case, it will b e argued, for considering that there is a 'dialogic unconscious', which is constituted in dialogue a n d which functions as a form of repression. If the repressive aspect of conversation is considered, t h e n discursive psychology a n d psychoanalytic theory m a y not b e utterly incompatible, at least if theoretical adjustments are m a d e on b o t h sides. I n developing these themes, the present p a p e r will attempt to argue within the basic assumptions of discursive psychology, especially the Wittgensteinian notion that psychological language stands in n e e d of outward, rather than inner, criteria. It will b e suggested that this position does n o t preclude psychoanalytic forms of explanation, but, u n d e r certain circumstances, encourages them. Never theless, to develop psychoanalytic themes based o n the idea of a dialogic uncon scious, it is necessary to e x p a n d the image of social interaction, which discursive psychology has inherited from conversation analysis a n d ethnomethodology. At present, conversation analysts h a v e revealed in depth h o w everyday morality is routinely accomplished in conversational interaction. If the unconscious, as well as the conscious, is constituted in interaction, then the everyday interaction is not only reproducing moral norms, but it also reproduces immoral temptations, which are routinely resisted a n d repressed. To investigate these issues, it will b e argued, it is necessary to examine the absences, rather than presences, in dialogue. If this is d o n e , t h e n one should b e able to investigate not only h o w everyday morality is accomplished within talk-in-interaction, as conversation analysts and ethnomethodologists have demonstrated; also it will b e possible to see h o w repression is routinely accomplished. I n so doing, discursive psychology will b e attending to some of the traditional concerns of psychoanalysis. Most importantly, the basic psychoanalytic process of repression, whose routine operation even Freud considered to b e mysterious, might b e observed direcdy as it is constituted through social, discursive interaction. Before discussing these themes, it is necessary to say a few words about the use of the terms 'psychoanalysis' and 'discourse analysis', for both terms cover a variety of different intellectual positions. I n discussing psychoanalysis, the emphasis will b e o n the writings of Freud, rather than o n later schools of thought. Of course, it should b e possible to link the notion of a dialogic unconscious with concepts derived from other psychoanalytic schools. Indeed, this needs to b e done, if the idea of a dialogic unconscious is to b e fully developed. However, this initial p a p e r concentrates u p o n Freud's writings. N o t only h a v e Freud's writings a unique historic and cultural significance, but arguably they still constitute the clearest presentation of psychoanalysis's central, enduring concepts. Just as there are a variety of psychoanalytic schools, so very different approaches to the study of language are to b e found u n d e r the b r o a d heading of 'discourse analysis' (see, for example, the c o m p e n d i a edited b y van Dijk, 1985,1996). T h e type of discourse analysis to b e explored in this p a p e r is that which has b e e n heavily influenced b y conversation analvsis. pHinn»«rfi—
100
DISCOURSE STUDIES
philosophy of Wittgenstein. This form of analysis concentrates o n studying actual utterances in their particular contexts. I n this respect, it differs substantially from Foucauldian a n d Lacanian approaches. Foucauldian studies tend to view discourses as cultural totalities, a n d they seek to identify these totalities rather than examine the ways in which people use utterances in specific circumstances. Foucault's notion of a 'discourse' is, to use Saussure's distinction, m o r e langue than parole - concerned with the hypothetical total structure rather than particular usage. Similarly, Lacan, w h o takes seriously the link b e t w e e n language a n d the unconscious, operates with a s o m e w h a t abstract t h e o r y of language, which explicitly gives priority to langue, or the total linguistic structure, rather t h a n to parole (see, for example, Lacan, 1979, p p . 2 0 - 2 1 ; for m o r e details, see Billig, 1996,1997c). By contrast, the present approach, in c o m m o n with m u c h discursive psychology, is firmly rooted in the study of the utterance a n d the particular, occasioned use of language. T h e present approach, therefore, differs from those psychologists w h o h a v e b e e n recendy investigating the links b e t w e e n psycho analysis a n d discourse in terms of Foucauldian and Lacanian ideas (see, inter alia, B u r m a n & Parker, 1993; Henriques, Hollway, Urwin, Venn & Walkerdine, 1984; Hollway, 1989; Parker, 1990, 1992; Wilkinson & Kitzinger, 1995). I n consequence, the notion of the dialogic unconscious, as outiined in the present paper, is neither Lacanian n o r Foucauldian. It does not refer to a presumed 'fallen' structure of signification. N o r does it refer to utterances, which cannot b e spoken because the speaker lacks the cultural a n d discursive resources for composing such a n utterance. O n the contrary, the dialogic unconscious com prises utterances which could well have b e e n spoken, but which remain unspoken. T h e s e are utterances which in some sense have b e e n dialogicalfy repressed in particular conversational settings. 1
D i s c o u r s e Analysis a n d t h e U n c o n s c i o u s
>i
Discursive psychology has set itself the project of investigating h o w psychological states are constituted through discursive activity. This m e a n s paying attention to the use of psychological language in conversation. T h e point is n o t to treat psychological language (such as 'I feel h a p p y ' or ' m y thoughts o n this matter a r e . . . ' ) as symptoms of inner processes. T h e analysis of psychological language, for discursive psychologists, is n o t a second-best strategy, as if the 'real' topics of psychology w e r e internal, and thereby unobservable, p h e n o m e n a . Quite the contrary, discursive psychologists assume that the traditional topics of psychology refer to p h e n o m e n a which are outwardly observable. If one wishes to study 'feelings', psychologists should b e paying attention to what p e o p l e are doing w h e n t h e y claim to h a v e feelings. T h e principles b e h i n d this a p p r o a c h can b e found in the later writings of Wittgenstein, w h o stressed that, because language is socially shared, there must b e public criteria for the use of psychological words. If there were not such criteria, speakers w o u l d n o t b e able to speak so readily of feelings, emotions or states of m i n d (see Shotter, 1993a,£, 1995, for discussions of Wittgenstein and
BILLIG
THE DIALOGIC UNCONSCIOUS
161
his contribution to discursive psychology). As Wittgenstein stated in Philosophical Investigations, 'an "inner process" stands in need of outward criteria' (1953, remark 580). H e claimed that 'the characteristic mark of all "feelings that there is expression of them, i.e. facial expression, gestures' (1967, r e m a r k 513). T h e implications are profound. Psychologists should not search for the ghostly, unobservable essences, which are presumed to lie behind the use of feeling-words. T h e y should b e examining in detail the ways people m a k e claims about psychological states a n d what they are doing w h e n they m a k e , or dispute, such psychological claims: they should b e asking 'what is the purpose of this language, h o w is it being used?' (Wittgenstein, 1967, r e m a r k 716). T h e techniques of conversation analysis a n d e t h n o m e t h o d o l o g y enable discursive psychologists to answer such questions, for they p e r m i t analysts to investigate the microprocesses of speech acts in which psychological language is used. Analysts can study how, for example, claims to h a v e particular emotions or psychological states are discursively accomplished. I n this way, emotions and psychological claims are seen to b e socially constituted a n d accomplished (Edwards, 1995a). For example, traditional psychologists assume that ' r e m e m bering' is something which takes place within the cognitive system of the isolated individual. By contrast, discursive psychologists treat r e m e m b e r i n g as a social and collective activity (Edwards & Potter, 1992; Middleton & Edwards, 1990; Potter & Edwards, 1990). T h e y e x a m i n e in detail the speech acts involved in m a k i n g claims a b o u t r e m e m b e r i n g a n d forgetting, a n d they ask w h a t such claims are accomplishing socially (e.g Billig & E d w a r d s , 1994; E d w a r d s & Middleton, 1986). Consequently, r e m e m b e r i n g , far from b e i n g treated as an unobservable, internal process, b e c o m e s a directly observable, social activity based u p o n speech acts. Discursive psychology's stance of exploring observable social action, rather than searching for hidden, inner states, is reflected in the analysts' own theoretical language. Their concepts tend to emphasize action and activity, rather than inner process. Following the example of H a r v e y Sacks, the word ' d o ' features promin ently in discursive analysis (Edwards, 1995b). 'Feelings' are activities which h a v e to b e ' d o n e ' (or accomplished socially a n d discursively). Discursive psychologists do n o t search for t h e feelings b e h i n d o b s e r v a b l e actions (or doings), b u t concentrate u p o n the doings, without which n o one, including analysts, could talk of feelings. It should b e little surprise that an approach, which discourages p h e n o m e n o logical language as a tool of analysis, has little theoretical space for the vocabulary of psychoanalysis. If inner consciousness is n o t an explanatory concept, then there would appear to b e even less reason to hypothesize an inner unconsciousness. The main texts of discursive or rhetorical psychology contain h a r d l y a mention of the 'unconscious', 'repression' or Freud. Occasionally, and somewhat briefly, psychoanalysis is discussed, b u t typically in ways which discount its significance (see, for example, Gergen, 1994, p . 169; or Shorter, 1993fl, p p . 138/? but see also Shorter, 1990, p p . 129j£ for a n a r g u m e n t that cognitive psychology has forgotten its Freudian heritage).
DISCOURSE STUDIES
m
Discursive psychology, if it turned its attention to psychoanalysis, might easily treat psychoanalytic language as a topic of investigation, rather than as a theoretical resource for analysing other forms of discourse. This was a tactic which Volosinov/ Bakhtin r e c o m m e n d e d years ago in his critique of Freudianism (Volosinov, 1994; see also Daelemans & Maranhao, 1990).J u s t as discursive analysts presendy study h o w people use psychological language in everyday conversation, so they could examine h o w the language of unconscious motives is used in b o t h lay a n d profes sional discourses. T h e distinction between the lay and the professional is not firm. As Moscovici (1976) has shown, psychoanalytic concepts are n o longer confined to professional, psychoanalytic circles, but h a v e b e e n represented, and altered, in ordinary language. In this respect psychology, including psychoanalysis, has a w i d e cultural a n d political significance in c o n t e m p o r a r y Western society (Parker, 1994; Rose, 1990). Discursive psychologists c o u l d specifically e x a m i n e the c u r r e n t use of psychoanalytic concepts, thereby examining in discursive detail the p h e n o m e n o n identified in b r o a d oudine b y Moscovici. Such an examination of psychoanalytic discourse would n o t necessarily d e m a n d special methodological or theoretical treatment. W h e t h e r analysts were studying the language of consciousness or unconsciousness, the methodological techniques could b e the same. T h e analysts would not themselves b e using psychoanalytic concepts as analytic tools. T h e y would see what is being ' d o n e ' with talk about 'hidden, unconscious motives'. T h e y would not claim that 'hidden, unconscious motives' stand behind, or cause, this talk. For example, they would study h o w people talk about 'repression', but they would n o t assume that the talk itself was based u p o n repression. I n this respect, such analyses would not b e examining the repressive aspects of discourse; n o r would they b e p r o b i n g the discursive aspects of repression, as understood b y psychoanalysts.
R e t u r n of t h e R e p r e s s e d R e p r e s s i o n Although the unconscious m a y b e largely absent from the topics studied by discursive psychology, it is not to b e easily eliminated. As psychoanalysts well understand, the repressed returns to h a u n t consciousness. Discursive psycholo gists, not to mention conversation analysts, might exclude the concept of repression from their theoretical vocabulary, but, occasionally, the repressed repression can b e detected, lurking o n the edges of analysis. Two examples can b e briefly mentioned. Wetherell a n d Potter (1992), in their analysis of racist discourse, present a n example of a 'benevolent racist', w h o disclaims being racist. This person, they suggest, fits K a r e n H o r n e y ' s image of the individual resolving a neurotic conflict through fantasies about the ideal self (p. 54). Billig (1992), analysing h o w English families talk about the British Royal Family, at one point has recourse to the notion of projection (pp. 106-109; see also Billig, 19970). Speakers are discussing whether the heir to the throne would b e allowed to m a r r y a non-white. T h e speakers disclaim their o w n racism, a n d they also deny
BILLIG THE DIALOGIC UNCONSCIOUS
163
that the Royal Family, with w h o m they claimed to 'identify', is racist. Billig suggests that the 'racism' is projected onto others: the speakers oppose the idea of a nonwhite m a r r y i n g the heir to the throne, b u t attribute this position to others - to 'the public' in general. T h e details of the analysis are n o t what matter for the moment, only the analyst's use of psychoanalytic terminology. What, o n e might ask, are the analysts 'doing', w h e n they use psychoanalytic terms? I n b o t h examples, analysts are considering the utterances of speakers, who are denying that they are racist, while expressing views which can easily b e heard as racist. T h e analysts, in presenting this material, distance themselves from the respondents' denials of racism (see also v a n Dijk, 1992, 1993a). If the respondents disclaim any racist motives, t h e analyst, in using psychoanalytic terminology, alerts readers n o t to accept those disclaimers: the possibility of racist motives, unacknowledged b y the holder of the motives, is left open. I n this respect, as Parker (1992) has suggested, psychoanalytic concepts can b e useful for a discursive psychology which has a n acknowledged critical, political stance. T h e analytic m o v e of using psychoanalytic terminology does not represent an a b a n d o n m e n t of the basic tenets of speech act theory, but is, in some respects, derived from t h e m . T h e possibility of using psychoanalytic descriptions follows from the assumption that there are outward criteria for the use of psychological concepts. As Wittgenstein continually argued, because there are public criteria for the use of psychological language, w e do not describe our own psychological states directly, whilst being forced to infer the states of others indirectly and with less certainty. O n the contrary, as Wittgenstein wrote, 'every day w e hear one m a n saying of another that h e is in pain, is sad, is m e r r y , etc. without a trace of doubt' (Wittgenstein, 1980, remark 138). T h u s , we can legitimately claim confidence in our descriptions of another's psychological state. We m a y b e con fident, even if the other person disagrees with our assessment; in this case, we can argue that the other's description of themself conflicts with the outward criteria b y which such descriptions m u s t b e m a d e . Wittgenstein, at one point, c o m m e n t e d that 'it also h a p p e n s that one says "I know that y o u felt like this then, even if you w o n ' t admit it n o w " ' (1980, r e m a r k 138). In this sort of situation - w h e r e one p e r s o n claims to k n o w the feelings of another better than the person themself - the possibility of psychoanalytic explan ations is raised. Such an explanation signals that the other's self-descriptions are unconvincing. N o t only is a m o r e convincing account of their motives offered, but also a n account w h y the self-descriptions are unconvincing - w h y the person is unable to admit something about themself which is a p p a r e n t to another and which should b e , but is not, apparent to themself. I n these circumstances, uncon scious motives m a y b e cited convincingly. Wittgenstein gives the example of someone w h o cannot explain w h y they suddenly climbed o n a chair: 'Perhaps in such a case', h e suggested, 'we should say that h e acted with unconscious intention' (1988, r e m a r k n o . 222). I n another example, Wittgenstein wrote about someone, attempting to draw the picture of one individual b u t actually producing a portrait of another, being 'guided b y forces in his unconscious' (1967, r e m a r k n o . 262). T h e r e are two points which can b e noted about these usages of unconscious explanations. I n the first place, such exnlanatinne A*™~J •
DISCOURSE STUDIES
I(!4
m o r e ' n o r m a l ' explanations. T h e psychoanalytic language g a m e can never replace in its entirety the language games of ordinary psychological explanations, for they are to b e used w h e n the ordinary games appear to b r e a k d o w n (Brearley, 1991; Cioffi, 1991; Hopkins, 1995). Second, the possibility for m a k i n g explan ations, which cite unconscious motivations, is n o t prohibited b y the assumptions of speech act theory, o n which discursive psychology is based. I n d e e d , the possi bility follows from t h e Wittgensteinian assumption that emotions are social constructs a n d that it is n o t inevitable that the individual actor is privileged in p r o d u c i n g m o r e convincing self-descriptions t h a n any outside observer can. U n d e r some circumstances, others m a y d o u b t a person's self-ascriptions, or may w o n d e r at the failure to p r o d u c e any self-descriptions; t h e n they m a y speculate w h y the person cannot admit to themself what appears apparent to outsiders. This sort of discursive m o v e is to b e expected in a society w h e r e psychoanalytic conceptions h a v e passed into c o m m o n sense (Moscovici, 1976). A s is illustrated b y the two examples taken from the analysis of speakers denying racism, discur sive analysts can m a k e the m o v e too.
C o n v e r s a t i o n a n d Constraint
i
A n important question n e e d s to b e addressed: if the discursive a p p r o a c h does n o t theoretically prohibit psychoanalytic speculations, at least in the w a y that it prohibits speculation about inner cognitive processes, then w h y are psychoanalytic notions so rarely employed? O n e might answer such a question historically, pointing out h o w psychology in general has repressed its psychoanalytic heritage (Parker, 1994; Shorter, 1990). I n this respect, discursive psychology would merely b e continuing a long established intellectual habit. O n the other h a n d , one might attempt to answer m o r e specifically b y showing h o w discursive psychology's routine procedures of theory a n d methodology c o m b i n e to inhibit discussion of the unconscious. It will b e suggested that discursive psychology has adopted practices from conversation analysis, which tend to draw attention to the presences, rather than the absences, in discursive utterances. I n particular, there are two principles of conversation analysis, which inform discursive psychology and which militate against psychoanalytic concerns with absences: (a) analysts should try to conduct their analyses from the participants' perspectives, as revealed in what the participants say; (b) analysts should use what participants say in order to reveal the structural organization of conversation. (a) Conversation analysts, as well as ethnomethodologists, frequendy clairrj that o n e of the strengths of their approach is that they focus ' o n h o w partici pants themselves p r o d u c e a n d interpret each other's actions' (Pomerantz, 1988. p p . 360-361). Conversation analysts d o n o t seek to explain interaction in terms of sociological structures, which lie 'behind the backs of the participants' (Boden. 1994; Heritage, 1984). Instead, they observe h o w participants m a k e sense of. and account for, the social world. In this respect, conversation analysis, to quote Deirdre B o d e n , assumes that ' h u m a n beings are knowledgeable agents in the produc tion (and reproduction) of their lives a n d their history' (Boden, 1994, p. 13'.
BILLIG
THE DIALOGIC UNCONSCIOUS
165
This assumption of knowledgeability forms the core of the ethnomethodological enterprise: 'By giving back to social agents their knowledgeability of their o w n social actions, it was t h e n possible to sit back a n d observe the structuring quality of the world as it happens' (Boden, 1994, p . 74). I n this way, ethnomethodologists, to quote B o d e n again, consider social order as 'a m e m b e r ' s order than an analyst's construct' (p. 65). T h e assumption of actors' knowledgeability is b o t h a methodological facilita tion a n d a restriction. O n the o n e h a n d , analysts are b i d d e n to take seriously what people actually say and do, rather than assume actors are simply re-enacting given roles within a hypothetical social structure. T h e r e is n o d o u b t that this principle h a s productively concentrated attention o n the fine detail of social life. O n the other h a n d , the assumption is theoretically restricting in that the analyst is discouraged from taking a stance, which observers of others often d o take a n d which is central to the whole project of psychoanalysis. This is the stance, which claims that ' m e m b e r s ' , or social actors, sometimes are u n a w a r e of the reasons for their actions a n d that their accounts, far from indicating know ledgeability, express a lack of knowledge a n d , indeed, repress self-knowledge. By assuming knowledgeability at the outset, the analyst cannot easily investigate how the repression of knowledgeability might b e accomplished. (b) T h e second principle aims to find within the utterances of social actors the structural organization of e v e r y d a y life (Boden, 1994; Heritage, 1984). According to Buttny, m u c h social interaction, such as adjacency pairings in conversation, has a 'normative organization' (1993, p . 39). Conversation analysts seek to discover this normative organization, which is p r e s u m e d to lie within, and not behind, what participants say. This is w h y turn taking and sequential organization h a v e b e e n so central to conversation analysis. According to B o d e n (1994), 'turn taking a n d the sequenced structuring of action' lie 'at the heart' of social interaction (p. 53). T h e organizational pattern is p r e s u m e d to b e discover able through understanding the sequential constraints o n speakers. D r e w (1995) claims that 'turns' in conversation are treated 'as the p r o d u c t of the sequential organization of talk' (p. 70). T h e organization is p r e s u m e d to b e present in the accounts which speakers themselves give. T h u s , the analyst seeks to n o t e con versational devices, which 'are demonstrably or observably relevantto the participants themselves' (Drew, 1995, p . 76; emphasis in original). By so doing, analysts h o p e to discover the 'stable and organized properties of conversational structures' (p. 76). These structures are so normatively powerful, that they shape to a large ex tent what is said in conversational interaction. D r e w (1995) claims that, according to the perspective of conversation analysis, participants' inner cognitive states are autonomous from their talk. W h a t people say 'is n o t determined by, or the automatic product of, the processes of the m i n d ' ; instead, utterances in con versation 'are shaped most proximately b y the sequential position in which the turn is produced' (Drew, 1995, p . 70). It is as if speakers find themselves inhabiting a normative structure which is m o r e powerful t h a n their individual feelings a n d to which they have to conform for interaction to proceed. This structure is present \» ithin the interaction a n d the analyst, therefnrp «»ot-c f~ —-•- - < -
lfifi
DISCOURSE STUDIES
a m o n g participants' actions to uncover the sequential ordering of accounting' (p. 59, emphasis in original). W h e n analysts have located an instance of accounting, they should search for the conversational antecedents a n d consequences of these accounts. Most importantly, they should 'examine w h a t connects these parts to gether as a coherent discursive practice of accounting' (p. 60, emphasis in original). T h e u p s h o t is that analysts tend not to search for absences in conversation, for the given conversation is to b e seen as coherent in itself. It is p r e s u m e d that participants are knowledgeable about t h e conversational structure, which is determining their 'conduct in talk-in-interaction' (Drew, 1995, p . 78). Similarly, it is p r e s u m e d that this knowledge is expressed in the utterances themselves and, thus, the structure is knowable t h r o u g h these expressions. I n consequence, the determination of particular utterances is to b e u n d e r s t o o d in terms of other utterances! Again, the emphasis is u p o n the uttered presences of conversation. Analysts work with the text (or transcription) of the interaction, rather t h a n using the text as a sign of something else, which is itself absent in the text, b u t which leaves traces. T h e m e a n i n g of a particular conversational turn is not to b e sought in the inner states of m i n d of the speaker; nor is it to b e sought in a hypothetical 'social structure'; it is to b e found in the relations with the preceding interventions, and checked b y the subsequent ones. W h a t was, and w h a t will b e spoken, provide t h e m e a n s for understanding the function a n d m e a n i n g of what is said. In this respect, the analysts put themselves in the same position as the participants. What this m e a n s is that analysts search for the connections between utterances which are present in the conversation. W h a t is absent from t h e conversation tends to b e absent from the analysis.
C o n v e r s a t i o n a l Morality a n d its Flaws By following the theoretical a n d methodological presuppositions of conversation analysis, researchers h a v e provided a n impressive corpus of knowledge about the intricate structures of contemporary conversations. T h e y h a v e demonstrated in detail the shared practices b y which participants m a n a g e everyday interactions and, thereby, r e p r o d u c e the normative structures which permit conversation to progress. For example, conversation analysts have claimed that there is a 'prefer ence for agreement', which facilitates expressions of agreement a n d militates against expressions of outright disagreement (BUmes, 1987; Pomerantz, 1984). T h e r e a r e ways of ' d o i n g politeness', w h i c h h e d g e requests i n indirectness, so ftxaii.'&Yey aietvotYvaaxdas cotvvetsa&ou-feeateT^ (Jkovm. & . L e M n s Q n . t
m d '\s\ s o dova.^, ftxe'j \ve\\> to trtamoVe
SsveaSkVj
too^es^CvNft fclmos^Mw
its rather syn.copa.ted a n d agreeably coWabotalwe ojaalvty' leaden, V. p. "1 Above au, there is politeness. As Chilton (1990) h a s written: 'Politeness formuUbsa are universal a n d pervasive in everyday intercourse largely c o n v e n t i o n a l a n d often benign. T h e preferred response is reciprocal politeness' (p. 222JH
BILLIG THE DIALOGIC UNCONSCIOUS
167
All this m i g h t b e taken as suggesting that conversation analysis reveals little m o r e than codes of etiquette. However, the theoretical importance of ethno methodology a n d conversation analysis is that they demonstrate h o w social order is r e p r o d u c e d (Heritage, 1984). At root, the codes of conversation a n d politeness are codes of morality. T h u s , Garfinkel (1967) claimed that ethno methodology was revealing the accomplishment of 'practical morality'. I n this respect, the micro-processes of conversational interaction reveal h o w every day life is infused with m o r a l concerns. Even the m o m e n t a r y pauses a n d frac tional hesitations in routine conversations can b e filled with m o r a l import. For example, turn taking is n o t m e r e l y a structure which permits a conversation to proceed agreeably: it is a m o r a l matter. As B o d e n (1994) suggests, b y learning the routines of turn taking 'we take o n that reciprocal m o r a l code that guides h u m a n action' (p. 54). If conversation analysis provides a n investigation of practical morality, it also conveys a n image of the nature of this morality. T h i s is a morality which is seen to be routinely and successfully accomplished: it is not a morality which d e m a n d s greater standards of ethics than can b e practically realized. Often, b u t b y n o means always, a somewhat upbeat, positive rhetoric is used. Analysts, in quota tions already presented, use terms such as 'benign', 'cooperative' a n d 'agreeable'. 'Members' are seen to 'collaborate', in order to 'accomplish' this morality; even the term ' m e m b e r ' suggests a belongingness. Because internal speculation is discouraged, the characteristics of the ' m e m b e r s ' appear to b e constituted through the talk-in-interaction. A n d this talk is seen to b e a collaborative accomplishment, which is structured b y the normative d e m a n d s which are shared b y the m e m b e r s . The result is that it is easy for ' m e m b e r s ' to a p p e a r analytically as cooperative turn takers, w h o do n o t possess a clouded hinterland. This positive image of social morality is very different from Freud's view. I n the Freudian account, another, m o r e dangerous, side of h u m a n nature is visible: immorality always lurks on the edge of o v e r d e m a n d i n g morality. For a Freudian, polite codes of civilized behaviour function to restrict impulses, which continually threaten social order. T h e m o r e that the d e m a n d s of politeness are h e a r d to fill conversational spaces, the m o r e the Freudian w o u l d suspect the repression of immoral impulses. T h e Freudian would not ask h o w the conversational structures of morality are accomplished, b u t what temptations these structures are trying to prevent For example, if agreement is the socially preferred n o r m , then perhaps disagreement and disruption constitute a h i d d e n temptation, whose delights must b e repressed from consciousness. Such questions tend not to b e asked b y conversation analysts, whose perspective io date has emphasized the achievement of conversational morality, rather than •he repression of immorality. This can b e illustrated b y considering the conver sational device o f repair', which has been m u c h studied b y conversation analysts i.e. Schegtoff.Jefferson & Sachs, 1977; Pomerantz, 1984). Introducing a discussion d conversational repair, Nofsinger (1991) states that 'conversation, like other farms of human behaviour, is not perfect'. People can start to say the wrong ibagi, or they can forget a word. According to Nofsinger, 'it is important for our ndostanding of each other that we fix these flaws before they lead to m o r e IOIOUI and fundamental nrnhlpmc'^ io^\ vr_i---
168
DISCOURSE STUDIES
is taken from Pomerantz (1984; the same e x a m p l e is also analysed b y Potter & Wetherell, 1995). O n e speaker suggests to another that the latter h a d not brought a great deal of fruit cake. T h e silence with which the r e m a r k is greeted leads the original speaker to 'back d o w n ' a n d reverse the original statement, suggesting that, after all, a little piece of cake can go a long way. T h e silence and the backing d o w n indicated that the original statement was being h e a r d as a possible criticism, which would disrupt the collaborative interaction. A n d so a repair was called for. M u c h effort has b e e n spent in detailing what Boden (1994, p . 75) calls the 'impressive range of repair', such as 'self-repair' versus 'other repair' etc. For pre sent purposes, two points are worth noting. I n the first place, repairs are seen to b e necessary in order to rectify 'flaws', which threaten the structure of conver sation. Second, the analytic attention is u p o n the repair rather than the so-called flaw. T h e flaws, which threaten the structure of conversation, cannot b e explained b y that structure. Analysts tend to demonstrate h o w and w h y the repair structure operates. W h y the flaw should h a v e arisen in the first place is typically left unexplained, even though flaws must occur as often, if n o t m o r e often, then the repair work which follows them. Significantly, there is n o technical terminology for describing h o w and w h y 'flaws' are 'accomplished'. T h e technical vocabulary is reserved for the repair work, which is seen as the accomplishment. Indeed, the very use of the w o r d 'flaw', as used b y Nofsinger, suggests that such utterances pose a threat to the conversational accomplishment. Given the intricacy of speech, especially in relation to the codes of indirectness a n d mitigation b y which politeness is routinely 'done', o n e can see the normative structures of conversation as highly constraining. Speakers are constrained to wait their turn, express d e m a n d s indirectly, mitigate refusals and so on. I n these ways, rudeness, or conversational immorality, is kept at bay. W h a t is not asked indeed, the standard ways of doing conversation analysis discourage the question - is w h y rudeness so frequently appears, or almost appears, before being quickly repaired. By contrast, the interruptions of social order - or, to use Freud's phrase, the 'parapraxes of everyday life' - h a v e often constituted the starting point for psychoanalytic inquiry. T h e flaws indicate a h i d d e n realm of desire, rather than chance accidents which h a p p e n because h u m a n s are not perfect beings. T h e moral codes constrain; their structures inhibit; b u t they d o n o t totally dominate. Their intricacy of restraint can b e read as a sign that the temptation to immorality stalks the practice of conversational morality, awaiting the opportunity to colonize vacated spaces.
Temptations of R u d e n e s s a n d the D i a l o g i c U n c o n s c i o u s According to psychoanalytic theory w h a t is prohibited beckons as temptation. If the forbidden w e r e n o t tempting, there would b e little n e e d for compla moral codes. For temptation to b e successfully curbed, it must not only b e resisted but also denied. Social actors must chase the i m m o r a l desire from the conscious m i n d , lest the surface of everyday life b e disrupted. All this is relevant to the practice of everyday morality, which can b e witnessed in the course of even the
BILLIG
THE DIALOGIC UNCONSCIOUS
169
most casual conversations. This morality suggests a realm of discursive immorality: self-repairs, successfully accomplished, indicate a temptation resisted. This can b e the temptation to act directly in ways which rudely throw off the c o m p l e x restraints of discursive indirection. O n e might stipulate that the possibility for 'doing politeness' depends u p o n the capability for 'doing rudeness' and, thus, that there is a p e r m a n e n t flaw in the apparendy agreeable, cooperative structure. Developmentally, politeness and rudeness are linked. If politeness is routinely taught, as it must b e for the skills of conversational turn taking to b e acquired, then so must rudeness. Each time the adult says 'you must say please . . . don't speak w h e n others are speaking . . . don't use that word - it's rude', the adult is telling the child what not to say, when speaking politely. I n addition, the adult is telling the child about rudeness; in pointing out what must n o t b e said the adult is equipping the child with dangerous, disruptive tools. Moreover, the parent, in instructing the child a b o u t these matters, tends to speak in direct ways which w o u l d n o t b e classed as polite in adult c o m p a n y (Aronsson, 1991). I n this way, the teaching of politeness, which is itself constituted through language, provides models of rudeness. If all forbidden acts promise the possibility of pleasure, because they are forbidden, t h e n rudeness and disagreement might b e thought to h a u n t all those routine, banal enactments of discursive morality. Each conversational exchange, which successfully unfolds a normative structure without b r e a k d o w n , might b e said to express that structure of discursive politeness a n d to repress the corres ponding o n e of rudeness. However, the notion of repression implies m o r e than that something is h i d d e n : it implies that what is h i d d e n is desired as a source of pleasure. W h a t evidence, one might ask, is there for supposing that rudeness (or politeness's repressed other) is a h i d d e n , tempting pleasure? I n Pomerantz's ex ample of successful repair, o n e might ask w h y the first speaker b r o k e the codes which prohibit direct criticism. Was the second speaker then tempted to challenge the first speaker, instead of remaining silent? Was the spontaneous self-repair a sign of temptation avoided? T h e s e questions are discouraged in conventional conversation analysis, whose theoretical vocabulary does not provide the words for describing a 'dialogical id' in battle with the 'conversational superego'. Thomas Scheff (1990) has suggested that the fear of ' s h a m e ' holds social interaction in play. Should the normative structure b e b r o k e n , a n unrepaired conversational b r e a k d o w n m a y ensue. T h e outcome is likely to b e recognized as 'shameful' and the speaker feels shame, which, according to Scheff, 'is probably the most intensely painful of all feelings' (1990, p . 169). T h e speaker in Pomerantz's example corrects herself, lest she b e heard as m a k i n g an inappropriate criticism of her fellow conversationalist. H a d she n o t d o n e so, a scene - or conversational breakdown - might have ensued with all its attendant embarrassment a n d shame. The use of psychoanalytic terminology is appropriate if there are reasons for thinking that what is shameful is also tempting. O n e might suggest that all social restrictions are accompanied b y the desire - whether or n o t conscious, and whether or n o t constant or intermittent - to transcend the restriction. If this is so. then one should expect impoliteness to b e a n object of temptation and to represent a desired freedom. This temptation is b o t h inhibited and created b y
170
DISCOURSE STUDIES
the fear of s h a m e a n d b y t h e routines of practice, which accomplish politeness. Moreover, these routines, b y their very accomplishment, regularly succeed in driving the disruptive thought from the m i n d . Sometimes, the r u d e insult might just slip out, despite ourselves. A n d regularly, too, and again without conscious forethought, the words of repair will spontaneously slip out. H o w e v e r tempting this line of argument, with its suggestion of dialogical id a n d superego, nevertheless it is necessary to offer some signs that the rudeness is a temptation, which must b e routinely repressed. O n e criterion for the existence of unconscious desires is that they regularly, but unpredictably, leak into conscious awareness a n d into the flow of social activity. Freud, especially in The Psychopathology of Everyday Life, claimed that the 'parapraxes' of social life, such as lapses of m e m o r y , indicated the workings of repressed desires. H o w e v e r , the case of h u m o u r is particularly interesting. According to psychoanalytic theory, in humour the socially repressed is n o t only expressed but is enjoyed. Freud claimed that 'our enjoyment of the j o k e ' indicates what is b e i n g repressed in m o r e serious talk {Jokes and their Relation to the Unconscious, p . 138). Mulkay (1988) has argued that h u m o r o u s discourse, b y being m a r k e d off from ' n o r m a l ' serious discourse, sustains ' o n e of the basic requisites of the serious m o d e : namely that disagreement should b e avoided as far as possible' (p. 79). T h e point is n o t merely that humorous discourse expresses disagreement a n d impoliteness, b u t that the latter, which are repressed in normal discourse, can b e enjoyed in the format of humour. Comic heroes are frequently figures w h o disrupt the patterns of restrictive politeness. F r o m Diogenes of Sinope, living in his t u b , through to Groucho M a r x and John Cleese there is a c o m m o n thread. Audiences find pleasure in the character who sweeps s h a m e a n d the social rules aside. Sometimes, as withJ o h n Cleese's char acter of Basil Fawlty, the constraints of politeness are turned into weapons of rudeness: 'thank-you-so-much', h e proclaims in a p a r o d y of politeness (see Brown & Levinson, 1987, p p . 2 2 9 - 2 3 0 for a discussion of the way that overpoliteness can b e c o m e a m e a n s of insult). It might seem curious that displays of rudeness, which ordinarily will be treated as unethical, are greeted with pleasure in the context of humour. Freud claimed that the pleasure of a tendentious j o k e expressed a repressed desire: 'To the h u m a n psyche all renunciation is exceedingly difficult, a n d so we find thai tendentious jokes provide a m e a n s of u n d o i n g the renunciation a n d retrieving what was lost' (1991, p . 145). This is true of the urge to b e impolite. According to Freud, 'the prevention of invective or of insulting rejoinders' occurs so commonly that 'tendentious jokes are especially favoured in order to m a k e aggressiveness or criticism possible against persons in exalted positions w h o claim to exercise authority' (p. 149). T h e comic h e r o b e c o m e s a figure for identification, to be admired for doing precisely that which we would fear to do. O u r laughter betray the repressed wish for disinhibition, whilst, because the joke is 'only' a joke ii confirms t h a t inhibition. Peter Sloterdijk (1988), discussing Diogenes, whc insulted Alexander the Great with impunity, talks of the 'dialectic of msinhibition (p. 103). Inhibitions are released within t h e joke; pleasure is performed safely, especially if we, the inhibited, can enjoy vicariously t h e insults. T h e joker, recog nized as such, reaffirms the b o u n d a r y b e t w e e n n o r m a l a n d h u m o r o u s discourse, 5
BILLIG THE DIALOGIC UNCONSCIOUS
171
thereby strengthening the inhibition. Even Alexander the Great could j o i n in the laughter at Diogenes. T h e disinhibition is localized, contained within the joke. Yet, the whole business of laughter a n d delight only makes sense if there is a wish which stretches b e y o n d the m o m e n t of h u m o u r : otherwise w h y would the comic figure b e greeted with delight rather t h a n outrage? For Freud, the forces of desire w e r e universal a n d biologically based; as such they threatened the very nature of social order. Polite dialogue, thus, d e p e n d e d upon the sublimation of forces, which owed their origin to non-civilized, nondialogical factors. T h e opposition between desire and civilization was, for instance, clearly expressed in later works such as Civilization and its Discontents a n d The Future of an Illusion. However, repression, a n d thereby the unconscious, m u s t b e formed in dialogue. Without m o r a l requirements, which must b e consciously expressible, there w o u l d b e n o reason to repress desires. O n e can go further. Language, which is necessary for the production of morality, is n o t m e r e l y a vehicle for transmitting m o r a l prescriptions. Language constructs a n d requires its own morality of talk, a n d t h e r e b y its o w n repressed desires. In consequence, the location of desire outside the processes of dialogue a n d social order is n o t necessary. T h e temptations of impoliteness d o not stand outside the dialogic process, but are constituted within it. T h e desires to b e r u d e , to contradict, 'to speak one's m i n d ' , to h a v e d o n e with the constraints of politeness are formed within dialogue. Such desires cannot antedate, n o r stand outside, the constrictions of politeness. I n this respect, it m a k e s sense to talk of the unconscious being dialogically constituted. If the impulses for rudeness are formed in dialogue, then, so it m a y b e argued, are their repression. Dialogue must create its own ways to p u s h the temptations of rudeness aside from routine consciousness, a n d to permit the conscious aims of dialogue to continue. T h e person's sense of their o w n self will b e involved, investing this repression with greater force than a social m i s d e m e a n o u r which can be lightly broken. I n talking politely a n d agreeably - in doing consideration of others' feelings, through repair, indirection a n d a thousand other conversa tional devices - w e are reproducing ourselves as tolerant, moral, polite selves. The intolerant, rude, hurtful possibilities (and desires) have to b e routinely pushed aside from conscious awareness. T h e m e r e thought of being r u d e - of criticizing direcdy, snapping back, discarding the constraints of discursive etiquette - raises temptation in ways which should not even b e thought of if interaction is to proceed routinely in its structured, reflexively m o n i t o r e d smoothness. It might b e said that the routine conversational intricacies, which are revealed by conversational analysts, d e p e n d u p o n a repression which is n o t routinely revealed. T h e pushing aside of temptation must b e itself pushed aside, for, as Freud stressed, the process of repression is itself to b e repressed. O n e might suggest that the theoretical stance of conversation analysis, as practised in the main, expresses this repression. Its image of the speaker, acting to agree, avoiding disruption, and sustaining the normative structures of dialogue, reflects the selfnage of the caring, considerate, tolerant speaker with n o further interior life. At .iie same time, residing within us is the capacity to h o o t with pleasure at uncaring, inconsiderate and intolerant actions.
DISCOURSE STUDIES
172
Dialogic Unconscious in Action If the foregoing arguments are accepted, then it should b e possible to analyse discourse, in order to examine the processes of repression in action. T h e dialogic unconscious will b e constituted within talk, a n d the analyst, seeking to understand h o w repression is being discursively accomplished, will n e e d to examine the absences of discourse. W h a t is n o t said, b u t could easily h a v e been, and, indeed, o n occasions is almost said but then r e m o v e d from the conversation, becomes of p r i m e significance. Such analyses would extend t h e scope of conversation analysis, while remaining within discursive psychology's project to show how psychological p h e n o m e n a are constituted within social activity. However, such analyses w o u l d not a d d a 'psychoanalytic dimension' in any simple way. T h e notion of the dialogic unconscious should leave neither discursive psychology n o r psychoanalysis untouched. I n the first place, an analysis of the dialogic unconscious in discursive action lends itself to ideological analysis. T h e focus can b e shifted from the individual to the social a n d to the construction of the unconscious. Repression, then, is not a universally static process, b u t something which is part of ideological a n d sociohistorical currents (Frosh, 1989). Particular codes of politeness differ culturally a n d historically (Brown & Levinson, 1987; Katriel, 1986). So d o the topics on which p e o p l e are expected to converse a n d remain silent. Freud's patients found it difficult, e v e n in the privacy of the psychoanalytic consulting-room, to talk of personal, sexual matters, which are n o w the c o m m o n themes of television discus sion shows (Carbaugh, 1988; Livingstone & Lunt, 1994). O n the other hand, his patients w o u l d b e unlikely to h a v e b e e n as troubled b y some of the themes of 'race' as w e r e the speakers studied b y Wetherell & Potter (1993) a n d Billig (1992). T h e latter two examples illustrate the conditions of ideological analysis. Analysts not only h a v e to stand back from what the particular speakers are saying; analysts also m u s t attempt to distance themselves from, a n d reflect u p o n , the social conventions a n d ideological assumptions which enable such accounts to b e given (Parker, 1992, 1994; van Dijk, 1993*). T h e analyst of the dialogic unconscious n e e d s to investigate h o w routines of talk can prevent the utterance of themes/accounts/questionings, which might seem reasonable to outsiders b u t which are collaboratively avoided b y the par ticular speakers as a localized form of politeness. To the outsider, including the analyst, it can look as if there is a joint conspiracy to achieve a collective refusal of knowledgeability. If, as B o d e n (1994) suggests, rationality is localized, then the local rationality can also b e seen as a n irrationality, with its collaborative talk accomplishing discursive repression. T h e traditional procedures of psychoanalysis are not, b y themselves, suited for observing the repressions of the dialogic unconscious. I n traditional psycho analysis, o n e form of dialogue - the psychoanalytic conversation - is privileged as a m e a n s for discovering repressions which are p r e s u m e d to h a v e taken place elsewhere. A s will b e suggested, Freud was uncertain h o w repression was actually accomplished, being m o r e concerned to show its incompleteness. W h a t he did not take seriously was the possibility that conversations - including those between
BIIXIG
THE DIALOGIC UNCONSCIOUS
173
the psychoanalyst a n d patient - repressed t h r o u g h their expression. As conver sationalists talk about o n e set of topics, they are keeping others from discussion. H o w possible topics routinely r e m a i n undiscussed - and h o w possible questions remain unasked - requires the sort of microanalysis pioneered b y conversation analysts, applied to discursive absences as well as presences. A s will b e suggested, the analysis of discursive absences might provide clues about the operation of repression. I n this respect, the notion of the dialogic un conscious might help to illuminate issues at the heart of psychoanalytic theory. After all, Freud himself believed that the idea of repression, rather than that of the unconscious, was the real discovery of psychoanalysis: h e claimed that 'nothing like it h a d ever before b e e n recognized in mental life' (1995c, p . 18). However, Freud was loth to conceive of repression in dialogic terms. A brief example can b e given of the sort of routine dialogic repression, which is overlooked in classic psychoanalytic theory. T h e example is taken from the famous ' D o r a ' case, presented in Freud's 'Fragment of a n Analysis of a Case of Hysteria'. T h e 'Fragment' is a classic piece of psychoanalytic writing, notable for the detail of Freud's exposition a n d his imaginative interpretations (Marcus, 1986). I n recent years, this case has b e c o m e an object of intense reinterpretation. Feminist critics, in particular, h a v e claimed that the case illustrates the dynamics of male power, which Freud overlooked a n d which, b y his overlooking, h e ex emplified (for critiques of Freud in relation to Dora, see inter alia: Cixous & Clemont, 1986; Gallop, 1986; Gearhart, 1986; Irigaray, 1989; J a c o b u s , 1987; Lacan, 1986; Masson, 1990; Moi, 1986; R a m a s , 1983; Rose, 1986). T h e back ground can b e briefly given (for details of the historical situation of D o r a herself, see Appignanesi & Forrester, 1993; Decker, 1991; Loewenberg, 1985; Rogow, 1978,1979). Dora, as a y o u n g w o m a n of 19, was brought to Freud for treatment by her wealthy, domineering father. She was suffering symptoms of hysteria and h a d r e c e n d y threatened suicide. T h e life of D o r a ' s family was intertwined with that of another family in a c o m p l e x pattern of deceit. Dora's father was having a protracted affair with a y o u n g married w o m a n , Frau K , although h e denied this to his o w n family. To facilitate the affair, h e encouraged increasing contact between the two families. Frau K. befriended Dora, often leaving h e r young children in D o r a ' s care. H e r r K h a d b e e n propositioning D o r a since she was fourteen, o n occasion grabbing her, trying to kiss her and, once, o n holiday entering her b e d r o o m while she slept. D o r a told h e r father about K.'s advances, but he refused to believe her, claiming that she, in h e r hysteria, was engaging in sexual fantasies. This was one of the symptoms, which Dora's father h o p e d Freud would cure. It was clear that Freud did not trust Dora's father, yet h e followed the remit to restore D o r a to psychological health. To this e n d , Freud sought to uncover repressed desires, which h e assumed lay b e h i n d h e r hysterical symptoms. D o r a was, according to Freud, really in love with H e r r K., a n d h e r r e a d y denials, w h e n Freud put the idea to her, were further proof of the desire a n d its repression. Freud, after the analysis was terminated, came to the further conclusion that Dora also h a d erotic desires for Frau K , a n d that she also repressed these from conscious awareness.
174
DISCOURSE STUDIES
Critics h a v e claimed that Freud fell into the trap of pathologizing Dora, the victim of a pathological, patriarchal family. Freud's m e t h o d o l o g y was to use the psychoanalytic dialogue as a m e a n s of uncovering repressions, which were p r e s u m e d to b e individual a n d non-dialogic. His evidence for Dora's repressed wishes c a m e from his interpretations of what D o r a was telling him. I n a later essay 'Repression', Freud specifically suggested that repression was 'individual in its operation' (19956, p . 572, emphasis in original). I n searching for the individual aspects of repression, Freud's attention was directed towards Dora's individual inner life - to her dreams, fantasies, private thoughts a n d so on. H e p a i d litde attention to the conversations i n which she would h a v e routinely participated. O n e might argue that these conversations were crucial for the situ ation, in which D o r a was caught. Moreover, if one takes the notion of the dialogic unconscious seriously, they might h a v e p r o v i d e d m o d e l s for the individual repression, in which Freud was interested. According to the account which Freud presented, H e r r K. sent D o r a flowers every day for a whole year, whenever h e was in the neighbourhood. H e con stantly b o u g h t h e r presents a n d spent 'all his spare time in h e r c o m p a n y ' (1995a, p. 188). Freud a d d s an astounding, b u t unelaborated, remark: H e r r K. pursued D o r a 'without h e r parents noticing anything in his behaviour that was character istic of love-making' (p. 188). How, one might ask, was this failure to notice achieved. It was, of course, a convenient failure for all concerned, except for Dora, the youngest a n d most vulnerable of the participants. W h a t was said b y Dora's parents w h e n the flowers were delivered yet again, or w h e n H e r r K. arrived with another w r a p p e d present for the adolescent? As h e stood in the hallway, with all the conventional d e m e a n o u r of a lover, what routines of greeting were developed? D i d t h e y m a k e jokes about his being a suitor? H e r e was a process of repression b e i n g socially p r o d u c e d in overt family activity, as all refused to draw the obvious conclusion or to ask the dangerous question. K.'s visits a n d flowers, b y virtue of their frequency, must h a v e taken o n the character of a habit. If jokes were uttered, did these b e c o m e family jokes, which b y their familiarity repressed the truth they expressed? O r did a silence, filled b y other words, b e c o m e routine? O n e might predict that a pattern was established, so that the repression was reinforced t h r o u g h routine, especially dialogic routine. I n short, o n e would want to k n o w h o w the routines of greeting, as well as the routine ways of talking about the visits w h e n the topic occurred in conversation, functioned to repress other, m o r e obvious interpretations. Similar discursive processes of avoidance can b e seen in the two examples of racist discourse, discussed earlier. T h e denial of prejudice is n o t challenged. Speakers use conventional expressions, which polite white speakers are not ex pected to dispute: ' I ' m n o t prejudiced b u t . . .' is rhetorically structured for a g r e e m e n t (Billig, 1991; Wetherell & Potter, 1992; v a n Dijk, 1992, 1993a). Questioning must b e absent for the conversation to proceed without 'flaw'. For instance, the English families, talking about royalty, w e r e collectively denying their own racism, while arguing for the impossibility of a non-white monarchy. A discursive projection was constructed: it was others - 'public opinion' or an amorphous 'them' - who would prevent any interracial marriage - not the Queen.
BILLIG
THE DIALOGIC UNCONSCIOUS
175
with w h o m the speakers w e r e claiming to identify as the figure of nationhood. In this way, the speakers were distancing themselves from the racist practices, which they were tacitly supporting b u t attributing to others. C o m p l e x discursive strategies were used to maintain the position. Speakers avoided the word 'white', with its connotations of racial purity and, thereby, of racism (for details, see Billig, 1997a). T h e discursive balance was preserved b y the avoidance of awkward questioning, as speakers collaborated to accomplish the collective projection. If repression is dialogically a n d socially constructed then topics of repression will v a r y culturally a n d historically. T h e bourgeois citizenry of Freud's V i e n n a might h a v e developed complex discursive codes to repress the topic of sexuality. However, the Viennese, whose elected mayor at the time of Freud's conversations with Dora was an openly racist demagogue, would b e surprised, perhaps shocked, by today's inhibitions about 'race'. T h e y would have difficulty adjusting to today's complex discursive codes of utterance and absence, b y which racism gains entry to respectable h o m e s with the d e m e a n o u r of denial, rather like H e r r K. being welcomed through the hallway with his b u n c h of flowers.
;
D i a l o g i c N a t u r e of R e p r e s s i o n
Finally, some implications from the notion of the dialogic unconscious can b e drawn in relation to Freudian ideas about the nature of repression. Necessarily these implications can only b e outlined briefly. Discursive psychologists have suggested that public discourse often provides the model for individual functioning. For example, Billig (1996) p r o p o s e d a rhetorical m o d e l of thinking, suggesting that internal debate is based u p o n public argumentation; children learn to think inwardly o n the basis of their participation in the outward rhetoric of debate (see also Shorter, 1993a, k, Wertsch, 1991). T h e same approach can b e applied to repression. T h r o u g h participation in dialogue, which represses as it expresses, people acquire the skills of repression. These dialogic a n d social skills, w h e n internalized, provide the basis for the sort of repression which Freud called individual in its operation. Although Freud m i g h t h a v e seen repression as the key concept of psycho analysis, his writings are somewhat sketchy about describing the routine oper ation of repression. Routines are to b e expected for, as Freud claimed, repression 'demands a persistent expenditure of force', for the repression of a n y particular thought is not 'an event which takes place once' (19955, p . 572). W h e n Freud attempted to describe h o w repression operates, h e often lapsed into conceptual difficulty or mechanical m e t a p h o r (see Soyland, 1993, a n d Gellner, 1985, for discussion of Freud a n d metaphor). I n ' T h e ego and the id', Freud envisaged the conscious part of the m i n d (the ego) repulsing unconscious forces from the id. Yet. the repulsion, accomplished b y the ego, must itself b e repressed. T h u s , the ego must b e u n a w a r e of itself a n d its accomplishments. Conceptually, therefore, the ego cannot b e identified with consciousness for in some respects it 'behaves just like the unconscious' (1995<£ p . 631). But h o w can consciousness b e just like the unconscious, prelinguistic id? Freud recognized that there were conceptual
176
DISCOURSE STUDIES
problems: 'We land ourselves i n endless obscurities a n d difficulties if w e k e e p to o u r habitual forms of expression (p. 631).' Part of the p r o b l e m lay i n Freud's insistence that repression was, at root, a non-linguistic process, which was as mysterious as the non-verbal, instinctual id. T h e r e were, h e admitted, linguistically constituted forms of repression, b u t these w e r e m o r e superficial, less basic forms. Freud, i n his essay o n 'Fetishism', spe cifically distinguished b e t w e e n repression p r o p e r (Verdrangung), w h i c h was non-linguistic, a n d disavowal (Verleugnung), which operated through language (Freud, 1977, p p . 352jff). Repression keeps affect away from consciousness, whereas disavowal is aimed at only the idea of the affect, rather than the affective forces themselves (see Laplanche & Pontalis, 1983, for further references to Freud's distinction). W h a t is clear is that Freud believed that t h e basic process, b y which the ego deflects t h e desires of the id, was t h e non-linguistic Verdrangung, rather than Verleugnung. A n d about t h e ways Verdrangung operated h e could say litde except b y w a y of the mechanical metaphor. It is possible to challenge the primacy of the non-linguistic forms of repression, while still keeping within the internal logic of Freudian theory. A s Freud stressed, the necessity for repression only arises as a consequence of social morality. In ' O n narcissism', h e wrote that 'for the ego t h e formation of an ideal would b e the conditioning factor of repression' (1995«; p . 558). Without a sense of conscience a n d a punitive ideal t o live u p to - t h e ego would n o t n e e d to censor desires. It can b e argued that, from a developmental point of view, t h e notions of morality only arise within language: indeed, as conversation analysis suggests, talk-ininteraction forms the basis of routine, practical morality. If language-in-interaction creates m o r a l imperatives, then it m a y also provide t h e m e a n s for their routine accomplishment. This w o u l d include t h e routines of dialogic repression. O n c e t h e issues are p u t like this, t h e n a step is m a d e towards clearing some of t h e mysteries of repression. If repression is thought t o b e individual, uncon scious a n d asocial, then t h e individual cannot learn (let alone b e taught) h o w to repress. Something mysteriously innate, a n d non-cultural, must spontaneously occur, w h e n cultural n o r m s m a k e their presence felt. O n t h e other hand, if language-in-interaction itself depends u p o n repression, then i n learning how to use language t h e child is learning lessons of repression. Polite talk will provide the models for individual repression, just as argumentation provides t h e models for internal deliberation (Billig, 1996). T h e child w h o learns t h e codes of discursive politeness, which are integral to learning to talk, learns h o w to change the subject, to avoid questioning ('it's r u d e to ask'), to turn threatening topics into humour, a n d so on. I n learning these dialogic skills, the child is acquiring routines, which can b e applied to their own individual mental life. Shockingly shameful thoughts m a y flash into consciousness; t h e individual knows h o w to change the subject, attribute thoughts to others, or tag t h e m with humour. A b o v e all, the person is acquiring routines of thought - or t h e buzz of internal chatter - which can keep the shocking waiting outside, pressing the bell a n d only gaining admission if it can appear, like H e r r K , as a friendly, respectable guest. If this line of argument is accepted, t h e n t h e routine conversations in Dora's household were not irrelevant to h e r 'condition'. They, a n d other conversations.
BILLIG
THE DIALOGIC UNCONSCIOUS
177
provide the models for a n y self-disavowals enacted within the interior mental life. T h e theoretical implications are considerable. O n the one h a n d , the notion of a dialogic unconscious serves to m a k e t h e basic processes of repression appear less mysterious a n d m o r e openly observable t h a n is posited in Freudian t h i n k i n g . O n the o t h e r h a n d , t h e dialogic u n c o n s c i o u s d e p i c t s o r d i n a r y conversations as m o r e psychologically a n d ideologically charged than they typically appear to conversation analysts. I n consequence, the idea of the dialogic unconscious opens the w a y to an unlikely combination of psychoanalytic a n d conversation analytic ideas in ways which could b r o a d e n a n d d e e p e n the project of discursive psychology.
Acknowledgements
w/**-
The author is grateful for the continuing support from members of the Discourse and Rhetoric Group at Loughborough University, and, in the present case, is particularly grateful to Derek Edwards andJonathan Potter for their unrepressed, productive disagree ment. The author also recognizes the helpful comments of Helene Joffe, John Shorter, Teun van Dijk and the anonymous reviewers.
Notes 1. Hollway (1989), in an interesting and original investigation, uses psychoanalytic ideas to investigate dialogue. In certain respects her work parallels some of the present arguments. However, she does not use conversation analytic notions, which stress the occasioned character of discourse. She tends to use Lacanian ideas of signification which, from a conversation analytic point of view, do not tie in closely with the details of the utterances, which she studies. 2. There has been a tendency for conversation analysts to study agreement rather than disagreement, or to see disagreement as a disruption to the joint accomplishment of social life. However, more analysts are beginningtosee the importance of argumentation in social life. To quote Schiffrin (1984), argument can be a 'form of sociability' (see also Billig, 1991, 1996; Grimshaw, 1990; Horowitz, 1996; Schiffrin, 1987; and see Gilbert & Mulkay, 1984, for analyses of argument in science). The effect of this stance on the theoretical vocabulary of conversation analysis remains to be seen. Already some analysts are counterposing the supposed 'preference for agreement' with an occasioned 'preference for disagreement' (Billig, 1991; Horowitz, 1996).
References \ntaki, C. (1994). Explaining and Arguing. London: Sage. Appignanesi, L. & Forrester, J. (1993). Freud's Women. London: Virago. Aronsson, K. (1991). Facework and control in multi-party talk: A pediatric case study. In I. Markova & K. Foppa (Eds), Asymmetries in Dialogue. Hemel Hempstead: Harvester Wheatsheaf. Billig, M. (1991). Ideology and Opinions. London: Sage.
I
J78
DISCOURSE STUDIES
Billig, M. (1992). Talking of the Royal Family. London: Roudedge. Billig, M. (1996). Arguing and Thinking, 2nd ed. Cambridge: Cambridge University Press. Billig, M. (1997a). Discursive, rhetorical and ideological messages. In C. McGarty & A. Haslam (Ed.), The Message ofSocial Psychology: Perspectives on Mind in Society. Oxford: Blackwell. Billig, M. (19974). Keeping the white queen in play. In M. Fine, L. Weis, L. Powell & M. Wong (Eds), Off-White. London: Routledge. Billig, M. (1997c). From codes to utterances: Cultural studies, discourse and psychology. In M. Ferguson & P. Golding (Eds), Beyond Cultural Studies. London: Sage. Billig, M. & Edwards, D. (1994). La construction sociale de la memoire. La Recherche, 25, 742-745. Bilmes,J. (1987). The concept of preference in conversation analysis. Language in Society, 17, 161-181. Boden, D. (1994). The Business of Talk. Cambridge: Polity Press. Brearley, M. (1991). Psychoanalysis: A form of life? In A. P. Griffiths (Ed.), Wittgenstein's Centenary Essays. Cambridge: Cambridge University Press. Brown, P. &Levinson, S. C. (1987). Politeness: Some Universals inLanguage Use. Cambridge: Cambridge University Press. Burman, E. & Parker I. (Eds). (1993). Discourse Analytic Research. London: Sage. Buttny, R. (1993). Social Accountability in Communication. London: Sage. Carbaugh, D. (1988). Talking American. Norwood, NJ: Ablex. Chilton, P. (1990). Politeness, politics and diplomacy. Discourse and Society, 1, 201-224. Cioffi, F. (1991). Wittgenstein on Freud's 'abominable mess'. In A. P. Griffiths (Ed.), Wittgenstein's Centenary Essays. Cambridge: Cambridge University Press. Cixous, H. & Clemont, C. (1986). The Newly Born Woman. Manchester: Manchester University Press. Daelemans, S. & Maranhao, T. (1990). Psychoanalytic dialogue and the dialogic principle. In T. Maranhao (Ed.), The Interpretation of Dialogue. Chicago, IL: University of Chicago Press. Decker, H. S. (1991). Freud, Dora and Vienna 1900. New York: Free Press. Drew, P. (1995). Conversation analysis. I n j . A. Smith, R. Harre & L. van Langenhove (Eds), Rethinking Methods in Psychology. London: Sage. Edwards, D. (1991). Categories are for talking. Theory and Psychology, 1, 515-542. Edwards, D. (1994). Script formulations: An analysis of event descriptions in conversation. Journal ofLanguage and Social Psychology, 13, 211-247. Edwards, D. (1995a). Two to tango: Script formulations, dispositions and rhetorical symmetry in relationship troubles talk. Research on Language and Social Interaction. 28, 319-350. Edwards, D. (1995 J). Sacks and psychology. Theory and Psychology, 5,579-596. Edwards, D. & Middleton, D. (1986). Conversational remembering and family relationships: How children learn to remember. Journal of Social and Personal Relationships, 5,3-25. Edwards, D. & Potter, J. (1992). Discursive Psychology. London: Sage. Edwards, D. & Potter, J. (1993). Language and causation: A discursive action model of description and attribution. Psychological Review, 100, 23-41. Freud, S. (1991). Jokes and their Relation to the Unconscious (1905). Harmondsworth: Penguin. Freud, S. (1977); Fetishism (1927). In On Sexuality. Harmondsworth: Penguin. Freud, S. (1995a). Fragment of an analysis of a case of hysteria (1905). In P. Gay (Ed.). The Freud Reader. London: Vintage. Freud, S. (19954). Repression (1915). In P. Gay (Ed.), The Freud Reader. London: Vintage.
BILLIG THE DIALOGIC UNCONSCIOUS
L
179
Freud, S. (1995c). An autobiographical study (1925). In P. Gay (Ed.), The Freud Reader. London: Vintage. Freud, S. (1995rf). The ego and the id (1923). In P. Gay (Ed.), The Freud Reader. London: Vintage. Freud, S. (1995*). On narcissism (1914). In P. Gay (Ed.), The Freud Reader. London: Vintage. Frosh, S. (1989). Psychoanalysis and Psychology. London: Macmillan. Gallop, J. (1986). Keys to Dora. In C. Bernheimer & C. Kahane (Eds), In Dora's Case. London: Virago. Garfinkel, H. (1967). Studies in Ethnomethodology. Englewood Cliffs, NJ: Prentice Hall. Gearhart, S. (1986). The scene of psychoanalysis: The unanswered questions of Dora. In C. Bernheimer & C. Kahane (Eds), In Dora's Case. London: Virago. Gellner, E. (1985). The Psychoanalytic Movement. London: Paladin. Gergen, K. (1994). Realities and Relationships. Cambridge, MA: Harvard University Press. Gilbert, G. N. & Mulkay, M. (1984). Opening Pandora's Box: A Sociological Analysis of Scientists'Discourse. Cambridge: Cambridge University Press. Grimshaw, A. D. (Ed.) (1990). Conflict Talk. Cambridge: Cambridge University Press. Harre, R. (1995). Discursive psychology. In J. A. Smith, R. Harre & L. van Langenhove (Eds), Rethinking Psychology. London: Sage. Harre, R. & Gillett, G. (1994). The Discursive Mind. London: Sage. Henriques, J., Hollway, W., Urwin, C , Venn, C. & Walkerdine, V. (1984). Changing the Subject. London: Methuen. Heritage, J. (1984). Garfinkel and Ethnomethodology. Cambridge: Polity Press. Hollway, W. (1989). Subjectivity and Method in Psychology. London: Sage. Hopkins, J. (1995). Wittgenstein, interpretation and the foundations of psychoanalysis. New Formations, 26, 54-73. Horowitz, A. D. (1996). 'A good old argument': The discursive construction of family and research through argumentation. Unpublished PhD thesis, Loughborough University. Irigaray, L. (1989). The gesture in psychoanalysis. In T. Brennan (Ed.), Between Feminism and Psychoanalysis. London: Routledge. Jacobus, M. (1987). Reading Woman: Essays in Feminist Criticism. London: Methuen. Katriel, T. (1986). Talking Straight. Cambridge: Cambridge University Press. LacanJ. (1979). The Four Fundamental Concepts ofPsycho-Analysis. Penguin: Harmondsworth. Lacan,J. (1986). Intervention on transference. In C. Bernheimer & C. Kahane (Eds), In Dora's Case. London: Virago. Laplanche, J. & Pontalis, J.-B. (1983). The Language ofPsycho-Analysis. London: Hogarth Press. Livingstone S. & Lunt, P. (1994). Talk on Television. London: Routledge. Loewenberg, P. (1985). Decoding the Past: The PsychohistoricalApproach. Berkeley: University of California Press. McHoul, A. W. (1987). Why there are no guarantees for interrogators. Journal of'Pragmatics, 11,455-471. Marcus, S. (1986). Freud and Dora: Story, history, case-history. In C. Bernheimer & C. Kahane (Eds), In Dora's Case. London: Virago. Masson, J. (1990). Against Therapy. London: Fontana. Middleton, D. & Edwards, D. (Eds) (1990). Collective Remembering. London: Sage. Moi, T. (1986). Representation of patriarchy: Sexuality and patriarchy in Freud's Dora. in C. Bernheimer & C. Kahane (Eds), In Dora's Case. London: Virago. Moscovici, S. (1976). La Psychanalyse: Son Image et son public, 2nd ed. Paris: Presses Universitaires de France. Mulkay, M. (1988). On Humour. Cambridge: Cambridge University Press.
180
DISCOURSE STUDIES
Nofsinger, R. E. (1991). Everyday Conversation. Newbury Park: Sage. Parker, I. (1990). The abstraction and representation of social psychology. In I. Parker & J. Shorter (Eds), Deconstructing Social Psychology. London: Routledge. Parker, I. (1992). Discourse Dynamics. Routledge: London. Parker, I. (1994). Reflexive research and the grounding of analysis: Social psychology and the psy-complex. Journal of Community and Applied Psychology, 4, 239-252. Pomerantz, A. (1984). Agreeing and disagreeing with assessments: Some features of preferred/dispreferred turn shapes. InJ. M. Atkinson &J. Heritage (Eds), Structures of Social Action. Cambridge: Cambridge University Press. Pomerantz, A. (1988). Offering a candidate answer: An information seeking strategy. Communication Monographs, 55, 360-373. Potter, J. (1996). Representing Reality. London: Sage. PotterJ. & Edwards, D. (1990). Nigel Lawson's tent: Attribution theory, discourse analysis and the social psychology of factual discourse. European Journal of Social Psychology, 20, 405-424. Potter, J. & Wetherell, M. (1987). Discourse and Social Psychology. London: Sage. Potter, J. & Wetherell, M. (1995). Discourse analysis. InJ. A. Smith, R. Harre & L. van Langenhove (Eds), Rethinking Methods in Psychology. London: Sage. Ramas, M. (1983). Freud's Dora, Dora's hysteria. InJ. L. Newton, M. P. Ryan &J. R. Walkowitz (Eds), Sex and Class in Women's History. London: Routledge and Kegan Paul. Rogow, A. A. (1978). A further footnote to Freud's 'Fragment of an analysis of a case of hysteria'. JourBa/o/ the American Psychoanalytic Association, 26, 331-356. Rogow, A. A. (1979). Dora's brother. International Review ofPsychoanalysis, 6,239-259. Rose, J. (1986). Dora: Fragment of an analysis. In C. Bernheimer & C. Kahane (Eds), In Dora's Case. London: Virago. Rose, N. (1990). Psychology as a 'social' science. In I. Parker & J. Shorter (Eds), Deconstructing Social Psychology. London: Routledge. Scheff, T.J. (1990). Microsociology. Chicago, IL: Chicago University Press. Schegloff, E. A.,Jefferson, G. & Sacks, H. (1977). The preference for self-correction in the organization of repair in conversation. Language, 53, 361-382. Schiffrin, D. (1984). Jewish argument as sociability. Language in Society, 13,311-335. Schiffrin, D. (1987). Discourse Markers. Cambridge: Cambridge University Press. Shotter.J. (1990). The social construction of remembering and forgetting. In D. Middleton & D. Edwards (Eds), Collective Remembering. London: Sage. ShotterJ. (1993a). Conversational Realities: Studies in Social Constructionism. London: Sage. Shotter, J. (1993b). The Cultural Politics of Everyday Life. Milton Keynes: Open University. Shorter, J. (1995). Dialogical psychology. InJ. A. Smith, R. Harre & L. van Langenhove (Eds), Rethinking Psychology. London: Sage. Sloterdijk, P. (1988). Critique of Cynical Reason. London: Verso. Soyland, A.J. (1993). Psychology as Metaphor. London: Sage. van Dijk, T. A. (Ed.) (1985). Handbook of Discourse Analysis, vols 1-4. London: Academic Press. van Dijk, T. A. (1992). Discourse and the denial of racism. Discourse and Society, 3,87-118. van Dijk, T. A. (1993a). Elite Discourse and Racism. Newbury Park: Sage, van Dijk, T. A. (19933). Principles of critical discourse analysis. Discourse and Society, 4, 249-283. van Dijk, T. A. (Ed.) (1996). Discourse Studies: A Multidisciplinary Introduction. London: Sage. Volosinov, V. N. (1994). Critique of Freudianism. In P. Morris (Ed.), The BakhtinReader. London: Edward Arnold.
I
BILLIG
THE DIALOGIC UNCONSCIOUS
181
Wertsch, J. V. (1991). Voices ofthe Mind. Hemel Hempstead: Harvester Wheatsheaf. Wetherell, M. & Potter, J. (1992). Mapping the Language of Racism. Hemel Hempstead: Harvester Wheatsheaf. Wilkinson, S. & Kitzinger, C. (Eds) (1995). Discourse and Feminism. London: Sage. Wittgenstein, L. (1953). Philosophical Investigations. Oxford: Blackwell. Wittgenstein, L. (1967). Zettel. Oxford: Blackwell. Wittgenstein, L. (1980). Remarks on the Philosophy ofPsychology. Oxford: Blackwell. Wittgenstein, L. (1988). Lectures on Philosophical Psychology, vol. 1. Brighton: Harvester.
Investigating Narrative Inequality: African Asylum Seekers' Stories in Belgium Jan Blommaert
1. Introduction
I
n a remarkable paper, Dell H y m e s a n d Courtney C a z d e n investigated 'the possibility that one form of inequality of opportunity in our society has to do with rights to use narrative, with whose narratives are admitted to have cognitive function' (Hymes a n d Cazden, 1980:126).' O n the basis of observations on speaking rights in university classes they concluded that t h e use of parti cular ways of narrating focused u p o n the expression of emotions a n d personal experience and voiced in an 'anecdotal' m o d e was easily dismissed, while other narrative m o d e s in which academic voicing a n d emotional d e t a c h m e n t were m o r e p r o m i n e n t were clearly privileged. Thus, 'contributions to class discussions b a s e d o n narratives of personal experience did n o t get the floor' (1980: 127) and 'the truth of the matter would b e that only the "anecdotes" of s o m e would count' (1980: 131). T h e rights to use particular narrative m o d e s are unevenly distributed, a n d this pattern of distribution disenfranchises those w h o have to rely on 'disqualified' narrative m o d e s for conducting their business in society.
M y aim in this article is to d o c u m e n t a n d discuss autobiographical stories told b y African asylum seekers in Belgium. I will try to highlight aspects of the structure and functions of narratives in a n attempt to show h o w they represent crucial communicative resources for asylum seekers. Without recourse to the long and detailed narratives about h o m e , escape a n d travelling, asylum seekers cannot m a k e their motives a n d causes for seeking asylum fully understood. I n high lighting these aspects, I display a concern for narrative-textual s h a p e inspired b y authors such as H y m e s (1981, 1996, 1998) and H a v i l a n d (1996, 1997), as well as one for narrative-textual dynamics inspired b y w h a t has c o m e to be
Source: Discourse & Society vol. 12, no. 4,2001, pp. 413-449.
BLOMMAERT
INVESTIGATING NARRATIVE INEQUALITY
183
k n o w n as a 'natural histories of discourse' a p p r o a c h (Bauman a n d Briggs, 1990; Briggs, 1997; Gal a n d Woolard, 1995; Silverstein a n d U r b a n , 1996) emphasizing entextualization practices (de- a n d recontextualization practices) as crucial ingredients of 'interpreting' a n d 'understanding' text. T h e s e approaches to text will b e set within a wider project that bears affinities to that of critical discourse analysis a n d can b e summarized as investigating language-as-social-process in order to gain a m o r e precise insight in p o w e r relations (see, e.g. Wodak, 1995). T h e p r o b l e m I wish to address through a n analysis of these stories is that of narrative inequality in the context of asylum applications in Belgium. T h e asylum procedure involves a complex set of discursive practices and language ideologies that are, in practice, being used as criteria for 'truth', 'trustworthiness', 'coherence' and 'consistency'. Such discursive practices require access to communicative resources that are often far b e y o n d t i e reach of African asylum seekers n o t only linguistically b u t also narratively a n d stylistically. Like so m a n y bureaucratic procedures, the asylum application p r o c e d u r e used to b e 'blackboxed'. H e n c e , direct evidence of the official p r o c e d u r e inter views is n o t (yet) available, a n d I h a v e b e e n forced to use the research tactic employed by, for example, L a b o v in his Logic of Nonstandard English (1970) (as well as, e.g. b y H y m e s reanalysing Labov's data in H y m e s , 1996: chapter 8); I have also h a d to use 'second best' data a n d look at preconditions for talk in these data. T h e asylum seekers' narratives u n d o u b t e d l y illustrate the w a y in which communicative resources are mobilized for 'making sense' of the asylum seekers' case; at the same time, t h e y suggest at least the possibility of narrative inequality w h e n measured against the n o r m s a n d expectations inscribed in the discursive patterns of the asylum application procedure. T h e lack of attention to the crucial functions of densely contextualized (and contextualizing) narratives is partly d u e to the particular treatment of text in bureaucratic procedures, particularly the shaping of textual trajectories in which 'original' stories are continuously reformed a n d reformulated. H e n c e it illustrates the fact 'that talk is often structured vis-a-vis m e d i a t e d relationships it bears to objects a n d texts that are dispersed in time a n d space' (Briggs, 1997:454-5) - in this case, vis-a-vis preconceived criteria of textuality a n d narrative appropriate ness that are inscribed in practices of noting, summarizing and reading narratives, making 'files', interviewing a n d interrogating, translating and so forth (together with t h e i r p r o d u c t s : specific t e x t s f o r m a t s a n d i n d i v i d u a l texts) w i t h i n administrative procedures, themselves part of a h u g e text tradition such as that of Belgian law a n d bureaucracy. I argue that this process of (re)structuring talk into institutionally sanctioned text involves a d y n a m i c of entextualization that is based o n p o w e r asymmetries. This process also involves the problematic of the availabihty a n d accessibility of linguistic-communicative resources - an often overlooked 'context' of talk, b u t o n e that will b e given p r o m i n e n c e in this article (see also Blommaert, 2001). Narrative inequality assumes a variety of shapes a n d has a multitude of dimensions. T h r e e dimensions will b e discussed in some detail h e r e : (i) the accessibility a n d availability of linguistic-communicative resources; (ii) particu lar contextualization demands; a n d (iii) the ways in which texts are shifted through 2
184
DISCOURSE STUDIES
the procedure. I n m y discussion, I focus o n preconditions for discourse: the sort of 'material' for communication that asylum seekers h a v e access to a n d what it m e a n s to b e confronted with a bureaucratic system in which particular 'materials' are expected a n d required. I n the n e x t section, I briefly provide some background o n the situation of asylum seekers in Belgium a n d o n the fieldwork project in which the data discussed h e r e w e r e gathered. I n section 3 , 1 turn to the issue of communicative resources a n d discuss some of the features of language competence observable in the talk of the asylum seekers. Section 4 presents a discussion of the particular type of sub-narrative that I call ' h o m e narratives'. H o m e narratives are often long a n d sometimes anecdotal stories on the situation in the refugees' h o m e societies, involving usually very detailed information o n local events, the crisis from which refugees fled and so on. H o m e narratives are contextualizing ac counts, a n d the particular contextualization trajectories they follow require close inspection. I n section 5 , 1 try to situate the h o m e narratives in the discursive patterns that occur during the asylum application procedure.
2. Asylum Seekers in Belgium T h e data discussed h e r e were gathered during a fieldwork project in which more than 4 0 African asylum seekers w e r e interviewed, most of t h e m b e i n g illegal or in limbo, their asylum application either being undecided or having b e e n rejected. T h e interviews were o p e n interviews in which asylum seekers w e r e invited to tell the story of their escape, their reasons for escape, their experiences with the Belgian asylum procedure a n d with life as a n asylum seeker in Belgium. The interviews w e r e conducted in French, English a n d (in a small n u m b e r of cases) Dutch. D u t c h was the m o t h e r tongue of the interviewers; thus in most cases the interviewers a n d the interviewees h a d to rely o n a language other than their native language to d o the interview, a n d as a rule non-native varieties of those languages were being used. 3
T h e interviews were obviously very different from the ones taken in the official application procedure. T h e m a i n difference was, however, n o t the open format of our interviews. T h e official interview also contains important 'open' slots. T h e s e o p e n interview parts obviously offer a lot of material for detailed scrutiny a n d cross-questioning b y the officials, a n d so b e c o m e a threat rather t h a n an opportunity for asylum seekers. A n o t h e r shared feature between our interviews a n d the ones held in the procedure was the use of non-native languages as a lingua franca in the interview, implications of which will b e discussed in section 3 of this article. T h e m a i n difference was probably the key in which the interviews were put: ours were obviously inconsequential (though some subjects either suspected or h o p e d that the interview they did with us would have some benefit), and overall they were held in a friendly a n d sympathetic atmosphere. I n Belgium as elsewhere in E u r o p e , asylum seeking has b e c o m e a topic of h e a t e d political debate in recent years. T h e last d e c a d e has witnessed a rapid — „ T k i c / - n m h i n e d w i t h the increased complexity 1
BLOMMAERT
INVESTIGATING NARRATIVE INEQUALITY
185
of the cases - m a n y asylum seekers d o not offer a 'typical' motive for seeking asylum, as we shall see further - has led to bottlenecks in the administrations in charge of asylum regulations. I n the past, it took years to reach a decision o n some applications. In the m e a n t i m e , applicants lived in the country in appalling conditions of poverty a n d marginalization (pushing t h e m into the crudest systems of labour exploitation, including prostitution); often the administrative treatment was reduced to a very superficial inspection of the application in attempts to speed u p the procedure a n d reduce the backlog of applications. T h e infrastructure for hosting asylum seekers rapidly (and chronically) p r o v e d to b e grossly inade quate a n d violence b e c a m e a constant ingredient of the treatment of asylum applicants b y the police force. This flooding of the administrative apparatus went h a n d in h a n d with negative stereotyping of asylum a n d asylum seekers (Blommaert a n d Verschueren, 1998: chapter 8), associating asylum seekers with crime a n d qualifying m o s t of t h e m as ' a d v e n t u r e r s ' w h o s o u g h t a s y l u m exclusively for economic purposes (in popular parlance: 'to take advantage of our generosity'). In September 1998, a y o u n g Nigerian w o m a n called Semira A d a m u died during h e r forced repatriation. H e r asylum application, based o n the a r g u m e n t that she would b e forced to m a r r y a n old a n d violent m a n in Nigeria, h a d b e e n rejected. She was p u t o n a plane for Togo a n d seated between two policemen; when she started to shout, the police officers p u t a small pillow over h e r m o u t h . Semira A d a m u went into a c o m a a n d died. H e r violent death caused a public outcry both a m o n g asylum seekers and a m o n g the public at large. Asylum seekers came out of h i d i n g a n d demonstrated, occupying churches a n d schools. A number of organizations were formed for the i m p r o v e m e n t of the living con ditions and the chances of asylum for asylum seekers, and the government ordered a public inquiry into the procedures of repatriation of rejected applicants. For a brief period, asylum was a priority o n the political agenda, a n d the negative stereotyping of asylum was transformed into a public image of victimhood. I n die spring of 1999, the protests gradually faded a n d the issue of asylum seekers resumed its previous shape a n d course. As shown later, the asylum p r o c e d u r e relies heavily on an investigation of applicants' stories. O n their arrival in Belgium, they are interviewed about the causes and motives for fleeing their country a n d seeking asylum. T h e general format applied in this procedure is that of a criminal investigation. Well over H) percent of the applications are turned down, often - as we shall see - o n v e r y doubtful grounds. PoUtically, the 'control' of immigration (i.e. the reduction of the number of effective immigrants) combined with a policy o f ' h u m a n e repatriation' f rejected applicants is a generally accepted doctrine. T h e n e w p h e n o m e n o n of massive asylum seeking is rarely seen as an issue in its o w n right, rather it is associated with the m o r e general issue of immigration and immigrant policies in Belgium, a n d it is usually presented as a further complication of an already thorny issue of 'integration'/assimilation of foreigners into the local cultural, linguistic and social communities. !
The discursive patterns b y m e a n s of which these political categories are being constructed and «ihia<-°'i — * " 1
r
DISCOURSE STUDIES
186
society are b e y o n d the scope of this article (see Blommaert, 1997; Blommaert a n d Verschueren, 1998); suffice it to state h e r e that the issue of asylum seekers is obviously o n e in which rhetorical accomplishments and discursive constructions of reality are crucial. T h e politics of asylum in Belgium is a politics of represen tation in which discursively constructed a n d disseminated gross categories (e.g. 'political', h e n c e 'legitimate' asylum seekers versus 'economic', h e n c e 'illegitim ate' asylum seekers) are crucial political instruments. T h e historical situatedness of our data deserves some attention. T h e interviews w e organized with asylum seekers were conducted between O c t o b e r / N o v e m b e r 1998 and March/April 1999, i.e. during the brief period in which asylum seekers c a m e out of hiding a n d w e r e eager to tell their stories to all those interested, and during which there was some positive m e d i a a n d political attention for these stories. This accounts for the fact that w e could collect data at all. D u r i n g these few m o n t h s , asylum seekers b e c a m e visible a n d their stories b e c a m e publicly accessible. Second, it explains the nature of their stories: they are apologetic a n d argumentative, using a discursive space shaped b y the Semira A d a m u crisis, in which n e w views of asylum seeking, motives a n d experiences could be articulated. For a brief period, asylum seekers h a d a legitimate voice, and their case was taken u p b y m a n y opinion makers a n d politicians.
3. Resources
i
O n e of the characteristics of the administrative world we live in is the unchal l e n g e d a n d a p p a r e n t l y u n c h a l l e n g e a b l e a s s u m p t i o n that b u r e a u c r a t i c and administrative clients would h a v e complete control over the m e d i u m a n d com municative skills in which bureaucratic a n d administrative procedures are being carried out. I n Belgium (as u n d o u b t e d l y elsewhere) administrative procedures require highly developed literacy skills as well as (some degree of) access to a standardized variety of a language (in Belgium: Dutch, French or German). It should b e noted, in passing, that literacy requirements seem to increase in size a n d scope the lower o n e gets into society. People on low incomes often have to go through a mass of c o m p l e x a n d very diverse p a p e r w o r k in order to get social welfare benefits, privileged access to social housing, medical treatment or edu cation. Thus, J i m Collins' comment that ' m o d e r n educational systems produce stratified literacies: elites are socialized to an interpretive relation to texts, and nonelites to a submissive relation to texts' (1995:84) becomes pressingly relevant T h e first requirement, literacy, is taken for granted and is rooted in a sociocultural tradition in which generalized schooling provides (stratified patterns of) literacy to a^Belgian citizens. Being literate is a sociocultural given in Belgium. T h e second requirement is m o r e controversial a n d multilayered: the emphasis o n standard varieties is rooted in political-linguistic struggles of the past, discus sion of which would take us outside t h e scope of this article. Standard varieties of languages were used as e m b l e m s of national identity in m a n y parts of the world; the outcome of the struggle in Belgium gave rise to a highly complex a n d politically very sensitive language legislation in which tolerance for other
BLOMMAERT INVESTIGATING NARRATIVE INEQUALITY
187
speech varieties (nonstandard forms as well as foreign languages) is restricted. Thus, nonstandard varieties are allowed in the domain of orality a n d non-official language use; as soon as the bureaucratic a n d administrative arenas are b e i n g entered, the standard (written) variety of the language is imposed. T h e fact is that the communicative requirements imposed o n clients are conditioned b y historically contingent p h e n o m e n a such as the social distribution of communicative resources through, for instance, the education system, the mass m e d i a a n d so forth. I n m o r e t h a n one way, the requirements thus presup pose m e m b e r s h i p of the society, or at least of social units with similar linguistic economies and communicative sociologies. Assumptions of choice in pragmatic theories (for instance in conversation analysis) are very often based o n the implicit acceptance of such m e m b e r s h i p , a n d tend to obscure significant inequalities in the range of possible choices some people can control. Constraints o n choice, anchored in inequality of speech repertoires, are certainly crucial features of our asylum seekers' data, a n d c o m p e t e n c e in t h e m e d i u m of narration is o n e issue. It m a k e s a significant difference whether the narrator uses a language or language variety over which she or h e has adequate control. Narrating in a second, third or other foreign language m a y considerably reduce the set of resources which speakers can select for structuring their story a n d thus for 'making their point'. As w e shall see, this does n o t pre-empt the fact that they m a k e their point; it does h o w e v e r influence the w a y in which interlocutors perceive their story and get the point. In what follows, I first d o c u m e n t some p r o b l e m s in language proficiency. After that, I turn to what I call 'ethnocoherence': patterns of coherence that can be discerned in such narratives, despite the sometimes considerable problems of language proficiency. 3.1 Problems of Language Proficiency Let us n o w take a closer look at some examples from the corpus. T h e y illustrate the difficulties experienced b y m a n y speakers in expressing themselves in French, English or Dutch. M o r e often than not, stories are told in hesitant styles packed with self-corrections a n d mid-sentence changes of tactic, footing or topic, less than adequate lexical selection, problems in verb inflection (e.g. tense a n d aspect marking), p r o b l e m s in selecting adequate p r o n o u n s , a n d so on. At the same time, the stories that are being told are topically a n d structurally quite complex, as we shall see in the n e x t section. T h e p r o b l e m s with language proficiency considerably complicate matters. Extract 1 illustrates the types of stories which result from the use of 'broken' French coupled with a detailed a n d complex h o m e narrative. T h e interviewee, a man from Angola, is explaining an episode of Angolan history to the interviewer. 4
Extract 1 oui/Pautre president.. .(xxxxxx)/on l'a empoisonne/c'est le president Mobutu/ qui a mis le poison retarde/ il est parti au russe/ l'URSS/ pour traiter/ il a retourne/ il est mort/ mais on a abandonne son corps hein/oui/ {{Question: c'etait un President
188
DISCOURSE STUDIES
de MPLA ?}} c'etait le meme mouvement MPLA/ dans le temps/ annee septantecinq/ quand il est mort on dit/ comme on = = i l est marxisme/ on a pris on a choisi =on= on a fait faux testament/ cette testament c'etait au temps du russe qui a fait ga/ comme toi tu = l e = le president il est mort/ il a decide Eduardo qui va me remplacer/ sans vote/ parce que il est toujours du meme parti/ Eduardo il est d'origine angolais/ mais il est des Cap Verdiens/ parce que ce sont des anciens prisonniers/ et Portugais il a mis a Pile hein/ nous sommes a l'ocean/ et on a mis une prison la-faas/ parce qu'il est venu pour commander l'independance/ c'etait une petite ville =une petite= une petite village/ on a mis au pouvoir/ maintenant le president/ c'est on dit/ il dit que non/ tous les gens/ qui parlent Lingala/ les gens du Nord/ ce sont des gens plus malins/ plus intelligents/ par rapport au gens du Sud/ en Angola nous sommes quatre couleurs/ comme le Bre=le Bresil. Translation: yes/the other president... (xxxxxx)/they have poisoned him/ it's president Mobutu/ who put the delayed poison/ he has left to Russian/ the USSR/ to treat/ he gave back/ he died/ but they have left his corpse, right/ yes/ {{Question: it was a president of the MPLA ?}}/ it was the same movement MPLA/ in those days/ year seventyfive/ when he died they say/ like they=he is Marxism/ they took they chose= they=they have made false testament/ those testament it was in the time of Russian that has made it/ since you you=the=the president is dead/ he decided Eduardo who is going to replace me/ without vote/ because he is always of the same party/ Eduardo he is of Angolan origin/ but he is of the Cape Verdians/ because they are former prisoners/ and Portuguese has put on the island, right/ we are at the ocean/ and they have put a prison over there/ because he had come to command the independence/ it was a small town= a small=a small village/ they have put to power/ now the president/ that is what they say/ he said that no/ all the people/ who speak Lingala/ the people from the north/ they are more clever people/ more intelligent/ in relation to the people from the south/ in Angola we are four colours/ like Bra=Brazil A s demonstrated in the next section, a tremendous amount of information is squeezed into narrative passages such as this one. T h e point I wish to m a k e here is that the narrative structure a n d the salience of w h a t the m a n tells us are over shadowed b y the code in which h e tells it: a variety of colloquial a n d informally acquired French in which grammatical, syntactic a n d lexical errors are frequent w h e n m e a s u r e d against normative standard French. T h e sequence of events is not m a r k e d b y the usual tense a n d aspect markers; n o u n s a n d adjectives are used as synonyms ("il est m a r x i s m e " instead of "il est marxiste"; "russe" instead of "Russie"); articles are deleted ("Portugais" instead of "le Portugais" or "les Portugais" and so on. T h e Angolan m a n is struggling with the code. Hence, contextually crucial and sensitive episodes such as the o n e in which h e describes the change of p o w e r in his country (crucial a n d sensitive because they form the basis of the reasons w h y h e escaped from his country) are narrated as: Extract 2 dans le temps/ annee septante-cinq/ quand il est mort on dit/ comme on ==il est marxisme/ on a pris on a choisi =on= on a fait faux testament/ cette testament r-'fltait mi b m n s An russe aui a fait ca/ comme toi tu =le= le president il est mort/ if
BLOMMAERT INVESTIGATING NARRATIVE INEQUALITY
189
a decide Eduardo qui va me remplacer/ sans vote/ parce que il est toujours du meme parti/ Such ways of narrating crucial a n d sensitive matters are offered to interviewers who are also non-native speakers of French. But there are various sorts and degrees of 'non-nativeness', having to d o with what sort of variety has b e e n acquired b y the speakers, through which types of channels a n d m e a n s (formal-informal learning, spoken a n d / o r written, which genres). I n this case, the interviewers were Belgian-Flemish highly literate university students with a m o r e or less developed c o m p e t e n c e in 'schoolbook French', i.e. a variety of the standard language rarely used in practice. Students such as the group of interviewers in this project d o n o t h a v e m u c h exposure to French, a n d they actively use it only on relatively rare occasions. So the interaction situation is o n e in which o n e party uses - a n d I adopt Fergusionian terminology for the m o m e n t - a very 'low' variety of French a n d another a v e r y 'high' variety, while b o t h parties have difficulties with the production of spoken French a n d n o n e of t h e m is 'fluent'. T h e passage in Extract 2 therefore appears very confusing, rambling and incomprehensible to the Belgian interviewer. T h e asylum seekers often acquire their varieties of E u r o p e a n languages in informal circumstances: outside school, through exposure to specific varieties of speech, with dialect a n d / o r j a r g o n influences. This, as well as the effect it has on the interaction with the interviewers, can b e seen in Extract 3 w h e r e the same Angolan m a n (P) explains the sort of currency used in Angola: 5
Extract 3 P: mais en Angola on mange les dollars/ nous avons les Oj on mange? P: oui on mange= =d=dollars/ au magasin on achete en dollars / a cause de (mon= au magasin on dit) c'est les magasins des francais/ Elf Translation: P: but in Angola we eat dollars/ we have the Qj we eat? P: Yes we eat==d'dollars/ in the shop we buy in dollars/ because (sho= in the shop they say) it's the shop of the French/ Elf The interviewer ^ i n t e r r u p t s P with a request for clarification o n the phrase "on mange les dollars". T h e phrase "manger" followed b y the n a m e of a currency is a colloquial Congolese French expression m e a n i n g 'to use a specific currency'. Interestingly, P does n o t recognize OJs request for clarification as referring to -manger" b u t understands it as a request to clarify w h y dollars axe being used as currency in Angola. So his answer elaborates on the fact that the shops are owned b y the French petrol c o m p a n y Elf. The next e x a m p l e illustrates h o w expressing sometimes even quite simple concepts or e x p e r i e n c e s c a n b e c o m e h i g h l y p r o b l e m a t i c b e c a u s e of t h e iimitations imposed b y the code in which the story has b e e n told. A w o m a n from Angola tells her experiences in a detention p p n i r o ac*A*,~ — i — rn6
DISCOURSE STUDIES
190
tries to explain that there is a problem with ventilation in the centre; the whole fragment in Extract 4 is a struggle to find the right way to express 'ventilation'. Extract 4 il y en a des families qui vivent ici/ des enfants/ mais.. II n'y a pas des choses de vapore==pour vaporiser des cigarettes/ parce que la cigarette la-bas ce n'est pas de==c'est de==de tabac/ les gens il va fumer du matin au soir/ il n'y a pas de • • fenetres pour sortir euh/ enlev==entrer Ie vent==il n'y a pas/ presque tout c'est ferme/ dans les toilettes c'est la meme chose/ des fenetres ga n'ouvre pas Translation: there are families who live here/ children/ b u t . . . there are no things to vapore= to vaporize cigarettes/ because the cigarette there it is not the== it's = tobacco/ the people he smokes from morning till evening/ there are no windows to get out ehr/ take aw=get wind in=there are none/ almost everything is closed/ in the toilets it's the same/ the windows they don't open Metapragmatic framing, speech act selection and other phenomena that belong to the sociocultural anchoring of language are also sources of com municative problems in our data. Often, the use of a common medium in inter views between native and non-native speakers triggers the assumption from the native speaker that the interlocutor not only shares the linguistic code, but also the pragmatic and metapragmatic codes. Giving metapragmatic instructions about the way in which the interview will be conducted (a common opening phase of research interviews) is one such locus of problems. The following frag ment is taken from the start of an interview with R, a male refugee from the Ivory Coast. The tape starts running, and the interviewer B begins b y giving some 'director's instructions' to R. Extract 5 B: Ik denk dat het gemakkelijkst zou zijn alsje begint/met hoe datje naar hier bent gekomen ehr R: =met vliegtuig B: ja/waarom wa=waarom je naar hier gekomen bent/ehr/en hoe je hier geraakt bent/dat jij gewoon/ we gaa=we gaan niet veel vragen stellen/'t is wat/dat jij gewoon praat/wat praat R: ja B: zeggen watje.. R: =ja het is heel moeilijk om te zeggen (...) 1
Translation: B: I think that it would be easiest if you start/with how you got here ehr R: =by plane B: yes/why wh=why you came here/ehr/and how you got here/that you just/ we won=we won't be asking many questions/it's what/that you just/talk a bit R: yes B: say what you., n • =irn
I
BLOMMAERT
INVESTIGATING NARRATIVE INEQUALITY
191
B's intention is to provide some metapragmati'c instructions for the interview. But mentioning one topic of inquiry "how y o u got h e r e " in line (1) triggers a response from R ("by plane", line 2). R h a d interpreted line 1 as a question, n o t as part of a set of preliminary instructions. B repairs the mistake in line 3 - 4 , R acknowledges this in line 5, a n d B h a n d s the floor to R in line 6. Two problems can b e identified with respect to R's misinterpretation of B's statement in line 1. First, the interview was in Dutch, B's m o t h e r tongue. Dutch, for R, is a new, foreign language which h e masters admirably but not completely. T h e original utterance in line 1, spoken b y B, was "hoe datje naar hier bentgekomen" ("how-thatyou-to-here-have-come"), in which the vernacular D u t c h 'hoi ('how') is m e a n t to stand for the total set of circumstances of R's c o m i n g to Belgium. It is n o t u n c o m m o n in varieties of Dutch to ask " h o w [Dutch hoe] did you get h e r e " with the intention of learning the reasons for one's b e i n g there. But R, a non-native speaker of Dutch, only picks u p the 'typical' instrumentalmeamngoi ' h o e ' ('how'), and answers that h e came h e r e "met vliegtuig" ("with-plane"). B's repair in line 3 consists of differentiating a n d specifying the precise semantic directions of his questions - 'how' as well as 'why'. So p a r t of the mistake is based o n differential competence in Dutch, m o r e precisely, it is based o n R's lack of familiarity with the semantic a n d pragmatic nuances of colloquial Dutch. Second, there is also a mistake in speech act interpretation. R fails to pick u p the declarative nature of B's utterance in line 1, a n d interprets it instead as a question to which h e has to reply. I n all likelihood, R interprets the circumspect formula tion "I think that it would b e easiest if y o u start/with h o w you got h e r e " as an indirect, deferent question, rather than as the neutral declarative statement intended by B. Both problems have to do with the intercultural nature of the communicative event. T h e p r i m a facie s h a r e d n e s s of the m e d i u m - R h a s a n a d m i r a b l e knowledge of D u t c h - can trigger a n illusion that n o t only the lexical and grammatical aspects of the language but also the complete set of sociocultural, pragmatic a n d metapragmatic aspects are shared b y the participants. This is not usually the case. I n this fragment we see that differential competence in a language can b e hardly noticeable, as it can lie in aspects of language that are less clearly visible than, for example, w r o n g lexical choices or grammatical (e.g. gender, article or inflectional) errors. 3.2 Ethnocoherence The various difficulties with competence in the m e d i u m of the talk, reviewed so far, all suggest that stories are told in 'simple' varieties of languages. But this does not m e a n that such stories are 'simple', n o r that narrators fail to bring about significant degrees of narrative structure or coherence in their stories. Even though the language m a y b e simple a n d plain, and despite 'errors' in the lan guage, stories can b e narratively complex a n d display obvious patterns of coher ence. But coherence seems to emerge in peculiar (ethno) ways, since speakers have no access to the full potential of g r a m m a r a n d syntax of the languages they have to use.
192
DISCOURSE STUDIES
Let m e illustrate this with an example from our data. The fragment is taken from the beginning of an interview with Habiba, a Somali woman. H is Habiba; A and B are interviewers. Extract 6 H:
A H A: H: A H A: H A B+ H
I'm from *Somalia and my name is Habiba Mohammed and 1=1 have *five childrens and I coming here before the children are coming=when I was euh when I=I'm arrive in Belgium I was *alone\ ah\ yeah\in sake of the war=the war of Somalia\ uhuh\ . •• And. I w=I'm. Twen=*thirty five years old\ uhuh\ and euh I was working in Somalia ICRC International red Cross that's ICRC *Red Cross\ ahOKOKjaja H: [acknowledge] and I was euh office assistant ja yeah. So Somalia is starting war "nineteen ninety one i uhuh\ so until ninety one to ninety five I was in Somalia uhuh\ and [baby starts crying] wa [laughs] and I have *four children at that time and euhm.. My husband comes from euh *north Somalia uhuh and I *south Somalia is fighting north at=at south is fighting\ uhuh so my=my husband and my children have no. *safety for their lives uhuh
H a b i b a clearly h a s difficulties speaking English. H e r statements contain Dutch caiques such as "in sake" (from D u t c h "in spice", i.e. 'concerning', erratic plural marking ("childrens") a n d verb inflection ('and I coming here') a n d so on. But let us re-transcribe Habiba's narrative, deleting the interviewer's backchannelling interventions as well as the clarifying insertion sequence (on ' I C R C ' , two turns). T h e s e backchannelling interventions are important, because they support the structuring of H a b i b a ' s narrative; b u t b y deleting t h e m we arrive at a n u m b e r of narrative statements: Extract 7 1. I'm from "Somalia and my name is Habiba Mohammed and 1=1 have *five childrens and I coming here before the children are coming=when I was euh when I=I'm arrive in Belgium I was *alone\ 2. yeah\in sake of the war=the war of Somalia\ 3. and. I w=I'm. Twen=*thirty five years old\
BLOMMAERT
INVESTIGATING NARRATIVE INEQUALITY
193
4. 5. 6. 7. 8.
and euh I was working in Somalia ICRC International red Cross J , and I was euh office assistant So Somalia is starting war *nineteen ninety one so until ninety one to ninety five I was in Somalia and wa= and I have *four children at that time and euhm.. My husband comes from euh "north Somalia 9. and I *south Somalia is fighting north at=at south is fighting\ 10. so my=my husband and my children have no. *safety for their lives There is a considerable degree of narrative structure and coherence in this frag ment, despite the ' b r o k e n ' English in which it is m a d e . First, H a b i b a brings n e w information in each of h e r statements. T h e statements d o not overlap; each of them introduces a n e w element in the story. Also, there is a clear break in this sequence of narrative statements. Although each of the statements adds n e w information, H a b i b a m a r k s a thematic b r e a k b e t w e e n s t a t e m e n t 5 a n d 6. Statements 1-5 all refer to H a b i b a herself, they identify h e r (name, age, country of origin, profession); statements 6-10 are a b o u t '[the war in] Somalia' a n d provide background to the reasons w h y she came to Belgium. T h e break between both thematic parts is m a r k e d b y cohesive devices: " a n d " in the first part, "so" (and "and") in the second. T h e use of these particles creates a complex pattern of information in the story: Extract 8 PARTI 1. I'm from "Somalia 2. and my name is Habiba Mohammed 3. and 1=1 have *five childrens 4. and I coming here before the children are coming {clarification} =when I was euh when I=I'm arrive in Belgium I was *alone\ yeah\in sake of the war=the war of SomaliaN 6. and. I w=I'm. Twen=*thirty five years old\ 7. and euh I was working in Somalia ICRC International red Cross 8. and I was euh office assistant PARTII 1. So Somalia is starting war "nineteen ninety one 2. so until ninety one to ninety five I was in Somalia 3. and wa= and I have *four children at that time 4. and euhm.. My husband comes from euh "north Somalia 5. and I "south Somalia is fighting north at=at south is fighting\ 6. so my=my husband and my children have no. "safety for their lives We can go a bit further. I n part I I , two levels can b e distinguished. N o t all the statements in part I I are equally relevant. Statements 1, 2 and 6 are ' m a i n ' statements, setting important argumentative a n d narrative frames. Statement 1 introduces the general historical frame of the war in Somalia; statement 2 places Habiba in that historical frame, and statement 6 draws a general conclusion from this general sketch of Habiba's family's situation in the war. Statements 3 , 4
DISCOURSE STUDIES
194
a n d 5 elaborate on statement 2: they clarify a n d specify statement 2, H a b i b a (and h e r family) living in Somalia during the w a r years. T h u s we arrive at the following structure for p a r t I I : Extract 9 PART II 1. So Somalia is starting war "nineteen ninety one 2. so until ninety one to ninety five I was in Somalia 3. and wa= and I have *four children at that time 4. and euhm.. My husband comes from euh *north Somalia 5. and I *south Somalia is fighting north at=at south is fighting\ 6. so my=my husband and my children have no. *safety for their lives T h e m a i n statements are m a r k e d b y "so" the subordinate ones b y "and". The structure is crystal clear, a n d H a b i b a accomplishes it b y m e a n s of only two cohesive markers: " a n d " a n d "so". T h e narrative patterns we find in stories such as that of H a b i b a n o t only contain a wealth of information, structured into patterns that reflect relevance and affect - Goffmanian footing - they often also contain intricate argumentative patterns in which theses are formulated, refuted, demonstrated b y m e a n s of evi dence a n d so on. Emphasis, logical or associative sequences, cause-effect relations, argument elaboration patterns are all m a r k e d b y speakers. I n order to make their point, speakers d r a w u p o n complex sets of related arguments, illustrations, conclusions, deductions a n d so forth. Let us take a look at another example. T h e following fragment is taken from an interview with an Angolese couple P (husband) a n d D (wife); the Belgian interviewer is G M . T h e interview was in French. T h e following fragment occurs as part of a n exchange o n the role of the m e d i a in the Semira A d a m u case. The Angolese couple h a d b e e n interviewed b y local T V reporters, a n d G M asks whether they believe that this m e d i a exposure would h a v e a beneficial effect on their asylum applications. T h e answer is negative: the authorities publicly claim that they would a d o p t a m o r e flexible attitude towards asylum seekers; yet as soon as the latter present themselves to the authorities, the answer they get is that "they will h a v e to follow the procedures". T h e Angolese w o m a n elaborates this further: Extract 10 1 4: 2 3 4 5
a la tele on dit/*ooh noon/on peut pas donner tous lees personnes apres cinq ans les==on a fait euh dix ans/ les choses comme 5a/ on va regarder les procedures /mais le premier jour on dit a la tele / ooh noon /on va *donner les gens qu'on a fail cinq ans dans notre pays/ parce qu'il y a eu des enfants qui etu=etudient ici depuis longtemps/ on a fait *six ans==*cing ans et les enfants ils vont a l'ecole==c'est le bourgemesne de Bruxelles
I
BLOMMAERT
6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
INVESTIGATING NARRATIVE INEQUALITY
195
qui a dit comme ca== GM: ==hmm= D: =ouais/ les enfants sont ""integres chez nous depuis longtemps=il part a l'ecole/mais maintenant je pense qu'on va les donner les = = p o u r donner les enfants==la a ses parents/ *mais apres quelques jours on a dit/OOH N O O O N on va euh==suivre les procedures [hits hands on legs] MAIS ""COMMENT ON PEUT SUIVRE LES PROCEDURES? [annoyed] depuis moi je suis ici en Belgique/ je suis sans papiers/ tu me suis?/ sans papiers/ ici/ exemple/ pas pour moi pour le moment/ il vient ici comme un==a une femme ici dans le journal==hier==on=a=lu/ sans papiers/ venue ici/ il a trois enfants==cinq enfants ici en Belgique/ sept a n s = s i x ans sans papiers/ alors/ a ce momentla vous disez/ faut suivre les procedures/ la femme la/ depuis Pete vous avez eu les proce=les papiers/ vous avez jete les papiers de la femme/ on sait pas qu'est=q procedure/ ""quelle procedure on peut suivre?/ pour la femme/ c'est *ga les problemes/ les *Belges peut pas donner les gens un papier comme ca/ ils pensent qu'on donne/ *aah tous les gens ils restent dans notre pays/ mais euh==*SI LONGTEMPS ils part dans le pays des gens pour rester la/a ""Portugal il y a des Belges==on a *fait quelque chose la-bas
Translation: 1 D on the telly they say/ *ooh nooo / we cannot give all the people after five years the==they 2 have done ehr ten years/ the things like that / they are going to look at the procedures / but 3 the first day they say on the telly/ oooh nooo / we will *give the people that has done five 4 years in our country / 'cause there have been children who stu=studied here since long / 5 they have done *six years==*five years and the children go to school== it's the mayor of 6 Brussels who says like that== 7 GM: = = m h m = = 8 D: =yeah/ they children they have integrated here since long=he leaves for school but now I 9 think that they'll give them the=to give to the children=there to their
1
10
parents/ hands few days on*but they legs]after said/a oohh nooo we will ehr=follow the procedures [hits
DISCOURSE STUDIES
196
11 12 13 14 15 16 17 18 19 20 21
BUT *HOW CAN WE FOLLOW THE PROCEDURES? [annoyed] *since I arrived here in Belgium I've got no papers/ do you follow me?/ no papers/ here/ example/ not for me for the moment/ there comes here like a=there's a woman here in the newspaper=yesterday=we have read/ people without documents/ came here/ he's got three children==five children here in Belgium/ seven years=six years without documents/ so/ at that moment you say/ got to follow the procedures/ that woman there/ since the summer you have e=u the proce=the documents/ you have thrown away that woman's documents/ we don't know what's=wh=procedure/ *what procedure one can follow/ for the woman? That is the problems/ the *Belgians cannot give to the people th=documents just like that/ they think that they give/ *aaahh all these people stay in our country/ but eh=*so long they leaves to countries of the people to stay there/ in *Portugal there are Belgians=they have done something there
T h e Angolese w o m a n tries to make a crucial point in understanding her condition a n d experiences as a refugee: the gap between what Belgians believe happens to refugees, a n d what really h a p p e n s to t h e m in Belgium. She m a k e s this point in 'broken' French, part of the features of which d o not a p p e a r in the English translation (e.g. ambiguous 3rd p e r s o n "on", m e a n i n g "they" as well as "us" and impersonal "one"). However, m o r e interesting is the elaborate argumentative structure of D ' s talk. P r o m p t e d b y G M ' s suggestion that T V exposure would create a m o r e favourable climate for asylum seekers to obtain their documents (i.e. to b e 'regularized' a n d given legal residence permits) she argues that this is n o t the case. She describes what h a p p e n e d . First, largely in r e p o r t e d speech ("ooh n o o " , performance m a r k e d b y lengthened vowels) she describes h o w the M a y o r of Brussels declared o n T V that people with children w h o g o to school in Belgium a n d w h o h a v e b e e n in the country for a n u m b e r of years would be regularized (lines 1-6). N e x t she describes h o w the Belgian authorities shifted their position a couple of days later, again using reported speech (line 10). She then states her own position: "but h o w can we follow the procedures?", and starts elaborating o n that t h e m e using a n anecdote picked u p from the newspaper. This t h e n leads to a conclusion a n d a coda, in which the situation of asylum seekers i n Belgium is c o m p a r e d to the freedom with which Belgians settle else where in the world. Despite h e r 'broken' French, D constructs a clear pattern of arguments, sche matically represented as follows. Note also h o w the central motif of the argument, the issue of h o w to follow procedures, is m a r k e d b y parallelisms (indicated by arrows «=):
BLOMMAERT
INVESTIGATING NARRATIVE INEQUALITY
Extract 11
197
*-
M I i
1. Background They will give the [documents] to the children to their parents because they have been integrated here (says the Mayor of Brussels) 2. Point of departure: [But] afterwards they say that we have to follow the procedure*
«=
3. Refutation 3.1. Core But how can we follow the procedures?
«=
3.2. Elaboration ' 1. Ever since I arrived here, I lived without documents 2. Anecdote: Report yesterday we have read in the newspaper A woman with no documents and 3/5 children No documents for 7/6 years
',
1
Comment And then they say that you have to follow the procedure <= [BUT] you have thrown away that woman's documents We don't know what procedure to follow «= 4. Conclusion and coda That is the problem The Belgians cannot give documents just like that [i.e. they don't give papers just like that, procedures must be followed, and this is problematic] [BUT] Belgians believe that everyone just stays in their country [WHILE] Belgians themselves can be found all over the world (Portugal) We see a well-organized narrative pattern emerge over and b e y o n d difficulties in handling the linguistic code. T h e core of the argument is clearly marked, a n d the refutation is accomplished b y m e a n s of comparison of D's situation with that of s o m e o n e else (the anecdote) a n d leads to a convincing point: staying in Belgium is n o t as simple as Belgians tend to believe it is; in fact, Belgians have an easier time w h e n they decide to go a n d live abroad. Stories told in 'simple' language are n o t necessarily 'simple' stories. But they may b e stories that are o p e n to all sorts of misinterpretation b y interlocutors who, often, h a v e a different (yet equally limited) competence in the m e d i u m in which the interview is done. It requires some effort to detect coherence in the stories told in ' b r o k e n ' varieties of Dutch, English or French, because coherence and structure h a v e to b e sought at levels of linguistic structuring n o t easily penetrable to non-specialists a n d difficult to pick u p while the story is told. We will turn to such forms of structuring in section 4. But before that, we h a v e to refer back to what we said at the beginning of this section: in our society, admin istrative procedures a n d the rights that are d e p e n d e n t of their 'correct' fulfilment require a n d assume the sharedness of linguistic a n d communicative resources. The fact of the matter is that these resources often lack from t h e repertoire of the asylum seekers interviewed b y us. Consequently, issues of resources are crucial elements in the forms of narrative inequality investifirated h f w Tn»««~i:*-*— 7
DISCOURSE STUDIES
198
in speech repertoires condition narrative inequalities a n d thus inequalities in the allocation of social rights. i
•
4. H o m e Narratives 4.1 A Contextual Account O n e feature in m a n y interviews is the w a y in which interviewees attempt to provide detailed contextual accounts of their lives, local circumstances, politics and conflicts in their h o m e society formulated in the shape of sub-narratives of the larger narrative. We call such contextual accounts ' h o m e narratives'. Often these narratives are triggered b y an awareness that the story cannot b e fully understood unless other p e o p l e k n o w some details about the society they come from, the particular events that caused their flight, a n d so on. Such exposes can b e highly complicated. T h e y can disturb a n expected pattern of sequential event narrating ('first this, t h e n that') a n d give a m u d d l e d impression (see O c h s and Capps, 1996, for a general survey of features of similar narratives). T h e details given b y interviewees can also backfire: details are o p e n to scrutiny and one inaccuracy, inconsistency or contradiction in the story can b e enough to disqualify asylum seekers, as will b e shown in section 5. T h e following fragment is taken from a n interview with R, a m a l e refugee from the Ivory Coast. T h e interview was in Dutch. At o n e point, the interviewers (B and T) ask "What were the precise reasons w h y you came here?". W h a t follows is a long contextualizing narrative, placed between two general framing state ments which are given in bold. Extract 12 1 R:
(sighs) ja/voor/alle mensen is h e t moeilijk de politiek v a n Ivoorkust/ te begrijpen/ 2 want/ er wordt nooit iets over gesproken/ maar bij ons is er/ is er nog het leger van 3 Frankrijk aanwezig/ onze/ euh onze luchthaven is de basis van het Franse leger/ en die 4 . mensen zijn toch daar/ en we=we hebben geen recht om zelf ons=onze regering te te te 5 kiezen/ ja die mensen worden door Frankrijk genoemd/ maar wij=wij zijn geen kolonie 6 m e e r / sedert achtendertig j a a r zijn w i j / e u h / h e b b e n wij onze onafhankelijkheid gekregen/ 7 maar tot nu toe worden onze ministers altijd door Frankrijk gekozen/.../ ons land ligt/ in 8 het midden van West-Afrika/ en dat is een strategische positie/ 9 B, T: ja ja 10 R: wij=wijzijnnietrijk/en/OK/achtjaargeleden/probeerdenwijpolitieke par=partijen op 11 te richten en zo/ maar de regering zelf heeft achtendertig partijen opgericht/ terwijl dat het
BLOMMAERT
12
INVESTIGATING NARRATIVE INEQUALITY
199
moeilijke is om/ euh/ gekozen te worden of een land te le=te leiden/ als je geen lid van de 13 • partij bent/ dan hebben zij=hebben wij andere/ euh/ toestemming van partijen te creeren 14 gekre=gekregen/ en daarnaast/ of bovendien heeft de regering zelf achtendertig partijen 15 gecreeerd/ naast de =plus de=deuh monopartisme zo 16 B: ja 17 R: de monopartisme die bestaat sedert zestig tot negentig/ dertigjaar/en ja/ zij hebben zelf 18 ' dertig=ja achtendertig partijen gecreeerd/ de mensen die niet voor de regering werken/ ik 19 bedoel/ in/ euh/ de=de prive/ het privewerk is heel weinig/ dan werkt al=iedereen 20 ' voor=voor de staat/ en als/ ja de mensen die voor de staat werken/ zijn zoals deuh/ 21 gevangenis/ zij zijn zoals gevangen/ want je/ je=je bent altijd bang/ jouw werk te 22 verliezen/ misschien in jouw familie met jouw neven en nichten en zo zijn jullie 23 misschien ongeveer honderd mensen en je bent alleen als persoon die werkt/ dat/ja/ al die 24 andere mensen zijn/ die=die zijn arm en/ die rekenen allemaal op jou/ dan moet je jouw 25 7 werk houden anders wordt iemand van jouw familie gekocht/ om te gebruikt te worden 26 tegen je/ ja zo'n dingen/ ja zo'n kleine dingen/ en wij// ja/ ja daar is onze=onze president 27 van onze partij [wijst naar foto in de huiskamer] wij hebben ook=wij proberen ons hier te 28 organiseren in Belgie in Nederland in Frankrijk/ de=de grote groep bevindt zich in 29 Frankrijk in Italie in Duitsland ook in andere landen/ in Nederland/ de mensen van 30 Nederland komen bij ons/ overmorgen hebben wij onze/ ja grote vergadering in 31 Brussel/.../ dat is het dan/ ja deuh reden was dat in vijf=vijf en negentig moesten wij/ een 32 verkiezing doen/ en in vierendertig/ euh vier en negentig moest een lijst van alle bewoners 33 gedaan worden/ maar de mensen die=die/ euh die=die/deden/ als die bij u komen en die 34 weten dat je een=een lid van de oppositie bent/ dan wordt je=jouw name=euh jouw naam 35 niet ingeschreven/ en als jouw naam niet ingeschreven wordt kan je niet stemmen/ en die 36 hebben zo gedaan/ja wij hebben geprobeerd/ die lijst terug te vinden/ja/ ja een andere te 37 maken/ 38 B, T: ja 39 R: maar dat was illegaal f
200
DISCOURSE STUDIES
40 B:
ja
41 R:
ja/ dan worden wij vervolgd/ e n zo ben ik moeten vluchten
Translation: 1 R:
(sighs) yeah/for "'everyone it is difficult to understand the politics of the Ivory Coast 2 because/ it is never discussed/ but over there we still have the *French army/ ehr ehr our 3 airport is a French army base/ and these people are there anyway and we=we=we have no 4 right to choose our=our own government/ yes these people are appointed by France/ but 5 we are *not a colony anymore...since thirty-eight years we ehr have obtained our 6 independence/ but until now our ministers have always been appointed by France...our 7/8 country is in the center of West-Africa/ and that is a strategic position 9 B, T: =yesyes 10 R: yes/so...yes there is=we are not rich/ and/ okay/ eight years ago we tried to create political 11 par=parties and so but the government itself has created *thirty-eight parties/ 12 while this is the difficult thing to ehr get elected or to lead a country.. If you are not a 13 member of a party/ *then they=we obtained different ehr permission to create parties/ and 14 apart from that or on *top of that the government has created thirty eight parties/ apart 15 from the=*plus the thehh monopartism like that 16 B: yes 17 R: yes the monopartism exists since sixty/ till ninety/ thirty years/ and yes they have 18 themselves created thirty=thirty-eight parties/ the people who *don't work for the 19 government/ I mean/ in ehr..the=the private the private sector is very small/ everyone 20 already works for=for the government/ and if., yes the people who work for the 21 government they are like ehr the prison/ they are like imprisoned/ because you..you=you 22 are always afraid of losing your job/ maybe in your family with your nephews and nieces 23 and so maybe you are about one hundred people and you are the only one who has a job/ 24 that/ yes/ all these other people are/ they=they are poor and they all count on you / then 25 you have to keep your job *if not *someone from your family will be bought to be used 26 against you/ yes things like that.small things like that/ and we...yes [sighsj yes there is
I
BLOMMAERT
INVESTIGATING NARRATIVE INEQUALITY
201
27
our president of our party [points to a picture on the wall] / we also have=we also try to 28 get organized here in Belgium in Holland in France/ the=the large group is in France in 29 Italy in Germany also in other countries..in Holland/ the people from Holland will come 30 here the day after tomorrow we have our/ yes large meeting in Brussels...that's it...yes 31 thehh reason was that in nine=ninety five/ we had to do an election/ and in thirty 32 four=ninety four a list of all the inhabitants had to be made/ but the people who did that 33 when they came to you and they know that you are a=a member of the opposition/ then 34 your name=ehr your name would not be written down/ and if your name is not written 35/36 down you can't vote/ and that's what they did/ yes we tried to find that list/ yes and make 37 another one 38 B, T: yes 39 R: but that was illegal 40 B: yes 41 R: if it/ yes/ then we are prosecuted/ and so I had to escape/ 1
Almost 40 lines of m o n o l o g u e separate the question as to w h y R h a d to escape from the Ivory Coast a n d R's conclusion to the answer "And so I h a d to escape". In this monologue, h e tries to capture some of the basic political mechanisms a n d problems of his h o m e country: French neocolonialism (lines 2-7), m o n o p a r t i s m and the difficult transition to multipartism (lines 10-15), political clientelism (lines 17-265), the opposition m o v e m e n t in the diaspora (lines 2 8 - 3 2 ) , a n d the direct cause of his flight: protest against the manipulation of the elections (lines 32-41). This fragment is again followed b y a very detailed account of the political and institutional customs of the Ivory Coast b l e n d e d with accounts of events leading to R's flight. 4.2 Patterns and Structures The m a n from the Ivory Coast starts b y saying "for ""everyone it is difficult to understand the politics of the Ivory Coast", indicating the n e e d to provide such detailed information o n life in the Ivory Coast: a precise understanding of w h y he escaped from his country would b e impossible for people w h o hardly k n o w anything about life and politics in his country (cf. "because/it is never discussed"). If we look a bit closer into the narrative structure of this fragment, we see the following patterns emerge. First, the narrative can b e divided into episodes marked b y particular narrative patterns:
202
DISCOURSE STUDIES
Extract 13 1. for *everyone it is difficult to understand the politics of the Ivory Coast because ... it is never discussed 2. But over here we still have the French army ehr ehr our airport is a French army base and these people are there anyway 3. and we=we=we have no right to choose our=our own government yes these people are appointed by France but we are *not a colony anymore... since thirty-eight years we ehr have obtained our independence but until now our ministers have always been appointed by Francetreason} our country is in the center of West-Africa and that is a strategic position 4. {preceded by reply from B, T} yes/so...yes there is=we are not rich/ and.. 5. okay.eight years ago we tried to create political par=parties and so 6. [louder] but the government itself has created *thirty-eight parties 6.1. while this is the difficult thing to ehr get elected or to lead a country.. If you are not a member of a party/ *then they=we obtained different ehr permission to create parties/ and apart from that or on *top of that the government has created thirty eight parties/ apart from the=*plus the thehh monopartism you see? 6.2. yes the monopartism exists since sixty/ till ninety/ thirty years/ 6.3. and yes they have themselves created thirty=thirty-eight parties/ 6.4. the people who *don't work for the government/ I mean/ in ehr..the=the private the private sector is very small/ everyone already works for=for the government/ and if., yes the people who work for the government they are like ehr the prison/ they are like imprisoned/ i because you..you=you are always afraid of losing your job/ i -maybe in your family with your nephews and nieces and so maybe you are about one hundred people -and you are the only one who has a job that/ yes/ all these other people are/ they=they are poor and they all count on you / -then you have to keep your job -*if not *someone from your family will be bought to be used against you/ yes things like that..small things like that/ 7. and we...yes [sighs] yes there is our president of our party we also have=we also try to get organized here in Belgium in Holland in France/ the=the large group is in France in Italy in Germany also in other countries..in Holland/ the people from Holland will come here the day after tomorrow we have our/ yes large meeting in Brussels... t
BLOMMAERT INVESTIGATING NARRATIVE INEQUALITY
203
8. that's it...yes thehh reason was that 8.1. in nine=ninety five/ we had to do an election/ and in thirty four=ninety four a list of all the inhabitants had to be made/ 8.2. but the people who did that when they came to you and they know that you are a=a member of the opposition/ then your name=ehr your name would not be written down/ and if your name is not written down you can't vote/ and that's what they did/ 8.3. yes we tried to find that list/ yes and make another one 8.4. but that was illegal 9. if it/ yes/ then we are prosecuted/ and so I had to escape/ i We can identify nine narrative episodes marked prosodically or b y means of markers such as "but" or "and". The most complex episode is episode 6, sub divided into at least four sub-episodes. If we reorganize this narrative pattern so as to structure it into an argumentative 'answer'-pattern to the question, we see the following (paraphrases are given between accolades): . , Extract 14 Why did you escapefrom the Ivory Coast? Point ofdeparture 1. For everyone it is difficult to understand the politics of the Ivory Coast General reason: neocolonialism 2. We still have the French army 3. And we have no right to choose our own government {reasons:} we are in the centre of West-Africa 4. {that is why the French are there, because) we are not rich Specific reason: membership of political opposition party 5. Eight years ago we tried to create political parties 6. {this is problematical, because} the government itself has created 38 parties {the government manipulates the democratization process: monopartyism plus 38 bogus parties} {iron grip of the regime on society} you are always afraid of losing your job 7. {our party also exists in Europe} 8. The {immediate} reason was {election fraud} {the regime tried to commit fraud in voter registration} {we tried to counterfeit our own voter registration list} but that was illegal Conclusion 9. Then we are prosecuted and so I had to escape This home narrative plays a crucial role in 'making sense' for the asylum seeker. The immediate cause of his problems is one issue, the wider context of neocolo nialism and an oppressive one-party regime is another. This man broke the law
DISCOURSE STUDIES
204
of the Ivory Coast in trying to counterfeit a voter registration list; b u t the jus tification for this act is the wider political context that makes clear that the Ivory Coast is n o t a democracy, that the laws are dictatorial, and that h e a n d his friends b r o k e that dictatorial law in their struggle for democracy. His asylum application in Belgium was jeopardized o n grounds that h e broke the law in his h o m e country; his h o m e narrative is an attempt to explain the justified nature of his asylum application. In sum: his h o m e narrative is n o t a side-track, it is right o n track. 4.3 Contextualizing Accounts Let us go somewhat d e e p e r into the explanatory a n d argumentative functions of these h o m e narratives. R, i n the fragment above, starts his m o n o l o g u e with a d e e p sigh a n d a reflection o n the ignorance of Belgian p e o p l e about the intricate realities of the politics of the Ivory Coast. I n other interviews, w e saw such h o m e narratives framed in terms of absence of attention from the 'international public opinion' - 'everyone should k n o w this' (a p h e n o m e n o n also noted b y Malkki, 1995 a m o n g H u t u refugees in Tanzania). I n either case, the accurate contextualization of personal experiences in terms of local political, economic, social a n d cultural terms appears to b e a crucial explanatory strategy for attaining understanding. Text-structurally, h o m e narratives are often framed b y utterances situating t h e m in relation to a particular argument or e x p l a n a n d u m : p r e c e d e d b y framing statements such as "you must k n o w t h a t . . .", "you see, in m y country . . .", "people h e r e don't k n o w t h a t . . a n d so on, a n d followed b y statements connec ting the h o m e narrative to a particular argument such as "that is w h y . . . " , "and so . . . " . So o n e of their functions is to provide a particular epistemic format: a metapragmatically framed 'aside' in which crucial referential a n d indexical 'fillings' are given of terms, concepts a n d features of the story. This epistemic format is shaped b y m e a n s of contextualizing discourses of time a n d place. T h e refugee experience is a n experience of displacement (similar to the one described for native American groups b y Collins, 1998) seen as crucial in understanding the w o r k of identity of such groups in light of changing living conditions, political allegiances, economic practices a n d physical migrations. Stories from such groups thus display a complex interplay of physical space (e.g. p l a c e n a m e s ) , social space (e.g. sociocultural values or ethical codes associated to certain places) a n d narrative space (discursive articulations through indexical links in a narrative), in which migrant, minority or other social identities are b e i n g shaped (cf. Collins, 1998: ch. 6). Malkki (1995) demonstrated how physical displacement - part of the condition humaine of refugees - gives rise to forms of r e m e m b e r i n g that include spatial as well as temporal trajectories of leaving, being transported, settling, staying a n d returning, all of which become characteristic of the exile identity of the H u t u refugees she investigated (see also Maryns a n d Blommaert, 2001). I n our data, too, displacement b e c o m e s a matter of identity, and (rebelling the story of escape a n d exile as well as the story of dealing with Belgian authorities gives rise to narratives that provide the core of " •«••! ,-r> i f 1.-™^. t i , n.it/-*k."/-ifrran1-i v n f rpfncrpps rakes the n
BLOMMAERT INVESTIGATING NARRATIVE INEQUALITY
205
shape of trajectory telling. Temporal and spatial elements are used in constructing 'refugeeness', as this identity relies o n the fact of having left o n e place, travelled across parts of the globe a n d setded in another place. Surely, the increasing structuration of narratives into 'fully formed narratives' testifies to the gradual and discursive practice-based construction of such refugee identities. I n the h o m e narrative in Extract 12, three places are articulated together with three time frames. T h r e e different time frames are used to qualify features or events associated to either Belgium or the Ivory Coast: the present (incorpor ating the act of telling), the past a n d a timeless, p e r m a n e n t state of affairs. T h e timeless a n d the past frames are used to qualify the Ivory Coast; the present is used to qualify Belgium. We can summarize these in the following schema: Time
Place
Present
HERE: Belgium • systemic observations: people don't understand politics in the Ivory Coast • political activism in exile THERE: Ivory Coast • monopartism, dictatorship • political activism, election incident THERE: Ivory Coast • systemic observations: neocolonialism, strategic location, poverty, clientelism
Past (i) Long (38 years) (ii) short (8 years, critical moment 1995) Timeless
Richard B a u m a n (1986, quoted in Briggs, 1997) defines narrative as 'a represen tation of a segment of social life in such a w a y as to create a series of sequentially ordered events'. I n the h o m e narrative discussed h e r e , the temporal sequence of events from past to present which would constitute a 'typical' narrative is 'broken' by a n u m b e r of systemic observations with respect to the general political a n d social situation in the Ivory Coast, narrated in a factual timeless present tense. Thus, in response to the question "why did y o u escape from the Ivory Coast?", the m a n from the Ivory Coast presents us with a classic narrative in the sense of Bauman's definition, the chronological unfolding of events of which is, however, 'scrambled' temporally. We get a sequence of narrative episodes in which the time frame shifts b e t w e e n present, past and timeless states. If we place the narrative units identified a n d n u m b e r e d in Extracts 13 a n d 14 above in a schema based on the three time frames (and keeping in m i n d the connection b e t w e e n time frames and places), the sequential-temporal ordering in the narrative appears in the table below (numbers refer to the episodes). The starting point of the narrative is the here-and-now: an observation of the ignorance of the Belgians about the p r e d i c a m e n t of the people in the Ivory Coast, considerably complicating life for Ivory Coast refugees in Belgium. Next, a timeless time frame is introduced in which general contextualizing 'facts' are being offered, linguistically articulated in factual a n d declarative expressions without concrete t e m p o r a l anchoring (episodes 2, 3 , 4). O n c e this timeless, general context has b e e n sketched, a m o r e concrete a n d temoorallv s w m p n K = i -
206
DISCOURSE STUDIES
timeless contextualizing c o m m e n t (6.1, 6.4) a n d once b y a n episode connecting the events in the Ivory Coast to the present in exile (7). T h e narrative concludes with a statement (9) that can b o t h b e seen as a closure of the sequential account of events a n d as an anchoring of the whole preceding story into the present situ ation of exile a n d asylum seeking in Belgium. It can, of course, also b e seen as the point at which b o t h aspects intersect a n d b e c o m e one: the present situation of being a refugee.
Present (here)
Past (there)
Timeless (there)
1 2 3 4 5 6 6.1 6.2 6.3 6.4 8 9X9 This pattern of temporal and spatial debds is both a generic and an epistemicargumentative matter. It is generic in the sense that it constitutes the genre of h o m e narratives: nonlinear references to h e r e / t h e r e a n d n o w / t h e n / a l w a y s make u p the considerable complexity of such stories, the trajectory telling. It is at the same time epistemic-argumentative in the sense that it is directed at a hearer (and at overhearers) w h o needs to b e instructed about these issues. This otherdirectedness of h o m e narratives makes t h e m fulfil the contextualizing functions they h a v e in the story at large: they provide settings, scenes, referential domains a n d indexicalities that n e e d to b e a d o p t e d b y the hearer. I n this m o v e , there appears to b e an acute awareness of the categories and interpretive resources of the hearer: the assumption used in telling h o m e nar ratives is that (i) Belgian hearers do not know this; (ii) consequently their view of the situation in the h o m e country is wrong a n d prejudges a correct interpretation of the causes a n d motives for the asylum application; a n d (iii) hence, the Belgian hearers need to know a series of things, so as to modify their categories a n d inter pretive resources in treating the story of the asylum seeker. H o m e narratives are based u p o n a perception of false or inaccurate contextualization of situations and events in the countries of origin - wars, political conflicts, poverty - and are aimed at recontextualizing or 'recentering' t h e m (Bauman a n d Briggs, 1990). T h e recentering is a process of localizing the conflicts, events a n d upheavals: instead of b r o a d , abstract a n d i m p e r s o n a l ('decentered') categories of war, conflict, crisis, p o v e r t y a n d so o n , c o n c r e t e a n d h i g h l y p e r s o n a l , 'local' indexicalities are offered: 'this is not the w a r in Angola; this is my war i n Angola'.
BLOMMAERT
INVESTIGATING NARRATIVE INEQUALITY
207
A n d the centering a n d localizing is performed b y m e a n s of stories that b l e n d personal chronological trajectories - autobiographies - with general framing statements on h o w life was a n d is over there a n d what people n e e d to k n o w in order to correctly understand the autobiographical narrative.
5. N a r r a t i v e s , D e n s e C o n t e x t u a l i z a t i o n s a n d t h e A s y l u m P r o c e d u r e 5.1 Asylum Procedure and Narratives We n o w m o v e back to the c o m m e n t s we m a d e at the outset of this article. I n the Belgian asylum procedure, the asylum seekers' stories assume a central and critical role. T h e basis o n which applications are being e x a m i n e d a n d sanctioned is the textual set of statements taken from the asylum seekers, in which they a n n o u n c e their desire to seek asylum in Belgium a n d provide reasons a n d motives for that desire. Inconsistencies in the stories are a major cause for refusing refugee status to asylum seekers. As will b e shown later, inconsistencies, ranging from contradic tions between two versions of the story over not r e m e m b e r i n g n a m e s , places or dates, a n d superficial or incomplete knowledge of things assumed to b e k n o w n to people from one country (e.g. the n a m e of the president, n a m e s of towns and cities, etc.) can all serve as cause for refusal. Strangely enough in a legal system backed b y sophisticated high-tech a n d advanced forensic sciences, this kind of legal p r o c e d u r e relies heavily o n narrative analysis. This narrative analysis is, of course, performed in ways that can only b e qualified as impressionistic. We h a v e old a n d time-tested rules of evidence used to investigate statements m a d e b y people living in a highly literate environment, whose memories are supported b y a (sometimes massive) literate 'archive' (diaries, notebooks, newspaper clippings, photographs, all kinds of small documentary souvenirs of events in one's life, see Radley, 1990), a n d whose statements are produced a n d interpreted in an interactional setting in which the interlocutors share the same linguistic resources a n d communicative skills (see the discussion in section 3 above). T h e s e rules of evidence stress textual consistency, linearity, logic, rationality and factuality; they require considerable attention to details; they rely on written language as the basic a n d most lasting format of declaring 'truth'; in short, they are highly culture- a n d society-specific a n d reflect local ideologies of language, literacy a n d communication. These rules of evidence are applied to statements m a d e b y people for w h o m few of these formats of language a n d communication can b e taken for granted. H e n c e , what h a p p e n s in the institu tional processing of asylum seekers' stories is often a battle with unequal arms, and the confrontation of different narrative conventions creates a h u g e p r o b l e m of justice a n d fairness. Attention to asylum seekers' storytelling conventions is scant. Yet their central position in the asylum procedure makes them into an important topic of research, both analytically a n d politically. T h e salience of looking at the w a y in which asylum seekers' stories are b e i n g told can p e r h a p s b e briefly illustrated with reference to the Semira A d a m u case. W h e n h e r death was a n n o u n c e d in the 8
208
DISCOURSE STUDIES
media, o n e of the striking ingredients of m a n y statements was the abundant reference to 'her story'. It was said that, indeed, forced marriage could b e suffi cient grounds for asylum, b u t that in her case 'her story did not fit'. T h e official verdict announcing the refusal of asylum argued: T h e story of the applicant is n o t precise o n m a n y counts. Thus, she could not specify the family n a m e of a childhood friend with w h o m she stayed. She also failed to specify the date o n which she was supposed to h a v e escaped from Lagos. (My translation, D u t c h original from De Morgen, 3 O c t o b e r 1998) O t h e r references to h o w Semira A d a m u interacted with p e o p l e - again, matters of style - include a statement b y the Catholic pastor of Brussels Airport, a m a n w h o m e t Semira A d a m u occasionally. To h i m as well, Semira A d a m u was not quite trustworthy: I n the [asylum] Centre n o b o d y believed her. I d o n ' t k n o w the real story either. I rely o n m y intuition. Sometimes, w h e n y o u asked her something, she just stood u p a n d walked away. O r she started to sing. 1 once asked h e r w h y she h a d travelled so m a n y thousands of kilometers to escape from a sixty-five year old m a n . 'Is Africa n o t b i g enough', I asked her. She didn't answer that question. (My translation, D u t c h original from De Morgen, 3 October 1998) Again, a lack of credibility seems to b e tied u p with the w a y in which Semira A d a m u c o m m u n i c a t e d . E v e n s y m p a t h e t i c voices refer to peculiarities in communicative style, l i s e Thiry, a former socialist senator w h o acted as foster p a r e n t for Semira A d a m u , qualified h e r as follows: A p a r t from that, Semira sometimes seemed British rather than African to m e . She didn't h a v e the outspokenness a n d extroversion some of these [African] w o m e n h a v e . (My translation, D u t c h original from De Moigtn, 3 O c t o b e r 1998) Communicative style (including narrative style) is always a source for characw assessment a n d character attribution. I n a field as sensitive as asylum regularjom, where officials' decisions m a y imply the difference between life or deadi, relative well-being or poverty, safety or danger, some attention to what actually happen in interaction m a y n o t just b e desirable, b u t simply imperative. 5.2 Textual Trajectories Let us begin with a brief survey of the interactional processes that constitute the core of the asylum application processes. A first observation is that in tile totalir* of the p r o c e d u r e , direct interaction b e t w e e n asylum seekers and officials it restricted to a n u m b e r of well defined occasions, viz. interviews on the asylum
209
seeker's story. Soon after their arrival in the country, asylum seekers are inter viewed by officials, sometimes (but b y no means always) assisted b y interpreters. The topic of the interview is their motivation to seek asylum in Belgium: w h y did they leave their country, how, when? What reasons do they offer for assuming that their lives or life chances are endangered in their country of origin? W h o are they, where do they come from? W h o assisted them in their exodus? A n d so on. The story is noted by the official and has the status of an affidavit. From that moment onwards, it is the source and reference text for all other steps in the procedure. The asylum seeker is admitted in the country, and his or her applica tion for asylum is being processed b y the authorities. In the process, various other interviews can take place. Apart from interviews, there are hardly any occasions in which asylum seekers themselves speak. Most of the communication on the case is written and legalistic and hence treated on behalf of the asylum seeker by lawyers, welfare workers or members of N G O s specialized in asylum affairs. So what we have is a handful of events during which the applicant directly produces oral narrative discourse. In between these moments we have tremendous text-production on that narrative resulting in written summaries, notes, translations, questions and replies written by lawyers or welfare workers, court rulings in which fragments of the narrative are being quoted and interpreted, and so on. It is an instance of the 'circulation of discourse' (Briggs, 1997:538 ff.) that characterizes legal and forensic procedures as well as those of welfare work and bureaucracy (see Sarangi and Slembrouck, M96). These patterns of circulation are biased b y inequalities in techniques of discourse representation; recontextualizations and re-entextualizations of the narratives in specific ('official' and authoritative) generic formats are usually privileged forms of discursive practice reserved to specific professional groups nich as lawyers, administrators, 'experts' and bureaucrats. This textual complex, in which an oral 'original' narrative is the input for a long scries ofgenerically differentiated replications of that original is characterized by an ideology of 'fixed text' (Blommaert, 1997; Collins, 1996; Urban, 1996), in which the difference between the 'original' and its 'copies' are assumed to be minimal (hence in which it is assumed that every translation, summary, quotation, reading of die story is correct and accurate) because 'procedurally correct' text U*. text produced or collected according to standard procedures) is supposed to be i transparent unambiguous set of signs. T h e story of the asylum seeker is remoulded, remodelled and re-narrated time and time again, and so becomes a kxl trajectory with various phases and instances of transformation. At the same time, the story is treated as a singular text, and responsibility for that text (and thus fcrall re-entextualizations in the text trajectory) is attributed to the asylum seeker. The asylum seeker is constructed as the responsible author for the whole tattnextualicy complex, despite the enormous differences in text-structure and test-modalitv. the genre and the code, the social spaces in which versions are being produced and used, and the power and authority attributed to different versions of the text. Commenting o n similar phenomena, Briggs (1997: 540) observes trial in such cases, the question seems to be 'one of a fundamental Asymmetry in the power to determine h o w utterances can circulate between
DISCOURSE STUDIES
210
contexts, epistemologies, a n d institutions, the w a y that narratives can b e struc tured, and what sorts of legal effects can accrue to particular discursive relations'. T h e result of this text trajectory is a suggestion ofjustice based o n text-ideologies e m p h a s i z i n g t h e 'correct t r e a t m e n t ' of the a s y l u m seeker's n a r r a t i v e into notes, summaries, translations a n d so forth, but obscuring a variety of forms of appropriation of discourse a n d of shifting it into domains of authoritative reentextualizations that are far b e y o n d the control of the asylum seeker. So when the letter to Semira A d a m u speaks of 'the story of the applicant', what is referred to is the total textual trajectory of the narrative. 5.3 A Contest of Contexts To illustrate this, let us h a v e a look at two re-entextualizations of the story of an Angolese asylum seeker. T h e interview we did with h i m and his wife took four hours; the story of his situation in Angola and of his escape to Belgium took m o r e than one hour, a n d well over 1500 lines of transcript. As always, it is a very complex story replete with h o m e narratives bearing the characteristics discussed in the previous section. T h e m a n applied twice for asylum, and twice his application was rejected. These rejections were a n n o u n c e d to the applicant in two official letters, p a r d y standard a n d partly filled in with specific data on the application. I n these documents, we find instances of re-entextualization of the m a n ' s story. I give fragments from the two texts of the letters (my translation, originals in Dutch a n d French, respectively): Extract 15
'
•
"The concerned was interrogated on November 23, 1993 at the CommissariatGeneral [for Refugees and Stateless Persons], in the presence of [name], his attorney. He claimed to be a 'political informant' of the MPLA. On October 18, 1992 however, he passed on information to UNITA. At the UNITA office, however, he met with Major [name], who works for the MPLA. Two days later, Major [name] had the concerned arrested. Fearing that the concerned would give the Major away at the trial, [name of the Major] helped the concerned to escape. The concerned fled to [locality] where a priest arranged for his departure from Angola. The concerned came, together with his wife [name and register number| and three children, through Zaire and by plane, to Belgium. They arrived on May 19, 1993. It has to be noted that the concerned remains very vague at certain points. Thus he is unable to provide details about the precise content of his job as 'political informant'. Furthermore the account of his escape lacks credibility. Thus it is unlikely that the concerned could steal military clothes and weapons without being noticed and that he could consequendy climb over the prison wall. It is also unlikely that the concerned and his wife could pass the passport control at Zaventem [i.e. Brussels Airport] bearing a passport lacking their names and their pictures. Furthermore, the itinerary of the concerned is impossible to verify due to a lack of travel documents (the concerned sent back the passports). -^t
BLOMMAERT INVESTIGATING NARRATIVE INEQUALITY
211
The statements of the concerned contain contradictions when compared to his wife's account. Thus he declares that the passports which they received from the priest [name] were already completely in order at the time they left Angola. His wife claims that they still had to apply for visa in Zaire." Extract 16 "The concerned, of Angolese nationality, has declared being a member of UNITA ('Union Nationale pour lTndependance Totale de 1'Angola'). He has forwarded a first application for asylum in Belgium on May 25, 1993. His application has been refused by the Commissariat-General on December 15,1993. After an illegal stay in Belgium of three years, he has forwarded a second application for asylum on November 20, 1996, on the basis of some documents from Angola which he had received. In two faxes (ofJuly 26, 1996 and September 4, 1996), one of his neighbors (calling himself'[name]') informed him that a certain '[name]', a friend of the concerned who was repatriated to Angola by the Belgian State, had been killed by soldiers. Prior to that, the latter had asked [name of the 'friend'] about the whereabouts of the concerned. From that moment onwards, the house of the concerned is supposed to be occupied [in French: serait occupee] by soldiers. In a letter of March 15, 1994, people are supposed to have already informed him [in French: on lui await deja communique] that the priest named '[name]', who had helped him leave Angola, had disappeared. Apart from that, he has offered a communique from UNITA (of October 23, 1996) as well as his MPLA ('Mouvement Populaire pour la Liberation de PAngola') veteran's ID." Let us n o w take a closer look at these two examples. T h e y are genetically germane and are sequential; Extract 16 obviously adds to Extract 15. T h r e e different types of discourse can b e distinguished in the texts: (a) reported elements from the asylum seekers' narrative in the shape of 'factual' summaries of the applicant's narrative (b) procedural statements referring to the applicant's status, the administrative actions taken in his respect and by him, and the sorts of evidence offered (c) explicit metapragmatic statements and comments on the applicant's narrative. Let us first bring the statements belonging to the first category (reported elements) together from b o t h texts: Extract 17 -He claimed to be a 'political informant' of the MPLA. On October 18, 1992 however, he passed on information to UNITA. At the UNITA office, however, he met with Major [name], who works for the MPLA. Two days later, Major [name] had the concerned arrested. Fearing that the concerned would give the Major away at the trial, [name of the Major] helped the concerned to escape. The concerned fled to [locality] where a priest arranged for his departure from Angola. The concerned came, together with his wife [name and register number] and three children, through Zaire and by plane, to Belgium. They arrived on May 19, 1993. -In two faxes (of July 26, 1996 and September 4, 1996), one of his neighbors (calling himself 'fnamel') informed him that a rortvir, T~--—1» — J -<• 1
DISCOURSE STUDIES
212
concerned who was repatriated to Angola by the Belgian State, had been killed by soldiers. Prior to that, the latter had asked [name of the 'friend'] about the whereabouts of the concerned. From that moment onwards, the house of the concerned is supposed to be occupied [in French: serait occupy] by soldiers. In a letter of March 15, 1994, people are supposed to have already informed him [in French: on lui aurait de'ja communique] that the priest named '[name]', who had helped him leave Angola, had disappeared. Obviously, w h a t we h a v e h e r e are extremely concise a n d highly selective 'sum maries' of the story, organized chronologically and marked b y tense-aspect markers in the verbs as well as b y explicit chronological markers (e.g. " o n O c t o b e r 18, 1992", "two days later", "July 2 6 , 1 9 9 6 " a n d "September 4,1996"). T h e account is not exclusively 'replicating' but contains significant amounts of evaluative modal qualifications (especially i n Extract 16), indicating doubts a b o u t the factual truthfulness of parts of the narrative (e.g. the quotes surrounding the names mentioned b y the applicant, the potentialist verbs "serait occupee", "aurait communi que"). T h e s e summaries reflect the portions of the applicant's narrative d e e m e d 'substantial' to his asylum application a n d qualified, in o n e discursive m o v e , in terms of truthfulness a n d plausibility from the perspective of the procedure. Importantly, this selection of text portions involves text-rewriting a n d textstructuring practices, a n d shifts the epistemic centre from the asylum seeker to the administrator processing the application. T h e second category (procedural statements) includes statements such as " T h e concerned was interrogated o n N o v e m b e r 2 3 , 1993 at t h e CommissariatGeneral [for Refugees a n d Stateless Persons], in the presence of [name], his attorney" a n d " H e has forwarded a first application for asylum in Belgium on M a y 2 5 , 1993. His application has b e e n refused b y the Commissariat-General o n D e c e m b e r 15, 1993. After an illegal stay in Belgium of three years, h e has forwarded a second application for asylum o n N o v e m b e r 2 0 , 1 9 9 6 , o n the basis of some documents from Angola which h e h a d received". It also contains refer ences to evidence offered: "two faxes", "a letter", "a c o m m u n i q u e from UNITA" a n d "his M P L A veteran's I D " . This second category is broadly contextualizing m u c h in the sense that h o m e narratives contextualize the story of the applicant. References to the location of these facts a n d claims in the p r o c e d u r e re-centre the narrative: the facts a n d claims offered b y the applicant h a v e to 'fit' a pro cedural context, a context of sequences of activities, criteria of relevance, and criteria of 'testing' the truth of stories. T h e third category directly connects to this: h e r e w e find explicit expressions of doubt such as (explicit metapragmatic statements in italics): Extract 18 "It has to be noted that the concerned remains very vague at certain points. Thus he is unable to provide details about the precise content of his job as 'political informant'. Furthermore the account of his escape lacks credibility. Thus it is unlikely that the concerned could steal military clothes and weapons without being noticed and that he could consequently climb over the prison wall.
BLOMMAEHT
INVESTIGATING NARRATIVE INEQUALITY
213
It is also unlikely that the concerned and his wife could pass the passport control at Zaventem [i.e. Brussels Airport] bearing a passport lacking their names and their pictures. Furthermore, the itinerary of the concerned is impossible to verify due to a lack of travel documents (the concerned sent back the passports). The statements of the concerned contain contradictionswhen compared to his wife's account. Thus he declares that the passports which they received from the priest [name] were already completely in order at the time they left Angola. His wife claims that they still had to apply for visa in Zaire.'' T h e qualifications pertain to elements from the narrative not included in the ' s u m m a r y ' given elsewhere in the texts. T h e emphasis is o n two things: general common-sense plausibility a n d documentary evidence. Certain arguments are qualified as unlikely o n the basis of a general perception of what is possible a n d what is not (e.g. the unnoticed theft of military uniforms and weapons, the fact that another double-agent, the Major, w o u l d b o t h arrest h i m a n d help h i m escape). T h e r e is n o evidence offered of the impossibility of these facts, they just M « K unlikely. O t h e r elements revolve a r o u n d the presence a n d absence of docu mentary evidence: the faxes, the letter, the veteran's I D , a n d "the itinerary of the concerned is impossible to verify due to a lack of travel documents (the concerned sent back the passports)". T h e pattern of entextualization becomes clear n o w : what we h a v e h e r e is a complete refocalization of the narrative towards n e w deictic centres. T h e story of the applicant is relocated in another space a n d time-frame: that of the admin istrative procedure a n d its pace, that of its standard categories, criteria and textual formats (chronological-sequential a n d d o c u m e n t a r y ) . W h e r e a s t h e h o m e narratives were shown to h a v e a localizing function - personalizing the story and anchoring it in particular spatial a n d temporal deictic frames - the highly modified versions of the story we find in the letters from the authorities show a completely different contextualization of the story: away from the local, away from the experiential, the affective, the emotional, the individual positioning of people in conflicts, towards generalizable categories a n d space-time frames. There is an official version of the conflict (e.g. the war in Angola), used n o t only in asylum procedures b u t also in international trade, development cooperation, foreign policy, and so o n . I n that version, the experiential contextualizations of places, social roles, parties in the conflict a n d so forth, so p r o m i n e n t in h o m e narratives, are replaced b y a rather rigid pattern providing general categorizations (e.g. 'In country X , h u m a n rights are consistency violated', 'country Y is a democ racy', 'in country Z there is a peace agreement between the warring parties, hence the situation has b e e n stabilized' - the latter is the case for Angola) as well as more or less fixed attributes for roles in the conflict (the good, the bad, 'official' parties versus 'unofficial' parties, the State, the army, civil society etc.). T h e r e is a general recentering of the biography of the asylum seeker: the procedurally relevant biography of the applicant is that portion of his or h e r life that can b e rewritten in the shape of a travelogue, starting with (an) event(s) that p r o m p t his or her escape from the country of origin, the process of escape a n d travel itself 9
DISCOURSE STUDIES
214
including details of time, duration, m e d i u m a n d itinerary, and his or h e r arrival in Belgium, all portions of which can b e d o c u m e n t e d b y m e a n s of place descrip tions and time frames. From this grid, deductive patterns of'plausibility' and 'likelihood' are derived; whenever there is doubt, documentary evidence is required to r e m o v e the implausibility. I n the extract given here, o n e of the crucial elements was the fact that the applicant h a d told h o w h e h a d b e e n arrested b y a n U N I T A Major, who was in fact a double-agent just like h i m ; that same Major h a d consequently h e l p e d h i m escape from prison. I n terms of general plausibility within this con textual frame, it is 'unlikely' that the person w h o arrested the applicant would b e the same as the one w h o organized his escape, the m o r e since there was no documentary evidence supporting the Major's role as a double-agent. Similarly, the absence of documentary evidence of the itinerary of the applicant makes that crucial part of his story "impossible to verify" a n d h e n c e unlikely or unreliable. 5.4 Noise and Inequality T h e recontextualizing m o v e s m a d e through the text-trajectories are, as noted b y Briggs above, connected to d e e p p o w e r differences: they involve differential control over contextualization procedures, a n d they involve differential access to skills used for reformulating, ordering, structuring discourse in such a way as to build a 'convincing' argument. T h e complex indexicalities of the stories as well as their general localizing-contextualizing functions basically d o not fit well with the textual and narrative preoccupations of the asylum procedure. Hence, the particular format of narrating imposed during the p r o c e d u r e is sensed to incapacitate asylum seekers a n d prejudice their attempts at m a k i n g sense of 'their' conflict - the basis of 'their' asylum application. A Congolese woman succinctly expresses the textual a n d contextual difficulties experienced during the interrogations that create the 'basic' version of the narrative: Extract 19
i
Mais ils sont trop durs la-bas parce que.. Tu peux raconter je sais pas/ tu peux raconter quelque chose.. Et la prochaine fois quand tu viens quand tu oublies un seul chose c'est fini tu as=tu es rejettee/ alors euh personne ne peut raconter deux fois la meme hist=m§me histoire exactement/ il y a toujours de petits details qu'on ne dit pas ou bien qu'on a=qu'on ajoute apres/ mais eux ils ne tiennent pacompte de ga/ des que tu ajoutes quelque chose ils disent ah la fois passee tu n'a pas dit ga alors c'est fini on te rejette/ il faut aussi comprendre les gens c'est. On=on n'arrive jamais a=a=a raconter une histoire de la meme facon deux foi.ga non Translation:
i
But they are too hard there because... You can tell I don't know what/ you can tell something.. And the next time when you come you forget just one thing if* over you are=you have been rejected/ well ehm nobody can tell twice the same sto=same story exacdy/ there are always small details one doesn't say or else one a=one adds afterwards/ but they don't take that into account/ as soon as you add
BLOMMAERT
215
INVESTIGATING NARRATIVE INEQUALITY
you/ one also has to understand the people it's... One=one never manages to=to=to tell a story the same way twice/ that no. C o m b i n e d with the p r o b l e m of resources discussed in section 3 , asylum seekers' stories a p p e a r to contain too m u c h 'noise' to b e easily inserted i n the asylum procedure. T h e y tell their story in non-native varieties of a language, sometimes to an interpreter w h o also has to use a non-native variety; this t h e n is h a n d e d over to s o m e o n e else, w h o translates a n d transforms it into a n o r d e r e d a n d pat terned written narrative, squeezed into the b o x e s of a standard form; conse quently, this story is treated several times b y still other people, w h o select from the story the parts that look 'txurjrful' a n d the ones that seem 'unlikely'. W h a t is lost i n the process is the narrative of place a n d time that is at the core of the stories: those parts of the story in which applicants bring international conflicts and p h e n o m e n a such as war, famine, poverty into their o w n experiential space, relate personal motives to them, a n d offer this as arguments for obtaining asylum. As soon as this 'noise' has b e e n cut off from the applicant's story, the chances of being understood, believed a n d supported are very slim. This is w h y the densely contextualizing narratives discussed in the previous section b e c o m e crucial a n d problematic i n the asylum procedure. T h e y are a necessary genre for asylum seekers, a n d at the same time a n u n m a n a g e a b l e (and a priori disqualified) genre in the procedure. T h e y are salient because they are a realization of a genre that does n o t fit the genre. O n the basis of what m a n y asylum seekers can mobilize as linguistic and communicative resources, they are the upper limit of what they can d o ; a n d that does not correspond to what they must or are supposed to do.
6. C o n c l u d i n g R e m a r k s Social structure percolates i n a n d is i n d e x e d b y t h e narrative conventions, ideologies and codes of the asylum procedure. I h o p e to h a v e demonstrated that there is at least a potential mismatch between resources a n d expectations which is at the h e a r t of the p o w e r asymmetry that characterizes encounters between asylum seekers and the state; it locates asylum seekers i n a position of conflict vis-a-vis the procedure. But b o t h the p o w e r asymmetry a n d the conflict are socially a n d culturally invisible because of two reasons, b o t h of which h a v e fo do with the pervasiveness of ideologies. The first reason is the widely shared ideology that clients of administrative procedures in a democratic society such as Belgium h a v e control over the basic inguistic-communicative resources n e e d e d to participate fully in the procedures and so to obtain justice a n d the benefits they are entitled to. T h u s , the p r o b l e m •Aharrativity in the asylum procedure can b e represented as a rather superficial, schnical problem a n d suggested remedies can range from training p r o g r a m m e s ir interpreters to m o r e support for asylum seekers from 'experts' such as lawyers. The point I h a v e tried to m a k e i n this article is that we are confronted with a iiinriampni-nl n r n K l o m
r.t
-~
=— '
*•
216
DISCOURSE STUDIES
shape w h o can talk when, in what ways, a n d with what effects' (Briggs, 1996:13), n o t o n e that can b e captured, as Briggs rightly points out in m e t a p h o r s such as 'negotiation', 'difference', or 'variation' which all suggest some degree of equality a n d sharedness a n d a facility of choice for the inferior party. T h e second reason w h y the p o w e r asymmetries a n d conflicts are invisible is their e m b e d d e d n e s s in administrative procedures that are normalized for m e m bers of the autochthonous m i d d l e class, a n d that are i m b u e d with great prestige as the symbolic custodians of a sociopolitical system qualified as just, egalitarian a n d democratic. Every political crisis caused b y incidents with asylum seekers resulted in a reaffirmation of the faith in 'our justice' a n d led to a tightening and increase of the sophistication of the administrative procedure: m o r e people would b e hired, they would b e better trained, the procedure would b e accelerated, handwritten documents would b e replaced b y computerized standard forms. The administrative text-making machinery was never questioned; o n the contrary, it was strengthened a n d enlarged. T h u s the capacity to shift narratives from one context to another and from one authoritative entextualization to another was increased, a n d with it the p o w e r asymmetry with regard to the production, treatment, ordering, a n d m a k i n g sense of narratives. This belief in the just, egali tarian a n d democratic nature of the state a n d state procedures reminds us of a remark m a d e long ago b y Perry Anderson (1976-7) in his celebrated commentary o n Gramsci: the assumption of equality in the face of democratic procedures a n d state institutions such as the law is at the core of capitalist ideology because it obscures fundamental inequalities in society. T h e state is assumed to represent everyone a n d to r e n d e r service to everyone in the same way; at the same time, administrative procedures of the state privilege elite literacies a n d narrativities a n d so shape a n d perpetuate d e e p social inequalities (a process complicated by the introduction of n e w technologies in bureaucracy). A l o n g with the concrete case I h a v e argued h e r e , I h o p e to have offered arguments for two m o r e general theoretical points, b o t h of which are inspired b y Briggs's (1997) highly relevant discussion of narrative a n d inequality in institu tional contexts. First, issues of resources deserve far m o r e attention than they h a v e received so far in the study of language in society. Assumptions of shared ness, as stated previously, seem to determine m u c h of what goes on in the critical analysis of discourse in western societies, whereas the highly layered and hier archical systems of literacy a n d communicative skills that dominate our societies seem to dictate a m o r e attentive stance towards p h e n o m e n a such as accents, differential competence, difficulties in writing and so on. Before people can em b a r k o n discursive w o r k captured u n d e r labels such as 'conversation', 'exchange' or 'negotiation', conditions of'sayability', 'expressibility' a n d 'mobilizability' ot resources n e e d to h a v e b e e n m e t (see the remarks in Briggs, 1997: 538-401, Every conception of 'context' in discourse should include such conditions, because they are a t r e m e n d o u s influence o n what h a p p e n s in discourse and on what h a p p e n s with discourse. Resources are contexts. Second, I share Briggs's concern for the circulation of discourse as a crucial ingredient of identity-forming social and political practices in our societies. 'Cases (administrative, legal, welfare, medical, educational a n d p r o b a b l y far more) are M
BLOMMAERT
INVESTIGATING NARRATIVE INEQUALITY
217
formed in the textual trajectories outlined in this article rather than in single instances of communication a n d single texts. We n e e d to follow the process of text-making-as-social-and-political-process; it is h e r e that p e o p l e and subjects are constructed, cases are j u d g e d a n d individual lives are b e i n g influenced. As soon as w e enter worlds in which talk a n d written text are seen as replicas of one another (and in which s o m e o n e else's notes of what I said can b e offered to m e as ' m y ' story), we enter a world of differential p o w e r relations, which needs to b e scrutinized in great detail. Michel Foucault's (1975) image of subjects being transformed into knowable objects of clinical observation b y m e a n s of a multilayered c o m p l e x of discursive a n d material practices is l o o m i n g large. T h e apparendy small shifts our stories undergo as soon as they enter institutional textmaking systems are instances of such practices of Foucauldian savoiria which social issues b e c o m e individual yet standardized 'cases'. T h e fact that we tend not to b e aware of these processes - a n d even attribute considerable prestige to the system in which they develop - is a crucial a n d unavoidable topic for critical studies of language a n d society.
Acknowledgements 1 am grateful to Dell Hymes, Ben Rampton, Monica Heller, Jim Collins,Jef Verschueren, Stef Slembrouck, Roxy Harris, Jiirgen Jaspers, Jan-Ola Ostman and Andreas Wagner for comments and suggestions. Katrijn Maryns contributed significantly to the final shape of the argument of this article, as did Chris Bulcaen, my most critical and most constructive reader. The second and fourth year students of the African Studies Program of the Univer sity of Ghent (1998-9) did an enormous amount of work collecting and shaping the data, and so I owe the greatest debt to them. A preliminary version of this article was circulated as 'Language, Power and Identity Working Paper no. 1' (see http://bank.rug.ac.be/lpi).
Notes 1. Hymes and Cazden's paper was originally published in 1978 in Keystone Folklore 22; it was reprinted as chapter 7 of Hymes (1980) and later reprinted again as chapter 5 of Hymes (1996). The paper is not really ajoint product; individual sections are attributed to either Hymes or Cazden; for the purposes of this article, I consider it to be a col laborative writing product authored by Hymes and Cazden. 2. In late 2000, Katrijn Maryns obtained permission to record procedure interviews. Fieldwork is underway and genuine 'cases' will be available for research very soon. The first bits of data can be found in Maryns (2000) and Maryns and Blommaert (2000). On the basis of first fieldwork observations, nothing contradicts the patterns and mechanisms of discourse production and reception proposed in this paper. X The interviews were conducted by students of the 2nd and 4th year of African Studies at the Ghent University, as part of a course project supervised by the author. The length of the interviews varied from approximately half an hour to four hours. The interviews were recorded between November 1998 and April-May 1999, at the height of the so-called 'Semira Adamu crisis'; see section 2.
218
4. I provide English translations, despite the fact that the 'broken' and hence very com plicated forms of expressions can hardly be projected in another language in an equivalent way. I use a highly simplified set of transcription conventions in the pre sentation of the examples, namely: * = /
stress on the following syllable latching, rapid succession of turns, or self-correction internationally marked phrase or sentence end pause CAPS high pitch, loud
5. Katrijn Maryns (2000) presents some samples of interviewer-interviewee interactions in the asylum procedure. In the data reported here, there are instances where the Belgian interviewer is wrong-footed by the interviewee for reasons that have to do with proficiency in French or English. Interestingly, interviewers then display a tendency to move into a cross-questioning tactic, insisting on details and accuracy. 6. This was, in fact, the second time that P used the expression " manger + currency" in the interview. The first time, Q, did not pick it up as problematic. Note in passing how this phenomenon - a man from Angola using Congolese French colloquial expressions - offers us a glimpse of the migration history of the speaker. We can now suppose that the man, after having fled from Angola, has spent some time in the Congo before moving on to Europe. See Maryns (2000) for arguments and case analyses on the issue of identifying asylum seekers. 7. The emergence of structure in these narratives may be an effect of the repeated tel ling of the story, so that 'fully-formed narratives' are created. Dell Hymes, in cor respondence about this analysis, remarks: It seems very likely that you have here what could in some cases be the early stages of 'fully formed' narratives of the sort I addressed in my own paper. That if those who tell them were in circumstances which led to them being told again and again, rehearsed, as it were, they might take on increasingly tight form... It is extremely relevant to find... that a contrast in initial elements, such as absence/presence of 'So' can be found to distinguish parts already at this stage. (See also Hymes, 1998.) The point I wish to make in this part of the article is that such structuring not only occurs during very early phases of the narrative development of such stories, as Hymes points out, but that it also occurs despite a very restricted repertoire of lin guistic resources. 8. Narrative inequalities based on differing narrative conventions and the differential distribution of communicative resources have been well documented with respeci to courtroom discourse. I would like to mention just two cases. Marco Jacquemei (1992) emphasized the role of metapragmatic strategies in attorneys' courtroom tactics against Italian mafia-/>eretfti; the witnesses' communicative style was commented upon in attempts to discredit their credibility, and the normativity of 'correct' courtroom discourse was opposed to the pentitfs use of the Neapolitan dialect. At a more general level, O'Barr and Conley (1996) distinguish between a professional and a lay ideology of law; the former can be characterized as 'rule-oriented', the latter as 'relational' Both ideologies shape and make relevant differences in courtroom talk, and the failure to respect these differences by lay litigants 'may limit their access to justice' (O'Ban
BLOMMAERT INVESTIGATING NARRATIVE INEQUALITY
219
and Conley, 1996: 118). For reasons explained earlier, the discussion in this section is conjectural. 9. In a letter about this case from the Department of Immigration of the Ministry of the Interior, it is for instance said that 'the situation in his country [i.e. Angola] cannot be considered to be exceptional so as to vindicate a regularization on the basis of art, 9§3\ There are no fixed criteria forjudging local situations to be 'normal' or ^abnormal', decisions on this point being the privilege of the Minister of the Interior.
References Anderson, Perry (1976-7) 'The Antinomies of Antonio Gramsci', New Left Review 100: 5-78. Bauman, Richard (1986) Story, Performance and Event: Contextual Studies of Oral Narrative. Cambridge: Cambridge University Press. Bauman, Richard and Charles Briggs (1990) 'Poetics and Performance as Critical Perspectives on Language and Social Life', Annual Review of Anthropology 19: 59-88. Blommaertjan (1997) 'The Slow Shift in Orthodoxy: (Re)formulations of "Integration" in Belgium', in Charles Briggs (ed.) Conflict and Violence in Pragmatic Research, Pragmatics 7(4), special issue: 499-518. Blommaert, Jan (2001) 'Context Is/As Critique', in J. Blommaert,J. Collins, M. Heller, B. Rampton, S. Slembrouck andj. Verschueren (eds) 'Discourse and Critique', Critique of Anthropology 29(1), special issue: 13-32. Blommaert,Jan and Verschueren, Jef (1998) Debating Diversity: Analysing the Discourse of Tolerance. London: Routiedge. Briggs, Charles (1996) 'Introduction', in Charles Briggs (ed.) Disorderly Discourse: Narrative, Conflict and Inequality, pp. 3-40. New York: Oxford University Press. Briggs, Charles (1997) 'Notes on a "Confession": On the Construction of Gender, Sexuality, and Violence in an Infanticide Case', Pragmatics 7(4): 519-46. Collins, James (1995) 'Literacy and Literacies', Annual Review of Anthropology 24: 75-93. Collins, James (1996) 'Socialization to Text: Structure and Contradiction in Schooled Literacy', in Michael Silverstein and Greg Urban (eds) Natural Histories ofDiscourse, pp. 203-28. Chicago, IL: University of Chicago Press. Collinsjames (1998) Understanding Tohwa Histories: Western Hegemonies and Native American Responses. New York: Routiedge. Foucault, Michel (1975) Surveiller et punit: Naissance de la prison. Paris: Gallimard. Gal, Susan and Woolard, Kathryn (eds) (1995) 'Constructing Languages and Publics', Pragmatics 5(2), special issue: 129-282. Havilandjohn (1996) '"We Want to Borrow Your Mouth": Tzotzil Marital Squabbles', in Charles Briggs (ed.) Disorderly Discourse: Narrative, Conflict and Inequality, pp. 158-203. New York: Oxford University firess. Haviland, John (1997) 'Shouts, Shrieks, and Shots: Unruly Political Conversations in Indigenous Chiapas', Pragmatics 7(4): 547-73. Hymes, Dell (1981) In Vain I Tried to Tell You: Essays in Native American Ethnopoetics. Philadelphia: University of Pennsylvania Press. Hymes, Dell (1996) Ethnography, Linguistics, Narrative Inequality: Toward an Understanding of Voice. London: Taylor & Francis. Hymes, Dell (1998) 'When is Oral Narrative Poetry? Generative Form and its Pragmatic Conditions', Pragmatics 8:475-500.
220
t
DISCOURSE STUDIES
Hymes, Dell and Cazden, Courtney (1980) 'Narrative Thinking and Storytelling Rights: A Folklorisfs Clue to a Critique of Education', in Dell Hymes (ed.) Language in Education: EthnolinguisticEssays, pp. 126-38. Washington, DC: Center for Applied Linguistics. Jacquemet, Marco (1992) '"If He Speaks Italian It's Better": Metapragmatics in Court', Pragmatics 2(2): 111-26. Labov, William (1970) 'The Logic of Nonstandard English', in Frederick Williams (ed.) Language and Poverty, pp. 153-90. Chicago, IL: Markham. Malkki, Liisa (1995) Purity and Exile. Chicago, IL: University of Chicago Press. Maryns, Katrijn (2000) 'Identifying the Asylum Speaker: Notes on the Pitfalls of Linguistic Identification', paper presented at the Conference on Applied Language Studies and Services into the Millennium, Cape Town, December. Maryns, Katrijn and Blommaert,Jan (2000) 'Pretextuality and Pretextual Gaps: On De/ Refining Linguistic Inequality', paper presented at the Conference on Applied Lan guage Studies and Services into the Millennium, Cape Town, December. Maryns, Katrijn and BlommaerLjan (2001) 'Stylistic and Thematic Shifting as a Narrative Resource: Assessing Asylum Seekers' Repertoires', Multilingua 20(1), in press. O'Barr, William and Conley, John M. (1996) 'Ideological Dissonance in the American Legal System', in Charles Briggs (ed.) Disorderly Discourse: Narrative, Conflict and Inequality, pp. 114-34. New York: Oxford University Press. Ochs, Elinor and Capps, Lisa (1996) 'Narrating the SelF, Annual Review ofAnthropology 25: 19-43. Radley, Alan (1990) 'Artefacts, Memory and a Sense of the Past', in David Middleton and Derek Edwards (eds) Collective Remembering, pp. 46-59. London: Sage. Sarangi, Srikant and Slembrouck, Stefaan (1996) Language, Bureaucracy and Social Control London: Longman. Silverstein, Michael and Urban, Greg (eds) (1996) Natural Histories of Discourse. Chicago, IL: University of Chicago Press. Urban, Greg (1996) 'Entextualization, Replication and Power', in M. Silverstein and G. Urban (eds) Natural Histories ofDiscourse, pp. 21-44. Chicago, IL: University of Chicago Press. Wodak, Ruth (1995) 'Critical Linguistics and Critical Discourse Analysis', injef Verschueren, Jan-Ola Ostman and Jan Blommaert (eds) Handbook of Pragmatics, Manual, pp. 204-10. Amsterdam: John Benjamins.
9 "You Gotta Know How to Tell a Story": Telling, Tales, and Tellers in American and Israeli Narrative Events at Dinner Shoshana Blum-Kidka
t
S
tory-telling in ordinary talk between intimates is o n e of the most c o m m o n enactments of narrative discourse. We tell stories to each other as a m e a n s of packaging experience in cognitively a n d affectively c o h e r e n t w a y s (Labov & Fanshel 1977, Sacks 1974), or in the terms of Bruner 1990, as a way to test the borderlines between the exceptional a n d ordinary. Fisher 1987 considers the narrativization of experience a basic h u m a n n e e d ; the essential nature of human beings is captured b y the m e t a p h o r of m a n as homo narrans. Extending Kenneth Burke's definition of m a n as a symbol-making and symbol-using animal, Fisher proposes a n all-encompassing definition for the role of narratives (63): T h e idea of h u m a n beings as storytellers posits the generic form of all symbol composition. It holds that symbols are created and communicated ultimately as stories m e a n t to give order to h u m a n experience a n d to induce others to dwell in t h e m in order to establish ways of living in c o m m o n , intellectual a n d spiritual communities in which there is con firmation for the story that constitutes one's life. 1
Although the narrative use of language, as insisted b y H y m e s 1982, is a uni versal function, such use can b e expected to vary culturally as d o other ways of speaking. Indeed, oral narrative styles vary b y culture for b o t h adults and chil dren, as shown b o t h b y work adopting a cross-cultural perspective e.g. Scollon & Scollon 1981, Tannen 1980) and b y comparative studies in the context of language socialization (e.g. H e a t h 1983, Schieffelin & O c h s 1986). M y goal h e r e is to apply a culturally sensitive analysis to narratives old in the course of dinnertable conversations in middle- to upper-classJewish-American and Israeli families.
DISCOURSE STUDIES
222
To apply such an analysis, w e n e e d first to understand the nature of the narrative event in question. Conversational story-telling can b e looked at from a social-interactionalist position, as interaction, with a focus o n h o w the narrative emerges in its context, a n d / o r as discourse, with a focus o n the textual e n d product abstractable from the context. T h e first a p p r o a c h is represented b y work o n conversational story telling from the perspectives of b o t h ethnomethodology (e.g. Jefferson 1978, Sacks 1974) a n d discourse analysis (Polanyi 1989, Schiffrin 1984, Tannen 1984); such work shows the m a n n e r in which the structure of oral narratives is conver sationally accomplished. Particularly relevant h e r e are studies focusing o n narratives during family dinner-table conversations (Erickson 1982,1988; Ochs et al. 1989; O c h s et al. 1992). A complementary, rich source of information on this dimension is provided b y studies in folklore that highlight the poetic and social interactional assets of performance (Bauman 1986, Briggs 1988, H y m e s 1981, Kirshenblatt-Gimblett 1975, S h u m a n 1986). By contrast, the discourse ap p r o a c h is best illustrated b y Labov's influential w o r k (Labov & Fanshel 1977, L a b o v & Waletzky 1967), which unveils the structural coherence of seemingly chaotic onversational renderings of personal experience. T h e narrative-events examined-here, performed b y b o t h adults and children, function as crucial socializing contexts for family interaction in general, as claimed b y Bernstein 1971. H e n c e a further perspective that n e e d s to b e a d d e d is crosscultural variation in practices of narrative socialization, as suggested b y crosscultural a n d cross-ethnic studies of language socialization (Blum-Kulka & Snow 1992, H e a t h 1983, Miller et al. 1990, Schieffelin & O c h s 1986). Yet n o n e of these approaches o n its own captures the u n i q u e nature of family narratives. We n e e d a n approach that accounts simultaneously for family story telling as a n event, a social action unfolding in real time, as well as (at the discourse level) a text about other events. T h e links a n d transitions between these two realms are provided b y performer/tellers. Dinner-table narrative events are unique: they represent a three-way intersection of the act of narration, the textual content a n d form of the narrative, a n d the persons responsible. Taking all three dimensions together, narratives b e c o m e narrative events. Like other speech events, narrative events h a v e their specific n o r m s governing the scene, participation rights, message content, message form, a n d rules of interpretation (Hymes 1974:55-8). I n narrative events, these features can b e seen as subsumed under three dimensions of narrativity: telling (narration), tales (narratives), a n d tellers (narrators). I n oral storytelling, the realm of telling is e m b e d d e d (in an opene n d e d fashion) in the realm of conversation, a n d the realm of tales within that of telling, as depicted in Figure l. 2
3
Telling is the act of narrating in real time, the actual performance of a story before a n audience. I n the terms of Goffman (1981:144-5), telling is enacted by the role of the speaker as Animator, the one responsible for the sounds that make intelligible speech c o m e into being. T h e central issue of performance in family narratives is a social-interactional one: w h o participates in whose stories, and how. To narrate "is to m a k e a bid for power" (Toolan 1988:6); entering the tell ing m o d e in the family context raises the issues of narrative participation rights.
BLUM KULKA
AMERICAN AND ISRAELI NARRATIVE EVENTS AT DINNER
223
TALE
(TELLER)
'
TELLING
1
CONVERSATION Figure 1: Dimensions of narrativity
Even w h e n such rights are assumed b y virtue of social role i n the family, as with parents, they still n e e d to b e renegotiated conversationally o n each specific oc casion (Jefferson 1978, Polanyi 1989, Polss 1990). For children, both participation rights a n d m o d e s of story entry remain goals to b e achieved with some difficulty. T h e division of telling-space m a y vary b y role in the family (children vs. adults) as well as b y role a n d culture, as w h e n children in o n e culture are granted story telling rights over a n d b e y o n d their rights i n another. Cultures also m a y differ in the framing of transitions from the realm of conversation to the realm of tell ing, as well as i n the importance of the telling with relation to the tales. Tale is the stuff from which narratives are m a d e . T h e t e r m refers to t h e two dimensions of narrative captured in the poetics of narrative fiction b y the distinction offabulavs. sjuzet (in the Russian formalists' terms) or story vs. narrative (Rimmon-Keenan 1983). T h e fabula or the story of narratives "designates the narrated events, abstracted from their disposition in the text a n d reconstructed in their chronological order, together with the participants in these events" (Rimmon-Keenan, 3). I n other words, in experience-based narratives, the fabula consists.of the real-world building blocks used for the construction of the story. The sjuzet, on the other h a n d , relates to the w a y in which the story is shaped in the making: the "spoken or written discourse that undertakes the telling." I n the actual discourse, events d o n o t necessarily appear in chronological order, a n d content is filtered through s o m e perspective, sometimes called a "focalizer" (Rimmon-Keenan, 74).Itismeresponsibmtyofthe^4a/Aor(Goffman 1981:144-5) to select the words in which the fabula is encoded in a w a y that is still retrievable for the audience. Cross-cultural variation in fabula choice m a y b e expressed in spatio-temporal orientation of narrative plots (e.g. recent vs. not-recent past) or in the type of protagonists that are foregrounded (self vs. other). However, cultural attitudes to the preferred style of sjuzet m a y b e expressed through critical com ments to children about the w a y they are telling a story. As o n e father told his
DISCOURSE STUDIES
224
son, w h o failed to give a convincing performance of a j o k e learned from the father, "You gotta k n o w h o w to tell a story." Tellers (performers) m a y or m a y n o t b e the persons accountable for the story. This is the role called t h e Principal b y Goffman (1981:144-5): the o n e committed to what the w o r d s say. I n personal narratives, all three speaker roles (principal, author, a n d animator) can m e r g e into one. I n a personal narrative, it is the Prin cipal w h o has actually experienced the events recounted a n d claims authorship for the narrative, acting also as Animator. But m o r e generally, a n d especially w h e n children are involved, the three speaker roles m a y b e realized b y different speakers. A m o t h e r w h o repeats to the father a story of personal experience, re counted to h e r earlier b y h e r child, m a y act only as Animator (if she quotes verbatim); or she m a y take authorship rights, if she edits the child's version. In either case, the child remains the principal for the story. F r o m a cross-cultural perspective, the relation of tellers to telling and tale raises issues of authorship vs. performance: H o w is authorship emically defined? W h o , in each culture, is entided to tell whose stories?
Background and Method T h e database for the study consists o f 2 6 4 narrative events that occurred during two dinner-table conversations with eight middle-class Jewish-American and eight Israeli families.* AllJewish-American families were residents of the Boston area; all Israeli families lived at the time in Jerusalem. T h e narrative events were examined as part of a larger project investigating the pragmatic socialization of children in the presence of a m e m b e r of the research t e a m w h o came from the s a m e cultural background as the family. Following initial contacts b y phone, the observer visited the h o m e and got acquainted with the family prior to record ings. T h e same observer stayed with the family throughout the research period. T h e families were told that w e w e r e interested in comparing Israeli to American family dinners, a n d they w e r e p r o v i d e d with details w h e n interested. T h e obser vers were invited to join the families for dinner as a matter of course. Hence the situation in which we are c o m p a r i n g the groups is that of families interacting with a semi-official guest. As will b e discussed later, interaction with the observer p r o c e e d e d differently in the two groups, revealing a cultural difference in this m o d e of family self-presentation. All families participating in the project came originally from a E u r o p e a n background (mainly from Russia a n d Poland) and h e n c e share a J e w i s h Eastern-European heritage. T h e p a r e n t generations are native-born A m e r i c a n or native-born Israeli. T h e A m e r i c a n a n d Israeli families w e r e m a t c h e d for n u m b e r a n d ages of children. Segmentation and Coding Narrative events w e r e defined b r o a d l y as conversations that recapitulate pasi events. Since the perception of w h a t constitutes a narrative m a y well differ from
BLUM-KULKA
AMERICAN AND ISRAELI NARRATIVE EVENTS AT DINNER
225
children to adults, or across cultures, we deliberately avoided imposing further structural criteria o n tales (e.g. n u m b e r of events mentioned). T h e segments extracted from full transcripts of the dinner conversations v a r y in length from brief exchanges, of a few seconds, to long elaborated happenings, of u p to 10 minutes. Boundaries of the segments w e r e decided textually, in agreement a m o n g three analysts, o n the basis of transition markers signaling entrances a n d exits from the telling realm. Analysis of the texts proceeded b y a set of quantifiable coding categories, designed to capture variation o n each dimension of narrativity. T h e categories used a n d the findings from the coding are integrated into the discussion. 5
I n the following I address the degree of cultural diversity between JewishAmericans a n d Israelis in their attitudes toward telling, tales, a n d tellers in family narrative events. I argue that, considered from this threefold frame work, family narrative events of the two groups share certain features attributable to the speech event in which they are e m b e d d e d . Yet the groups differ culturally in the struc turing of each dimension of narrativity, a n d i n the relative importance granted to each:Jewish-American narrative events foreground tellers and the act of telling, while Israeli narrative events prefer to focus o n tales a n d tellers. T h e argument * is developed as follows: the degree of cultural diversity is discussed, first, in terms of the division of narrative space b e t w e e n m e m b e r s of the family and styles of story initiation; second, in terms of the spatio-temporal orientation of tales, a n d the degree of conventionalization i n the transformation of tales into telling; a n d third, in terms of the relationships between ownership of the tale and participation in the telling.
Issues of T e l l i n g Gaining Access to Narrative-space: Children and Adults Family dinner-table narratives t e n d to b e jointly constructed affairs (Erickson 1988, Ochs et al. 1989). I n our case, collaboration takes several forms: stories are co-narrated, constructed through question/answer sequences, or told with sporadic b u t meaningful contributions from a n active audience. I n all these, telling is shared b y several m e m b e r s of the family, sometimes including the observer. Collaborative process is at work in b o t h Israeli and American narrative events: in almost half of the narratives, m o r e t h a n four m e m b e r s participate. By contrast, less t h a n 5% are performed b y just o n e participant; see Figure 2. What these n u m b e r s mask, though, are cultural styles of participation; as will be shown, Israeli a n d Jewish-American families differ greatly in the w a y they draw lines of demarcation between teller(s) a n d audience. Israeli and A m e r i c a n families differ in their attitudes toward the division of telling-space, as well as in m e m b e r s ' m o d e s of participation in narrative events. Consider attitudes t o w a r d the division of telling-space b e t w e e n adults a n d children. In both groups, dinner time is perceived as a prime occasion for spot lighting children as narrators. To reveal attitudes toward children, we first divided
226
1
DISCOURSE STUDIES
i
Four or more 42% Figure 2: Number of participants in narrative events
narrative events i n terms of identity of the p r e d o m i n a n t mainteller(s), as either child (or children) or adults); see Figure 3 . I n a further analysis, w e identified the initiator of the narrative event b y role i n the family as child, father, mother, or observer, see Figure 4. T h e degree to which children are considered conversational partners in adult c o m p a n y is noteworthy across the two cultures. I n discussing determinants of cultural variability in conversation, Schieffelin & Eisenberg (1984: 382) note:
PERCENT
ISRAELI
AMERICAN TYPE
[ 3 CHILD MAINTELLER • Tflemrp 3 ? M a i n r p l l p r s - division of tellinff SOace
ADULT MAINTELLER
BLUM-KULKA
AMERICAN AND ISRAELI NARRATIVE EVENTS AT DINNER
227
PERCENT 60 50.4 40 A
35.8 28.5
30 A 20
23.5
r
A
24£
E^i
10 A
•
0
ISRAELI
16.3
11 H
AMERICAN
r
ROLE •
FATHER B MOTHER •
CHILD •
OBSERVER
Figure 4: Narrative initiation Cultures vary along a continuum ranging from societies in which children are not allowed or expected to talk to adults or strangers a n d / o r are not thought of as co-conversationalists to those in which children are actively encouraged to talk to adults, w h o in turn support this interaction a n d try to understand it. I n some societies adults think it is important to elicit speech from children, socializing t h e m in terms of production a n d inter action skills (e.g., Kaluli in Papua New Guinea, white middle-class American); b u t in other cultures children are encouraged to b e quiet a n d speak only w h e n spoken to (e.g., rural Louisiana Blacks, L u o in Kenya).
t
We found that, across the two groups, children take u p 4 2 % of all narrative space (see Fig. 3). T h e case of the middle-class families studied h e r e well exem plifies a cultural pattern which treats children as rightful co-conversationalists; in both the Jewish-American a n d Israeli families, children a r o u n d the dinner table share, in the terms of Goffman (1981:131-2), a n official status as ratified participants. ' T h e pattern is most salient in the A m e r i c a n families (see Fig. 3). American children act as maintellers in 6 6 % of all narrative events, leaving the adults as maintellers of 34%. But the adults in the Israeli families take u p a m u c h higher proportion of narrative space, playing the d o m i n a n t role in 5 4 % of narrative events. I n line with the overall attitude toward the division of narrative space between adults a n d children, A m e r i c a n children are also m o r e active story-initiators (by 15%) than Israeli children. 1
This difference points to a cultural distinction in m o d e s of narrative social ization at dinner. Notions of tellability a n d cultural styles of telling are acquired by children b o t h b y engaging in the telling a n d b y being exposed to stories told b y others. T h e A m e r i c a n families tend to emphasize socialization b y allow ing for the display of narrative practice; b u t in the Israeli families, children are
DISCOURSE STUDIES
228
This does n o t necessarily m e a n that Israeli children engage generally in less display of narrative t h a n do A m e r i c a n children; stories m a y b e told b y children at other points during the day, to siblings or o n e parent. But it does m e a n that dinner-table conversations, w h e r e b o t h adults a n d children jointly participate, are perceived b y the A m e r i c a n families as an occasion to focus o n the chil dren as narrators; in the Israeli families, narrative space is divided between adults and children. From the children's point of view, different gains are involved in each practice. Being encouraged to tell stories m a y b e important in developing confidence in performing rights a n d the skills of narration (Blum-Kulka & Snow 1992); however, acting as p r i m a r y or e v e n secondary audience to adult stories allows access to the experience of significant others, thereby e x p a n d i n g the bases for identification (Miller et al. 1990) a n d shaping cultural notions of teliability or reportability (cf. H y m e s 1981). 7
Narrative Initiation: Insiders and Outsiders, Women and Men Adult roles also are differentiy distributed across the two groups, most notably in regard to the observer. Observers in the Israeli families initiate 15% more narrative events than d o the observers in the A m e r i c a n families (see Fig. 4). I n other words, the observers in the Israeli families s e e m m u c h m o r e confident in their story-telling rights than their A m e r i c a n counterparts. O r , as suggested b y D e b o r a h Tannen (personal communication), the Americans m a y h a v e dif ferent ideas about their obligations, rather than their rights: with their scientific tradition of "objectivity in social science, p e r h a p s they feel it i n c u m b e n t upon t h e m n o t to participate a n y m o r e t h a n necessary. This result ties in with other observations o n the relations b e t w e e n observers and families in the two groups. I n both groups, observers (with o n e exception) w e r e invited to j o i n the dinner table as a matter of course; b u t the nature of their interaction with the family differed in key, and consequently in rules of interaction. I n terms of the con tinuum of formality p r o p o s e d b y Irvine 1979, the generally prevailing key of family discourse (because of intimacy a m o n g participants) is that of informality. Yet there are differences of degree between the two groups: Israelis seem to celebrate the outermost informal end of the continuum. Consequently, observers are drawn into the circle of conversation from the onset, a n d interact with all family m e m b e r s , n o t only in the role of familiar guests b u t rather as actual or potential friends. This is in line with the ethos of solidarity in Israeli culture, which tends to symbolically minimize social distance (Blum-Kulka et al. 1985, Katriel 1986). O n e of the privileges granted to friends is that of b o t h self-directed and other-directed narrative initiation. Observers in the Israeli families accord ingly share initiation almost equally with other m e m b e r s present. Being a familiar guest at the American family's somewhat m o r e formal dinner table does n o t bestow o n the observer the privileges granted to friends; though partaking in the interaction, the observer does so cautiously, u n d e r the rules of interaction governing communication between nonintimates. Not being overdy active in narrative initiation is o n e w a y in which such caution is exercised.
I
BLUM-KULKA
AMERICAN AND ISRAELI NARRATIVE EVENTS AT DINNER
229
H e n c e it is n o t surprising to find that observers in the A m e r i c a n families are engaged in less t h a n 10% of story initiations. A n o t h e r difference between Israelis a n d Americans is in the relative parts played b y mothers vs. fathers in narrative initiation. I n the Israeli families, mothers have twice as m a n y initiations; in the A m e r i c a n families, fathers play a slightly m o r e active role. This finding is in line with general patterns of participation rates in dinner-table talk. Following the categories p r o p o s e d b y Bublitz 1988 for topical action, w e h a v e analyzed the proportional role of fathers vs. mothers as topic initiators throughout the entire conversation: out of 47 parent-introduced topics, Israeli m o t h e r s introduced 76.5%; b u t out of 67 parent-introduced topics in the A m e r i c a n families, fathers introduced 66%. O n e reason for these findings m a y b e the A m e r i c a n perception of the oc casion as having formal overtones because of the presence of the observer. Though the serving of food is accomplished mostly b y the mothers, the fathers in these families take it u p o n themselves to entertain the guest, e.g. to introduce topics a n d to use narratives to k e e p the conversation going. By contrast, in the Israeli families, the mothers seem to take charge of the event, both in instrumental terms (here too it is mostly the w o m e n w h o serve the food) a n d b y keeping the conversation going. 8
Styles ofNarrative Initiation The solidarity ethos of Israeli society, as manifested in attitudes toward the ob server, finds further expression in styles of narrative-event initiation. Consider story entry. I n an independent study of the same corpora, Polss 1990 has analyzed in detail the types of devices used b y initiators a n d story recipients in the course of story-entry talk. G e r m a n e to the discussion of attitudes toward telling is her analysis of the types of devices used in responsive utterances. Following Tannen's distinction (1984, 1985, 1989) between high-involvement a n d high-considerateness conversational styles, Polss distinguishes between high-involvement and lowinvolvement narrative response strategies. High-involvement responses focus o n the tale a n d the teller; in Tannen's terms, they show active "participatory listenership" (1984:30). These include devices such as request for information, confirmation of information, a n d listener contribution to the narrative. Low-involvement responses focus on the telling; they signal message reception, thereby confirming the teller's success in aligning story-recipients. T h e s e include different types of uptakers (Edmondson & H o u s e 1981:62-3), e.g. neutral back-channeling responses (mhmm, uhhuh, yeah, right, okay, etc.) a n d emotively colored ones [really?, good, for heaven's sake). Polss found interactive, high-involvement style m o r e characteristic of story entry in Israel than in America: the proportion of high to low involvement styles is 82% to 18% in the Israeli narratives, c o m p a r e d to 6 9 % to 3 1 % in the American stories. Thus Israeli story-opening exhibits a higher emphasis on coparticipation and demonstration of personal involvement. Israeli coparticipants frequently interpret story-initiator information, showing their concern with tale a n d teller.
DISCOURSE STUDIES
230
(1) Israeli family 7: The conversation takes place as the family is getting ready to sit down at the table. Present also are the daughter (16) and the two sons (12 and 10). 1 Observer: etmol hayinu\//] hayiti ecelpnina Yesterday we were [//] I was at the ve + . . . home of Pnina and + . . . 2 Mother: ve-cvika? And Cvika? Observer: ve-cvika ken. And Cvika, yes. Mother: nu # ve-ex halax? So how did it go? Observer: haya meod nexmad. It was very nice. [Story] 9
T h e fast rate of speech, lack of interturn pauses, a n d dialogic unfolding of this story entry places it high o n the involvement continuum. T h e story recipient's high engagement can further b e seen b y h e r use of what Tannen (1984:118) calls "cooperative promptings'' at every turn (e.g., "So h o w did it go?"). Americans, b y contrast, exhibit less relative focus o n interpersonal involvement, devoting their efforts to floor-management tasks aimed at securing the telling. (2) American family 2: The children are Martin (8m), Daniel (6m), and Tamara (4f). The conversation takes place in the middle of dinner. Martin: My best friend got about + . . . You see me and my best friend were studying rockets-? Observer: Uhhuh. [Story] T h e shared features noted here are high degrees of collaboration a n d of inclusion of children in narrative events. Cultural diversity is revealed in attitudes toward tellers - the question of w h o participates - a n d in styles of story entry. This trend for cultural diversity of a gradient nature, against a background of shared orientations, is also noticeable in attitudes toward tales.
Transforming Tales to Telling Cultural Variation in Spatio-Temporal Orientation of Tales W h e r e do tellers find the tales for constructing narratives at dinnertime? Certainly n o t in fiction; the vast majority of narratives in b o t h groups (90%) are derived from real-life experience. I n p r e d o m i n a n t l y adult narratives, fictional topics do not exceed 4%. I n child-involved narratives, fiction does play a role (14%). Israeli children mention fictional characters from story books (Aladdin is one) a n d tell the con tents of movies and books. A m e r i c a n children talk about fictional characters from television: Sesame Street, Bugs Bunny, etc. We h a v e analyzed spatio-temporal framing b y c o d i n g the two dimen sions independently. First, since our definition of narratives included only stories of the past (see O c h s et al. 1989 for a different notion of temporal framing in family narratives), we distinguished narratives b y temporal reference, including
BLUM-KULKA
AMERICAN AND ISRAELI NARRATIVE EVENTS AT DINNER
231
today ("I finished m y assignment today in u m a n d . . ."), recent past ("I m e t an interesting m a n o n the beach last week"; "Last night D e b b i e . . . " ) , or distant past ("that h a p p e n e d about five years ago.") Cases which were either timeless (such as jokes a n d fiction), or lacking in cues for reliable assignment to a "recent" or "distant" time frame, were excluded from this analysis, leaving a corpus of 195 clearly time-framed marked narrative events (out of264). Second, we defined narrative spatial orientation as geared either toward the home, the world of school and/or work, or any other location in the world. For example, a narrative about h o w the substitute teacher b e h a v e d that d a y is considered a "today/school" narrative, a visit to the m u s e u m last w e e k is a "recent-past/world" story, a n d a teenager's early c h i l d h o o d m e m o r y a b o u t a family pet is a "distant-past/ h o m e " story. I n terms of their spatio-temporal framing, family dinner narratives defy the expectations of literary critics. Toolan (1988:1-2) defines narratives as "a recounting of things spatio-temporally distant." This might be true for fictional narratives; but most family narratives are definitely n o t temporally, and only partially spatially, r e m o t e from tellers a n d audience. Considered together for both groups, the majority of temporally m a r k e d narratives (63%) concern events of the very recent past - today, yesterday, or last week - leaving 4 2 % to focus on events from the distant past (n = 195); see "Past 1" a n d "Past 2 " in Figure 5. I n terms of location, half the narratives analyzed fall into the third group, being located in the world. A third c o n c e r n school or work, with the rest (12%) being located in the h o m e s ; see Figure 6. Within this general framework, however, we find cross-cultural preferences. The distribution b y story time in Israeli a n d A m e r i c a n narratives shows that story time is very differentia the two groups (Fig. 5). T h e most striking difference is revealed in regard to the "today" frame: in the A m e r i c a n families, almost half •i
PERCENT 60 50-i
46.2
45.9
38.8
ISRAELI
AMERICAN FRAME
•
TODAY H P A S T l
Figure 5: Temporal tale frames
•
PAST 2
DISCOURSE STUDIES
232
60
PERCENT 47.9
50 40-| 32.5 30 20 10
19.5
1
ISRAELI
AMERICAN TYPE
I HOME B WORK/SCHOOL O WORLD Figure 6: Spatial tale frames the time-marked narratives focus o n today (46%). Telling about the happenings of the day thus stands out as the m o s t important single time frame for American narratives. By contrast, in Israeli narratives, "today" stories take u p only onequarter of narrative space (24%). For Israelis, the recent a n d nonrecent past are the preferred time frames, taking u p over three-quarters of narrative space (76%). T h e difference lies in general preference for time frames, n o t in choice of topics within the frames chosen. For example, for b o t h Americans a n d Israelis, stories of the distant past cover a rich variety of topics. T h e s e include a series of humor ous anecdotes about house painters that the family employed over the years, the story of a n exceptional shopping expedition, several c a m p i n g and travel stories, a n d anecdotes from the children's earlier years. Cultural preferences for the tale's spatial frame are revealed in the choices of h o m e orientation vs. school or work orientation (see Fig. 6). For both groups, about half the narratives are world-oriented, concerning spatial frames such as m u s e u m s , c a m p i n g g r o u n d s , or s h o p p i n g malls. B u t Israeli narratives are markedly m o r e home-oriented t h a n American. I n a considerable proportion of Israeli narratives (20%), the locus of the tale is at the h o m e ; b y contrast, home stories are quite rare (6%) in the A m e r i c a n corpus. T h e Israeli h o m e stories include stories about birthday celebrations, family pets, grandparents visiting, or specific incidents that link the world a n d the h o m e , such as the story of a mother's anxiety o n coming h o m e a n d n o t finding h e r two (early teen) sons there when she expected them. A n interesting corollary to Israeli h o m e orientation is provided b y analysii. of the types of protagonists foregrounded in family narratives (Figure 7). We h a v e divided protagonists as self, other, or us. For b o t h Israelis a n d Americans, attention is divided almost equally between stories that involve the self as prot agonist a n d those concerning others. T h e p r o m i n e n c e of the self as protagonist
BLUM-KULKA
AMERICAN AND ISRAELI NARRATIVE EVENTS AT DINNER
233
PERCENT
ISRAELI
AMERICAN TYPE
I SELF • O T H E R
QUS
Figure 7: Tale protagonists is not surprising, given that family dinners provide a unique opportunity for each m e m b e r to use the narrative m o d e for raising issues of personal concern before a p r e s u m a b l y supportive audience. Stories about others often involve the self too, e.g. the self as critical observer, as w h e n the narrative concerns a teenager recounting a n incident between the teacher a n d another student. The third category, "us stories,'' are quite rare in b o t h groups; b u t they are m o r e likely to a p p e a r in the Israeli narratives than the American. Furthermore, the nine Israeli "us narrative" events are rather long a n d elaborated happenings, lasting u p to seven to eight minutes each, while the two A m e r i c a n "us narrative" events are m u c h shorter (3-4 minutes) a n d less elaborated. We h a v e seen that the groups differ culturally o n the dimensions of spatiotemporality. For the Americans, the process of transformation from tales of today to the actual telling is enacted in a culturally specific, ritualistic w a y - t h a t is un paralleled in the Israeli narrative events. "Today" Rituals: Who will I Tell How My Day Goes? In an imitation of domesticity, a call girl in a n A m e r i c a n m o v i e from 1990 {Pretty Woman) greets h e r customer with " H o w was your day, dear?" But j u d g i n g by our dinner-table conversations, this conventional query functions socially in much richer ways t h a n merely a sign of wifely concern. Narratives about the day's h a p p e n i n g figure in both Israeli a n d A m e r i c a n dinner-table talk, b u t it is only in the A m e r i c a n families that such narrative events take o n the features of a proper interaction ritual (Goffman 1976). "Today" narratives seem to c o m b i n e three ritualistic features: the recurrent nature of the activity type, a n d the role expectations that it entails; the formulaic, repetitive language of the opening phase; and the ritual constraints governing the type of conversational contribu tion expected.
234
DISCOURSE STUDIES
As an activity type, "today" narratives resemble early childhood formats of interaction, which are "standardized . . . interaction patterns between adult and infant that contain demarcated roles and eventually become reversible" (J. Bruner 1983:120). Within the family context, the roles of demarcation cut across insiders and outsiders: thus the observer at the family dinner table has in this case no participation rights, never asks (and is not being asked) about other's or his/her own day. This is not surprising, given that "today" rituals, like early childhood interaction formats, are based on a very high level of shared assum ptions of both a cognitive and an affective nature. Members of the family have basic cognitive scripts about each other's activities during the day, and they act on the assumption that deviations from such scripts (the stuff narratives are made of) are a matter of mutual interest and concern. Theoretically, within the family circle, all members have reciprocal rights and duties to participate; but in practice, reciprocity is limited to spouses, since children either self-initiate participation or are invited b y parents to do so, but as a rule fail to show the same initiative toward their parents. Whereas childhood interaction formats act as a language acquisition support system (J. Bruner 1983), "today" narratives act as a critical socializing context for the acquisition of narrative skills. As the analysis of the texts will show, the ritual can b e performed with varying degrees of success. The processes involved in causes of success or failure provide contexts of socialization in regard both to the choice of acceptable topics and to appropriate ways of telling. The opening phase of "today" rituals is marked by clear discourse boundaries at the point of initiation. The stylistic features of the opening gambit are of a formulaic and repetitive nature, allowing for only a limited degree of lexical and syntactic variation. Consider modes of initiation. Transition from any other topic to the "today" narrative is enacted either b y an other-initiated formulaic question (some vari ation on How was your day?), or by a self-initiated today + action verb phrase ("I had lunch at the Parka today."). (3) American family 6: The children are Andrew (10m), Jessica (8f), and Joshua (3m). This "today" story is the first, to be followed by several others at the same meal. 1 Father: Jessie, how was your day? <2 Joshua: Ooooh aaah 3 Mother: What was the best part of your day, Jessie? «, 4 Jessica: After lunch. 5 Joshua: I get xxx 6 Jessica: xxx 7 Father: You were out playing in the rain. 8 Jessica: Uh-huh. [=! affirmative] 9 Mother: Do you have your templates,Jessie? 10 Jessica: xxx i 11 Father: [=! talking to Andrew] Really? What happened after lunch? You f left for the beach yesterday?
BLUM-KULKA
AMERICAN AND ISRAELI NARRATIVE EVENTS AT DINNER
235
The transition from the previous topic is often minimally marked by the use of a discourse marker (cf. Schiffrin 1987), e.g. so. (4) American family 2: Present are also the children: Daniel (6m), Marvin (8m), and Tamara (3f). 1 Father: So, Tamara, what did you do today? <— 2 Mother: <xx help yourself> [>] i 3 Father: <What'd you do today, Tamara? > [<] End@ . . . (5) American family 7: The children are Aaron (9.5m) and Abigail (7f). 1 Mother: So what did you accomplish today? 2 Father: xxx , 3 Aaron: Uh, uh. 4 Father: Yes 5 Mother: Yes [= laughing] 6 Father: Well* [Story]
«—
1
The salience of the ritual b e c o m e s particularly evident through selfnomination. The right for telling your day is implicidy felt to b e equally shared, as voiced b y 4-year-old Sandra. With no preliminaries, Sandra at some point in the middle of the dinner turns to her mother and asks a question. (6) American family 4: The children are Jordan (7.5m) and Sandra (4f). Sandra's initiation takes place half an hour into the dinner. 1 Sandra: Mommy to who will I tell how my day goes? <— 2 Mother: OK let's hear your day. 3 Sandra: Well # I xxx played puzzles xxx I made xxx [continued] Sandra's question shows that she is already aware of her rights for displaying her day, and moreover that she offers her day as a gift to be received. It is the duty of her family to appoint a receiver for the gift; and indeed, her mother acknowledges the gift, accepting it on behalf of all present [Okay, let's hear your day), thereby giving the signal for the ritual to begin. In other words, the ritual requirement in the case of self-nomination is for recipients to display positive acceptance - not only yielding the floor for the teller, as is the case in all nar ratives, but also paying homage to the specific offering made. In other-initiated narratives, the opening gambit creates a slot that the re cipient is expected to fill with a narrative of the doings and accomplishments of the day. In this way the ritual constraint that operates determines the type of con tribution expected from both initiators and respondents. In ex. 4, the child is probed to conformto this conversational demand (ci. Dascal 1983:109), while in ex. 5 the adult provides the expected narrative. What distinguishes the family "today" ritual from similar rhetorical routines in other types of interaction (e.g., How was your weekend? at the office) is apparently the notion of accountability. In the family, a question like How was your day ?, especially when addressed to children,
DISCOURSE STUDIES
236
implicitly invites a narrative n o less than does an explicit query about the day's happenings. I n other contexts, of course, a narrative m a y b e neither invited n o r m u c h wanted. It is through the notion of accountability that w e can understand h o w somebody's day can b e topicalized, objectified, distanced, and contemplated with care. (7) American family 8: The children arejared (11m) and Robin (9f). The mother's move follows a request from Robin to change the subject discussed previously. 1 Mother: I want to talk to Jared about his day, because he said it was so horrible. <— 2 Jared: It was not horrible, it was just boring. <— 3 Mother: Why was it boringjared? 4 Jared: It was really [//] actually that's not quite true. This person who studies lungs came in for science + \ 5 Mother: Lungs? 6 Jared: Yeah, and she said [//] she showed us some slides and brought in a plastic dog's lung [continued] As these examples show, b o t h children a n d spouses are invited to participate in the ritual. But as noted, nomination rights are n o t quite equally distributed. We encountered only o n e instance w h e r e a child tries to n o m i n a t e a parent, rather than vice versa. (8) American family 7: The children are Aaron (9.5m) and Abigail (7f). Andrew: What happened at work today Mother? Mother: Well I bet you, one power trouble at work today. Andrew: Oh really. [=! laughs] Mother: [=! laughs] Oh God. Abigail: You should drop your jobbie. Mother: I know. Drop your jobbie. End® . This example shows one way in which the ritual m a y fail: neither party (certainly n o t the mother) seems to consider the question as a serious attempt to initiate a "today" ritual. W h e n n o narrative comes forth, another child offers a personal c o m m e n t (T5), which serves to change the topic. In the case of successful "today" narratives, the same slot often is filled b y p r o m p t s to continue, worded in no less formulaic ways t h a n initiations. (9) American family 4: Sandra is 4; also present is Jordan (7.5m). 1 Father: So what else did you today Sandra? 2 Sandra: Um xx beads, puzzles and I played clock [continued] (10) American family 4. 1 Mother: What else did you do today, dear? n
t?_il
r
n,„j' , I
0
11
BLUM-KULKA
AMERICAN AND ISRAELI NARRATIVE EVENTS AT DINNER
237
Such constraints o n initiation a n d participation rights, as well as o n m o d e s of telling, turn "today" narrative events into a clearly delimited speech events that impose specific rights a n d duties o n all participants. For the gift of a today story to b e well received, it must b e a substantial gift. T h e first operating constraint is one of selection. N o t all the day's happenings are worthy of telling: adults explicitly call u p o n n o m i n a t e d tellers to exercise criteria of interest in regard to the tale before launching into the telling; thus 8-year-old Jessica (T3 in ex. 3) is asked to tell the best part of her day. A long account of a soccer g a m e b y an 8-yearold b o y is interrupted as follows. (11) American family 4: Jordan is 7.5. Father: Jordan, would you like to tell us something? Other than soccer, what happened today? Child tellers are required to order the day's happenings b y relevance [Tell us the worst/best part ofyour day; What were the highlights of your day?) prior to fore grounding one particular chain of events as a narrative topic. Children are also explicitly socialized in m o d e s of telling. T h e y are required, in the terms of Genette 1980, to turn stories into true narratives. A n extreme example of this process is illustrated b y the narrative event I refer to as "Everybody's Day." (12) Everybody's day: American family 5: The family has three children: Dorothy (13f), Beatrice (lOf) (also called Harriet), and Matthew (4m). 1 Father: So how was your day Harriet? [>] 2 Bea: [<]. 3 Father: What? 4 Bea: Cut that out. [%com: Bea does not like to be called Harriet] 5 Father: You're supposed to say "Oh it was wonderful xxx" 6 Bea: Daddy [=! annoyed] no lozzie [unclear?] [/] [?][>] [ =! laughs]. 7 Father: <=!laughs>[<] 8 Mother: xx Why don't you tell us about your day now? 9 Matthew: My [//] your day. 10 Bea: Mine? 11 Matthew: xxx. Father: Your first? 12 13 Matthew: And your tenth. 14 Father: O h [ = ! laughs]. 15 Matthew: Bea tell your day. 16 Bea: I woke up and I got dressed and xxx [=! speaks very softly]. 17 Matthew: What? Father: Speak up. 18 19 Bea: # um # I woke up and got dressed and went to xx xx 20 Father: You didn't bother to eat any of breakfast or lunch? 21 Bea: Nope. 22 Father: And I made you such nice french toast too. 23 Matthew: No then <my day!> [>]. 24 Father: [<| 25 Matthew: After Bea comes me [=! laughter]# my day # washed and
238
26 Father: Matthew! 27 Matthew: Oh [=! laughs] 28 Father: You can do any of that # whatever it was. 29 Dorothy: Matthew now # can I go? 30 Matthew: No. First I wake up # then go to bed # then wake up [>] [=! laughing]. 31 Father Oh [=! laughs] [<] 32 Matthew: First I wake up # then + \ Bea Matthew! 33 34 Father: No, stop. 35 Matthew: X X X . «36 Bea: Matthew you had a very £>]. 37 Matthew: [<] Father: First I wake up # then I go to nursery school 38 °/ocom: [=! imitates the way Matthew speaks] 39 Mother: Listen to this family. 40 Mother First I woke up and then had breakfast. 41 Matthew: Then you had nothing [=! laughs]. 42 Mother: Then I had my shower. 43 Matthew Then you had nothing [=! laughs]. 44 Mother; Then I did the wash. I made the lasagne+ . . . «45 Dorothy: Now it's my turn Matthew. Father: First I ate breakfast # then I got up. 46 47 Bea: Daddy [==! annoyed tone]. 48 Mother Oh [=! laughs]. «49 Father: Then I didn't take a shower. Then I brushed my teeth # then I went outside # then I got dressed + \ 50 Bea: Daddy [=! shouting]! 51 Mother: Oh [=! laughs]. 52 Father: Then I took my shoes off and put my pajamas back on # went to the basement # did some work # caught a pigeon # <made the pigeon for lunch> [>] 53 Matthew: [<] 54 Father: [<] [/] then I [/] then I went to # Woolworth's bought something that I had to return later in the day as usual because I always have to return everything I buy. 55 Bea: You can't buy things. 56 Father: urn # then we went to the library. 57 Bea: And found nothing. 58 Father: And I went with a girl who found nothing. Then we went out [>] 59 Bea: [=! shouted]! [Turns into a discussion about whose books are on whose library card] "Everybody's day" displays the typical features of the "today" ritual in its in sistence o n equality in rights of participation (on turn-taking, see particularly T 2 4 , T 2 5 , T 2 9 , T45), o n the repetition of the formulaic questions ( T l , T 8 , T15), a n d o n the type of sequential coherence built u p from "your day" questions
BLUM-KULKA
AMERICAN AND ISRAELI NARRATIVE EVENTS AT DINNER
239
responded to b y action verb series. But it is a ritual g o n e sour, negative rather than positive in affective outcomes. Several indicators in the discourse combine to show us w h a t is going wrong, and thereby also reveal underlying n o r m s for h o w it should have gone. First, there is a double message in the father's initiation: h e uses the formulaic How was your day ?, but prefixes it with an unusual term of address (which his daughter does not like), and follows it with a metacommunicative statement (You 're supposed to say...). H e thus sets a playful tone for the interaction, implying that it is not to b e taken seriously. But the m o t h e r shifts the key back to a serious tone (T8), and the children spend a considerable a m o u n t of metacommunicative energy in line with the general tendency of A m e r i c a n family dinner talk to topicalize turn-taking (Blum-Kulka & Sheffer 1993) - arguing for their share in the family's "today" narrative space. By shifting the a r g u m e n t constantly from the realm of telling, in which the issue at h a n d is a share in the "today" ritual, to the realm of conversation, in which the issue is floor management, the speakers indicate most clearly a dissatisfaction with the unfolding of the event. T h u s , in T 2 3 , M a t t h e w claims his share in the ritual of telling one's day; his utterance is a n attempt to enter this specific telling realm. His father's response, Oh, is it your turn now?, switches back to the realm of conversation, t h e r e b y implicitly inviting talk for talk's sake rather than com pliance with the demanding task (in terms of content) of telling one's day. In T24, the framing of the event shifts from the frame of telling one's days, in which it is meaningful to exchange and evaluate personally owned days, to the loosely defined frame "talk about the day," in which the focus is on having a share in the talk, rather t h a n o n telling o n e ' s o w n day. Subsequently, the realm of the tale is affected as well, yielding extremely p o o r "today" stories in terms of information. T h e first child to speak on topic, Beatrice (T16-22), fails to provide an inter esting narrative; she is interrupted b y h e r younger brother, Matthew (T23), who is dismissed (T33-34) as having defied expectation for m a k i n g a substantial contribution of any kind. While Beatrice still conforms to the rules, though n o t very successfully, M a t t h e w steps out of the rules completely. Deliberately or not, his a c c o u n t actually p a r o d i z e s b o t h t h e r e q u i r e m e n t for informative substance (T25, T31) a n d relevance. His father dismisses h i m b y m o v i n g from the realm of telling to the realm of conversation, negating a conversational turn that fails the "today" requirement (Youjust lost your turn); but Beatrice's evaluation of the tale (Matthew you had a very boring day) imply a critique of the teller as principal (T36). T h e mother's m e t a c o m m e n t in T 3 9 (Listen to thisfamily) suggests an awareness that things h a v e gone w r o n g ; yet her o w n attempt at a change of footing (Goffman 1974:124-60) goes back to the earnest, but n o t good enough, report m o d e earlier initiated b y h e r daughter (compare T16 to T 4 0 and T44). The event culminates with the father giving an echoed, free indirect speech, mocking imitation of Matthew's account (T38), only to go o n (T46ff.) to offer a parodized version of a "today" report of his own. Framing the report as a p a r o d y of the real thing is indicated b y an unlikely reversal of chronology (First I ate breakfast then I got up), followed b y the negation of an event (then T
240
DISCOURSE STUDIES
a shower). Audience response ( T 4 7 , 4 8 , 5 0 , 5 3 ) wavers between clear annoyance, which in the case of the daughter m a y m e a n taking offense (see T47), and laughter, which seems to indicate in this case that b o t h m o t h e r and son align with the father's choice of telling-frame. T h e overall message of this failed narrative event relates to tales, telling, and tellers. I n regard to the tale, it evokes the requirements of "today" tellability: to provide a significant contribution to the ritual, you h a v e to select, order, and dramatize the events of the day. A failure to d o so shifts t h e focus of the narrative event from the realm of telling back to the realm of conversation, a n d changes its key; these transformations result in the disintegration of the "today" ritual. Tellers, encompassing in this event b o t h the role of animator and a n accountable principle (Goffman 1981), are highly vulnerable to such transformations, finding themselves b e i n g challenged b o t h o n account of their tales (e.g. boring day) and their participation in the telling. W h a t is the overall function of the "today" rituals in the A m e r i c a n families? I n terms of discourse goals, they occupy a curious place between transactional a n d interactional speech (Brown & Yule 1983:1-2). T h e role of the ritualistic question How was your X c a n set off different types of speech events. I n case it is completed in a two-move exchange, as might b e the case a m o n g co-workers in an office situation, w e h a v e before us a n interaction ritual (in Goffman's sense) that will tend to serve mainly phatic interactional goals, aimed predominandy at the maintenance of social relations, rather than the transmission of information. But in the family the same question sets off very different kinds of expectations. H e r e the transmission of (narratively filtered) information is n o t only tolerated, but actually required, a n d the interactional goal achievable is n o t just social h a r m o n y b u t rather "affective convergence" (Aston 1988:255). T h e ritualistic m o d e seems to satisfy this variety of expectations; it provides conventionalized ways for the show of reciprocal interest a n d affect in the family, while simul taneously serving as a socializing context for transmitting cultural notions of appropriate ways to transform tales into telling. I n contrast, the Israeli time-framed narratives exhibit only few of the features of the "today" ritual. W e d o find that the activity type is enacted, especially by parents asking their children about the activities of the day. As in the American families, such questions expect a response in the form of a narrative, rather than phatically. Failure to provide a narrative is r e s p o n d e d to b y further probing (see T 4 , 5, 8 in ex. 13). (13) Israeli family 4: The family has two daughters, Ruti (12f) and Naomi (8f), ant* one son, Yaron (4m). 1 Mother: Yaron # tesaper lanu <ma asita ha- Yaron, tell us what you yom ba-gari> [>] [>| 2 Ruti: <ma [//] ex haya ba-gan> [<] [What [//] how was school> )< 3 Yaron: naim Pleasant. 4 Mother: naim? ma haya naim? tesaper lanu Pleasant? What was pleasant? • ex haya naim? Tell us how it was pleasant. *
Wt> n l a v e d .
BLUMKULKA
AMERICAN AND ISRAELI NARRATIVE EVENTS AT DINNER
7 Yaron: be-misxakim. 8 Mother: e&? ba-xuc? ba-xacer? [continued]
241
With games. Which? outside? in the yard? <-
But initiation of such narrative events varies with child a n d family. Transition to a narrative concerned with the day's happenings is initiated either b y a variant of What did you do in school (today) ?, as in ex. 13, or m o r e typically, b y a topically specific question, as in ex. 14. (14) Israeli family 6, with twin girls aged 6: Lilax and Iris. Father: lean halaxtem hayom be-shiur teva? Where did you go today during your "nature" lesson? Lilax: la-giva hazpti she-pa'am she-avra To this hill that we went to last halaxnu time. [continued] T h e time m a r k e r today tends to a p p e a r in stories told about children, rather than b y children in response to "today" elicitations. A n example is a story told mainly b y a mother, with some participation from a child, about the adventure her 4-year-old daughter h a d at school w h e n a rooster j u m p e d o n her; or a story, again told b y a m o t h e r , about h o w 6-year-old R a m i missed the bus to school. The need to m a r k narratives temporally tends to b e reserved b y Israelis for stories from the past. O n e device is the use of pa'am; literally pa 'am means 'once'; and as in English, it figures in conventional story beginnings. E x a m p l e s from the family narratives vary in topic, as seen in ex. 15-17. (15) Israeli family 2: The children are Shlomit (12f), Riki (lOf), and Mika (5f). Father: hayitipa'am. be-shuk aravi ve-ani I visited once an Arab market and I halaxti liknot kishuim vehayu sham went to buy some zucchini and there [continued] were (16) Israeli family 2. Mika: ani roca lesaper bedixa. pa'am axat I want to tell a joke. Once a man halax ish went... [continued] (17) Israeli family 1: Yoash is a guest, a friend of the family. Yoash: shamatipa'am et ex korim la. hag- I once heard, what's her name, Mrs. veret Milo Milo [continued] Other time-marking devices used include concrete specifications of time last week/yesterday/last year/about a month ago), as well as sequential ordering of events relative to self (the first time I went to the University). Talk about the day's activities lacks the ritual constraints o n participation structure: we found n o indication in the Israeli discourse of a n expectation for all to participate, as in the A m e r i c a n families. W h a t seem to b e missing are b o t h the accountability requirement, in regard to the most recent time framp anH
240
DISCOURSE STUDIES
a shower). Audience response ( T 4 7 , 4 8 , 5 0 , 5 3 ) wavers between clear annoyance, which in the case of the daughter m a y m e a n taking offense (see T47), and laughter, which seems to indicate in this case that b o t h m o t h e r and son align with the father's choice of telling-frame. T h e overall message of this failed narrative event relates to tales, telling, and tellers. In regard to the tale, it evokes the requirements of "today" reliability: to provide a significant contribution to the ritual, you h a v e to select, order, and dramatize the events of t h e day. A failure to do so shifts the focus of the narrative event from the realm of telling back to the realm of conversation, a n d changes its key; these transformations result in the disintegration of the "today" ritual. Tellers, encompassing in this event b o t h the role of animator a n d a n accountable principle (Goffman 1981), are highly vulnerable to such transformations, finding themselves being challenged b o t h o n account of their tales (e.g. baring day) and their participation in the telling. W h a t is the overall function of the "today" rituals in the A m e r i c a n families? I n terms of discourse goals, they occupy a curious place between transactional a n d interactional speech (Brown & Yule 1983:1-2). T h e role of the ritualistic question How was your Xcan set off different types of speech events. I n case it is completed in a two-move exchange, as might b e the case a m o n g co-workers in an office situation, w e h a v e before us an interaction ritual (in Goffman's sense) that will tend to serve mainly phatic interactional goals, aimed p r e d o m i n a n d y at the maintenance of social relations, rather than the transmission of information. But in the family the same question sets off very different kinds of expectations. H e r e the transmission of (narratively filtered) information is n o t only tolerated, but actually required, a n d the interactional goal achievable is n o t just social h a r m o n y b u t rather "affective convergence" (Aston 1988:255). T h e ritualistic m o d e seems to satisfy this variety of expectations; it provides conventionalized ways for the show of reciprocal interest a n d affect in the family, while simul taneously serving as a socializing context for transmitting cultural notions of appropriate ways to transform tales into telling. I n contrast, the Israeli time-framed narratives exhibit only few of the features of the "today" ritual. We do find that the activity type is enacted, especially by parents asking their children about the activities of the day. As in the American families, such questions expect a response in the form of a narrative, rather than phatically. Failure to provide a narrative is r e s p o n d e d to b y further probing (see T 4 , 5, 8 in ex. 13). (13) Israeli family 4: The family has two daughters, Ruti (12f) and Naomi (8f), and one son, Yaron (4m). 1 Mother: Yaron # tesaper lam <ma asita ha- Yaron, tell us what you yom ba-gan> [>] [>] 2 Ruti: <ma [//] ex haya ba-gari> [<] [What [//] how was school> [<| 3 Yaron: naim Pleasant. 4 Mother: naim? ma haya naim? tesaper lam Pleasant? What was pleasant? *ex haya naim? Tell us how it was pleasant. 5 Yaron: sixaknu We played.
BLUM-KULKA
AMERICAN AND ISRAELI NARRATIVE EVENTS AT DINNER
7 Yaron: be-misxakim. 8 Mother: eze? ba-xuc? ba-xacer? [continued]
241
With games. Which? outside? in the yard? <-
But initiation of such narrative events varies with child and family. Transition to a narrative concerned with the day's happenings is initiated either b y a variant of What did you do in school [today) ?, as in ex. 13, or m o r e typically, b y a topically specific question, as in ex. 14. (14) Israeli family 6, with twin girls aged 6: Lilax and Iris. Father: lean halaxtem hayom be-shiur teva? Where did you go today during your "nature" lesson? Lilax: la-giva hazoti she-pa'am she-avra To this hill that we went to last halaxnu time. [continued] T h e time m a r k e r today tends to appear in stories told about children, rather than b y children in response to "today" elicitations. A n example is a story told mainly b y a m o t h e r , with some participation from a child, about the adventure her 4-year-old daughter h a d at school w h e n a rooster j u m p e d o n her; or a story, again told b y a mother, about h o w 6-year-old R a m i missed the bus to school. The n e e d to m a r k narratives temporally tends to b e reserved b y Israelis for stories from the past. O n e device is the use oi pa'am; literally pa'ammeans 'once'; and as in English, it figures in conventional story beginnings. Examples from the family narratives vary in topic, as seen in ex. 15-17. (15) Israeli family 2: The children are Shlomit (12f), Riki (lOf), and Mika (5f). Father: hayitipa'am be-shuk aravi ve-ani I visited once an Arab market and I halaxti liknot kishuim vehayu sham went to buy some zucchini and there [continued] were (16) Israeli family 2. Mika: ani roca lesaper bedixa. pa'am axat I want to tell a joke. Once a man halax ish went... [continued] (17) Israeli family 1: Yoash is a guest, a friend of the family. Yoash: shamatipgZam et ex korim la. hag- I once heard, what's her name, Mrs. veret Milo Milo [continued] Other time-marking devices used include concrete specifications of time last week/yesterday/last year/about a month ago), as well as sequential ordering of events relative to self [the first time I went to the University). Talk about the day's activities lacks the ritual constraints on participation structure: we found n o indication in the Israeli discourse of a n expectation for all to participate, as in the A m e r i c a n families. W h a t seem to b e missing are both the accountability requirement, in regard to the most recent time f r a m ° eXDectarinn
fnr-
-> A:.—»---
<•
DISCOURSE STUDIES
242
T h e difference between the two groups in regard to "today" narratives reflects to s o m e extent a wider cultural difference in the w a y verbal elements of social interaction are conventionalized. Relative to Israeli society, A m e r i c a n social interaction seems m u c h m o r e governed b y situationally conventionalized scripts (see Blum-Kulka 1992 for elaboration), which i n turn pale b y comparison with the degree of linguistic conventionalization observed forJ a p a n e s e social behavior (Ide 1987). From this perspective, How was your day, apart from its b r o a d e r cultural and familial functions, is yet one m o r e instance of the rich repertoire of routinized interaction rituals i n the U n i t e d States.
i
Cultural Styles of C o l l a b o r a t i o n Issues of Co-ownership vs. Co-performance
T h e degree of participation in t h e telling raises t h e issue of h o w access to the information contained i n the tale, i.e. the tale-ownership, is related to t h e entitle ment to tell the story (Shuman 1986:137-41). I n other words, whose stories in the family are told b y w h o m and to w h o m ? I n discussing narratives that call for audience response, Sacks (1978:261) notes that it is the "involvement in it [the story] that provides for t h e story's telling." Personal experience grants owner ship, a n d shared experience grants joint ownership: "Parties w h o have experienced a n event together a r e j o i n d y in a position to describe it to someone else" (Goodwin 1981:159). A s noted b y Miller et al. (1990:298), " T h e conditions giving rise to joint ownership are m o r e likely to occur in intimate relations than in non-intimate ones." M e m b e r s of the family, being p a r t of an intimate network, c o m e to the dinner table with knowledge of b o t h shared a n d unshared events. Although the conditions for claiming j o i n t ownership m a y b e met, the fact that several people are in a position to tell a story does n o t result necessarily in joint performance: Mothers regularly tell stories about children in the presence of the children (Miller et al. 1990), claiming sole performance rights o n jointly owned stories. T h e reverse m a y b e true as well; at least during t h e dinner-table con versations studied here, highly involved audience response to a story of personal experience (or knowledge) m a y turn the telling into ajoint performance, implying a process-motivated claim to joint ownership. H e n c e ownership rights through access t o t h e tale have n o one-to-one correspondence to performance rights through access to the telling. As the first step in exploring the interrelations b e t w e e n these two dimensions. I consider each independently, a n d then explore t h e culturally preferred inter sections observed in the family narratives. O n e w a y to approach story ownership is b y considering access to the know ledge of t h e events recounted: is t h e n a r r a t e d event (the fabula) k n o w n to the teller only, o r is it shared b y any other participants? Abstracting the nature of the narrated events from the verbal narratives b y considering textual indicators, we follow Labov & Fanshel (1977:62-4) in distinguishing individually known A-events (known to teller on\v\ from t w n - n a r t v s h a r e d A-R menu fhnn^m in tallo*-
BLUM-KULKA
AMERICAN AND ISRAELI NARRATIVE EVENTS AT DINNER
243
a n d one other participant) a n d generally k n o w n O-events. Specific to the family scene is a fourth type of fabula, F-events, which are shared b y all m e m b e r s of the family. A breakdown of the data using these categories, as presented in Figure 8, suggests that Americans and Israelis only partially share attitudes regarding the degree of novelty expected from narratives a r o u n d the dinner table. I n b o t h groups, A-events dominate (66% a n d 69%). This is not surprising, given that dinner time provides a unique opportunity for all to share personal experiences with intimates. Family dinners with y o u n g children do not seem to occasion stories concerned with the state of the world; O-event narratives are almost absent from this speech event. For Americans, the n e x t most frequent category following A-events is A-B events (30%). For the most part, these recount events experienced jointly b y a child a n d o n e of the parents. For the A m e r i c a n families, there is a marginal 3.5% of F-event stories. O n t h e other h a n d , for the Israelis, F-event narratives prove a viable category, representing 9% of all stories told. It is interesting to n o t e that, for Israelis, the percentage of F-event narratives in child-involved narratives rises to 15%, while in the equivalent set of A m e r i c a n stories it drops to 2%. Consider n o w m o d e s of performance. We can distinguish three major m o d e s of telling: monologic, dialogic, a n d polyphonic. Even though family narrative events are jointly constructed affairs, styles of collaborations vary from low to high participation b y participants other than the p r i m a r y narrator(s). At the dominantly single-voice end of the continuum, w e find monologic narratives, in which one primary narrator remains in control of the floor throughout the event. T h e audience at such events responds indirecdy, sustaining the telling but not involved in the tale. At the multivoiced, polyphonic end, we find narratives that defy the distinction of p r i m a r y vs. s e c o n d a r y narrator(s), being constructed in close 10
PERCENT 80 70
A
60
A
50
A
69.1
66
40 J 30
A
20
A
10
A
|
3.5
0 ISRAELI
0.7
AMERICAN TYPE
•
A EVENTS •
A-B EVENTS •
F EVENTS •
O EVENTS
DISCOURSE STUDIES
244
collaboration between several participants. Between these two ends, we find dialogic narrations, constructed typically through a question/answer format. A multiplicity of voices at the level of telling (in the polyphonic mode) can also transform relations between tellers. I n the terms of Goffman (1974:127), the audience at such events is transformed into "fellow performers" w h o become "inhabitants of the same realm." For T a n n e n (1989:12), it is an issue of involve m e n t : " a n internal, e v e n emotional connection individuals feel which binds t h e m together to other p e o p l e as well as to places, things, activities, ideas, mem ories a n d words. H o w e v e r , . . . [it is] n o t a given b u t a n achievement in conver sational interaction." I n Tannen's analysis of a dinner conversation between friends (1984), she found this achievement of high involvement through conver sational style to characterize the three Eastern-EuropeanJewish speakers present. Monologic narratives t e n d to b e self-initiated, t h o u g h they m a y follow a question from another participant. For example, in the R o b b e r y Story analyzed b y Polanyi (1989:66), the topic is p u t forward b y a participant other than the storyteller, appealing to the appointed participant's expertise o n the events to b e narrated. Turning to the two w o m e n w h o got r o b b e d , their friend asks: "I h e a r d secondhand or whatever that y o u go r o b b e d . - Yeah. - W h a t hap p e n e d ? " N o r d o tellers of monologic narratives necessarily h a v e to claim experi ential warrants for the tale - although, a r o u n d the dinner table, tales do tend to b e stories of personal experience. T h e distinctive feature of monologic narratives is the recognizability of a single narrator's voice. I n the dialogic m o d e , narration proceeds through a question/answer format, whether the story is self- or other-initiated. This is the m o d e that children use to tell stories from the very early stages of language acquisition (Ninio 1988, Sachs 1979). A well-documented subset of such stories is the j o i n t evocation of shared events, in which the caretaker a n d the child jointly recount the story (Heath 1982, S n o w 1991). But whether or not the events to b e recounted are k n o w n to both adult a n d child, elicited narratives with y o u n g children tend to r e m a i n collaborative, with story contributions distributed between adult and child (Snow & Goldfield 1982). Polyphonic narration is enacted in principle t h r o u g h b o t h co-performance a n d co-ownership. Co-performance requires access to the tale, or at least shared access to its social context. This access takes into account a n individual teller's information state (cf. Goffman 1974:133) as to "why events h a v e happened as they have," as well as participant's familiarity with the underlying sociocultural scripts that m a k e narratives "ring true with the stories they k n o w to b e true in their lives" (Fisher 1987:63). Co-ownership m a y b e culturally interpreted as entitiing all co-owners with story-telling rights (but n o t necessarily so; see Miller et al. 1990, S h u m a n 1986). T h e o u t c o m e will b e a j o i n u y constructed narrative, w h e r e division lines between p r i m a r y a n d secondary narrators are blurred. The issue of telling rights in such cases m a y vary culturally in regard to all partici pants, or only in regard to children. If children are included in the circle of J right-holders, a child m a y feel licensed to contribute o n topics felt relevant, I whether the story is initiated b y a n adult or another child. J 11
BLTJM-KULKA
AMERICAN AND ISRAELI NARRATIVE EVENTS AT DINNER
245
Israelis and Americans seem to differ in preferences for these three m o d e s of performance. Quantitative analysis of a subsample of the narratives included here (Blum-Kulka & S n o w 1992) shows that Americans prefer the monologic, display m o d e in 6 0 % of t h e cases, with a preference for the dialogic m o d e in 29%. Israelis, b y contrast, m a k e use of all three styles, preferring b o t h the dialogic (49%) a n d the polyphonic (30%) to the monologic (24%). But the cultural difference in regard to b o t h ownership a n d performance rights is revealed mainly in the diverse ways the groups interpret the relationship between these two dimensions. Excluding dialogic, typically adult/child narration from this analysis, w e can detect four types of interaction between access to the tale a n d participation in the telling. I n the following I argue that the two groups differ in their preferences for these, as s h o w n in Table 1. 12
Table 1: Styles of participation Mode Event-type
Polyphonic
Monologic
Shared Unshared
Israeli/American Israeli
American Israeli/American
A m o n g the four possible configurations between event type a n d m o d e of telling, Israeli a n d A m e r i c a n narratives seem to share two: telling shared ex periences collaboratively in the polyphonic m o d e , a n d allowing predominantly solo performances of personally k n o w n stories. However, the A m e r i c a n families support monologic performances even w h e n tales are k n o w n to m o r e t h a n o n e participant; and Israeli narration m a y unfold in a highly collaborative way, even in unshared tales. T h e distribution a m o n g these four m o d e s shows these preferences as repre senting a gradient p h e n o m e n o n ; see Figure 9. I n b o t h groups, shared experiences told collaboratively a n d single experiences told monologically constitute the bulk of narrative events (77% for Israelis, 84% for Americans). For Israelis, the third choice is to tell unshared events in a polyphonic m o d e (18%); for Americans, it is to tell shared events in the monologic m o d e (12%). I n the following, I illus trate the w a y cultural styles are manifest within each of the two major m o d e s of telling. The Monologic Mode: Telling Shared and Unshared Events A culturally shared a n d highly prevalent m o d e of telling (43% for b o t h groups) is the dominantly monologic performance granted occasionally to sole tellers who recount (mosdy b u t n o t exclusively) personal experience. I n one such case, an Israeli w o m a n describes in great detail the nightmarish d r e a m she h a d about getting lost o n o n e of the campuses of the H e b r e w University in Jerusalem. I n another instance, a n A m e r i c a n teen-age girl recounts a confrontation she h a d with a teacher in school. I n the following examole. thp s f n r v i— ^
PERCENT
ISRAELI
AMERICAN TYPE
•
SHARED & POLYPHONIC
•
UNSHARED & POLYPHONIC
•
SHARED & MONOLOGIC
•
UNSHARED & MONOLOGIC
Figure 9: Ownership vs. performance in telling tales observer terminates a series of stories concerning m e m o r i e s of food from child h o o d , all e m b e d d e d in m e m b e r s ' attitudes towards the religious practices of the parent a n d grandparent generations. , 18) Israeli family 1: The children are Nadav (Urn) andjonathan (10m). The segment follows several exchanges concerning food and eating habits. 1 Observer: aval ha-ax sheli Avner haya But my brother Avner was noda be-bareranuto be-oxel. huwell-known for his pickiness in mamash hayu shlosha dvarim food. There were actually three things he ate and that's it she-hit axal ve-zehu. What? 2 Mother: ma? 3 Observer: chips ve-stekze haya ha-xx.ve- Chips and steak it was xx # ani zoxeret she-ad basar bishvil And I remember that more she-hu yoxal basar notnim lo. meat so he'll eat meat they Sonya hayta notenet lo avatiax would give him. Sonya be-onat ha-avatixim xotexet lo [%com: Sonya is the teller's stepmother] would serve him she-yaani kaxa + \ watermelon in season # would cut it so + \ 4 Mother: ++lo yargish. bis me-ze ve-bis (that) he won't notice; a bite here and a bite there. me-ha-hu. 5 Observer: ve-hshe-hu higia le-cava hu And when he got to the army na'asa gorme haze oxel ve- he became such a gourment, mitanyen bemisadot. hayom eating and interested in shuv ani lo yoda'at hi hu dos restaurants. Now again I don't ' yesh lahem kol mine isurim. know, since he became aval ba-tkufa she-hu od haya observant, they have all these
-
L U M - K U I J K A A M E R I C A N AND ISRAELI NARRATIVE EVENTS AT DINNER
247
xiloni hu na'asa axlan bilti ragil rules. But during the period ve-hu amar "axshav ani mictaer al while he was still nonobservant, kol ha-shanim she-haya oxel kol he turned into a great eater and kax tov [>] ve-ani lo said, "Now I'm sorry for all these years when there was such good neheneti me-klum." food and I, and and I didn't enjoy it at all." Mother: [<] Nadav: ani afpa'am. lo ectaer. I shall never regret it. Observer: lo # kol ze ba lomar lax she-ata lo All this means that you can yaxol lada'at ma yihye be-od +\ never know what will happen in another + \ ©End T h e story of Avner, the non-eating child transformed into a gourmet adultonly to lose access (from the teller's point of view) to g o u r m e t food, through newly acquired religious practices - is exceptionally rich in cultural themes, negotiating transitions between life cycles in the Israeli context (child/soldier/ adult) andjewish life styles (non-observant/observant); see Polanyi 1989 for analysis of cultural themes in A m e r i c a n stories. For the purposes of the present analysis, the m o d e of telling is of interest. T h e example illustrates that, even in the m o n o logic m o d e , w h e n r e c o u n t i n g u n s h a r e d e v e n t s , Israeli a u d i e n c e r e s p o n s e goes b e y o n d brief uptakers (Edmondson & H o u s e 1981). I n T 4 , the evaluative detail a d d e d b y the m o t h e r is h e r o w n logical inference, never m e n t i o n e d b y the m a i n teller. T h e child's c o m m e n t in T 7 indicates h o w carefully h e h a d b e e n listening to the story; in the Israeli families, children typically act as active audience to all stories told, regardless of tellers (adults or children) a n d topic. I n the A m e r i c a n families, child participation in the role of active audience seems more restricted to child-related topics (Blum-Kulka & Snow 1992). I n the A m e r i c a n families, monologic telling is n o t restricted to u n s h a r e d events. I n 12% of the American (non-dialogic) narrative events, a sole performer is granted telling rights for a tale; this h a p p e n s in only 5 % of the parallel cases in the Israeli families. Consider the following personal experience narratives told with the explicit purpose of entertaining. (19) The bug: American family 3. The children are Samuel (10m) Joshua (6m). Talk on hikes precedes this segment. 1 Mother: Daniel loves that. %com: [Daniel is her husband] 2 Samuel: Really? 3 Mother: I on the other hand [//] He loves to commune with nature. Me, I can take nature through a glass window. 4 Samuel: xxx 5 Mother: A glass window # with the bugs out # and anything that's <— more than four feet removing themselves from my presence I can't deal with them [=! laughs]. We had a bug once in the shower # and Samuel called me in on a Friday afternoon # wouldn't you know it # and he screams "Ma!" and there is this thing # if it wasn't two inches big # I thought it was the
248
DISCOURSE STUDIES
6 7
Observer: Mother:
8 Observer: 9 Mother: 10 Observer: 11 Mother:
12 Observer: 13 Mother:
14 15
Father: Mother:
16 Observer: 17 Mother:
most disgusting and somehow as disgusting as they are small # they get worse when they get larger # and there I was # it wasn't even afraid # it wasn't running # it was taking its time # it was sort of taking a little walk across the shower. I would at least appreciate it if they were afraid +/ But no. +"But no, it sat there, very territorially, so I said "Get rid of it," he said "I'm not getting rid of it, you get rid of it," and of course Daniel was nowhere around. Of course, right So we had to shpritz it to death. Of course. Oh, it was horrible, and the darn thing was so resilient it wouldn't die, you know I find that absolutely appalling +\ It wasn't xxx That's right, well by the time we finish with them, they would be + . . . Cockroaches are going to inherit this earth one of these days. Well, you weren't using insecticide though. I was using mildew spray xxx and that bug had the nerve to walk around in it yet # but finally it gave up # the ghost # and then we had to dispose of it So I said "Go ahead # Samuel # dispose of it." He says "I'm not touching it # you touch it." So I took it out with thirty layers of tissue # so that I wouldn't even feel its shape and I picked it up and I threw it into the bathroom toilet. She hasn't gotten over it. I'm still thinking about it. I can see that bug # and I have visions of it rising up yet like a phoenix to haunt me. End®
T h e Bug Story is initiated b y the mother, w h o remains the primary narrator for its duration. T h o u g h b o t h father a n d older son (Samuel) are familiar with the events, audience response is limited to sustaining the act of telling through appropriate backchanneling and clarification comments (T6, T 8 , T10, T12, T16). T h e teller's c o m m e n t o n h e r relations with nature (lean take nature through a glass window, T3) serves as bridging talk to introduce h e r dramatic encounter with a spider in her b a t h r o o m o n a Friday afternoon. She t h e n m a n a g e s to entertain her listeners b y using several evaluative devices to highlight the turning points of h e r narratives (e.g. direct quotes, a shift to the use of the present tense; see T 5 , T7). H e r efforts are fully appreciated - as can b e seen b y audience reactions in T 6 , T 8 , a n d T16. Finally, she concludes b y providing a Labovian coda (Labov & Waletsky 1967:39); visions of the b u g rising like a p h o e n i x clearly link the past to the present. T h e Bug Story shows that co-ownership of the tale does not necessarily entail equal rights for the telling. Narrative events can distinguish tellers from experi encers. I n the Bug Story, the teller is the chief b u t n o t sole experiencer in the
BLUM-KULKA
AMERICAN AND ISRAELI NARRATIVE EVENTS AT DINNER
249
events narrated; yet she is granted full telling rights, foregrounding h e r story telling skills as individual performer. The Polyphonic Mode: Telling Shared and Unshared Events As listeners, w e tend to expect collaboration in the telling w h e n access to the tale is shared b y several of the participants. T h u s stories about a couple's trip abroad m a y b e told to friends jointly, or in a monologic style, b u t with frequent interceptions from the other "knowing" participant. Shared ownership of the tale in the family m a y cut across generations, or create ownership affiliations between any g r o u p of m e m b e r s , either b y true experience or b y claimed famil iarity with the events. Events that h a p p e n to parents jointly, prior to their chil dren's birth, are o n e source for generationally defined ownership. I n o n e case, the story of the parents' courtship is told in collaboration b y both parents in response to a 10-year-old's question: ex at ve-aba higatem le-ahava me-aruxat-erev pshuta? ' H o w did you and D a d c o m e to b e in love from a simple dinner?' T h e story of the parents' courtship is an e x a m p l e of the type of narrative which m a y b e told again and again, because of its relevance to the history of the family. Memories from a shared past carry the potential to b e c o m e p a r t of the familial narrative repertoire, their telling triggered b y the presence of a new audience. O u r presence at the dinner table might h a v e occasioned the retelling or construction of several such, family fables. Consider the potential of the following recollection of a family reunion as a candidate of being or becoming a family fable. (20) American family 7: Children are Aaron (9.5m) and Abigail (7f). 1 Father: Yeah # anyway we had a big family reunion. 2 Abigail: What's a family reunion? [=softly] 3 Mother: That was all the members of the family [//] all relations. 4 Father: Well # you can't be more specific. 5 Mother: Not all. 6 Abigail: All the members of his side of the family. 7 Father: Yes # it was all of my father's [//] my father and both of his sisters and all their children # and all the grandchildren. 8 Aaron: We were one of the grandchildren # right? 9 Father: Right. 10 Mother: Um hm. 11 Father: So all these people were at this big reunion. And there wasn't enough room to sleep inside the house. So a lot of people bought # had or rented. 12 Mother: Rented. Nobody had # everybody rented. 13 Father: Yeah? Everybody rented RVs, which are these motor homes which you sleep in. And are not real comfortable. 14 Mother: And they're not real private. 15 Father: xx didn't seem to mind. 16 Mother: Well you just [//] you were so shy Abigail # you you wouldn't talk to anybody. And you just stayed inside the
4—
«_
<—
250
DISCOURSE STUDIES
17 Aaron: 18 Mother: 19 Aaron: 20 Abigail: 21 Aaron: 22 Mother:
»
23 Father: 24 Abigail: 25 Father: 26 Abigail: 27 Father:
RV and you never wanted to come out # see everybody. You just wanted to stay inside and read books and play with your toys. She could read then? Well she had some books that she looked at the pictures # you know. Yeah. Picture book? Picture book? Pretended to read # pretended she was grown up. Uh huh. And when anybody tried to talk to you # you'd run away and hide in the RV. Except for Pearson. Yeah in the RV I would go. You visited xxx P. He picked you up, xxx feet and sat you on his back xxx started asking you about the pictures. Huh You succumbed to his charms, xxx have [=! laughter] End®
«-
T h e polyphonic m o d e of telling is particularly suitable for stories that have the family "us" as protagonist, a n d which are relevant in terms of contribution to the building of familial a n d individual identities. T h e Family R e u n i o n story is n o t about self or other, b u t rather reflexively about "us," the family, a n d the w a y we were then. Its relevance stems from stressing family continuity: it seems to b e saying " H e r e is something w e did together, even if y o u children do not r e m e m b e r it all." Revealing several sides of Abigail's younger personality, the story is particularly relevant for the child protagonist: it provides h e r with a sense of the continuity of the self, c o m b i n e d with the opportunity to reflect on h e r younger self as c o m p a r e d with h e r current self. T h e suggested key for re flection is a h u m o r o u s o n e ; the child is invited to join the others in laughing at her o w n shyness then, a n d the attraction that broke it. T h e text contains indicators that the story has b e e n told previously. Abigail's first contribution to the narrative (all the members of his side of the family, T6) suggests familiarity with the tale, or at least parts of it. A later r e m a r k (yeah in the RV I would go, T24) can b e taken either as h e r recollection of the events, or as h e r acceptance of h e r parents' ver sion. T h e story is thus o n its w a y to b e c o m i n g a family fable. Telling is carried b y b o t h parents, with a shift in appointed audiences. The father seems to b e talking first to t h e Observer; b u t the m o t h e r deliberately shifts perspective, appointing Abigail, t h e h e r o i n e of t h e story, as primary audience (you were so shy Abigail, T16). Aaron, the older brother, takes the role of the challenger, doubting his m o t h e r ' s version (She could read then?, T17) - and once reassured, insisting o n his sister's limitations ([She] pretended to read, T21). T h e mother aligns with h e r son b y picking u p the 3 r d person reference to Abigail (Well she had some books, T18); then, treating the exchange with A a r o n as a "side sequence" (Jefferson 1972), she shifts back to addressing Abigail directly. Thus, although in different narrative roles, eventually all family m e m b e r s collabor ate in constructing a narrative that m a y well b e c o m e part of the family's fable repertoire. ^
BLUM-KULKA
AMERICAN AND ISRAELI NARRATIVE EVENTS AT DINNER
251
I n the Israeli narratives, b y contrast, sharing the tale is n o prerequisite for participation in the telling. Even w h e n the specific events are strictly of the A-event, "only teller knows" type, in the family they are told against the back g r o u n d of shared sociocultural scripts which in turn p r o v i d e occasions for nonexperiencer participation. Israeli dinner-table participants seem to take full advantage of their familiarity with such scripts. I n 18% of the (non-dialogic) Israeli narrative events, singular tales get a polyphonic telling. I n the A m e r i c a n families, w e detected only three such examples (4%). I n the following narrative, the fact that the story concerns a n A-event does n o t p r e v e n t other m e m b e r s from taking a n active p a r t in the construction of the story. (21) Saving a watermelon: Israeli family 4. The family has two daughters (Ruti, 11, and Naomi, 8) and one son (Yaron, 4). The story follows an account of a car accident the mother was involved in the same day. 1
6
Father: ani etmol [/] ani etmol hicalti avatiax. Observer: Oh [=! laughs] ' Naomi: ex [/] ex? Ruti: ex hu hicil? Father: atem to ta'aminu. ani nasati li # hayiti ba-boker # zf haya etmol # ken # hayiti hare etmol ba-boker ba-bank < lifhei hacaharayim. > l>] Mother: £>]
7 8
Father: Observer:
9
Father:
2 3 4 5
(
10 Observer: 11 Father: 12 Mother: 13 Father:
14 Observer: 15 Father:
I
16
I yesterday [/] I saved a water melon yesterday.
Ruti:
How? How? How did he save it? You won't believe it. I was driving, in the morning, it happened yesterday, yes, I was at the bank yesterday before noon.
Watermelon by the knife, [an expression used by watermelon vendors] lo zfh alha-sakin. mazal. No, it's not by the knife. Lucky. ani mekava she-lo <sikanta et [=! laughs] I hope you didn't acmexa> [>] endanger yourself [=! laughs] [<]me-ha-super And I'm driving from the be-giva hacarfatit # ma at supermarket on French Hill # what do you expect? mekava? That you didn't risk your life she-lo sikanta et xayexa + \ almost. kimat. oy va-avoy li. [=! exclamation] be-super lemata le-kivun ha-ramzorNear the Super in the direction of ve-pitom ani roe holexetsham isha the traffic lights, and suddenly I im ezoyalda ve-pitom ve-ezp sakit see a woman walking with a child matxila lehitgalgelba-moradft ve- (female) and suddenly a bag starts he-yalda roca laruc [>] ve-ha-ima maxzika child wants to run there and is ota. yelling and her mother is holding her.
17
Father:
18
Ruti:
19
Mother:
raiti [/] raiti she-lo keday la'acor I saw that it's not worth stopping et ze im ha-oto, maker acarti etha- it with the car, so I stopped the oto ve-racti ve-hiclaxti litfos etha- car quickly and ran, and managed avatiax she-hitgalgel be-tox sakit to catch the watermelon that was # ve-lo kara lo shum davar # bari rolling in the bag, and it came to ve-shalem hexzarti oto le-zro'otha-no harm, I returned it safe and sound into the arms of the little yalda. girl. +> . The sobbing child.
[>] 20 21 22 23 24 25 26 27 28 29 30 31 32
33
34 35
[>] Mother: <ma at medaberetxx> [<>] What are you saying, things like this . . . dvarim kaele + ... Father: [<]. You have in the States. hem meod adivim ba-dvarim ha- They are very polite in these Ruti: ele. things. Father: aval avatiax, im ha-yalda hayta But a watermelon, if the child have chased the Observer: raca la-avatiax az haya yaxol would watermelon things might have lihyot nora mesukan. become very dangerous. Father: lo. hayta sham beaya. zot omeret No. There was a problem. I mean ha-ba'aya shel ha-isha hayta o ha- the woman's problem was either the watermelon or the child. avatiax o ha-yalda. Mother: ve-hi hexlita ha-yalda [>]. Father: [<] # aval She decided (for) the child, but the child decided (for) the water ha-yalda hexlita avatiax. melon. END®
This story is offered as a h u m o r o u s counterpoint to the preceding narrative b y the mother, recounting her near escape from a car accident. T h e father embeds the u p c o m i n g story in the ongoing conversation b y repeating the verb saw
BLUM-KULKA
AMERICAN AND ISRAELI NARRATIVE EVENTS AT DINNER
253
(in the previous story, savedfrom the incident} in the n e w a n d u n e x p e c t e d context of saving a watermelon. Unlike the Bug Story, the watermelon incident is b a s e d on events k n o w n to the teller only. Yet, in the Israeli family, participants take an active part in the performance from the onset. T h e high level of involvement (Tannen 1984, 1989) can b e glimpsed b y just scanning the n a m e s of the par ticipants making comments during the event (wife, Observer, both older children) and b y noticing the high proportion of overlapped talk. T h e nature of audience participation changes with different phases of the narrative event. T h e event is composed of three phases: the opening ( T l - 1 3 ) , which provides the abstract a n d the general setting (time a n d place); the m a i n b o d y of narrative (T14-18); a n d the discussion of its point (T19-35); cf. Polanyi 1989. I n the o p e n i n g phase ( T l - 1 3 ) , audience response takes the form of w h a t Tannen (1984:118) calls "cooperative prompting": b o t h children display interest (T3-T4), the wife debates jokingly the kind of watermelon in question (T6), a n d the observer expresses e m p a t h y for the teller experiencer (T8). T h e teller presents the m a i n events of the story - including setting, com plication, a n d resolution in Labovian terms - in turns T13, T15, T17. Audience response at this stage is diminished, recipients confining themselves mainly to backchanneling. But o n c e the h a p p y e n d i n g b e c o m e s evident, the a u d i e n c e takes o n a highly active part b o t h in embellishing the story and in debating its point. In T18, Ruti uses a phrase b o r r o w e d from the language of written H e b r e w fairy tales ('the sobbing child'), which serves to dramatize the scene of the story's denoue ment. T h e child's sobbing is implied in an earlier statement b y the teller ('she is yelling', T13), b u t is never m e n t i o n e d again. This evaluation of the tale is offered because it coheres with the rest, not because the speaker has first-hand or vicarious proof for it being "true." T h e point of the narrative emerges in collaboration a m o n g several par ticipants. I n response to his wife's clarification question (T21), the teller reformu lates the point of the story as concerning 'the saving of a family' rather t h a n •saving a watermelon' (T22). I n the side-sequence that follows, wife a n d h u s b a n d disagree about the plausibility of the event taking place anywhere but in Israel. The mother's attempt to explain the narrative's coda, in terms of cross-cultural variability as regards n o r m s of politeness (T26, T28), is rejected b y b o t h father and daughter (T29, T 3 0 , T31). Note that the girl shows her alignment with h e r father's position b y cooperatively overlapping talk (Tannen 1984:118) that actually completes the father's utterance (T30). In T 3 2 - T 3 5 , t h e three adults return to the issue of the point of the story. By rephrasing the "true" nature of the complication as a case of real danger (T32), the Observer reinforces the transformation of the narrative's point, from b e i n g an entertaining anecdote about "saving a watermelon" to being a serious story about saving lives. T h e discussion foregrounds the status of the narrative as a moral construct (cf. Fisher 1987), giving it m e a n i n g through negotiating the nature of the m o r a l d i l e m m a at h a n d . H u s b a n d a n d wife collaborate in under scoring this n e w angle, b y shifting perspectives (T34-35) to the viewpoint of the
DISCOURSE STUDIES
254
w o m a n with the child, for w h o m the p r o b l e m was o n e of protecting h e r child rather t h a n the watermelon. T h i s e x a m p l e shows that Israeli participants d o n o t feel restrained b y non-access to actual experience to claim authorship for the story. T h o u g h the fabula of the watermelon is derived from the personal experience of the teller, the other participants in the event take a highly active part in the story's construction, especially in negotiating its macrolevel point. T h e process culmin ates in a j o i n t agreement as to the point of the story; this agreement, as well as the high degree of participation throughout, reveal a preference for a multivoiced m o d e of performance in this group. But collaboration goes b e y o n d performance: through the process of b e c o m i n g fellow performers in the telling, participants claim joint ownership of the tale. I n other words, in monologic u n s h a r e d event narratives, ownership rights are reconfirmed through the telling; b u t in the pro cess of a joint performance of an (initially) unshared event, ownership rights may b e generated performatively through the v e r y act of participation in the telling. T h e s e examples show that the two groups differ in attitudes toward the relationship b e t w e e n ownership a n d p e r f o r m a n c e . Israeli families reveal a flexibility in regard to b o t h dimensions; they strive toward joint ownership for all narratives. This trend is manifested b y the high degree of participation in the telling of b o t h personal narratives a n d shared family events. By partaking in the construction of the story, m e m b e r s of the Israeli family use narrative author ship (Shuman 1986:174-8), to m a k e claims for j o i n t narrative ownership. I n the Jewish-American families, however, participation in the telling seems preconditioned b y joint access to the tale. T h u s the recollection of shared family m e m o r i e s (the Family Reunion story) is accomplished with the help of several family m e m b e r s . But, in contrast with the Israelis, story authorship is highly valued. H e n c e w e n o t e a trend to allow for the display of individual story-telling skills e v e n in cases w h e r e the tale is k n o w n to m o r e than o n e participant (the Bug Story). For these families, authorship through performance is used to assert a n d / o r achieve individual ownership. •I
i
i
Conclusions I h a v e g r o u n d e d m y analysis of family narrative events in the threefold frame work of telling, tales, a n d tellers. T h e distinction between the first two dimensions is certainly not m y discovery. U n d e r different guises, it concerns students of b o t h narratology (e.g. Genette 1980, R i m m o n - K e e n a n 1983) a n d folklore (e.g. B a u m a n 1986, Briggs 1988, Young 1987). M y inclusion of tellers within the same paradigm is m e a n t to emphasize the social constitutive nature of oral story telling - a n d within it, the role of individual selves (e.g. in the presentation of self as protagonist a n d / o r performer) in relation to the other two realms. This m o d e l permits us to isolate the shared a n d unshared properties of JewishA m e r i c a n a n d Israeli family narrative events. W e have found similar patterns with respect to multiple participation, the prevalence of personal experience stories, a n d a respect for children's story-telling rights. These patterns probably
LUM-KULKA
AMERICAN AND ISRAELI NARRATIVE EVENTS AT DINNER
255
derive as m u c h from a c o m m o n Eastern-European oral story-telling tradition as from narrative practices prevalent in middle-class families in the Western world. As documented b y Kirshenblatt-Gimblett (1974:283), oral story-telling has b e e n a cultural focus in J e w i s h society from time immemorial; it was frequent a n d i m p o r t a n t in traditional E a s t e r n - E u r o p e a n c o m m u n i t i e s , it w a s egalitarian (everybody could tell), and was not limited to specific speech events. By contrast, features like the dialogic nature of m a n y oral narratives (Polanyi 1989) or the deliberate involvement of children in story-telling (McCabe & Peterson 1991) characterize W h i t e N o r t h A m e r i c a n mainstream practices. However, the two groups also differ in m a n y respects. T h e differences show u p b o t h in the ways of construing each of the three narrativity dimensions inde pendently, a n d in the interplay between them. Access to telling is less available to Israeli than A m e r i c a n children, but is m o r e available to Israeli than American observers. Spatio-temporal framing of tales in the A m e r i c a n families locates stories outside the h o m e , b u t close in time; Israeli families favor stories m o r e distant in time, b u t located in the h o m e . While most narratives in b o t h groups recount individually k n o w n events with self as protagonist, Israeli families are more likely than Americans to recount family-shared events that center a r o u n d the family ("us") as protagonist. In the transitions between the realm of conversation a n d the realm of telling, American families search for clear demarcation lines, occasionally ritualized, as in the "today" stories. T h e high-involvement style of Israeli story entry tends to blur the boundaries between the two realms. Israeli narrative events often begin and e n d in a highly polyphonic m o d e ; while the opening phase estab lishes shared access to the tale, the closing phase focuses o n negotiating a shared interpretation of the story's meaning, seamlessly m o v i n g back to the realm of conversation. T h e relations of the tellers to the tales a n d / o r telling is also perceived in cul turally distinct ways. Americans support tellers b y attending to individual telling rights, but Israelis t e n d to support tellers b y attending to the tale. Access to story ownership in the A m e r i c a n families is asserted through familiarity with the tale, but in the Israeli families it is also achievable through participation in the telling. As a result, monologic m o d e s of telling in the A m e r i c a n families extend to shared events, while telling of u n s h a r e d events is celebrated b y Israelis in the polyphonic m o d e . T h e proposition unique to the Americans seems to b e "Let me tell our story"; for Israelis, it is "Let us (all) tell your (singular) story." Yet all these families share a n Eastern-European background. As such, all could b e expected to manifest in narratives the high-involvement conversational style found b y Tannen 1984 to typify J e w i s h N e w Yorkers from the same back ground. Spolsky & Walters (1985:64) argue that this high-involvement style has an analogy in Eastern-European learning styles of the Yeshiva, w h e r e par ticipatory listenership a n d rapid turn-shifting " m a r k the discussion of equals." Another analogy is provided b y styles of worship in the synagogue: the EasternEuropean t e m p l e is seemingly chaotic, c o m p a r e d to the Asian or Western European, yet this "chaos" is governed b y underlying ideological principles, which allow for a higher involvement o n the part of the individual worshiper.
DISCOURSE STUDIES
256
Ideology is also a strong motivating factor in understanding Israeli con versational style. As elaborated b y Katriel 1986, the Israeli style emerged against the background of a strong ideological opposition to all things associated with Eastern-European diaspora traditions, including ways of speaking. O n m a n y dimensions of language use, the Israeli style has i n d e e d b e e n shown to break away from traditional m o d e s (e.g. Blum-Kulka 1992, Katriel 1986). But simul taneously it seems to show strong traces of cultural continuity. We first noted this trend in the prolific use of affective nicknames a n d endearments, which echo in their suffixes Yiddish a n d Slavic sound patterns (Blum-Kulka & Katriel 1991). T h e study of family narrative events provides a second example. Com p a r e d to the Jewish-Americans, the Israelis are m o r e involved in all aspects of narrative discourse, sharing b o t h tellings a n d tales. From this comparative perspective, the impact of Eastern-European traditions is less noticeable in the narrative discourse of Jewish-American families. I n their case, it is A m e r i c a n culture which seems to h a v e played the formative role. Thus the emphasis on individual rights a n d self-accomplishment, seen in attitudes toward the telling, could b e expected from sociological accounts of American society (e.g. Bellah et al. 1985). T h e ways that Jewish-American narratives differ from Israeli narratives echo A m e r i c a n ways of speaking, although they may well differ in style from narratives in other A m e r i c a n speech communities.
Notes An earlier version of this article was presented at the Boston University Child Language Conference, October 1989. The research reported was funded by Grant no. 87-00167/1 from the Israel American Binational Science Foundation (BSF). I am indebted to Leslie Polss, Naomi Mazoz, Talya Miron-Shatz, and Sigal Ravina, who participated in coding and analyzing the data. Special thanks are due to the par ticipants at the 1990 summer Dubrovnik Inter-University Graduate Seminar on CrossCultural Pragmatics - especially Guy Aston,Juliane House, Gabi Kasper, Tamar Katriel, Susan Paulston, Catherine Snow, and Elda Weizman - for many insightful comments on the family narratives. I also thank Deborah Tannen, who graciously identified herself as the reviewer for the manuscript. Her comments were extremely helpful. 1. Other approaches to narrative stress the socioculturally constitutive role of narrative practices (Bauman 1986, J. Bruner 1986, E. Bruner & Gorfain 1984). As formulated by Bauman (p. 113), "narrative... is not merely a reflection of culture, or the external character of social institutions, or the cognitive arena for sorting out the logic of cultural codes, but is constitutive of social life." 2. I borrow the term from Bauman 1986, who follows RomanJakobson in distinguishing the narrative context of the situation, namely the narrative event, from the storyworld evoked through the telling, namely the events narrated. 3. See Young 1987, especially chapters 1-2, for a philosophically attuned discussion of the phenomenological framework for the analysis of the different narrative realms involved (in her terms story worlds and takwords). 4. The families were selected by the snowball technique: all adults had to be collegeeducated, professionally employed, and native-born Israelis or Americans. The full
BLUM-KULKA
5.
6. 7.
8.
9.
10.
11.
12.
AMERICAN AND ISRAELI NARRATIVE EVENTS AT DINNER
257
sample includes three groups: native-born Israelis, native-born Jewish-Americans, and native-born Jewish-American immigrants to Israel. Three dinner conversations were recorded (twice by audio and once by video) for each family, within a period of two to three months; all families (including children) participated in an ethnographic interview following the recordings. For further details on the project see Blum-Kulka 1990, Blum-Kulka & Katriel 1991, Blum-Kulka & Sheffer 1993, Blum-Kulka & Snow 1992, Olshtain & Blum-Kulka 1989. The common practice in studies of children's narrative development is to insist on two consequent and causally or temporally related events (Labov & Waletzky 1967) as a minimal requirement for a segment of text to be considered a narrative. On one occasion, we used this definition to allow for comparability across data sets collected under different conditions (Blum-Kulka & Snow 1992). But we found the definition inadequate for capturing the richness of conversational narratives at dinner. From a young child's perspective, recounting a single past event during one short speakingturn (not an "extended turn," as in conversation-analytical accounts of story-telling) may very well count as "telling a story." Operational definitions of what constitutes a narrative need to accommodate the type of discourse in which the narrative is em bedded, as well as variation in participants' emic perspectives. Goffman (1981:131-2) distinguishes official, ratified participants vs. unofficial (unintentional) eavesdroppers and (intentional) over hearers. I use "reportabilicy" here in the sense of Hymes (1981:82): the knowledge that com petent members of a culture or community have as to what behavior is reportable in that community. An interesting issue, beyond the scope of this article, is the gender differences between the parents, noted at the family dinner table, as related to possible gender differences in speech in the two respective societies at large. Transcription follows the CHILDES system (MacWhinney 1991:122-5) as follows: < > overlap; [>] overlap follows; [<] overlap precedes; # short pause; [/] retracing without correction; [//] retracing with correction; + \ interrupted utterance; + . . . trailing off; + quick uptake; +, self-completion; ++ other-completion; [=! text] paralinguistic material; [%com] contextual information. Punctuation marks are used to mark utterance terminators. Some deviations from CHILDES were introduced: the use of capitals and quotation marks for reported speech, to ease reading; and the segmentation of the text by turns (the relevant units here), rather than utterances. Par ticipants are identified by role (for adults) and by name (for children). Age and sex of child are given in parentheses, in that order: Andrew (8m) = Andrew, aged 8 years, male. Conversational features (e.g. interruptions and overlaps) are marked ap proximately on English translations from Hebrew. Leftward arrows to the right of the transcript point to the passages being discussed. Examples of textual indicators for event type would be: A fanny thing happened to me today (A-event); Remember the time I walked you to school? (A-B event); Remember our last camping? (F-event); / was surprised by the results ofthe election (O-event). As Shuman (1986:31) notes: "One must have information in order to talk about something. However, people with the information are not necessarily entitled to tell what they know." Included in this analysis were all narrative events from one meal of 10 American (five middle-class and five working-class) families and five Israeli middle-class families (•-73).
DISCOURSE STUDIES
258
References Aston, Guy (1988). Learning comity: An approach to description and pedagogy ofinteractional speech. Bologna: Cooperativa Libraria Universitaria Editrice Bologna, Bauman, Richard (1986). Story, performance and event. Cambridge & New York: Cambridge University Press. Bellah, Robert N.; Madsen, Richard; Sullivan, William M.; Swidler, Ann; & Tipton, Steven (1985). Habits of the heart. New York: Harper & Row. Bernstein, Basil (1971). A socio-linguistic approach to socialization, with some reference to educability. In Dell Hymes & John Gumperz (eds.), Directions in sociolinguistics. New York: Holt, Reinhart & Winston. 465-98. Blum-Kulka, Shoshana (1990). You don't touch lettuce with your fingers: Parental politeness in family discourse. Journal ofPragmatics 14:259-88. (1992). The metapragmatics of politeness in Israeli society. In R. J. Watts et al. (eds.), Politeness in language: Studies in its history, theory and practice. Berlin: Mouton de Gruyter. 255-81. ; Danet, Brenda; & Gerson, Rimona (1985). The language of requesting in Israeli society. In Joseph Forgas (ed.), Language and social situations. New York: Springer. 113-41. , & Katriel, Tamar (1991). Nicknaming practices in families: A cross-cultural per spective. In Stella Ting-Toomey & Felipe Korzenny (eds.), Cross-cultural interpersonal communication. Newbury Park, CA: Sage. 58-78. . & Snow, Catherine (1992). Developing autonomy for tellers, tale and telling in family narrative-events. Narrative and Life History 2:187-217. , & Sheffer, Hadas (1993). The metapragmatic discourse of American-Israeli families at dinner. In Gabriele Kasper & Shoshana Blum-Kulka (eds.), Interlanguagepragmatics. Oxford & New York: Oxford University Press. 196-223. Briggs, Charles (1988). Competence in performance: The creativity of tradition in Mexicano verbal art. Philadelphia: University of Pennsylvania Press. Brown, Gillian, & Yule, George (1983). Discourse analysis. Cambridge & New York: Cambridge University Press. Bruner, Edward, & Gorfain, Phyllis (1984). Dialogic narration and the paradoxes of Masada. In E. Bruner (ed.), Text, play and story: The construction and reconstruction of self and society. Washington, DC: American Ethnological Society. 56-80. Bruner, Jerome (1983). Child's talk: Learning to use language. New York: Norton. (1986). Actual minds, possible words. Cambridge, MA: Harvard University Press. (1990). Acts of meaning. Cambridge, MA: Harvard University Press. Bublitz, Wolfram (1988). Supportivefellow-speakers and cooperative conversations. Amsterdam: Benjamins. Dascal, Marcelo (1983). Pragmatics and the philosophy ofmind, 1. Amsterdam: Benjamins. Edmondson, Willis, & House, Juliane (1981). Let's talk and talk about it: A pedagogic interactional grammar ofEnglish. Munich: Urban & Schwarzenberg. Erickson, Frederick (1982). Money tree, lasagna bush, salt and pepper: Social construction of topical cohesion in a conversation among Italian-Americans. In Deborah Tannen (ed.), Analyzing discourse: Text and talk. Washington, DC: Georgetown University Press. 43-71. —— (1988). Discourse coherence, participation structure, and personal display in famih dinner table conversation. Working Papers in Educational Linguistics 4:1-26.
BLUM-KULKA
AMERICAN AND ISRAELI NARRATIVE EVENTS AT DINNER
259
Fisher, Walter (1987). Human communication as narration. Columbia: University of South Carolina. Genette, Gerard (1980). Narrative discourse. Ithaca, NY: Cornell University Press. (French original published 1972.) Goffman, Erving (1974). Frame analysis. Harmondsworth, England: Penguin. (1976). Interaction ritual: Essays on face to face behavior. Garden City, NY: Anchor. (1981). Forms oftalk. Philadelphia: University of Pennsylvania Press. Goodwin, Charles (1981). Conversational organization: Interaction between speakers and hearers. New York: Academic. Heath, Shirley Brice (1982). What no bedtime story means: Narrative skills at home and at school. Language in Society 11:49-76. (1983). Ways with words: Language, life and work in communities and classrooms. Cambridge & New York: Cambridge University Press. Hymes, Dell (1974). Foundations in sociolinguistics: An ethnographic approach. Philadelphia: University of Pennsylvania Press: (1981). "In vain I tried to tell you": Essays in Native American ethnopoetics. Philadelphia: University of Pennsylvania Press. Ide, Sachiko (1987). Strategies of "discernment" and "volition" for linguistic politeness. Paper delivered at the International Pragmatics Conference, Antwerp, August. Irvine, Judith (1979). Formality and informality in communicative events. American Anthropologist 81:773-91. Jefferson, Gail (1972). Side sequences. In David Sudnow (ed.), Studies in social interaction. New York: Free Press. 294-338. (1978). Sequential aspects of storytelling in conversation. In Jim Schenkein (ed.), Studies in the organization ofconversational interaction. New York: Academic. 219-48. Katriel, Tamar (1986). Talking straight: Dugri speech in Israeli Sabra culture. Cambridge & New York: Cambridge University Press. Kirshenblatt-Gimblett, Barbara (1974). The concept and varieties of narrative per formance in East-EuropeanJewish culture. In Richard Bauman &Joel Sherzer (eds.), Explorations in the ethnography of speaking. Cambridge & New York: Cambridge University Press. 263-83. (1975). A parable in context: A social-interactional analysis of story-telling per formance. In Dan Ben-Amos & Kenneth S. Goldstein (eds.), Folklore: Performance and communication. The Hague: Mouton. 106-30. Labov, William, & Fanshel, David (1977). Therapeutic discourse: Psychotherapy in conversation. New York: Academic. , & Waletzky, Joshua (1967). Narrative analysis. In June Helm (ed.), Essays on the verbal and visual arts: Proceedings ofthe 1966Meeting ofthe American Ethnological Society. Seatde: American Ethnological Society. 12-44. MacWhinney, Brian (1991). The CHILDES Project: Tools for analyzing talk. Hillsdale, NJ: Erlbaum. McCabe, Allyssa, & Peterson, Carole (1991). Getting the story: A longitudinal study of parental styles of eliciting narratives and developing narrative skills. In A. McCabe & C. Peterson (eds.), Developing narrative structure. Hillsdale, NJ: Erlbaum. 217-53. Miller, Peggy; Potts, R.; Fung, H.; Hoogstra, L.; MintzJ. (1990). Narrative practices and the social construction of self in childhood. American Ethnologist 17:292-309. Ninio, Anat (1988). The roots of narrative: Discussing recent events with very young children. Language Science 10:35-52. Ochs, Elinor; Smith, Ruth; & Taylor, Carolyn (1989). Detective stories at dinner-time: Problem solving through co-narration. Cultural Dynamics2:238-57.
2fiO
DISCOURSE STUDIES
Ochs, Elinor; Taylor, Carolyn; Rudolph, Dina; & Smith, Ruth (1992). Storytelling as a theory-building activity. Discourse Processes 15:37-72. Olshtain, Elite, & Blum-Kulka, Shoshana (1989). Happy Hebrish: Mixing and switching in American-Israeli family interactions. In Susan Gass et al. (eds.), Variation in second language acquisition: Discourse andpragmatics. 59-83. Clevedon, England: Multilingual Matters. Polanyi, Livia (1989). Telling the American story. Cambridge, MA: MIT Press. Polss, Leslie (1990). Starting the story: A cross-cultural analysis of conversational story entrance talk. M.A. thesis, Hebrew University, Jerusalem. Rimmon-Keenan, Shlomit (1983). Narrativefiction: Contemporary poetics. London: Methuen. Sachs, Jackie (1979). Topic selection in parent-child discourse. Discourse Processes 2: 145-53. Sacks, Harvey (1974). An analysis of course of ajoke's telling in conversation. In Richard Bauman &Joel Sherzer (eds.), Explorations in the ethnography ofspeaking. Cambridge & New York: Cambridge University Press. 337-53. (1978). Some technical considerations of a dirty joke. InJim Schenkein (ed.), Studies in the organization of conversational interaction. New York: Academic. 249-70. Schieffelin, Bambi, & Eisenberg, Ann R. (1984). Cultural variation in children's con versation. In Richard Schiefelbusch & Joanne Pickar (eds.), The acquisition of com municative competence. Baltimore: University Park Press. 378-420. Schieffelin, Bambi, & Ochs, Elinor, eds. (1986). Language socialization across cultures. Cambridge & New York: Cambridge University Press. Schiffrin, Deborah (1984). How a story says what it means and does. Text 4:313-46. (1987). Discourse markers. Cambridge & New York: Cambridge University Press. Scollon, Ron, & Scollon, Susan (1981). Narrative, literacy and face in interethnic communication Norwood, NJ: Ablex. Shuman, Amy (1986). Storytelling rights: The uses oforaland written texts by urban adolescent Cambridge & New York: Cambridge University Press. Snow, Catherine E. (1991). Building memories: The ontogeny of autobiography. In Dante Cicchetti &Marjorie Beeghly (eds.), Theselfin transition:Infancy to childhood. Chicago: University of Chicago Press. 213-42. , & Goldfield, B. (1982). Building stories: The emergence of information structures from conversation and narrative. In Deborah Tannen (ed.), Analyzing discourse: Text and talk. Washington, DC: Georgetown University Press. 127-41. Spolsky, Bernard, & Walters, Joel (1985). Jewish styles of worship: A conversational analysis. Intemationaljournal ofthe Sociology of Language 56:51-65. Tannen, Deborah (1980). A comparative analysis of oral narrative strategies. In Wallace Chafe (ed.), The pear stories. Norwood, NJ: Ablex. 51-87. (1984). Conversational style: Analyzing talk amongfriends. Norwood, NJ: Ablex. (1985). Relative focus on involvement in oral and written discourse. In David R Olson et al. (eds.), Literacy, language, and learning. Cambridge & New York: Cambridge University Press. 125-47. (1989). Talking voices: Repetition, dialogue and imagery in conversational discourse. Cambridge & New York: Cambridge University Press. Toolan, Michael J. (1988). Narrative: A critical linguistic introduction. London: Routiedge. Young, Katharina (1987). Taleworlds and storyrealms: The phenomenology of narrative. Dodrecht: Nijhoff.
Dialogue and Confrontation in Venezuelan Political Interaction Adriana Bolivar
Introduction h e beginning of Venezuela's democratic political system in the m o d e r n sense is often associated with two major events: the e n d of M a r c o s Perez -JL J i m e n e z ' s dictatorship on 23rd J a n u a r y 1958, a n d the signing of what is k n o w n as the Punto Fijo Pact, which established the rules of participation in the democratic g a m e the same year. Between 1959 a n d 1998 most of the changes were the responsibility of two major parties, Social Democrats (AD) a n d Christian Democrats (COPEI) which, every five years, participated in elections that brought either o n e or the other to p o w e r (Bolivar, 1992). After 1998 a n e w p a r t y was a d d e d to the p i c t u r e , Movimiento Quinta Republica (Fifth Republic M o v e m e n t - M V R ) . This is n o w the party that controls Venezuelan's social, political and cultural life within the frame work of a n extreme polarization that separates followers a n d adversaries of President H u g o Chavez a n d his government, which began in 1999, seven years after C h a v e z h a d led a military c o u p . T h e reasons for this division can b e found in political a n d economic factors in Venezuela's history, notably the predomin ance given to presidential styles that favor authoritarianism (Caballero, 2 0 0 3 , 2004) a n d the transformation from a primarily rural country to a n oil-producing one in the third decade of the twentieth century. C h a v e z ' s gaining of p o w e r in 1999 can b e u n d e r s t o o d as a reaction of Venezuelan people against the traditional parties that h a d g o v e r n e d during 40 years based o n personal authority rather than o n true dialogue a n d participa tion (Bolivar & Kohn, 1999a, 1999b; Bolivar, 1999c, 2001c), as well as a serious
Source: AILA Review vol. 18, 2005. an. 3-17
262
DISCOURSE STUDIES
economic a n d m o r a l crisis caused b y p o o r administration a n d corruption (see Bolivar & K o h n , 1999a; Bolivar, 2001a). M y aim in this p a p e r is to focus o n s o m e m o m e n t s during Chavez's govern m e n t before a n d after 11th April 2002, which is considered b y m a n y as the most conflictive period of the last few years a n d which sparked further conflict and violence in the struggle for political a n d economic power. M y interest is to find out h o w violence a n d confrontation intensify in the course of a n interaction in which participants initiate, r e s p o n d a n d close communicative cycles around issues a n d topics that feed group confrontation. Also, I w a n t to find out h o w different groups of p e o p l e evaluate verbal aggression individually from their own experience. I shall argue that in order to critically describe a n d interpret confrontation in political discourse we h a v e to focus o n the participants in the interaction with a view to assigning responsibilities a n d evaluating effects. O n e way of doing this is to k e e p a detailed record of events identifying the macro-exchanges that emerge in the interaction (Bolivar, 2001a). We can say that what we examine is conversation or dialogue at a higher level, in the macro-social text that can be described o n the a u t o n o m o u s plane that concerns the recording of experience, a n d o n the interactive plane that accounts for the relation between participants (Sinclair, 1981; Bolivar, 1986,1994b; Hunston, 2001). M y claim is that although b o t h planes are simultaneous, the choices of m e a n i n g d e p e n d o n the interactive plane. I h a v e to acknowledge the influence of the British School of Linguistics, particularly Firth (1951), Halliday (1967,1968,1970,1985), Sinclair & Coulthard (1975) a n d H o e y (1979,1983), as well as the influence of research d o n e b y critical discourse analysts, mainly those within the framework of Hallidayan Linguistics (Fowler et al., 1979; Fairclough, 1989, 1992a; a n d Wodak, 2001). T h e analysis I present h e r e takes the viewpoint of the linguist w h o tries to explain h o w meanings are constructed in everyday interaction, b u t w h o is aware that the interpretation a n d explanation will always b e a n approximation to what actually happens because "the knowledge obtained b y research is partial, situated (i.e. specific to particular situations and periods rather than universally applicable) and relative (i.e. related to the researcher's world view and value system)" (Taylor, 2001:12). I adopted the critical discourse perspective after m y first analyses of political dialogue (Bolivar, 1992) w h e n I realized that the semantic and pragmatic levels were n o t enough to understand the dynamics of political interaction, or to under stand sufficientiy the forces that shape its dynamics. Critical discourse analysis (CDA) follows general principles that give ample opportunity to explain meaning and society (see v a n Dijk, 1993; Fairclough & Wodak, 1997), allowing us also to face the twofold c o m m i t m e n t w e h a v e as researchers and h u m a n s w h o want to live in a better world (Bolivar, 1997). While there is n o unique m e t h o d in C D A , m e t h o d itself is very important in order to legitimize research in the field. This is w h y C D A analysts take great care to define their central theoretical notions and methods/procedures (see Meyer, 2001:14), just as all discourse analysts should d o (see Wetherell et al., 2001). The 1
"TMn-rricrniHve.
BOLIVAR
DIALOGUE AND CONFRONTATION
263
discourse-historical, dispositive analysis a n d mediated discourse analysis (see Wbdak & Meyer, 2001); and although all of t h e m h a v e m a d e important contri butions, I have adopted the label Interactional Analysis as m o r e accurately reflect ing m y own approach.
Theoretical F r a m e w o r k The Definition of Context M e a n i n g is constructed in social context, and this m a k e s context the category at the highest level, as studies for over thirty years prove (Duranti & Goodwin, 1997). We h a v e to distinguish b e t w e e n the global context a n d the context of situation. We take it that the global context can b e studied at the historical and cultural level. T h e definition of context of situation in Firthian terms serves our purposes because it embraces all the elements n e e d e d to carry out an analysis of dialogue a n d confrontation in political discourse or any other discourse. So we take into account the following elements (Firth, 1951:43): A. the relevant features of the participants: persons and personalities 1. the verbal action of the participants 2. the non-verbal actions of the participants B. the relevant objects C. the effects of verbal action. These categories allow us to describe language events a n d their effects, which m a y b e verbal or non-verbal. Verbal effects m a y materialize in the production of verbal texts, b u t non-verbal effects m a y express themselves in other languages such as b o d y language, or actions such as stone throwing, clapping, blowing whistles, hitting pots a n d p a n s in demonstrations, laughing, the use of colors (for instance, white for A D , green for C O P E I , r e d for M V R ) a n d other symbols. T h e interactions can c o m b i n e sequences of verbal and non-verbal action in the construction of meaning. I n the study of political interaction the context has to take into account the global context of history a n d culture in order to interpret the meanings constructed today, and it has to examine particular m o m e n t s in the present to see h o w theyfitin the social text that is created. Evaluation as the Central Notion I start from the assumption that discourse is fundamentally about evaluation a n d that we must try to m a k e the connection b e t w e e n the evaluating activity in the world a n d the evaluation in text. Evaluation is studied in language b y examining how interpersonal meanings are constructed, a n d b y the systems of m o o d a n d modality in g r a m m a r (Halliday, 1994). Initially I studied evaluation as motivation for intratextual change in newspaper editorials (Bolivar, 1986, 1994a, 1994b, 1996a), b u t later I extended the notion to other genres a n d to the description of lntortevh.nl
J il- -
T
-
DISCOURSE STUDIES
264
(Bolivar, 1999c, 2001a, 2001b, 2003a). M y aim as a discourse analyst has been to show h o w evaluation, defined initially as the type of information that concerns the expression of feelings a n d opinions, finds its place a n d function in text by negotiation which has its origin in the real world of p h e n o m e n a w h e r e the pur poses for the interaction are determined (Bolivar, 2001e:131). I believe that to b e able to give a n account of texts w e h a v e to study t h e m in the context of events in the world, as Firth argued m a n y years ago: I n linguistics as in a n y other social science, w e start with m a n ' s active participation in the world we are theorizing about. A n d we are all partici pants in those activities which linguistics sets out to study. Speaking, and listening, writing a n d reading, are simply accepted as meaningful in h u m a n life in society. I n brief, linguistics accepts speech a n d language texts as related to the living, a n d therefore to the m e a n i n g of life, and applies its theory a n d practice as far as it is able, to the statement of meaning in strictiy linguistic terms- that is b y employing the restricted language of linguistics set in its o w n theoretical framework (Firth 1957, i n Palmer 1968:169). It is o n the basis of these arguments that I would claim that linguistic and discourse categories can b e extended in order to u n d e r s t a n d social processes, since " O u r schematic constructs must b e judged with reference to their combined p o w e r in dealing with linguistic events in the social process" (Firth, 1951:42). T h e implications for the study of evaluation in discourse, then, are that, in the same w a y that we can study exchangesin conversation or triads in written texts (see Bolivar, 1986,1994a, 1994b, 2001a), we can identify communicative cycles at a higher level in the flow of social events where w e are constantly evaluating the actions of others, our own discourse and the discourse of others. I assume that in democratic political confrontation, the major responsibility for the interaction lies with the leaders w h o h a v e b e e n legitimized b y vote to exercise power. These leaders are constantly being evaluated in terms of what they say a n d d o , or don'i d o , so it is m o r e likely that they will perform the role of initiators a n d that they will close cycles with major evaluations regarding w h a t is g o o d or b a d , positive or negative, desirable or undesirable for their p e o p l e a n d their country. Corpus and Procedures For the purpose of studying political confrontation, the above-mentioned points imply that we must collect a corpus of written a n d spoken texts during the course of events; we have to identify the issues a n d topics, b u t focusing primarily on the actors/participants w h o introduce them. T h e aim is to see w h o initiates and responds to interactions a n d w h o closes cycles with the voice of the "primary knower" (Berry, 1975) so that w e can discover the types of patterns that emerge in the interaction. I n this way, we can describe interactions between participants in the spoken dialogue (speaker, hearer) a n d participants in written text (writerreader), a n d m a k e w a y for other voices or participants in the texts, as represented
BOLIVAR
DIALOGUE AND CONFRONTATION
265
b y the participants/actors in the real world. Focusing o n the participants in events is m o r e than a methodological decision, it is o n e that affects the collection of the texts a n d t h e interpretation and explanation of the context of culture a n d situation. T h e description focuses first on the macro-exchanges in which political actors (or those w h o assume such a role) confront, a n d t h e n o n the effects this confrontation has on those w h o act as observers or participate in it as citizens. I n order to analyze the macro-exchanges that emerge at particular m o m e n t s in the interaction we follow the steps summarized below: • A particular problem/conflict is chosen in order to observe h o w par ticipants introduce changes in the situation. • T h e events are followed in a sequential order starting with those par ticipants w h o initiate a topic or conflict. T h e texts are collected in the chronological order in which they are p r o d u c e d . • M a c r o a n d micro-exchanges are collected. For example, micro-exchanges in conversation or in the news and the macro-exchanges that emerge later in the political interaction. These are given a function in the sequence as Initiation (I), Follow (F) or Closing (C). We must b e aware that an Initiation of o n e exchange might also work as the Closing of a previous one. • Actions a n d topics are traced in chronological order. • T h e effects in language a n d actions are described, with attention to the a c c o m p a n y i n g use of symbols (flags, whisdes, banners). • T h e linguistic resources a n d discourse strategies are e x a m i n e d with emphasis o n the interpersonal value (mood a n d modality, (im)politeness). The Participants in the Confrontation The m a i n participant in the confrontation since 1998 has b e e n H u g o Chavez because h e started a n e w period in Venezuela's political history. Although from the outset his style was reminiscent in some ways of previous populist presidents like Betancourt (see Madriz, 2000) a n d Perez i n his second g o v e r n m e n t (Bolivar, 2001c), there was a striking difference in terms of the high level of verbal aggres sion in Chavez's electoral campaign, with a slogan threatening 'to fry the heads of Adecos' in oil, which h a r k e d back to the Federal War in the middle of the nineteenth century. In this campaign h e p r o p o s e d radical change a n d a return to the values of the past to rescue the Republic. I n the face o f ' d e c a y ' a n d 'death', he proposed 'reconstruction' a n d 'rebirth' (Molero d e Cabeza, 1999). H e broke the rules of protocol from the very day h e took office (2nd February, 1999) b y modifying the oath p r o n o u n c e d b y his predecessors: "I swear to G o d , to m y Motherland, to m y People, o n this m o r i b u n d Constitution, that I will encourage die necessary democratic transformations for the Republic to h a v e a n e w M a g n a Carta suited to the n e w times. I swear i f (Alocuciones presidenciales, 1999, m y translation) (see Bolivar, 2001c). T h e reference to this m o r i b u n d constitution broke the pragmatic rules of quantity, quality a n d m a n n e r (Grice, 1975), b u t it emphasized the relevance of his promises in the electoral campaign. Since then, Venezuelans h a v e witnessed m a n v chances- a n«>«, r w - < ^ - ^ — '
266
DISCOURSE STUDIES
Chavez t e r m to six years a n d allows h i m to b e reelected, a n e w n a m e for the country, Republica Bolivariana de Venezuela (Bolivarian Republic of Venezuela), a National Assembly instead of a Congress, a n e w p o w e r structure that has added M o r a l Power a n d Citizens Power to the branches of government, and n e w laws, a m o n g t h e m the recently a p p r o v e d L a w of Social Responsibility of the Media. Also, there is a n e w internal composition of the S u p r e m e C o u r t of Justice (comprising predominantly g o v e r n m e n t supporters), which gives Chavez more control of p o w e r than a n y other president before him. F r o m 1999, President C h a v e z a n d those close to h i m h a v e privileged a threatening rhetoric a n d impoliteness (threats, insults, discrediting nicknames, war metaphors, accusations) which in m a n y ways h a v e contributed to create a n d reinforce polarization (see M o n t e r o , 2 0 0 3 ; Bolivar, 2001b, 2002a, 2002b, 2005a, 2005b, a n d forthcoming); Bolivar et al., 2003). T h e results of a rhetorical analysis of 2,480 front pages of national newspapers between 1999 a n d 2002 (Montero, 2003) serve as evidence that the attitude of the opposition went first through a phase of perplexity a n d astonishment, to another of implicit or moder ate protest that introduced cacerolazos (the noise m a d e b y banging pots a n d pans), a n d finally to o n e of explicit response with increasing verbal aggression against the President a n d criticism of the g o v e r n m e n t (see Lozada, 2003). At the e n d of 2001 insults were c o m b i n e d with actions in the streets: demon strations, cacerolazoshy the opposition, fireworks b y Officialists (the n a m e given to g o v e r n m e n t followers) a n d physical attacks o n Opposition demonstrators. T h e violence reached its climax on the 11th of April 2002 w h e n m o r e than a million demonstrators took the streets asking for Chavez's resignation. During this demonstration 19 p e o p l e were shot dead b y snipers. Venezuelans witnessed that and the reading of Chavez's resignation o n television b y General Lucas Rincon. Chavez was out of power for 48 hours, during which time Pedro Carmona Estanga was appointed president. I n his brief government, C a r m o n a Estanga decreed the dissolution of powers and rescued the country's n a m e , changing ii back to Venezuela. Apparentiy the military changed its m i n d , a n d Chavez was reinstalled. W h a t actually h a p p e n e d in those 48 h o u r s (12th a n d 13th of April) remains uncertain (see Lafuente & Meza, 2004). T h e confrontation continued in 2 0 0 2 , a n d is still present, with Chavez's followers defending their leader's revolution a n d his adversaries insisting o n the loss of legitimacy of the govern m e n t because the democracy was b e i n g forced into a revolution people did not vote for in 1998. T h e opposition is n o w weakened b y Chavez's success in the 2004 referendum that will keep h i m in office until 2006, despite all the accusations of fraud, hh p r e s u m e d contact with C o l o m b i a n guerrillas and his alliance with Fidel Casta T h e political confrontation has involved the participation of the media, the Catholic Church, the political parties that oppose h i m (from both right and left wings), the Confederation of Workers (CTV), the Confederation of Trade and C o m m e r c e , the National Oil C o m p a n y , private schools a n d the universities. The verbal confrontation has even g o n e b e y o n d the national borders to include i President George Bush and some of his collaborators. J
BOLIVAR
DIALOGUE AND CONFRONTATION
267
The Patterns of Interaction We only h a v e r o o m to briefly cover some aspects of the confrontation with the media. We have identified several general patterns of interaction that, apparentiy, show h o w polarization and violence worked before a n d after 11th April 2002. T h e m e d i a w e r e the target of the president from the beginnings of his term, especially in his radio and television program AidPresidente, which lasts an average of five hours every Sunday. Indeed, in p r o g r a m n u m b e r 100 o n 17th M a r c h , 2002, C h a v e z claimed that the p r o g r a m was created because the m e d i a would not allow t h e m to work (see Bolivar, 2003b, 2005a). For their part, newspapers and television channels h a v e criticized Chavez harshly for his personalism a n d militarism (Bolivar, 2001c, 2001d, 2005b). I n the confrontation with the media, two patterns w e r e found, o n e that involves physical aggression before April 11th 2002, a n d another that shows h o w alignment against the g o v e r n m e n t grew after this date. Macro-exchange 1: between the President and journalists (from March 2000 to January 2001 - see Bolivar, 2001b). Initiation: The president exhorts the military to fight for the truth and insults the media (he calls them "all-time deceivers" (EINacional, 24 March, 2000, front page)). Follow (1): Journalists respond with a demonstration demanding freedom of expression and verbal reaction, returning the insult by calling Chavez gallina (chicken) by way of a reminder of the failure of his attempted military coup in 1992 (El Universal, May 2, 2000, p. 1-10). Follow (2): Chavez's followers insult the demonstrators ("You are enemies of the people, You only tell lies, Chavez tells the truth, Out with Journalists! Out!") and attack journalists physically Cjournalist Eugenio Martinez was hit in the face with the Sunday edition of El Universal. Most media vehicles were the target of stones and blows" - El Universal, 2000, front page). Closing: Journalists accuse Chavez of provoking and repressing journalists and of putting their lives at risk (El National, 5 May, 2000, D/2) Macro-exchange 2: between humorists and the President (between 20th and 30th October, 2002) Initiation: Humourist Pedro Leon Zapata criticizes Chavez' militarism in a caricature (A mi la sociedad civil me gusta firme y a discretion - I like Civil Society standing to attention and ready to obey orders - EINacional, 20 October, 2000). Follow (1): President Chavez responds on the same day from Margarita Island, late at night on a national broadcast, Pedro Leon, ccudnto te pagan por eso? (Pedro Leon: how much are they paying you for that?). It should be noted that Zapata's caricatures have been critical of all Venezuelan governments before Chavez. Follow (2): Humorists from four widely read national newspapers publish caricatures making reference to Zapata's text and picture (e.g. a sword imitating Chavez's face). They all implicitly refer to militarism and condemn the President's insult.
DISCOURSE STUDIES
268
Closing: Zapata publishes a new caricature using the same words pronounced by the President, ironically thanking him for the free promotion: y hablando como los locos, Hugo Rafael: icudnto tepago Zapatapor esepropaganddn? (And since we're talking nonsense, Hugo Rafael, how much did Zapata pay you for all this advertising? - El Nacional, 30 October 2000). T h e National U n i o n ofJournalists publishes a c o m m u n i q u e that gathers all the caricatures in one text h e a d e d Al Maestro con carino. . . (To Master Zapata with love . . . - El Nacional, 5 N o v e m b e r 2000, D / 7 ) in a gesture of soUdarity b y all Venezuelan journalists a n d artists. I n 2002, in the absence of the traditional parties that h a d b e e n discredited by Chavez, the m e d i a took the lead as political actors a n d w e r e largely responsible for organizing demonstrations against the government. This is evidenced by the way in which the private television channel Globovision (Global Vision) was nicknamed Plomo Vision (Bullet Vision) b y the President, and Globoterror (Global Terror) b y his followers. Venevision, another private channel, b e c a m e Venenovision (Poisoned Vision). M o r e t h a n 200 journalists, several newspapers a n d television channels (29 in all) received some kind of aggression in 2002, from damage to equipment to b o m b explosions (El Nacional, 3 J a n u a r y 2003 p . A/1-2). Between December 2002 and February 2003, the country was involved in one of the longest strikes in its history, called b y the Federation of Workers (CTV). T h e National Oil I n d u s t r y c o n d e m n e d C h a v e z ' s attempts to politicize the industry a n d p u t it in the h a n d s of a n e w Board of Directors integrated b y gov e r n m e n t supporters. T h e Opposition c o n d e m n e d the increasing violence, Fidel Castro's influence on Venezuela's decisions, the failure of the government's social a n d economic p r o g r a m s and Chavez's reluctance to accept a recall referendum. D u r i n g the strike a n e w pattern emerged, which included verbal aggression as well as violence, and which involved the issue of cultural values. This was initiated b y General Acosta Carles w h o , following orders given b y the President, b r o k e into the warehouses of Coca Cola a n d national b r e w e r y Polar in Carabobo State to confiscate the products with the intention of giving t h e m to the people affected b y the strike. Macro-exchange 3: between the military (General Carles), the people, and the President (between 17thJanuary 2002 and 11th February 2003) In this macro-exchange two non-verbal actions were evaluated by Venezuelans: (a women demonstrators were beaten to the floor by the soldiers accompanying the general and by the General himself; (b) the General (G) belched in front of die television cameras when a woman journalist (J) from Globovision tried to interview him. Burping in public is considered impolite in Venezuelan culture, so this generated evaluations from all sectors in society. Initiation (the violence and a micro-exchange on television): J:
General, today was the deadline of 48 hours given by Indecu (the government consumer protection body) to the company t o . . . G: (belches, holding a can of Malta, a soft drink) Excuse me, excuse me miss (belches again)
BOLIVAR
DIALOGUE AND CONFRONTATION
269
J: Isn't that rude? G: No, no, you know it was instinctive, right? And it came out because this has a lot of gas, and it was hot, right? J: Excuse me General, today was the deadline that Indecu had given the company to distribute the products . . . G: . . . do you remember this baseball player who did that, he has a refreshment and he goes (belches) J: This is a serious job and I am asking you a serious question, General... in relation to . . . Follow (1): The newspapers report the event and reject both the violence and the belching as grotesque. They also imply that the action was robbery rather than confiscation (El Universal, p.l EINacional, p. 2-4,18,January, 2003). Follow (2): The General is accused of dishonoring the uniform and he is reminded of the military code of ethics (El Universal, 18, January, 2003, p.2-4). Follow (3): Demonstrators take to the streets to condemn the abuse of power, the violence and the belching. The main slogan reads Ni tin eructo mas (No more belches), which can be interpreted as the condemnation of sexism and machismo as well as violence and abuse of power (El Universal, 19 January, 2003 p. 1-2) Follow (4): National Guards demand the removal of General Carles (El Nacional, 21 January, 2003, p.l) Follow (5): Chavistas (Chavez's followers) attack demonstrators in El Tuy ("Chavist ambush", El Nacional, 21 January, 2003, p.l) Follow (6): A Zapata caricature makes fun of the cultural revolution (EINacional, 22January, 2003). Follow (7): Lawyers claim that Carles's visit to the Coca Cola and Polar premises was illegal. (El Universal 22 January, 2003, p. 1-4). Closing: President Chavez ratifies the General in his position during a demonstration organized by the government on 23rd January 2003, to celebrate the anniversary of the 1958 uprising. Chavez's followers sing Carles Carles eructales otra vez (Carles, Carles, burp again), thus approving what the Opposition evaluated as grotesque. No apologies were made for the General's conduct. On February 11th 2 0 0 3 the general was decorated b y the President for being a General and for his patriotic services during the strike (EINacional, 11 February, -••"3, p . A / 3 ) . A graffiti a c c o m p a n i e d this decision: Acosta Carles General del ftteblo (The people's General). The strategic use of the President's threatening language after terrorist attacks was highlighted b y t h e m e d i a in headlines a n d editorials (Verbobomba - Bombverb. Tal CuaL. 2 5 February 2 0 0 3 : El hnmmi, A<»I Z » ~ ~ J — »
DISCOURSE STUDIES
270
violencia - T h e President's l a n g u a g e c o m m a n d s a n d legitimizes violence, El National, 2 March 2003, front page). These were tense and uncertain m o m e n t s for all Venezuelans since both sides were accusing each other for the violent attacks. T h e opposition n i c k n a m e d the General "General Belch", a n d insulted and harassed h i m i n protest against his behavior. O n 2 3 r d J a n u a r y 2004 the General was appointed m a i n speaker in the National Assembly o n the occasion of the anniversary of the 1958 uprising, a Bolivarian School was n a m e d after h i m (number 3,200 in SanJ u a n d e Los Morros, 6th M a y 2004). Acosta Carles is n o w Governor of C a r a b o b o , o n e of the most important states in Venezuela. The Effects of Confrontation T h e effects in the minds and self-esteem of Venezuelan citizens are worth reporting. In a study that involved 50 persons from m y own university (students, colleagues a n d workers) w h o were asked to write d o w n a list of the insults they recalled in exchanges between the Government a n d the Opposition (Bolivar, 2002a, 2003a), it was found that 2 6 3 different words h a d b e e n evaluated as insulting. Women a n d m e n in this group (Group 1) tended to evaluate differently b u t they all agreed o n a few words which indicated their perceived values a n d attitudes in terms of affect, j u d g m e n t a n d appreciation (see Martin a n d Rose, 2 0 0 3 : 15). T h e insults attributed to the Opposition against the G o v e r n m e n t (in decreasing order of strength) w e r e about: (a) m o r a l j u d g m e n t : 'assassins' (everybody), 'assassin' (Chavez), 'corrupt', 'thieves', 'liars', (b) appreciation of social value: 'shantytowners' (marginales), 'scum', ' m o b s ' , 'hordes', (c) appreciation of political value: 'communists', 'Bolivarians', 'dictator', 'Chavist'; (d) appreciation of mental and intellectual value: 'loony', 'nut', ' d u m b ' (all against Chavez); (d) appraising feeling: 'resentful' (resentidos). As for insults attributed to the G o v e r n m e n t against the Opposition w e found: (a) m o r a l j u d g m e n t : 'rotten in-groups' (cupulas podridas, to refer to A D and C O P E I ) , 'conspirators', 'fascist', 'traitors', 'liars', 'unpatriotic', 'corrupt', 'thieves'; (b) appreciation of political value: 'adecos' (AD people), 'anti-Chavez', golpistas, (involved in a coup d' etat); (c) appreciation of social value: 'squalids' (escudlidos) a n d 'oligarchs' (to m e a n high class or privileged); a n d (d) appraisal of feeling: 'in anguish' (afligidos). I n another study, still in progress, to evaluate the insulting repertoire after the referendum, three m o r e groups h a v e b e e n interviewed (100 persons): Group 2, from Universidad Experimental Libertador (UPEL-Maracay), G r o u p 3 from the recently created Universidad Bolivariana, and G r o u p 4, from Universidad Central de Venezuela in Caracas. It has b e e n found that w o m e n a n d m e n remember a large n u m b e r of o n e w o r d insults and a wide variety of complex expressions (721 in all). T h e most revealing point is that the words agreed u p o n b y women a n d m e n in these three groups are fewer b u t consistently the same. Only three words are recalled b y both w o m e n a n d m e n in their attributions to the Opposition against the G o v e r n m e n t : 'assassins', 'thieves' a n d 'communists', which indicate moraljudgment and political appreciation. Four words stand out in the attributions
BOLIVAR
DIALOGUE AND CONFRONTATION
271
to the Government against the Opposition: 'squalids' (escudlidos), golpistas, 'oligarchs' a n d 'fascists', which indicate appreciation of social and political value. M a n y n e w offensive expressions appeared after D e c e m b e r 2004. Of these, the three groups agree o n three words that are attributed to the Opposition against the G o v e r n m e n t : malandros (scoundrels), 'ignorant' a n d 'illiterate'. O t h e r words mentioned, but without agreement a m o n g w o m e n a n d m e n , are: 'monkeys', 'blacks', 'Indians', 'gorillas', 'homosexuals', 'dirty' a n d 'revolutionaries', all ex pressing negative appreciation of the people w h o follow Chavez, w h o is referred to as a 'tyrant'. Two words are r e m e m b e r e d b y all the w o m e n a n d m e n in Groups 2 and 4 in their attributions to the G o v e r n m e n t against the Opposition: 'terrorists' a n d 'assassins'. I n group 3, however, the attributions are different. W h a t w o m e n and m e n recall is plastas (pieces of shit) and pendejos (schmuck). W h e n we examine the textual value of these words, we realize that most of those that w o m e n a n d m e n r e m e m b e r form p a r t of a cohesive link with words introduced b y the President (escudlidos, oligarcas,fascistas, golpistas, plasta, pendejo) (see Bolivar, 2003a for the use of plasta). Also, most words attributed to the Opposition have b e e n used b y Opposition leaders, the m e d i a a n d the people in demonstrations. W h a t strikes our attention is that the group that might b e closer to Chavez, because they study at the university created b y the President, perceive more insults that lower their self-esteem (blacks, Indians, m o n k e y s , ignorant). After D e c e m b e r 2004, racism a n d sexism seem to h a v e d e e p e n e d as well as despise for the other - a n d a hatred that can only b e harmful for everybody.
Conclusions In 1999, before Chavez b e c a m e President of Venezuela, we w e r e concerned about the lack of dialogue a n d participation in our democracy (Bolivar & K o h n , 1999b, Bolivar, 2001d). N o w the concern has b e c o m e alarm because, in the political confrontation, violence a n d death have found a place. T h e analysis presented here has shown how, in ongoing interaction, actions a n d words are evaluated b y the participants, a n d h o w meanings are created in the course of events where verbal a n d non-verbal actions combine. We have also seen that words h a v e effects o n people's m i n d s a n d feelings because offensive words only serve to feed the confrontation a n d hate. Although I claim that, in discourse, initiations are important to identify those w h o start conflictive macro-exchanges, I believe that in everyday life, the responsibility rests o n us all because, as citizens, we must preserve respect for the other and enrich cooperation and understanding. However, the major responsibility lies with those w h o h a v e b e e n legitimized in power b y the vote, a n d w h o must not forget their c o m m i t m e n t to govern for everybody, bearing in m i n d that in the long r u n cooperation and persuasion work better than stressing the differences, threatening, a n d using fear to reach polit ical goals. H a v i n g seen h o w hate language works in practice it is m a n d a t o r y to reinforce peace language for the sake of the country a n d all those involved in the confrontation.
DISCOURSE STUDIES
272
References Alocuciones presidenciales. 1999. www.venezuela.gov.ve/ns/aloc.asp 2 de febrero de 1999. Berry, M. 1975. Introduction to Systemic Linguistics I: Structures and systems. London: B.T. Batsford. Bolivar, A. 1986. Interaction through Written Text. A discourse analysis of newspaper editorials. PhD dissertation. University of Birmingham. Bolfvar, A. 1992. The Analysis of political discourse, with particular reference to the Venezuelan political dialogue. English for Specific Purposes, 11, 158-175. Bolivar, A. 1994a. The structure of newspaper editorials. In Advances in Written Text Analysis, M. Coulthard (ed.), 276-294. London: Roudedge. Bolivar, A. 1994b. Discurso e interaction en eltexto escrito. Consejo de Desarrollo Cientifico y Humanistico. Caracas: Universidad Central de Venezuela. Bolivar, A. 1996a. The discourse of British newspaper editorials. LABSA Journal 1(1): 75-79. Bolivar, A. 1997. El analisis critico del discurso: teoria y compromisos. Episteme NS 17(1-3): 23-45. Bolivar, A. and C. Kohn (eds). 1999a. El discurso politico venezolano. Un estudio multidisciplinario Caracas: Comision de Estudios de Postgrado y Fondo Editorial. Tropykos. Bolivar, A. and C. Kohn, 1999b. Dialogo y participacion: iCual dialogo? icual participacion? In El discurso politico venezolano. Un estudio multidisciplinario, Bolivar, A y C. Kohn (eds), 103-115. Caracas: Comision de Estudios de Postgrado y Editorial Tropykos. Bolivar, A. 1999c. The linguistic pragmatics of political pronouns in Venezuelan Spanish. In Language and Ideology,]. Verschueren, (ed.) Vol. 1,56-69, Antwerp: International Pragmatics Association. Bolivar, A. 2001a. Changes in Venezuelan political dialogue: The role of advertising during electoral campaigns. Discourse & Society 2(1): 23-45. Bolivar, A. 2001b. El insulto como estrategia en el discurso politico venezolano. Oralia 4: 47-73. Bolivar, A. 2001c. El personalismo en la democracia venezolana y cambios en el dialogo politico. Revista Iberoamericana de Discursoy Sociedad3(l): 103-134. Bolivar, A. 2001d. El acercamiento y el distanciamiento pronominal en el discurso politico venezolano. Boletin de LingUistica 16: 86-146. Bolivar, A. 2001e. The negotiation of evaluation in written text. In Patterns of Text. In honour ofMichaelHoey, M. Scott and G. Thompson (eds), 130-158. Amsterdam:John Benjamins. Bolivar, A. 2002a. La violencia verbal en la interaccion politica y sus efectos en la ciudadania. Paper presented at the XI Jomadas Venezolanas de PsicohgCa Social, Caracas. Venezuela, 30 May 2002. Bolivar, A. 2002b. Violencia verbal, violencia fisicay polarization a traves de los medios. In El discurso politico en las Ciencias Humanasy Sociales, L. y A. Molero, Franco (eds). 125-136. Caracas: Fonacit. Bolivar, A. 2003a. La descortesia como estrategia politica en la democracia venezolana. In La perspectiva no etnocentrista de la cortesia:Identidadsocio-culturalde las comunidades hispanohablantes, D. Bravo, Actas del primer Coloquio EDICE, CD Rom, 213-226. Stockholm: Stockhohns Universitet. Bolivar, A. 2003b. Nuevos generos discursivos en la politica. In Andlisis critico del discurso. Perspectivas Latinoamericanas, L. Berardi (ed.), 101-130. Santiago de Chile: Frasfe editores.
273
Bolivar, A. 2005a. The President and the Media. In Diahganalyse VIII/Dialogue Analysis VIII-Dialogue in Literature and the media. Referate der 9 Arbeitstagung der IADA, Salzburg 2003/'Selected Papers from the 9th IADA Conference, Salzburg 2003, Betten, A. and M. Dannerer (eds) Tubingen: Niemeyer (2 volumes). Bolivar, A. 2005b. El analisis critico de la (des)cortesia. In Cortesia Lingiiisticay comunicacidn en espanoL D. Bravo (ed.), 273-297. Buenos Aires: Dunken. Bolivar, A. Forthcoming. La descortesia en la dinamica social y politica. In Actos de habla y cortesia en distintas variedades del espanol: perspectivas teoricasy metodoldgicas. Adas del II Coloquio International del Programa Edict, J. Murillo y D. Bravo (eds). San Jose: Universidad de Costa Rica. Bolivar, A., I. Chumaceiro and F. De Erlich 2003. Divergencia, confrontation y atenuacion en el dialogo politico. RevistaBeroamericana deDiscursoy Sociedad. Vol 4(3): 121-151. Caballero, M. 2003. Gdmez, un tirano liberal. Caracas: Alfadil Ediciones (fifth edition). Caballero, M. 2004. Romulo Betancourt, politico de nacidn. Caracas: Alfadil & Mexico: Fondo de Cultura Economica. Duranti, A. and C. Goodwin (eds). 1997 [1992]. Rethinking Context. Language as an interactive phenomenon. Cambridge: CUP. Fairclough, N. 1989. Language and Power. London: Longman. Fairclough, N. 1992a. Discourse and Social Change. Cambridge: Polity Press. Fairclough, N. and R. Wodak 1997. Critical Discourse analysis. In Discourse Studies. A multidisciplinary introduction, Vol. 2. Discourse as Social Interaction, T.A. van Dijk (ed.). 259-284. London: Sage. Firth, J.R. 1951. Personality and language in society. Sociological Review 42: 37-42. Firth, J.R. 1957. A synopsis of linguistic theory, 1930-55. In Selected papers ofJ.R. Firth, 1952-59, Palmer, F.R. (ed.), 1968, 168-205. London: Longman. Fowler, R., B. Hodge, G. Kress and T. Trew. 1979. Language and Control. London: Roudedge and Kegan Paul. Grice, H.P. 1975. Logic and conversation. In Syntax and Semantics, Vol 3: Speech Acts, P. Cole and J. Morgan (eds), 41-58. New York: Academic Press. Halliday, M.A.K. 1967. Notes on transitivity and theme in English./oaraa/ of Linguistics, Part 1: 3(1): 179-215. Part 2: 3(1): 37-81. Part 2: 3, 199-244. Halliday, M.A.K. 1968. Notes on transitivity and theme in English. Joaraa/ of Linguistics Part 3: 4(2): 179-215. Halliday, M.A.K. 1970. Functional diversity in language, as seen from a consideration of modality and mood in English. Foundations ofLanguage 6(3): 322-361. Halliday, M.A.K 1994 [1985]. An Introduction to Functional Grammar. London: Edward Arnold. Hoey, M. 1979. Signalling in Discourse [Discourse Analysis Monographs 6]. University of Birmingham: English Language Research. Hoey, M. 1983. On the Surface of Discourse. London: Allen and Unwin. Hunston, S. 2001. Evaluation and the planes of discourse: Status and value in persuasive texts. In Evaluation in Text. Authorial stance and the construction ofdiscourse, S. Hunston and G. Thompson (eds), 176-207. Oxford: OUP. Lafuente, S. and A. Meza. 2004 [2003]. El acertijo de abril. Relato periodistico de la breve caida de Hugo Chavez. Caracas: Debate. Coleccion Actualidad. Lozada, M. 2003. El lenguaje de la red: El discurso del ciberciudadano. Revista Iberoamericana deDiscursoy Sociedad4(3): 77-97. Madriz, M.F. 2000. Los demonios del comandante (La violencia como estrategia discursiva). Akademos 2(2): 65-86.
274
DISCOURSE STUDIES
Martin,J. and D. Rose. 2003. Working with Discourse. Meaning beyond the clause. London: Continuum. Meyer, M. 2001. Between theory, method, and politics: positioning of the approcahes to CDA. In Methods of critical discourse analysis, R. Wodak and M. Meyer (eds), 14-21. London: Sage. Molero de Cabeza, L. 1999. Analisis de dos discursos del proceso electoral de 1998 bajo un enfoque semantico pragmatico. In Eldiscursopolitico venezolano. Un estudio multidisciplinario, A. Bolivar and C. Kohn (eds), 145-157. Montero, M. 2003. Retorica amenazante y crisis de gobernabilidad en Venezuela 2002. Revista Iberoamericana deDiscursoy Sociedad4(3): 37-56. Sinclair,J. McH. and M. Coulthard. 1975. Towards an analysis of discourse. Oxford: OUR Sinclair, J. McH. 1981. Planes of discourse. In The Two-Fold Voice: Essays in honour ofRamesh Mohan. S.N.A. Rizvi, (ed.), 70-89. Salzburg: University of Salzburg. Taylor, S. 2001. Locating and conducting analytic research. In Discourse as Data. A guidefor analysis, M. Wheterell, S. Taylor and S. Yates (eds), 5-48. London: Sage & Open University. Van Dijk, T.A. 1993. Principles of discourse analysis. Discourse & Society 4(2): 249-283. Wetherell, M., S. Taylor and S. Yates (eds). 2001. Discourse as Data. A guide for analysis. London: Sage and Open University. Wodak, R. and M. Meyer (eds). 2001. Methods of Critical Discourse Analysis. London: Sage. Wodak, R. 2001. The discourse historical approach. In Methods ofCritical Discourse Analysis. R. Wodak and M. Meyer (eds), 63-94.
11 Personal Web Pages and the Semiotic Construction of Academic Identities Carmen Rosa Caldas-Coulthard i i
1. I n t r o d u c t i o n
V
isual communication is central to the information society and to the practices of Late Modernity. Pictures, p h o t o s , illustrations are everywhere - they permeate our academic work, everyday lives, and activities. It is c o m m o n nowadays to refer to brand images, corporate images, national images, and self-images. These are interwoven with our professional and personal identities, and they are signposts to lifestyles, cultures, and societies. In his seminal book Ways of Seeing, John Berger (1972: 7) points out that: Seeing comes before words - the child looks and recognizes before it can speak. But there is also another sense in which seeing comes before words. It is seeing which established our place in the surrounding world. We explain that world with words, but words can never undo the fact that we are surrounded by it. The relation between what we see and what we know is never settled. The way we see things is affected b y what we know or what we believe. To look is an act of choice. Scollon & Scollon (2003b) suggest that humanity has undergone at least four ages' in its development: the oral, the literate, the imagistic, and currently, the information age. For them, oral societies were constructed through oral narratives; literate societies, through the standardisation of genres and documentation essays, plans, forms, scripts, charts, tables, etc.); while the still-current imagistic
Source:
C. R. Caldas-CoultharH » ^ \*
TX_I—
/ - -i »
DISCOURSE STUDIES
276
society constructs itself through the m o v i n g picture - the camera, the analogue and the digital video. T h e information society, b y contrast, is constructed through the digital p a t h w a y that m a k e s its m e a n i n g through the combination of m a n y semiotic resources. We take this for granted. Meaning-making is d e p e n d e n t o n the interplay of different resources, a n d these surround us in our everyday life. N e w technologies, especially the Internet a n d the World W i d e Web, are pushing forward, almost daily, n e w communication strategies and people interact through these n e w resources. According to N U A (a resource for Internet statistics r u n b y 1NQUIRA, N U A internet surveys - http://www/nua.com/surveys), an exact estimation of h o w m a n y people are online throughout the world is not possible. H o w e v e r , they say, there are m a n y surveys that use different m e a s u r e m e n t p a r a m e t e r s and from observing m a n y of t h e published surveys over t h e last two years, they claim that u p to September 2002, approximately 605.6 million people h a d online access worldwide. J u s t one year later, Global Reach (Global Internet Statistics b y Language www.glreach.com) gives us the following results: O n l i n e L a n g u a g e P o p u l a t i o n s Total: 649 M i l l i o n (March 2003) Language Dutch Russian Portuguese French Italian Korean German Spanish Japanese Chinese English
, . ,
'
:
, ,..
•
, (
Percentage 1.9 2.7 2.8 3.3 3.6 4.2 6.5 8.1 10.3 11.9 35.2
T h e sheer n u m b e r s of people using virtual communication a n d the affordances of the n e w m e d i a are p r o d u c i n g a visual revolution, in which everything from the structure of books to the layout of pages, distribution of images such as photographs, illustrations or digital backgrounds, a n d the use of typography a n d colour are brought to bear o n conveying specific meanings. Representation, modality a n d multimodality therefore enter into the very constitution of things since an object's or idea's m e a n i n g is s h a p e d b y the very process of representing it b y way of language or images or other semiotic resources. R e p r e s e n t a t i o n for Hall (1997:61) is the process b y which m e m b e r s of a culture use signs to p r o d u c e m e a n i n g - 'things, objects, people or events d o not have in themselves a n y fixed, final or true meaning. It is us, in society, who makes things m e a n , w h o signify'.
CALDAS-COULTHARD
PERSONAL WEB PAGES
277
Representation, as a cultural process, establishes individual a n d collective identities, and symbolic systems provide possible answers to the questions: w h o a m I? what could I be? w h o d o I want to be? (Woodward 1997: 14). M o d a l i t y is the concept used b y communication specialists to talk about different m o d e s or codes of communication - speaking, writing, using images, gestures, sounds, colour, drawings, etc. to communicate a specific message. M u l t i m o d a l i t y (Kress & v a n L e e u w e n 2001: 20), the concept used b y semioticians a n d discourse analysts, o n the other h a n d , is 'the use of several semiotic m o d e s in the design of a semiotic product or event'. M o d e s are resources for making meaning. T h e same meaning, in a n y cultural domain, can b e expressed in different semiotic m o d e s - (music can e n c o d e action, images can e n c o d e emotion). We would miss out important aspects of a film without considering the role of music in the ongoing narrative, for example, or of a magazine without interpreting its pictures a n d illustrations. Therefore, for Kress a n d v a n Leeuwen, the particular ways that m o d e s are combined (they m a y , for instance, reinforce or c o m p l e m e n t each other) will determined the ways we receive a n d interpret a message. T h e authors claim that all semiotic products a n d events are com municative since they undergo processes of articulation, interpretation, and ultim ately, use a n d action. A n important point to m a k e is that all messages are multimodal since all mes sages m a k e use of multiple m o d e s . Web pages are archetypal examples of multi modal messages. W e b authors use m a n y semiotic resources in their pages to communicate their identities - the language used, the photos, the colours, the objects chosen a n d displayed, all signify symbolically different aspects of their identities. Multimodal analysis therefore is crucial for understanding a n d interpreting the world. O u r tendency so far as linguists, however, has b e e n to privilege written and spoken texts a b o v e all other m o d e s , a n d to consider objects, actions, a n d people as simply m a k i n g u p 'the environment in which the text comes to life', as Halliday (1978: 25) suggests. T h e internet and the World W i d e W e b h a v e changed our views dramatically. As Kress & van L e e u w e n (2001) h a v e pointed out, the desire for crossing bound aries inspired 20th-century semioticians to develop n e w theoretical frameworks that are applicable to all semiotic m o d e s . In this chapter, therefore, I will e x a m i n e academics' personal W e b pages in order to consider the multimodal discursive strategies used b y Web page creators to present themselves in terms of: 1
1. the narrativisation of their lives; 2. the multimodal choices that help to create a particular identity. I will also want to point out the implications of this kind of one-sided inter action for discourse analysis - the Web-page addresser sends a message n o t to an implied audience b u t to a n y b o d y out there in the world w h o h a p p e n s to open her or his page. T h e strategies used to address a n unspecified audienrp
278
DISCOURSE STUDIES
M y interest here is in looking specifically at h o w participants in this new genre recontextualise (Bernstein 1981), in the public sphere, a fictionalised repre sentation of themselves. T h r o u g h what seems to b e a light a n d inconsequential 'message in a bottle' to the world, social activity is presented, evaluated, and possibly n e w n o r m s of behaviour are b e i n g created.
2. Personal W e b P a g e s - T h e Personal i n t h e Public S p h e r e R e a l or Fictional Story Telling? O n a day of internet surfing, I came across a page, http://www.cs.bham.ac.uk/ ~ a x s / # w h o , b y o n e of t h e m e m b e r s of the University of Birmingham staff, a computational scientist n a m e d Aaron. At first sight, this was an academic page, giving the internet surfer information about the professor's career, his publica tions, courses given, etc. However, there was a link to family issues, which I pursued. This hyperlink seduced m e in m a n y ways, especially for the force of the narration. For m a n y minutes, I followed attentively the story being told. This was t h e 'real', n o t fictional, story of a y o u n g m a n w h o died of cancer, Ben, retold b y his father, the professor. I saw Ben's w e d d i n g photos, t h e n the events of his illness, a n d finally, the events surrounding his death at 34 years of age. M a n y different semiotic resources were posted o n this W e b page - photos of different occasions of the family life, a m a p of the cemetery w h e r e Ben was buried, the speech delivered b y his y o u n g wife during his funeral, a p o e m read b y a friend, etc. I was very touched b y the telling, so m u c h so that I wrote to Aaron, w h o then wrote back to m e a n d sent m y message to Ben's wife, who then wrote back to m e . T h e r e has never b e e n a n y face-to-face interaction with the participants of this virtual exchange. This experience p r o m p t e d m e to look m o r e closely at the question of personalisation of t h e public space of the Uni versity. M y data for this p a p e r are therefore restricted to academic Web pages, especially those of the University of Birmingham - pages of teaching staff, post graduate students, a n d some technicians. Academic pages should b e in principle public statements, b u t as Aaron's p a g e demonstrates, academics t e n d to use the public site to include their pri vate sphere in this kind of communication. H e n c e , m y interest in investigating m o r e t h e question of h o w t h e personal intrudes in t h e public world of academia a n d t h e m u l t i m o d a l r e s o u r c e s u s e d b y ' w e b b e r s ' to a c h i e v e their goals. (According to a n e w report from www.nua.com - Internet Surveys b y Category, b yJ a n u a r y 2000,19 million U . S. internet users, or 29 percent of the total number of users, h a v e personalised Web pages, a n d 88 percent of users believe thai personalisation is the best w a y for companies to learn about consumers.) T h e University of Birmingham i n d e x of individual Web pages linked to departments listed, in April 2002, a total of 4 2 5 of which only 2 3 belonged to female academics. This seems surprising since I would expect that many m o r e w o m e n were actively involved in communicating virtually. Unfortunately, this is n o t the case. N o t surprisingly, scientists, especially engineers, produce the* mainritv nf nae-es. T h e arts academics tend to b e m u c h m o r e parsimonious
CAIDAS-COULTHARD
PERSONAL WEB PAGES
279
Before I start examining s o m e examples of Web pages, I w a n t to consider briefly the question of computer mediation as communication a n d the question of recontextualisation.
3 . C o m p u t e r M e d i a t i o n as C o m m u n i c a t i o n , Narration, a n d t h e Construction of Identities Over the past ten years, as R o d n e yJ o n e s (2002) has pointed out, m a n y areas of investigation, especially in sociology, anthropology, linguistics, cultural a n d communication studies b e c a m e interested i n Computer-Mediated C o m m u n i cation ( C M C ) . T h e internet a n d the World W i d e Web are p r o d u c i n g n e w forms of interaction, b o t h in the business a n d personal spheres as well as new forms of social organisation. N e w discursive practices h a v e therefore b e e n developed. J o n e s points out that different professionals like linguists a n d communication scholars concentrate o n linguistic features (registers and genres), sociologists a n d anthropologists are interested in the d e v e l o p m e n t of online ' c o m m u n i t i e s and cultures', while psychologists focus o n the cognitive aspects of computermediated communication. J o n e s continues (2002: 3) that although there are many perspectives o n C M C , the kinds of data that are gathered to answer these varied questions are remarkably limited. Nearly all of the studies look for answers to their questions in the w o r d s typed o n the screen. D a t a typically take the form of d o w n l o a d e d 'logs' of computer chat sessions or corpora of emails or U s e n e t postings, a n d analysis is generally confined to those words. Interactions are assumed to exist in a kind of virtual v a c u u m , iden tities are hardly ever linked to the lives of the people p r o d u c i n g the words, a n d communities a n d cultures are generally seen to stop at the screen's edge. The m a i n p r o b l e m with looking at computer-mediated communication in this way, h e suggests, is that it ignores the importance of online behaviour as well as the importance of semiotic meanings. Another p r o b l e m with such studies, J o n e s (2002: 3) suggests, is that they tend to focus o n the mediational means (Scollon 1998) 'at the expense of looking at the kinds of actions that are taken, leading t h e m into a kind of technological determinism in which actions are interpreted only as "effects" of the media'. For mediational discourse analysts (Wertsch 1991; Scollon 1998, 2 0 0 1 ; Scollon & Scollon 2003a, 2003b), the m a i n focus of interest is o n the complex a n d indirect connections between discourse a n d action - they propose the notion that so cial identities a n d practices are not only realised in texts since w h a t goes o n in social interaction is m u c h m o r e complex: A mediated action is defined as a social artinn
.—•'-T
"
280
DISCOURSE STUDIES
mediated actions, it being definitional that 'social' means socially mediated. T h e principal mediational m e a n s (or cultural tool) of interest is language or discourse, but the concept includes all objects in the material world including other social actors. Within M D A [Mediated Discourse Analysis] there is n o action (agency) without some mediational m e a n s (i.e., the semiotic/material m e a n s of communicating the action) a n d there is n o mediational m e a n s without a social actor (agency). (Scollon & Scollon 2 0 0 3 b : http://www. gutenbergdump.net, valid in April 2002 ) 2
If action is ignored in C M C then, researchers are examining culture in the cultural took (Wertsch 1991) instead of examining culture through its relationships with action, agents, a n d tools. T h e p r o b l e m with the concentration purely o n 'mediational m e a n s ' , as J o n e s (2002) demonstrates very well in his research o n Chinese learners of English using ICQj, a chat a n d instant messaging software, is that often what is sought by users oflCQjand other interfaces of this type is n o t 'communication' at all, but rather 'connectivity', 'social presence', or 'play'. M y computer conversations with A a r o n exemplify this point. We were connected somehow. I n order to overcome these limitations, I want to propose, based o n Jones' research, another way of approaching m y W e b pages, o n e which focuses not just on the texts people produce w h e n they are sitting in front of their screens, but o n the multimodal resources used to construct identities which the n e w media n o w m a k e possible. These identities are inevitably connected to the world outside the computer screen. By representing themselves in particular ways, a narrator 'gives off, to use Goffman's terms (1959), personal clues about the social role performances he or she is m a k i n g a n d this representation indexes these invisible psychological or social cultural states. O f this vast brave n e w world, I will concentrate o n the question of how identities are presented through multimodal aspects. These are the questions: • A r e m y webbers constructed as 'characters' or 'idealisations' differing from the actual person b e h i n d the keyboard? (For example, the 2003 case of an American G I w h o , pretending to b e 19 years old, engaged through a chat r o o m with a supposed 19-year-old English girl - h e was 30 years old a n d she was 12 years old. H e was then prosecuted for a criminal offence.) • Alternatively, are they 'hybrid extensions' of their offline identities? For the most part, I want to claim, online identities are not purely reflections of offline identities or c o m p l e t e r e c o n s t r u c t i o n s , b u t recontextualisations. convenient hybrids of b o t h processes whose uses often h a v e as m u c h to do with what is h a p p e n i n g in the offline social life of users as their online social life. 3.1. Webbers' Identities and the Question ofRecontextualisation T h e discourse of personal Web pages is a discourse about social practices which takes place outside the context of that practice a n d within the context of another
CALDAS-COULTHARD
PERSONAL WEB PAGES
281
one - the virtual context. T h e process of including one social practice into another is a recontextualisation, or ...a sequence of communicative activities which m a k e the social practices explicit to a greater or lesser degree. Social practices are things that people d o , with greater or lesser degree of freedom, fixed b y custom or pre scription, or some mixture of these two. (van L e e u w e n 1993: 30) Texts a n d images are representations of given practices, n o t the practices themselves. As soon as one writes or speaks about a n y social practice, o n e is already recontextualising. T h e m o m e n t w e recontextualise, we are transforming and creating other practices. Web pages are recontextualisations that n o t only represent social practices, they also h a v e to explain and legitimate - in other words, they h a v e to m a k e explicit the 'why' of their representations. In the corporate world of today's universities, there is a pressure o n academics to communicate values, to advertise a n d 'sell' themselves. T h e y h a v e to re contextualise their self-identities in order to d o this. Giddens (1991) says that in p o s t m o d e r n societies self-identity b e c o m e s a reflexive project - an effort that we constantly work and reflect on. We p r o d u c e and revise a set of biographical stories - the telling of w h o we are, a n d h o w we came to b e where we are now. Self-identity, then, is n o t a set of traits or observable characteristics. It is a person's o w n reflexive understanding of their biography. Self-identity has continuity - that is, it cannot easily b e completely changed at will but that continuity is only a product of the person's reflexive beliefs about their o w n biography. (Giddens 1991: 53) A person's identity is not to b e found in behaviour, n o r - important though this is - in the reactions of others, but in the capacity to keep a particular narrative going. T h e individual's biography, if she is to maintain regular interaction with others in the day-to-day world, cannot b e wholly fictive. It must continually integrate events which occur in the external world, and sort t h e m into the ongoing 'story' about the self. (Giddens 1991: 54) For Giddens, then, the self is not something we are b o r n with a n d it is not fixed. The self is reflexively m a d e , constructed b y the individual (Gauntlett 2002): W h a t to do, h o w to act, w h o m to b e ? These are focal questions for everyone living in circumstances of late modernity - the ones which o n some level or another, all of us answer, either discursively or through day-to-day social behaviour. (Giddens 1991: 70) However, as L e m k e (http://www-personal.umich.edu/~jaylemke) points out identity construction i« al«"-» « •— — * 1
1
DISCOURSE STUDIES
282
beliefs, a n d values). People participate in activities through interactions and practices, a n d this participation, h e argues, constitutes and shapes a) identity choice, b) identity display, a n d c) identity construction. Identity is thus a positioning of self or other in a system of meaning-relations a n d in a network of material practices, including a system of p o w e r relations..., a n d discourse in general serves to construct identity a n d positioning with respect to this system of meanings. W h a t webbers do w h e n they create their personal pages is to choose a set of self-characteristics that are important for t h e m (life m o m e n t s a n d achievements, relations, places and moments) a n d construe a persona. F r o m the multitude of identities that academic users h a v e available to t h e m , I will discuss next two m a i n k i n d s of identity formation in virtual recontextualisations: narrators and characters. 3.2. Narrators as Extensions of Self and as Characters For the psychologist K o h u t (1971), according to Barry O ' C o n n o r (in www. selfpsychology.com), a sense of 'self is a fundamental aspect of our life experience, a n d integrated with the total ity of our life. It not only encompasses w h o a n d w h a t T a m ' , at critical stages of our growth, b u t in addition it is the totality of the inner workings of our true-self. O n e could say, that within the development experience of the h u m a n , which I would argue lasts our entire lives. T h e r e are certain 'way points' that n e e d to b e arrived at in terms of realizing 'self-potential'. Significant others are a n essential aspect of our self development. This of course c o m m e n c e s at the very earliest stages of our existence with the critical relationship with o u r mother, a n d with this the influence of our father o n b o t h m o t h e r a n d ourselves. This is further complicated in the m o d e r n environment with the diminishment of the nuclear family, a n d t h e re-designation of w h a t comprises a family, with or without societal support. T h e principal description of the 'self, i n ' m y data, follows the p u r e a n d minimal forms of identity representation: • • • • • • • •
I exist. H e r e I am. M y n a m e is so a n d so. I c o m e from such a place. This is m y family. Personal Web Pages T h e s e are m y things. These are m y interests.
••••!•
CALDAS-COULTHARD
PERSONAL WEB PAGES
283
Consider these examples (given the space constraints, I will illustrate with only a few pages):
Welcome to the
Who A m i ? I a m just a n o r m a l (?) British bloke w h o h a p p e n s to b e called R o b Branston, b u t I h a v e nothing to d o with the famous Branston Pickle. N o w I h a v e that sorted out a n d just o n the off chance that the o d d person is continuing to h a v e a look, I will tell the rest of you a bit m o r e a b o u t m e . I a m a 24-year-old Economist, researching for a P h D in Industrial Economics at the University of Birmingham in the U K . I a m approxi mately 6 feet 4V4 inches tall (195cm to m y metric orientated friends), h a v e short b r o w n hair a n d g r e e n / h a z e l coloured eyes. If you really want you can see some pictures of m e , but I w a r n y o u that I a m n o t at m y finest!! At the m o m e n t I a m looking for a j o b , so I h a v e b e e n spending a great deal of m y spare time working o n a page containing a n academic orien tated C V a n d information about P h D . I h a v e also m a d e a m o r e 'city' orientated C V as I am not sure w h a t m y future direction will b e . If you like what you read, w h y n o t send m e a n email? I h a v e also b e e n spending a lot of t i m e updating, e x p a n d i n g a n d e n h a n c i n g the suite of pages I created for L'institute - Institute for Industrial D e v e l o p m e n t Policy a n d the pages for L'institute - Ferrara G r a d u a t e School in Industrial Eco nomics. A s m y friend D o n Fazio says, I've got s o m e w o r k to d o n o w , so we'll continue this chat a bit later .... Webpage 1 How is R o b B r a n s t o n c o n s t r u i n g his ' s e l f ? T h e a n s w e r to t h e q u e s t i o n 'Who am I' in this case, is linked, interdiscursively, with material goods of a given society - you n e e d to b e at least Anglo-Saxon, to recognise the illustration
DISCOURSE STUDIES
284
(Branston pickle) a n d u n d e r s t a n d the semiotic value a n d p u n of the n a m i n g ap proximation. R o b Branston t h e n b e c o m e s just a n o r m a l British bloke w h o h a p p e n s to b e called Branston. T h e representation of the pickle bottle is crucial for determining n o t only the self-identity (normal) but also the national one (British). Some other examples from the data reaffirm p l a c e of origin as an important aspect of their identities. T h e webbers are 'outsiders' a n d n e e d to assert their foreignness: I a m a V e n e z u e l a n plant biologist. I hail from a s e r e n e t o w n Kovilpatti the 'matchless town of matches' in Tamil Nadu. N a m i n g is a very important form of identification a n d webbers tend to m a k e sure their n a m e s are foregrounded, like these examples: According to the Keirsey T e m p e r a m e n t Sorter I've got a E N T P (Extraverted iNtuitive Thinking Perceiving) personality type - this m e a n s I ' m an Inventor - w h o a m I to argue? I've a n unusual s u r n a m e - Pryke a n d so I've set u p a w e b p a g e listing other Prykes o n the net. M y n a m e is A l b e r t o and I ' m Italian. I ' m an Aerospace Engineer a n d for the time being I ' m working as Research Associate at the School of Electronic a n d Electrical Engineering of the University of Birmingham. If y o u are interested in knowing something m o r e about m e (for profes sional reasons only please, I ' m the h a p p y h u s b a n d of a wonderful y o u n g lady...) you can download m y Curriculum Vitae... O n e thing that I w o u l d l i k e to m a k e clear is m y surname, is it Cirre or Cirre-Torres? Well, in Spain we use 2 surnames, the 1st is the surname of your father a n d the 2 n d of your mother. W h e n I c a m e here, I wrote m y 2 surnames linked b y a h y p h e n , so m y 1st surname is Cirre a n d the 2 n d is Torres. For some webbers the use of the language of 'affect' (Martin 2000) is a strategy used to identify their selves: I started writing this page during o n e lonely, homesick September night, b e t w e e n a Volterra integro-differential equation a n d a multi-domain boundary element approximation. W a r m e d only b y a strong Italian coffee, with Elton J o h n filling the dark of m y r o o m , I was d r e a m i n g m y d r e a m s of far planets a n d u n k n o w n realms, the depth of the ocean, the mystical p o w e r of ancient oriental traditions.
CALDAS COULTHARD
PERSONAL WEB PAGES
285
This page is nothing but the pale shadow of m y thoughts of that night, b u t if even only one a m o n g y o u will find in it a reason to look further into the sky a n d into himself, then this is one of the m o s t i m p o r t a n t things I could ever do. This kind of representation, in fact, reaffirms stereotypical constructions of Latin people - emotional, p e r h a p s 'out of control' - the w a r m 'Italian coffee, the homesickness, the darkness of the r o o m ' construe the identity of the lonely student in a cold a n d foreign place. Others identify themselves through h u m o u r generally using self-deprecation: Welcome to Derek's H o m e Page. I ' m D e r e k Carter, a n d at present I ' m a third year P h D student studying in the D e p a r t m e n t of Electronics a n d Electrical Engineering at T h e University of Birmingham. Yes I'm y e t a n o t h e r of t h o s e b o r i n g e n g i n e e r s , a n d yes I do s p e n d most of m y time in front of computer terminals. M y field of research is parallel hetero geneous c o m p u t e r architectures, sounds very impressive doesn't it! This negative appraisal (Martin 2000) is legitimised through a cartoonesque representation of self:
Webpage 2 Kohut (1971) refers to self-object experiences, experiences (usually with other people) that nourish the self a n d which define the experience of the self a n d self-esteRm
286
DISCOURSE STUDIES
Significant others, like family relations, a r e a n essential aspect of o u r selfdevelopment. T h a t is w h y so m a n y family relations are presented in W e b pages the self of the w e b b e r is being represented through self-objects that reinforce her or his o w n performances. Photographs showing the w e b b e r within the family context are therefore extremely c o m m o n in m y examples. Marriage p h o t o s or simple stills of family groups recontextualise t h e w e b b e r as n o t only a m e m b e r of a g r o u p b u t also as belonging to a group. N o t only parents, siblings, partners, b u t even pets are introduced as parts of the self, as the example b e l o w shows:
INFORMATION
ii.
YOU REALLY WANTED TO KNOW ABOUT ME! J
i~
This is us on o u r B I G D a y ! ! Hi.' My name is Cathy Lowe, and up until 24th June 199S, I was known as Cathy Smart. I am 28 Years Old and married to Tony. We live in Birmingham and have done so for the last 5 years, well I say we, Tony has lived in Birmingham all his life, I moved all the way from Solihull (for my sins) in 1992.1 am a very lazy individual, much to my husband's dismay, but I occasionally get off my backside to go Swimming and the odd Step Aerobics Class. 1 enjoy pottering around in the garden, especially since w e had our Garden Pond re-built. I also enjoy reading, once I'm into a good book I'm oblivious to everything around me, even my husband!! We are both also Dog Lovers and up until a two years ago we had a BEAUTIFUL Alsatian called RADAR, unfortunately w e had to have him put to sleep as he was very ill, he was only 8 years old.
Here he is
Webpage 3
CALDAS-COULTHABX)
PERSONAL WEB PAGES
287
Physical appearance and e m b o d i m e n t s e e m to b e one of the other ways webbers recontextualise themselves - all m y examples present either a p h o t o or an illustration of 'self, especially of the 'face'. Laurie, below, shows us a sort of 'passport p h o t o ' of himself, b u t h e self-evaluates his o w n appearance w h e n h e says, revealingly, at the end of the page: 'By the way, I h a v e h a d a haircut'. Laurie's Homepage
lit
^ ^ ^ ^ ^
...
.^^^^H>
ii»m«te«i»».-
Tntr"rliTtf*tHtttirlTtiT*'i^-Tiir| in nuniiiiHwi tm nni)imrnii|iimiifc«liiiTlMllllBim iltniUlJT) 'UMll'l At^MefairfTnuiHIII 1.1. I HO.rt•-!««« ml HII u m i ^ p . H t m . J
Webpage 4 W h a t makes the 'self a supposedly 'real self in such situations is n o t necessarily contingent o n the realm of 'facts' or 'personal information', but rather on the opportunities for users to create the 'real selves' they thought they 'should b e ' or 'could b e ' , anchored in the semiotic representations chosen, either verbal or non-verbal. W h a t is interesting about participants' conception of their 'online' selves is that the complexity and multiplicity of these selves are seen as features that actually m a k e t h e m ' m o r e real' rather t h a n 'less real'. Computer-mediated inter action for these participants, because of the choices m a d e , seems to b e something which did n o t fragment their sense of their 'real selves', b u t e x p a n d e d it. We h a v e to see then, as J o n e s suggests, cyberspace, not as the site of 'cyber society' (Jones 2002), b u t regard it m o r e as w h a t Scollon (1998) calls a site of engagement, a 'field' u p o n which multiple actions, identities, a n d communities can b e enacted. T h e second identity type of construction I want to concentrate o n is character. 3.3. Character Many of the concepts developed in narrative t h e o r y are n o t specific to lan guage. I want to point out h e r e that the w a y narrators construct characters (and consequently their life stories) is b y the choice of h o w they recontextualise themselves, in particular spaces, in particular settings (like in a literary text). The spaces are created through the representation of locations a n d objects. T h e dimension of the 'activity' being engaged in (what participants are doing w h e n recontextualised) is also very important - some characters are recontextualised as moving in spaces, others as static and in n a r H m W
288
DISCOURSE STUDIES
By using especially t h e m u l t i m o d a l resource of p h o t o g r a p h s , narrators construct through their choices of h o w to present themselves, a particular version of the self, creating scenarios, spaces, a n d attributes. T h e visual landscape is part of this characterisation. By interpreting the w a y the narrators construe them selves via image, w e recreate a narrative of their lives - what they d o , where they d o it, w h o they position themselves as in the social practices a r o u n d them. Photos are m e m o r a b l e - they represent a n e a t slice of time, n o t a flow. Susan Sontag (1999) suggests that photos furnish evidence, a m o r e accurate relation to visible reality t h a n other mimetic objects. C o l e m a n (1998: 57) adds that Photographs h o w e v e r d o n o t show h o w things look. W h a t a p h o t o g r a p h shows us is h o w a particular thing or person could b e seen or could b e m a d e to look at a specific m o m e n t in a specific context, b y a specific p h o t o g r a p h e r employing specific tools. Photos are the most plausible a n d straightforward representations of identity part of o u r identity is a n awareness of ourselves, as w e would want others to see us. T h e very artificiality of the pose (smiling or not, looking at t h e camera or not) demonstrates the g a p b e t w e e n t h e w a y w e are a n d the w a y w e would like to appear as Barthes (2000) suggests in his Camera Lucida. Certainly, the smiling face is an attempt n o t only to produce a positive evaluation from gazers, b u t also to interact with them. H o w e v e r , albeit being faithful to t h e w a y w e 'look', w e have to remember that, as Kress & v a n L e e u w e n (2001) point out, in s o m e domains certain visuals are 'truer' than others. T h e authors call this the coding orientation a n d this has to do with modality. Modality, here, in t h e linguistic a n d semiotic sense, as Scollon & Scollon (2003b, www.gutenbergdump.net) explain, is the word that linguists a n d logicians use to talk about alternate realities. 'Irrealis' (not real) contrasts with 'realis' (real). Ought contrasts with is, was, will be. Realis is something 'real', 'definite', 'perceived b y the person speaking'. Irrealis is 'imagined', 'unreal', 'indefinite', ' u n k n o w n ' or 'unknowable'. T h e lin guistic a n d logical point of the distinction between realis a n d irrealis is that there is no external truth or reality that is being referenced. It is the m o d e of representation which is telling us w h a t t h e p r o d u c e r of the language or image wants us to believe. Images can m a k e the same distinctions in modality. T h e y can represent a person, a thing, or a space according to a scale of 'it might look like this' to 'it looks like this'. This is d o n e through choices of colour, h u e , distance, focus, a n d vectors. However, a p h o t o g r a p h tends to m a k e us believe that what w e see is the reality in this sense. Visual resources tend to b e m o r e 'realis' orientated than texts, which are m o r e easily 'irrealis' orientated, as Scollon & Scollon suggest (2003b. www.gutenbergdump.net). M y argument here, w h e n examining the photos used b y webbers, is that in fact, although these photos p u r p o r t to b e 'realis' representations of self, whai
CALDAS-COULTHARD
PERSONAL WEB PAGES
289
authors d o is to construct, through choices in modality, a 'characterisation' of the self in terms of the ways their identities are being displayed. T h e interesting point a b o u t the p h o t o g r a p h s b e l o w is that although these W e b owners are academics a n d the photos are placed in academic W e b pages, the visual/virtual landscapes chosen in a sense construe specific characters a n d the discourses that surround t h e m - the discourse of action, the discourse of adventure. T h e
Me, on my way to work! Webpage 5
Mark Erickson's Grateful Dead Page
Mark's Tape Trade List
Webpage 6
DISCOURSE STUDIES
290
modalities chosen, especially colour ( i n the original pages, blues, reds, greens), p r o d u c e 'irrealis' effects. W h a t we have h e r e are therefore projections of ideal selves: the sportsman, the association with a particular musician, are facets of hybrid 'private' identities that also h a p p e n to b e public (and academic). T h e examples below illustrate these points:
4. R e c o n t e x t u a l i s a t i o n of i n t e r a c t i o n I want briefly to look n o w at recontextualisations of verbal interactions. A c c o r d i n g to J o n e s (2002), the vast majority of p e o p l e w h o engage in c o m p u t e r - m e d i a t e d c o m m u n i c a t i o n r e g a r d it as a n e x t e n s i o n ( M c L u h a n 1994) of their 'real-life' social interactions rather than as separate from them. Therefore,J o n e s concludes, the effects of C M C are to 'ground t h e m m o r e firmly within their existing material communities a n d circumstances' (Jones 2002: 5). Several of the participants in his research, in fact, were quite a d a m a n t in reject ing the traditional d i c h o t o m y of 'actual' a n d 'virtual' reality, insisting that computer-mediated communication is as 'real' as anything else is ('as real as a telephone call'). Contrary to m y initial belief that W e b page messages were sent to u n k n o w n addressees, Web narrators in fact try to facilitate the exchange of information a m o n g colleagues a n d peers a n d the sense of social 'connectivity'. Narrators are therefore agents acting u p o n the world - they send messages to particular (not well-defined) people, h o p i n g to influence t h e m in some w a y or other. W h e n asked about the recipients of their Web pages, o n e of the writers I contacted replied: I don't h a v e a clear idea, b u t I w o u l d include, in n o significant order: colleagues in the school, research students a n d undergraduates in the school; colleagues i n o t h e r universities a r o u n d the world; potential collaborators in m a n y places, several types, in several disciplines; potential research students, people w h o m a y b e able to learn something from m y research or contribute to it b y c o m m e n t i n g or criticising; journalists w h o contact m e to find out about artificial intelligence or cognitive science; school teachers; school leavers exploring the school's w e b site and wanting to k n o w what's here; p e o p l e a r o u n d the university interested in man agement issues; people w h o h a v e h e a r d m e give presentations a n d would like to see the slides (usually extended versions); people w h o m a y b e interested in a b o o k I wrote in 1978, which went out of print a n d is now free of charge online.... This is the testimony of another w e b b e r : At first there was n o audience, actually, b u t o n e h a s appeared for many of the things there...[T]he idea was to m a k e a brief visual essay using photos we h a d to organize our o w n reflections about our pasts a n d see if
CALDAS-COULTHARD
PERSONAL WEB PAGES
291
that would help us find connections...I h a v e told a few students about t h e m (the different pages) as I try to introduce myself as b r o a d l y as pos sible to them thinking there's n o w a y they could understand the academic slices of m y life w / o knowing something of the rest. But I d o n ' t like to r a m b l e o n in a seminar a b o u t h o w I wrangled horses as a teenager, so if it's o n the net I just say, 'go take a look' a n d get o n with the subject. So I guess you could say that m u c h of it, m a y b e most of it is this: I m a k e litde essays, highly personal, about things I or we think about or h a v e d o n e , put t h e m on the net so our o w n kids a n d family a n d a few friends can see them. ...what's o n m y websites is the stuff that isn't in m y formal academic vita • b u t which is essential to understand what's in the vita. Incidentally, these pages are n o w found b y search engines and visited b y w h o knows w h o . I average about 60,000 visits a year. This shows that the message is addressed to targets, a n d n o t virtual beings, out there in cyberspace. A n interesting aspect of these interactions however, is their spoken conversational features. Most of the pages examined here have greetings, openings, closings, apologies, a n d the address to a 'you'. Consider for example: • H i , I a m (name).... H o w are you m y friends all over the world? • Hello. • Welcome! • H i , thanks for visiting this page! • W a n t to k n o w m o r e about m e ? Well, then...come a n d browse t h r o u g h then.... • H i , h a v e you found something? • Welcome everybody. I ' m sorry it's taken such a long time, b u t here at last are the beginnings of m y w e b page, with pictures showing highlights of the Big Day. • T h a n k y o u for stumbling u p o n m y h u m b l e website. I h o p e y o u find something of interest while y o u ' r e here. • D o n o t forget to c o m e back. See you n e x t time. • H o p e you're happy. Be nice to animals. A n d although this site hasn't got m u c h , you never know. • That's all, folks. In Late Modernity, the n e e d to communicate/talk (although through virtual media) continues to b e p a r t of people's lives, even if p e o p l e are stuck in their houses in front of their computers. This is not very different from old times, w h e n people use to visit each other and chat on verandas, at least in w a r m countries like m y own. O n e of the o p e n i n g messages of Paul M c l l v e n n y ' s p a g e (http://ntpaul. sprog.auc.dk) summarises extremely well the i n t e r d e p e n d e n c y between the interactants in this virtual communication:
I
292
DISCOURSE STUDIES
You b r o w s e , therefore I a m T h r o u g h pseudo-interactions webbers seem to continue traditions of diary or biography genres where p e o p l e tell 'whoever is out there' about their personal experiences, their relationships. T h e informal aspect seems to diminish the distance between writer a n d reader - you are there, I d o n o t k n o w w h o y o u are. But somehow, like m e a n d A a r o n , you browse and I exist. T h a t is why, I think, there are so m a n y 'hi, hellos, goodbyes', o p e n i n g a n d closings in virtual inter actions. T h e material page is the mediational m e a n s for social action.
5. Final R e m a r k s T h e University of Birmingham has n o w m a n a g e d to standardise Web pages to m a k e t h e m compatible with the corporate image of the University. A corporate t e a m p r o d u c e d a set of guidelines so that webbers should maintain the same institutional 'image'. T h e photos should look the same, nothing that is 'personal' should b e inserted into the pages. This is the discourse of the institution trying once again to standardise, paradoxically, presentations that h a v e to do with people's lives and identities. By doing this, the institution tries once m o r e to exert control over its m e m b e r s . H o w e v e r , from the examples discussed above, it seems that n e w creative formats resist corporate discourses. Self-presentation has to do with what we are, w h o m we deal with, a n d the choices we m a k e in life. People, as L e m k e (2002) suggests, choose the ways they want to display themselves a n d construct their identities. Identity construction is a complex p h e n o m e n o n a n d people project in public spaces, idealisations of what t h e y 'think' they are. T h e multitude of formats, photographs, and colour choice point to the hybrid formation of identity. T h e online construction is an extension of the offline persona. A n d institutions cannot control these representations. T h e r e will always b e a w a y out of the institutionalised corporate 'image'. To conclude, I want to emphasise the importance of the study of multimodality in the n e w communicative genres. Web pages combine language, image, a n d graphics into a single integrated text's structure. T h e y are therefore: • i n t e r d i s c u r s i v e : 'positioned within multiple, o v e r l a p p i n g a n d even conflicting discourses' (Scollon & Scollon 2 0 0 3 b , www.gutenbergdump. net) - the discourses of the family, the institution, a n d emotion; • intertextual: they h a v e links with other mass media, especially with clas sical a n d m o d e r n narratives of all sorts - books, films, comic strips, paint ings, a m o n g others; • dialogical: a m e a n s through which different kinds of 'conversations' can b e established. Web pages are also sites of engagement. I n s o m e ways, as D e b o r a h Tannen (2002, cited in J o n e s 2002: 9) points out, 'Asking linguists to p a y less attention to texts
"CALDAS-COULTHARD
PERSONAL WEB PAGES
293
is like asking astronomers to stop looking up so much'. It is, however, only through breaking down the traditional hierarchical separation of different modes of representation and focusing more onto the social actions, social identities, and socialpractices that we will achieve a true understanding of the society we live in and virtuality is here to stay. In the virtual world w e live in, there is no more looking back. i.
Notes 1. All the Web pages used in this chapter were in colour. Unfortunately, the illustrations in this book can only be reproduced in black and white. 2. One of the essential features of Web pages is their ephemerality. The authors alter and delete items and even whole pages, as you will find if you attempt to follow this reference, Scollon & Scollon (2003b, www.gutenbergdump.net). This page no longer exists.
References Barthes, Roland. 2000. 'Camera Lucida: Reflections on Photography' in Thomas (2000): 54-62. Berger, John. 1972. Ways ofSeeing. London: BBC and Penguin Books. Bernstein, Basil. 1981. 'Codes, Modalities and the Process of Cultural Reproductions: A Model' in Language and Society 10: 327-63. Caldas-Coulthard, Carmen Rosa. 2002. 'Cross-cultural Representation of "Otherness" in Media Discourse' in Weiss & Wodak (2002): 272-296. Caldas-Coulthard, Carmen Rosa, and Theo van Leeuwen. 2001. 'Baby's First Toys and the Discursive Construction of Childhood' in Wodak, Ruth (ed.) Critical Discourse Analysis in Post Modern Societies. Special issue of Folia Linguistica 35(1-2): 157-182. . 2002. 'Stunning, Shimmering, Iridescent: Toys as the Representation of Gendered Social Actors' in Litosseleti, Lia, andJane Sunderland (eds) Gender Identity and Discourse Analysis. Amsterdam: John Benjamins: 91-110. Caldas-Coulthard, Carmen Rosa, and Malcolm Coulthard (eds). 1996. Texts and Practices: Readings in Critical Discourse Analysis. London: Routledge. Coleman, Allan D. 1998. Depth of Field: Essays on Photography, Mass Media and Lens Culture. Albuquerque: University of New Mexico Press. Evans, Jessica, and Stuart Hall (eds). 1999. Visual Culture: The Reader. London: Sage. Gaundett, David. 2002. Media, Gender and Identity: An Introduction. London: Roudedge. Extracts available at www.theory.org.uk. Giddens, Anthony. 1991. Modernity and Self-Identity. Cambridge: Polity Press. Goffman, Erving. 1959. The Presentation ofthe Selfin Everyday Life. New York; Doubleday. Hall, Stuart (ed.). 1997. Representation: Cultural Representations and Signifying Practices. London: Sage and The Open University. Halliday, M. A. K. 1978. Language as Social Semiotic. London: Arnold. Jones, Rodney H. 2001. 'Beyond the Screen - A Participatory Study of Computer Mediated Communication among Hong Kong Youth'. Paper presented at the annual meeting of the American Anthropological Association (Nov. 28-Dec. 2, 2001).
294
DISCOURSE STUDIES
Jones, Rodney H. 2002. 'The Problem of Context in Computer Mediated Communi cation'. Paper presented at the Georgetown Roundtable on Language and Linguistics (Georgetown University, March 7-9, 2002). Kohut, Heinz. 1959. 'Introspection, Empathy and Psychoanalysis: An Examination of the Relationship Between Modes of Observation and Theory' inJournal of the American Psychoanalytic Association 7: 459-83. . 1971. The Analysis of the Self New York: International Universities Press. Kress, Gunther, and Theo van Leeuwen. 1996. Reading Images; The Grammar of Visual Design. London: Roudedge. . 2001. Multimodal Discourse: The Modes and Media of Contemporary Communication. London: Arnold. Lemke, Jay L. 2002. http://academic.brooklyn.cuny.edu/education/jlemke/papers/ polcarthtm. . 2003. 'Texts and Discourses in the Technologies of Social Organization' in Weiss & Wodak (2002): 130-49. Martin, James. 2000. 'Beyond Exchange: Appraisal Systems in English' in Hunston, Susan, and Geoff Thompson (eds) Evaluation in Text. Oxford: Oxford University Press: 142-75. McLuhan, Marshall. 1994. Understanding Media: The Extensions of Man. Cambridge, MA: The MIT Press. Scollon, Ron. 1998. Mediated Discourse as Social Interaction: A Study of News Discourse. London: Longman . 2001. Mediated Discourse: The Nexus ofPractice. London: Roudedge. Scollon, Ron, and Suzie Wong Scollon. 2003a. Discourses in Place: Language in the Material World. London: Routledge. . 2003b. Globalization: The Multimodal Shaping of Public Discourse. http://www. gutenbergdump.net. Sontag, Susan. 1999. 'The Image-World' in Evans & Hall (1999): 80-94. . 2000. 'In Plato's Cave' in Thomas (2000): 40-53. Tannen, Deborah. 2002. Personal conversation with Rodney Jones. March 9. Thomas, Julia (ed.). 2000. Reading Images. Basingstoke: Palgrave. van Leeuwen, Theo. 1993. 'Recontextualisation of Social Practice'. Unpublished manuscript. Weiss, Gilbert, and Ruth Wodak (eds). 2002. Critical Discourse Analysis: Theory andlnterdisciplinarity. Basingstoke: Palgrave. Wertsch, James V. 1991. Voices of the Mind. A Sociocultural Approach to Mediated Action. Cambridge, MA: Harvard University Press. Woodward, Kathryn. 1997. 'Identity and Difference' in Woodward, Kathryn (ed.) Culture, Media and Identities. London: Sage: 10-22.
12 Role and Position of Scientific Voices: Reported Speech in the Media Helena Calsamiglia andXJarrnen Lopez Ferrero
1. Introduction
I
n r e c e n t years the presentation of scientific knowledge, u n d e r s t o o d as scientists having regular interaction - direct or mediated - with the general public, has h a d a significant place in the media, a p p a r e n d y seeking to fill the traditional gap that exists b e t w e e n scientific c o m m u n i t y a n d p e o p l e in general. As M o i r a n d (1997) states, the press h a s b e c o m e , in specific topics of interest, the 'meeting point' for scientists with the average citizen. I n fact, scien tists, as a very specialized community of scholars a n d researchers, h a v e through out the 20th century b e e n secluded in universities a n d research centres, enclosed as it were b e h i n d the walls of professional worlds, a n d lacking contact with the needs and requirements of the general public. T h e development of d e m o c r a c y and communication, in t a n d e m with the increasing consciousness of the im portance of science for the distribution of political p o w e r a n d a better quality of life through technology, h a v e finally b r o k e n d o w n these walls. This has b e e n achieved through the generating of journalistic practices w h e r e scientists a n d science are represented and called u p o n to contribute to the formation of public opinion. This has b e e n d o n e n o t without distrust and misgiving o n the p a r t of scientists a n d experts b u t it can b e said that n o w a d a y s the n e e d for scientific culture is claimed as a social right. Nevertheless, bridging the two worlds, the world of the specialist a n d the world of lay people, remains a p r o b l e m , n o t only from the cognitive perspective but also from the perspective of the representation of science and scientists outside their own domain. Developments in discourse analysis h a v e involved approaches to the use of language in different contexts. T h e o n e that deserves our attention is precisely Source:
Discourse Studies vol 5. ™ o onnn —
—»
296
DISCOURSE STUDIES
the way in which progress in scientific knowledge reaches the general public in the age of information and communication. Scientific/academic use of language has b e e n recognized as a specific register with its o w n n o r m s , patterns a n d style, affecting n o t only terminology b u t ways of presentation a n d reasoning through particular discourse genres and procedures. N o w o n d e r , then, that the w a y a unit or a piece of knowledge is selected a n d transformed to b e presented and explained to non-experts can b e a very rich topic of research, since it d e m a n d s rigorous recontextualization conveyed through discourse procedures affecting b o t h global a n d local levels of text. I n this article we focus attention o n the w a y in which scientific voices are referred to in the context of a particular case, involving a serious health a n d diet p r o b l e m t h a t h a s b e e n t h r e a t e n i n g t h e basic n e e d s of t h e p o p u l a t i o n in industrialized countries in Europe. This is the case of the Bovine Spongiform E n c e p h a l o p a t h y (BSE) outbreak - c o m m o n l y k n o w n as ' m a d cow disease' which in the past 15 years has affected different countries, creating a state of public alarm and desorienting public opinion owing to the widespread consump tion of m e a t a m o n g the population a n d the subsequent fear of transmission of the disease to h u m a n s . We pursue our search at a time w h e n , after a p e r i o d of d o u b t a n d insecurity, the possibility of transmission to h u m a n s was openly stated in a session of the British Parliament, in M a r c h 1996. It is from this date o n that the ' m a d cows' affair explodes in t h e media, causing all kinds of diplomatic, commercial, political and social reactions. It has b e e n a fin-de-sieclehealth crisis that has deeply worried a n d disturbed governments d u e to its consequences affecting a variety of social groups and agencies. Its importance can b e gauged b y the immediate attention which it received from sociologists a n d other analysts interested in this critical situation (see, for example, D a b a n a n d Grau, 1996; Porta, 1996; Donadieu, 1997; Jasanoff, 1997; Zander, 1997; Ratzan, 1998). N e w episodes of the crisis h a v e b e e n again unfolding (2000-1). T h e disease's genesis has b e e n re-explained in the press in detail, especially during the last m o n t h s of 2000, w h e n n e w ' m a d cow' cases spread across several countries in E u r o p e , mainly in France, Germany, Portugal a n d Spain. So the debate a n d the doubts still remain, with politicians a n d scientists in a state of alert as to how to solve a p r o b l e m that has b e c o m e of great relevance d u e to the complex texture of its consequences. T h e aim of this article is to analyse the role of scientists as social agents people with something to say a n d d o in the case u n d e r study - as they are repre sented in the Spanish press. O u r analysis takes journalism into account as the frame of communication (at the macro-level), a n active d o m a i n of writing for a wide general public, with genres a n d conventions of its own; and, as the specific focus of analysis (the micro-level), citation as a form of poliphony, a concept taken from the theory of enunciation (Bajtin, 1979; see Voloshinov, 1929; Benveniste, 1971/1974; Ducrot, 1984). Within this framework, w e identify reported speech not only through the way that g r a m m a r differentiates between direct and indirect style, b u t widening this view to include what in journalistic practices constitutes a continuum of reference to 'other' voices which pragmatic perspective has taken
CALSAMIGLIA A N D LOPEZ FERRERO
ROLE AND POSITION OF SCIENTIFIC VOICES
297
into account (Reyes, 1993/1994) a n d in relation to journalistic practices (Waugh, 1995; Meridez, 1999a, 1999b). W e consider b o t h levels as a whole from a critical perspective that favours the view that citation m e a n s m a n a g i n g the w o r d s of others to convey a n d serve the p u r p o s e of the writer, giving a slant to what is said. This contrasts with other views, such as those presented in journalism training, arguing that citation n o t only makes the writer's discourse m o r e objective a n d credible, b u t frees h i m / h e r from any responsibility.
2. B a c k g r o u n d Information: Brief H i s t o r y o f t h e B S E C a s e T h e aim of our study is scientists' reported speech in a corpus of news which reports the popularly known ' m a d cow' disease, technically called Bovine Spongiform Encephalopathy. Although the p e r i o d w e analyse in the present article is the first eruption of this affair in the press in 1996, we emphasize the relevance of this scientific/social topic as a recurrent o n e in the media. But there are scientific topics that enjoy punctual or acute interest ('patron de interes puntual o agudo', Informe Quiral, 1999: 74), m e a n i n g that they generate a great a m o u n t of news in a brief period of time. Such is the case of the beginning of the ' m a d cows' affair, generating 75 n e w s items in the period of M a r c h - A p r i l 1996 in the most widely read general Spanish newspapers. I n order to p u t the n e w s in context, w e n e e d to refer to s o m e key scientific data o n Bovine Spongiform Encephalopathy as a degenerative pathology of the brain. T h e origins g o b a c k to an infection k n o w n as scrapie, identified in 1732 in sheep a n d goats. I n 1920, the h u m a n disease Creutzfeldt-Jakob (CJD) is identified, with intriguing similarities to scrapie. In the 1970s, Stanley Prusiner, the 1997 Nobel Prize winner in Medicine, identified the agent which causes the disease. It is called prion, a special variant of a protein (healthy or diseased) found in all brains. T h e altered protein (prion) causes changes in the n o r m a l proteins, a n d it triggers a chain reaction which finally leads to the brain disease. I n 1985, in Great Britain, Bovine Spongiform E n c e p h a l o p a t h y (BSE) was described as a disease closely related to scrapie. T h r e e years later, BSE was connected to the ingestion of animal feeds derived from m e a t a n d b o n e s of infected sheep. Political data are also relevant to understand the case. We will note the follow ing events: in 1986 the British g o v e r n m e n t officially declares BSE's existence. From then on, various action is taken b y the government: organic animal feeds are forbidden, a n expert report on B S E is d e m a n d e d (finally published in 1989), animals are slaughtered. But people are persuaded b y the g o v e r n m e n t agencies not to worry about eating meat. T h e r e is widespread conviction that there is n o connection at all between h u m a n and animal brain pathologies. At the same time, powerful economic interests e m e r g e in the face of a n uncertain b u t openly pre dicted possible crisis of enormous consequences. T h e British Government officials of different ministries are responsible for indecision a n d lack of transparency. In 1990, a Committee appointed b y the British Government is set u p to monitor the incidence of the disease. It consists of experts in contact with a t e a m of medical researchers in Edinburgh. O n 20 March 1996 the British Health s-*
298
DISCOURSE STUDIES
Parliament that a variant of the h u m a n encephalopathy (Creutzfeldt-Jakob) could b e related to BSE. T h e evidence is not clear but generates widespread bitter conflict b e t w e e n institutions, g o v e r n m e n t , c o m m e r c i a l organizations a n d t h e food industry, causing diplomatic relations to b e b r o k e n off a n d alarming the average citizen. T h e focus of the crisis is Great Britain, but other countries in E u r o p e are also u n d e r the threat of the disease a n d its transmission between species. This conflict is taken u p b y the m e d i a w h e r e scientific information is m i x e d with political statements, commercial reactions a n d citizens d e m a n d i n g clarification. A s regards scientific progress in brain pathologies, in spring 1996 there were some significant scientific advances which reached the mass media. But the scientific c o m m u n i t y was going at its o w n p a c e in the midst of the press storm. I n fact, research on such a rare disease as Creutzfeldt-Jakob h a d b e e n very sparse, only receiving a boost after the B S E outbreak in cattle. Nevertheless, 14 years after the b o v i n e disease was first identified, ultimately causing 92 h u m a n victims - m a i n l y in G r e a t Britain - lack of scientific certainty a n d political dithering u n d e r m i n e d E u r o p e a n consumers' confidence. D u r i n g the period studied in this article, M a r c h a n d April 1996, two scientific articles w e r e published; these were n o t generated b y the s u d d e n crisis and consequendy they w e r e n o t intended to provide a clear-cut response to people's concerns. T h e scientific information p r o v i d e d b y these two articles was taken u p in the news report in our corpus. T h e first article was a medical report pro d u c e d b y the E d i n b u r g h t e a m : 'A n e w variant of Creutzfeldt-Jakob disease in U K ' b y R.G. Will, J.W. Ironside et al., published in the medical magazine The Lanceton 6 April 1996. It contains n o clear proof b u t considers that exposure to B S E is the most plausible cause for transmission of the disease across species. T h e second article referred to the ongoing resarch o n the 'prion' protein, after Prusiner's findings in the 1970s. T h e article a p p e a r e d o n 2 5 April 1996, in the scientific magazine Nature, written b y D . C . Krakauer, M . Pagel, T.R.F. Southwood a n d P.M. de A. Zanott, entitled 'Phylogenesis of p r i o n protein'. T h e first article was leaked to the press b y the governmental committee before publication in The Lancet. T h e second was the direct source of information appearing in the press. T h e press presented the facts summarized h e r e , their interest sparked off after the British parliamentary session in M a r c h 1996, m e n t i o n e d above. T h e m e d i a brought into their discourse reported speech from different sources which supply reference points for the interpretation of the facts. This use of reported speech is the object of our study.
3 . B r o a d C o n t e x t of t h e Study: D i s c o u r s e a n d M e d i a T h e a p p r o a c h that w e take considers reported speech in the context in which it appears. T h e case u n d e r study has newspapers as its context or domain. Thus, in order to control the object of our study w e can locate a n d define it in the light of what v a n Dijk (1988) a n d C h a r a u d e a u (1997) h a v e p r o p o s e d concerning different aspects of informative discourse in the press. To begin with, Charaudeau provides a general semiotic approach from the perspective of the communicative contract, as represented in Figure 1.
CALSAMIGLIA A N D L6PEZ FERRERO
i
ROLE AND POSITION OF SCIENTIFIC VOICES
299
COMMUNICATIVE CONTRACT — , Space for strategies
MEDIATIC < INSTANCE
EVENT
>
I NEWS | CONSTRUCTED EVENT
<
RECEPTION INSTANCE
>
INTERPRETED EVENT
Process of transformation
Process of transaction Figure 1: General semiotic approach (Charaudeau, 1997: 72) T h e event in the m e d i a goes through a process of transformation that results in the construction of the event b y its representation in words. So the process of transaction allows for strategies within the frame of the communicative contract. T h e discourse m o d e s of a m e d i a event are viewed as shown in Figure 2. Given the discourse m o d e s of the m e d i a event, schematized b y Charaudeau, it is clear that the object of our study is the first phase, that of the reported event, w h e r e w h a t is being said predominates over reference to fact. If we rely o n Van Dijk's schema for the superstructure of news in the press (Figure 3) we see that if the m e d i a story is m a d e u p of situation a n d comments, the situation in the ' m a d cows' case is a n oral episode that triggers a variety of c o m m e n t s (oral reactions) from different sources. As the first focus of attention is a statement b y the British Health Minister a n d the subsequent debate in the iMEDIAEVENTl iREPORTEDEVENTl 1 COMMENTED EVENT| |EXTENDEDEVENTj
REPORTED FACT
REPORTED SPEECH
OPINION INTERVIEW DEBATES LEADERS
Reaction
statement
exposition
explanation
F i m i r e 2 : ni^rnnr^n
Words ~r - —
1
'
actions
DISCOURSE STUDIES
300 N e w s Report
Summary
Headline
Lead
Situation
Episode Expectations M a i n Events
Consequences
Context
Circumstances
Evaluations
History
Previous events
Figure 3: Hypothetical structure of a news schema (Van Dijk, 1988: 55) H o u s e of C o m m o n s , it is a case w h e r e words b e c o m e the core of information. This was the origin of the sensational treatment of the affair in the press. At the microlevel, journalistic practices h a v e established conventions on procedures of citation because of its constant presence in the news. I n style books we consistendy find instructions on the p r o p e r way to present citations. As an example: All alien words, statements or opinions transcribed literally in a n informa tive text h a v e to b e signaled with quotation marks, without changing the sense or intention with which they were uttered. . . . T h e author will always b e identified, through the p r o p e r n a m e or a n y other characteristic that supports, justifies or gives authority to h i s / h e r words. T h e identifi cation will b e d o n e at least once, in the begining, in the middle or at the e n d of the transcribed sentence. (Abe, 1993: 5 2 ; authors' translation) Prescriptive guidelines for journalists stress the use of citation to give authority a n d legitimacy to what is said through the strict (literal) quotation of words spoken b y others. It is o n e of the procedures b y which to obtain credibility for the facts u n d e r c o m m e n t as well as o n e of the m e a n s of certifying knowledge of what is being said. Literal quotation is intended to b e objective knowledge, far r e m o v e d from the subjectivity of the journalist.
4. T h e o r e t i c a l F r a m e w o r k Different studies h a v e b e e n carried out in the past d e c a d e o n specific aspects of t*. o - . r a r i o t v nf nprsnectives. o n v a r i o u s d o m a i n s and - -i - J nm
CALSAMIGLIA A N D LOPEZ FERRERO
ROLE AND POSITION OF SCIENTIFIC VOICES
301
languages: see, in particular, the compilation of articles b y Lucy (1993) a n d J a n s s e n a n d Van der Wurff (1996), including Haverkate's o n reported speech in Spanish literary works. We h a v e based our approach o n the conception of the enonciation theory as initiated b y the Soviet scholar Bajtin (1979), formulated b y the French linguist Benveniste (1971/1974), developed b y different representative authors such as Kerbrat-Orecchioni (1980), Authier-Revuz (1982), Ducrot (1984), M a i n g u e n e a u (1987, 1991, 1993) a n d applied to journalism b y Waugh (1995) and M e n d e z (1999a, 1999b). I n its first formulation, the theory puts forward the existence of a Speaker that actualizes the virtual system of language (enonciateur) in relation to a H e a r e r (enonciataire), two axes from which choices of textual items are m a d e a n d from which a particular variety of text (enonce(s)) is derived. T h e two main characters of the communicative act (enonciateur, enonciataire) are present in the text t h r o u g h different kinds of lexical a n d deictic traces whose interpret ation gives the key for a degree of personal and interpersonal implication. Most of the studies in this framework h a v e focused o n the emergence of subject (speaker/ writer) in texts, m a k i n g its basic assumption o n the fact that the subject is built through the process of enonciation. Ducrot, following the line of the fundamental heteroglossy of language as it is claimed b y Bajtin, breaks the assumption of a oneto-one relationship between Speaker a n d text. His hypothesis is that the Speaker can b e unfolded in several voices {polyphony) a n d postulates m a k i n g a distinction between (1) the text author or empirical subject (jfire-discursive identity); (2) the construction of the discourse subject (& voice created in text); a n d (3) the triggering of other voices which contribute to the perspectives adopted b y (2) to build a particular position. I n every situation of enonciation the author creates a voice taking a position, choosing a role according to the possible addressees. This position and role can b e enriched a n d m a d e precise b y the inclusion of different other voices that (2) adopts to confirm or contrast the position taken. T h e discourse analyst is interested in (2) a n d (3), that is, in the particular way the author chooses to b e concerned with the text (degree of incorporation, modality - voice-building) a n d in the variety of other voices that can b e integrated into the discourse. O n e ' s own voice can be unfolded, as for example in the case of irony. O t h e r voices can b e included, either explicitly, through citation, or implicitly, through what is called concealed citation. T h e identification of the presence of various voices and the way they are called u p in the text are o n e of the m o s t challenging analytical problems: to find the line taken b y the discourse subject in any text. This theoretical framework allows us to consider quoting as a strategy of the writer not only to convey information b u t also to give a n orientation to the text through the choices m a d e in the linguistic constituents of the quoting procedure. An integrated approach to a specific discourse procedure such as citation would appear to b e very useful in accounting for the sense, direction a n d effects of the insertion of words from different sources into the m a i n discourse of the journalist. We address the following questions: w h a t is the role of scientific voices in cases where they are entided to talk? Which are the forms of quoting? Is quoted speech a w a y of absolvingjournalists from responsibility or a m e a n s of orientating their position o n the topic of reference? H n w is all tV.;o -• n...^. «•
DISCOURSE STUDIES
302
crisis? To give an answer to the questions that arise in the study of the use of this discourse procedure we c o m b i n e different levels: (a) the description of grammatical construction and markers of explicit citation (Giron Alconchel, 1989; Reyes, 1993; M a l d o n a d o , 1991, 1999); (b) the function accorded b y the theory of enonciation to reported speech as a form of polyphony; (c) the interpretation accorded b y critical discourse analysis to the representa tion a n d presentation of agents in reported speech (Caldas-Coulthard, 1994; Van Leeuwen, 1996). From this perspective, the three levels contribute to a discoursive explanation that includes the written press context, the discourse functions accomplished and the linguistic units chosen. G r a m m a r s h a v e defined explicit citation in terms of direct/indirect styles. Nevertheless, it is interesting to note, as observed b y scholars like M e n d e z (1999a, b) w h o study quotations in the press, that the clear-cut differentiation between the two m a i n styles is not entirely relevant for the journal istic d o m a i n since there are a n u m b e r of different ways of using the words of others b y combining traits of direct a n d indirect style - considered as the basic axis defining reported speech. T h e increasing presence of words from different sources w h e n conflict is projected in news in the press gives rise to a variety of ways of formulating them. If the study of citation was based traditionally on its occurrence in literature, where free indirect speech is typically found, its use in other contexts (daily conversation or newspapers, for example) leads to the incorporation of other parameters in the identification of citation formulae. Taking into account that we are dealing with written data, we can distinguish different styles: •
1
direct citation: there is a. fracture between the syntax of D 1 and D 2 because it entails the maintenance of two different deictic centres (affecting tense, space a n d time adverbs a n d person-reference words), as a result of the two different enunciations being put in relation o n e to another; the two seg m e n t s are connected through juxtaposition a n d they are signalled by graphic markers such as (:) • indirect citation: there is only one discourse, D l , with a single deictic centre, a subordinate clause introduced b y a conjunction, and the correspondent agreement of tenses. • integrated citation: it has the form of indirect citation but with segments of greater or lesser extension - signalled as being cited direcdy/literally with clear graphic or typographic marking, mainly with quotation marks or m a r k e d fonts (boldface or italics). This type of citation allows mixing syntactic traits of direct a n d indirect style, a p h e n o m e n o n rejected by grammarians b u t frequently used b y journalists. • inserted citation: words of W 2 are brought into the m a i n discourse by means of markers such as 'segun X ' or ''para X', 'in the words ofX\ 'according to X* , . r k , v i , V i a w rtip f u n c t i o n of assienine explicit words to a particular ageni
CALSAMIGLIA A N D LOPEZ FERRERO
ROLE AND POSITION OF SCIENTIFIC VOICES
303
(literal or non-literal, d e p e n d i n g o n t h e use of graphic signs of quotation) without a n y communicative v e r b . T h e narrative style of the news makes the appearance of direct speech rarer while the rest are very frequent, with a b r o a d a n d loose acceptance of either o n e o r two deictic centres. Pragmatically speaking, t h e idea that is brought to the fore is that in a n y of the explicit occurrences of D 2 , in any of the forms referred to here, the possibility of obtaining stricdy the same original sense through literality is seriously called into question. Even if the exact words are reproduced, they are extracted from the original context a n d are placed i n a n e w context that can modify the sense. T h a t is t h e conclusion drawn b y Reyes (1993) a n d b y Caldas-Coulthard (1994) and suggest we should question a m a x i m apparently accepted b y the standards of journalism: that explicit quoting m e a n s fidelity to the text. A s argued: Direct style is often o p p o s e d to indirect style, somewhat ingenuously, saying that it seeks to r e p r o d u c e literally the cited words. It would b e m o r e exact to view it as a sort of dramatization of a previous expression, rather than a totally similar one. I n other words, it is n o m o r e or less reliable than indirect speech: they are two different strategies to refer to previous speech. (Maingueneau, 1987: 6 0 ; authors' translation)
,
T h e question is p u t in another way: it is fidelity to the voice which is u n d e r con sideration, so to speak, t h e original intention a n d purpose, t h e original partners of communication. T h e presence of different voices evoked i n a text leads us to consider the polyphonic character of news in journals (Ducrot, 1984; Authier-Revuz, 1982; Maingueneau, 1987). I n that sense, the idea of one text/one voice is n o t at issue. Rather, different voices axe assumed or direcdy evoked b y the writer, w h o builds the representation of a world of reference through a combination of a variety of voices. T h e relative weight, frequency a n d value of the different voices will, b y means of the observation of the empirical data, provide us with a profile of p r o tagonism taken o n b y the social actors whose words are called u p o n in relation to the topic. Finally, studies from a critical perspective of the ways social agents are pre sented (Caldas-Coulthard, 1994; Van Leeuwen, 1996) allow us to question the neutrality of the writer a n d pick o u t the argumentative trend that arises from the fact of choosing w h o is cited, h o w h e / s h e is represented, w h a t is cited a n d for what purpose.
5. D e l i m i t a t i o n o f t h e Linguistic-Discursive Object: F o r m a n d Function o f Explicit Citation Even if we are conscious that voices can also b e integrated in m o r e subtle ways, considered b y various scholars as 'hMfa-naineiiSJ / A ,i • ~
304
DISCOURSE STUDIES
as 'narrative m o d e of citation' (Charaudeau, 1997) or as 'echoes, h i d d e n citations' (Reyes, 1994), in this article we are concerned with explicit citation. This kind of citation consists of the m a i n text (Dl) a n d the quoted text (D2). D l is sequen tially the point of departure from which the writer (Wl) designs the development of his/her own discourse. Before inserting the cited voice (2), W l introduces a preparatory frame (Gir6n Alconchel, 1989), which is formed by: (a) T h e presentation of the identity of the cited voice. It involves t h e presen tation of the agent, b y such linguistic m e a n s as n a m e s , proper names, tides or honorifics, designation of status a n d public position, relational adjectives, etc. A n y of the choices m a d e are significant. (b) T h e communicative verbs (verba dicendi) that W l attribute to the voice of cited agents (W2). (c) O t h e r elements that are n o t necessary b u t possible: narratives, descrip tions, modalities. T h e frame is the presentation of W 2 m a d e b y W l , w h e r e W l is responsible for creating a n e w context for the cited words. T h a t is w h y we focus o n this precise portion of citation, the pre-citation, which plays the role of interface between D l and D2. For the study of the pre-citation segment we have d r a w n o n the suggestions in the Van Leeuwen's schema (1996:66) concerning different m o d e s of inclusion of agents: I n the framework of Van Leeuwen's description, we h a v e considered the five categories stressed b y boldface font in Figure 4: (1) genericization; (2) indi vidualization; (3) collectivization; (4) aggregation; a n d (5) objectivization. I n our data, — r—Activation 1—4 '— Passivization
1
•
ii
' i . i
1
• ..
i
,
r— Participation 2-*i Circumstalizarion '— Possessivation
,
i I
.
i
.
.
,. • i
- Determination • Personalization
•': i
Indetermination 3-H
1
i
. •
1
Genericization 5 r— Individualization Specification—"\ r - Collectivization '— Assimilation —»I— Abstraction '—Aggregation g
'— Impersonalization
8
- Objectivization Figure 4: The representation of social actors in discourse: Inclusion system network (adapted from Van Leeuwen, 1996)
CALSAMIGLIA A N D LOPEZ FERRERO
ROLE AND POSITION OF SCIENTIFIC VOICES
305
those are the most frequent forms of representation of m e m b e r s of the scientific community. Van L e e u w e n (1996: 46) considers an important factor in the presentation of social actors to b e the choice between generic and specific reference: actors can b e represented as 'classes' (genericization) or as 'specific' identifiable individuals (specification). T h e specification can b e singular (individualization) or plural (assimi lation). E x a m p l e s of generic forms in our corpus are forms such as 'scientists', 'doctors' without a specific reference to a concrete group of experts. T h e process of singular specification in the corpus is expressed b y authority voices (individual scientists); as for assimilation, V a n L e e u w e n (1996: 49) distinguishes two major forms: aggregation a n d collectivization. Aggregation quantifies groups of participants, treating t h e m as 'statistics' (example: ' a n u m b e r of critics'), collectivization does not. I n collectivization, social actors are presented as 'com munity', as a h o m o g e n e o u s , consensual g r o u p ; this is the case, for example, of ' W H O experts' in our corpus. Genericization, individualization, collectivization and aggregation are four personal ized forms in the process of agent representation. O n the contrary, objectivization is a form of impersonalization, also frequent in our corpus a n d realized b y metonymical references, such as 'texts'. T h e five forms considered can b e determined or undetermined T h u s , we h a v e considered the Van Leeuwen's distinction determination/indetermination: Indetermination occurs when social actors are represented as unspecified, ' a n o n y m o u s ' individuals or groups, determination w h e n their identity is, one w a y or another, specified. Indetermination is typically realised b y indefinite p r o n o u n s ('somebody', ' s o m e o n e ' , 'some', 'some people') used in nominal function Differentiation explicitly differentiates an individual social actor or group of social actors from a similar actor or group, creating the difference between the 'self a n d the 'other', or between 'us' a n d 'them'. (1996: 51-2) As for the reporting verbs related to the agents, while it is necessary to accept the wide variety of 'verba dicendi' and the impossible task of classifying t h e m , it is also clear that they can confer a content-oriented trend or an evaluative trend. We find interesting suggestions in the way T h o m a s and H a w e s (1994) a n d H y l a n d (2000) h a v e indicated types of content or evaluative approaches channelled through the choice of verbs of communication (see Figure 5). I n content-oriented process, verbs related to findings (observe, discover, notice, show) or to procedures (analyse, calculate, describe, assay, explore) refer to research acts of science agents. Popularizing acts (explain, describe, compare, define, exemplify) are used to clarify a n d facilitate understanding of specialized concepts. A n evaluation v e r b 'allowfs] the writer to ascribe a view to the source author, reporting h i m or her as positive (advocate, argue, hold, see), neutral (address, cite, comment, look at), tentative (allude to, believe, hypothesize, suggest), or critical (attack, condemn, object, refute)' (Hyland, 2000: 28). 2
806
- •* t V
DISCOURSE STUDIES
Content-Oriented Process ' Popularizing acts
Research acts
Findings
Procedures Evaluation
Factive (Writer agreement)
Author positive
Non-factive (No clear signal)
Author neutral
Author tentative
Counter-factive (Writer disagreement)
Author critical
Figure 5: Critical analysis classification of reporting verbs T h r o u g h an accurate analysis of the framing of the quote we can indicate the interpretive cues that are given b y the writer, because of the important role of choices in determining the argumentative orientation: •
choice of type of agent (conveying authority or credibility in a particular domain); • choice of his/her representation in discourse (as a n individual or a col lective, determined or undetermined, personalized (author) or objectivized (text), etc.; • choice of reporting verb (locating the agent in a position towards the topic). I n the case o f ' m a d cow' disease, types of agents that are called u p o n can tell us which are the sources that are presented as m o r e relevant, those of the scientific community or those with political or civic-social standing. This is a sociological distribution that gives external configuration to the voice. But their representation is achieved b y internal configuration t h r o u g h l a n g u a g e u s e ; t h e image of the agent is constructed a n d related through a specific reporting v e r b . O u r claim is that close analysis of the pre-citation segment, the frame of the cited words, will give u s important clues to the interpretation of the role a n d position of scientific voices.
6. Corpus a n d D a t a Treatment We h a v e b a s e d our study o n a corpus of 75 n e w s items from six different newspapers: La Vanguardia, El Pais, ElMundo, ElPeriddico, ABC a n d Avui These are the most widely r e a d general newspapers in Spain (Avui circulates in the Catalan-speaking area). T h e n e w s items considered w e r e published in March
CALSAMIGLIA A N D LOPEZ FERRERO
ROLE AND POSITION OF SCIENTIFIC VOICES
307
3
a n d April 1996 (between 22 M a r c h a n d 27 April 1996, to b e exact). This corpus contains the ' r e p o r t e d s p e e c h ' in t h e p r o c e s s of c o m m u n i c a t i o n of n e w s (Charaudeau, 1997). T h e 75 news articles (Dl) provided the 4 4 5 reporting structures (D2) that constitute the microlevel units analysed, b o t h reported speech from non-scientific voices (349 statements) and from scientific voices (96 statements). We h a v e focused the present study on the 96 scientific reporting structures, 21.57 percent of the reported speech in the corpus selected. We compiled a n extensive database to allow us to study quoted speech both from a quantitative and a qualitative point of view. Every citation has a corresponding register which details information on the macro-level of discourse (newspaper, press genre, press section, news author) as well as o n the textual micro-level (identification of W 2 , form of citation, reporting speech choice, text of citation).
7. Analysis After the delimitation of the object of our study a n d the presentation of the corpus, in the present section we proceed to study the role a n d position of scientific voices in the ' m a d cows' affair as they are c o m m u n i c a t e d via the Spanish press. First, we analyse the results of the study of the 'quoting frame' of scientific citations: representation of agents a n d choice of reporting verbs. T h e n w e present a brief analysis centred o n the contrast between scientific a n d non-scientific reported speech; this contrast allows u s to obtain relevant data which helps to explain the role and position of scientific voices. 7.7 Analysis of Scientific Voices We h a v e divided our corpus into three chronological information 'peaks' or phases, given the concentration of the news a r o u n d each peak, a n d the silence or 'pause' in between. Note that in phase 1 a n d phase 2, sources from scientific articles are n o t available, the source b e i n g governmental agencies that interpret the report issued b y the scientists (see Section 2). O n l y in phase 3 is the infor mation related to a n article o n the prion (from the microbiology field) which appeared in Nature. T h e three phases, including the percentage of citations, are the following: Phase 1: 22-24/3/1996 (44.30 percent) Phase 2: 28-31/3/1996 (33.04 percent) Phase 3:24-27/4/1996 (22.60 percent) For every phase we will give some representative examples of the dominant scientific voice a n d of the associated reporting verbs, so that we are able to see
DISCOURSE STUDIES
308
w h e r e there are modulations of t h e general stance detectable in the distribution of the different ways of representing scientific voices (Van Leeuwen, 1996). Table 1 presents a s u m m a r y of the process of representation and position of scientific voices in the different phases considered: w e present the most frequent forms of representation of actors a n d the reporting verbs associated with each kind of representation. Table 1: Forms of representation of identity and reporting verbs Phase
Representation of identity
Reporting verbs
1) 22-24/3/1996
a) Genericization b) Collectivization
a) Tentative b) Positive
2) 28-31/3/1996
a) Individualization b) Assimilation: collectivization and aggregation c) Objectivization
a) Research + markers b) Controversial: positive, tentative c) Positive, research
3) 24-27/4/1996
An only source represented as: a) Collectivization b) Objectivization c) Individualization
a) Research, tentative + markers b) Research c) Popularizing + markers
We h a v e characterized every phase b y a title that shows the differences found between the communication of scientific knowledge in each case: from caution (tentative point of view) to assertiveness (positive perspective) in the first phase; diversification of the scientific voices a n d their representation (individualization, assimilation a n d objectivization) in the second phase; presence of a single source in the third phase, which is collectivized, individualized or objectivized. 7.2 First Phase: Caution versus Assertiveness T h e systematic observation of citations in this first p h a s e shows that social actors are represented b y forms of genericization and collectivization (a form of assimilation). Scientists are presented as a class in genericization without a specific referent, as w e can see i n the following examples: 4
(1) L o s cientificos d i c e n a h o r a que esos cuatro afios fueron decisivos para que se saltara la b a r r e r a d e las especies y se desarrollara u n a version h u m a n a de la enfermedad. (La Vanguardia, 2 3 / 0 3 / 9 6 ) (1) Scientists n o w say that those four years were decisive in t h e breaching of the species barrier a n d the d e v e l o p m e n t of a n e w h u m a n version of t h e disease. (La Vanguardia, 2 3 / 0 3 / 9 6 ) (2) Los m e d i c o s c r e e n saber distinguir estos ultimos ' p o r q u e se producen e n m e n o r e s d e 42 afios, presentan u n a analitica diferente d e s d e el punto d e vista anatomopatologico y son d e evolucidn mas lenta, algo mas de ~ "
•
I r a n . . ' .
no;no/(K\
CALSAMIGLIA A N D LOPEZ FERRERO
ROLE AND POSITION OF SCIENTIFIC VOICES
309
(2) D o c t o r s b e l i e v e they can distinguish the latter 'because they occur in people less than 42 years old, are analytically different from an anatomopathological point of view, a n d are slower to develop, taking somewhat m o r e than 2 0 months'. (El Pais, 2 3 / 0 3 / 9 6 ) 'Scientists' a n d 'doctors' are genericized, a n d so symbolically r e m o v e d from the readers' world of immediate experience, treated as distant class. A s we will see in 7.5, other types of voices, non-scientific voices such as politicians, are not genericized. W h e n the process of genericization is chosen, the scientific cited words in press are presented b y a tentative v e r b , such as 'believe'. This type of reporting verb is o n e of the most frequent verbs used in our corpus, always associated with a generic or a n assimilate voice, as E x a m p l e s 1 and 2 shows, to express caution in front of a critical situation that relates b o t h to social perception a n d to the limitations of scientific knowledge. O n the other h a n d , in this first phase, scientific voices are also reported as m e m b e r s of an international organization (World Health Organization, W H O ) , as a form of collectivization: (3) Per la seva part, e x p e r t s d e I ' O M S v a n assegurar ahir que n o v e u e n justificacio cientifica per prohibir la importacio de earn de bovi procedent de G r a n Bretanya. (Avui, 2 3 / 0 3 / 9 6 ) (3) For their part, W H O e x p e r t s yesterday assured us they saw n o scien tific justification for prohibiting m e a t imports from Great Britain. (Avui, 23/03/96) In E x a m p l e 3 w e can see that W H O experts are presented as a 'community', as an h o m o g e n e o u s , consensual g r o u p which h a v e the responsibility of evaluating findings in science and its consequences for the general public. W H O experts are put u n d e r a process of collectivization. Their influence and political representa tion in society justify the reporting verbs chosen to present their speech: verbs associated with this collective are always positive ('assure'), a type of reporting verb which is n e v e r used in our data to introduce scientific class. T h e collective of W H O scientists is characterized b y their assertiveness versus the caution showed b y scientists in general. As we have said, in the computation of reporting verbs, the verbs m o r e frequendy used in our corpus are positive ('assure'), popularizing acts ('explain'), neutral ('say', 'state') a n d tentative ('believe'). I n this first phase - as in the other two -positive evaluation of facts is never in the generic voice of science. Assertive ness is always related to political representation. All these observations are completed b y the content of the quotations, which shows two distinct trends: the medical experts show 'caution' about their findings (Example 2) while the organization reveals 'assertiveness' in mitigating social alarm (Example 3). Knowledge of the context of these news items is crucial, be cause scientific information is channeilpd hv cm™™™— • ~ - -
310
DISCOURSE STUDIES
in parliament. So, in this first phase, the d o m i n a n t source is h o m o g e n e o u s and it is presented in agreement. 7.3 Second Phase: Diversification of Scientific Voices I n the second phase of the event, m o r e voices are called u p o n to inform about ' m a d cow' disease. D o m i n a n t scientific voices are reported in three different m o d e s : individualization, assimilation a n d objectivization. As for individualization, t h e linguistic form of p r e s e n t a t i o n is t h r o u g h determination b y use of n a m e and surname, professional identity and institutional status a n d origin, as E x a m p l e 4 shows: (4) S e g u n afirma S t a n l e y B . P r u s i n e r , p r o f e s o r d e N e u r o l o g i a y b i o q u f m i c o e n l a U n i v e r s i d a d d e California, e n el Scientific American, 'Es u n a teoria, p e r o p u e d e que los cambios estructurales de la proteina alterada, el prion, sean tambien responsables d e otras enfermedades degenerativas m a s c o m u n e s c o m o el Alzheimer. Y esta posibilidad no d e b e ignorarse'. (ElMundo, 2 8 / 0 3 / 9 6 ) (4) A c c o r d i n g to S t a n l e y B. Prusiner, a t e a c h e r o f n e u r o l o g y a n d bio chemistry at t h e University of California, in the Scientific American, 'It's a theory, but it m a y b e that structural changes in the altered protein, prion, might also b e responsible for other degenerative diseases more c o m m o n t h a n Alzheimer. A n d that possibility should n o t b e ignored.' (ElMundo, 2 8 / 0 3 / 9 6 ) I n this phase, as in the first one, it is noticeable that reporting verbs are distributed differently d e p e n d i n g o n the identity. Individual authorities are presented frequendy with markers of reported discourse such as 'for', 'according to' ('para', 'segun' in E x a m p l e 4), contributing thus to the neutrality or distance established b y the writer. Reporting verbs used to introduce authorities are above all research verbs - 'describe' ('describir'), 'conclude' ('concluir') - n o t very frequent in the corpus. Two examples of this kind of structures are the following 5 ('described') a n d 6 ('concludes'), with individualized voices ('Vincet Zigas' a n d 'Carleton Gadjusek' in E x a m p l e 5, 'Jeffrey W. A l m o n d ' in E x a m p l e 6): (5) V i n c e t Zigas, d e l Servicio d e Salud Publica d e Australia y Carleton Gadjusek, d e l Instituto N a t i o n a l d e la Salud d e E E U U , describieron en 1957 c o m o m i e m b r o s d e u n a tribu d e P a p u a N u e v a G u i n e a sufrian u n a e n f e r m e d a d m o r t a l caracterizada, p r i m e r o , p o r la p e r d i d a de coordination y, despues, p o r demencia. (ElMundo, 28/03/96) (5) V i n c e t Zigas, of t h e Australia P u b l i c H e a l t h S e r v i c e a n d Carlton Gadjusek of t h e N a t i o n a l H e a l t h Institute of t h e U n i t e d States, d e s c r i b e d in 1957 h o w a tribe in Papua N e w Guinea suffered a lethal . . . disease characterized firsdy b y loss of coordination and then b y dementia. (ElMundo, 28/03/96)
I
CALSAM1GLIA A N D LOPEZ FERRERO
ROLE AND POSITION OF SCIENTIFIC VOICES
311
(6) E l cientifico Jeffrey W. A l m o n d c o n c l u y e e n u n articulo publicado en el British Medical Journal: 'Es obvio que la forma d e p r e s e n t a t i o n d e la E C J e n G r a n Bretafia, e n comparacion con otros paises libres d e la enfermedad d e las "vacas locas", es u n a causa que d e b e ser estudiada'. (ElMundo, 28/03/96) (6) ScientistJeffrey A l m o n d c o n c l u d e s , in an article published in the British Medical J o u r n a l that 'It is obvious that the form in which E C J presents in Great Britain, in comparison with other countries free from "mad cows' disease", is a cause which should b e studied.' (ElMundo, 28/03/96) As far as assimilation is concerned, collectives are presented in a general determined or u n d e t e r m i n e d form a n d are n o w dispersed across different fields (neurology, veterinary, medicine, etc.), institutions (hospitals, health services, etc.) and countries (Great Britain, U S A , Australia, etc.). It is n o t the case, then, of a generic scientific reference as it was during the first phase but of heterogenous collectives of scientists: in E x a m p l e 7, 'Las autoridades sanitarias' a n d 'los propios medicos del hospital de C a n t e r b u r y ' h a v e specific references: (7) L a s a u t o r i d a d e s s a n i t a r i a s y l o s p r o p i o s m e d i c o s d e l h o s p i t a l d e Canterbury, donde murio A n n e Pearson, no h a n querido hacer c o m e n t a r i o a l g u n o s o b r e el caso, m a s a l i a d e d e c i r que 'es imprudente especular sobre como la victima contrajo el C J D , p o r q u e su caso esta siendo todavia estudiado en u n laboratorio d e E d i m b u r g o y n o h a y conclusiones definitivas'. (La Vanguardia, 30/03/96) (7) H e a l t h a u t h o r i t i e s a n d t h e d o c t o r s t h e m s e l v e s a t t h e C a n t e r b u r y Hospital w h e r e A n n e Pearson died did n o t wish to m a k e any c o m m e n t concerning the case, b e y o n d s a y i n g that 'It is unwise to speculate over h o w the victim contracted C J D , as her case is being studied at the Edinburgh laboratory a n d there are n o definitive conclusions.' (La Vanguardia, 30/03/96) Forms of aggregation are also frequent in this phase, showing the different points of view in the groups of scientists: in E x a m p l e 8 the determiner ' s o m e ' ('Algunos') in front o f ' o t h e r s ' ('Otros', ' Y otros') signals different groups of experts w h o did not agree o n the scientific explanation of the p h e n o m e n o n : (8) A l g u n o s i n v e s t i g a d o r e s c r e e n que, a d e m a s del prion, d e b e d e h a b e r otro agente infeccioso a u n n o identificado. O t r o s n o d e s c a r t a n que aparezca u n virus. Y o t r o s e s t a n s e g u r o s d e que la n u e v a enfermedad detectada en G B n o es E C J sino la p r o p i a E E B d e las vacas. (ElMundo, 28/03/96) (8) S o m e r e s e a r c h e r s b e l i e v e that, in addition to prion, there must b e some other unidentified agent of infection. O t h e r s d o n o t r u l e o u t the possibility of a virus appearing. A n d o t h e r s a r e s u r e that the n e w disease detected in G B is n o t ECJ, b u t n o n e other than the E E B occurring in cows. (ElMundo, 2 8 / 0 3 / 9 6 )
DISCOURSE STUDIES
312
In. this example, scientists are presented in aggregationin terms of Van Leeuween (1996), identified as groups that show disagreement in the evolution of science research. This disagreement is also reflected in a variety of verbs indicating scientific debate a n d different points of view: in E x a m p l e 8 a positive expression 'are sure' is used next to a tentative v e r b such as 'believe'. Finally, 'texts', a form of objectivization, are in this second phase a notable source for citation, being presented in a general determinate form or b y the name of the publication, with n o identification of authorship. Personal reference is lost through a m e t o n y m i c reference (as 'Studies' in E x a m p l e 9) that gives cred ibility to the written words - related to the accepted knowledge of the scientific community: (9) L o s estudios h a n d e m o s t r a d o que e n n i n g u n a d e ellas [enfermedades neurodegenerativas], los leucocitos - el 'ejercito' del sistema inmune se infiltran e n el cerebro. (ElMundo, 28/03/96) (9) Studies h a v e d e m o n s t r a t e d that in n o n e of t h e m [neurodegenerative diseases] do leucocytes - the i m m u n e system's ' a r m y ' - infiltrate into the brain. {ElMundo, 28/03/96) Citations of texts are presented with significant research verbs such as 'demonstrate' ('demostrar') in E x a m p l e 9 or 'reveal' ('revelar'). As can b e seen, in this second phase sources are heterogeneous, in their reference as well as in their m o d e of representation. Experts c o m e from different fields: neurologists, biologists a n d microbiologists, veterinarians, etc. bringing about an expansion of information that complements the purely medical data. Journalists begin to quote research o n microbiology, looking for a possible ex planation of the causes of the infection, trying to confer credibility based on the results of studies. 7.4 Third Phase: A Scientific Move . This phase is typified b y a particular instance of scientific communication as it appeares in the British j o u r n a l Nature, about a m o n t h after the outbreak of diplo matic a n d economic conflict generated b y the first wave of news. It represents progress m a d e in scientific information. We find different forms of representation of this source alone: first, as the text published in Nature, a form of objectivization exemplified in the following citation: (10) U n estudio q u e p u b l i c a h o y l a revista N a t u r e d e m u e s t r a que la e s t r u c t u r a genetica d e los p r i o n e s d e las vacas tiene u n parecido m u y significativo con los priones d e los seres h u m a n o s . (El Mundo, 25/04/96) (10) A study p u b l i s h e d today i n N a t u r e m a g a z i n e demonstrates that the genetic structure of prions in cows is very significantly similar to prions in h u m a n beings. (ElMundo, 25/04/96)
CALSAMIGLIA A N D LOPEZ FERRERO
ROLE AND POSITION OF SCIENTIFIC VOICES
313
Second, as the group of scientists w h o carried out t h e research, a m o d e of collectivization represented in E x a m p l e 11: (11) L o s cientfficos d e Oxford, incluidos T.R.F. S o u t h w o o d , antiguo m i e m b r o d e l a c o m i s i d n asesora d e l G o b i e r n o britanico s o b r e l a e n f e r m e d a d d e las v a c a s locas, h a n descubierto que estas proteinas [priones] d e vacas y personas comparten dos rasgos distintivos e n su estructura que estan ausentes en la protema de oveja. (ElPais, 25/04/96) (11) Scientists at Oxford, i n c l u d i n g T.R.F. S o u t h w o o d , former m e m b e r of t h e British g o v e r n m e n t a d v i s o r y c o m m i s s i o n o n m a d c o w s ' d i s e a s e , h a v e d i s c o v e r e d that these proteins [prions] in h u m a n s a n d cows share two distinctive traits in their structure that are absent in the sheep protein. (El Pais, 2 5 / 0 4 / 9 6 ) A n d third, as the authorities o n the matter, in the process of expressed in E x a m p l e 12:
individualization
(12) 'Las ovejas padecen u n a enfermedad similar desde hace doscientos afios p e r o n u n c a se h a transmitido a las personas. L a enfermedad d e las vacas, en cambio, si p u e d e haberse transmitido. Si asi es, debemos buscar u n a caracteristica que tengamos e n c o m u n con las vacas p e r o n o con las ovejas. Y es precisamente lo que h e m o s encontrado', e x p l i c o ayer D a v i d K r a k a u e r , d i r e c t o r d e l a i n v e s t i g a c i o n , e n entrevista telefonica. (La Vanguardia, 2 5 / 0 4 / 9 6 ) (12) 'For the last two h u n d r e d years, sheep h a v e suffered a similar disease that has n e v e r b e e n passed o n to h u m a n s . T h e disease in cows does, however, seem to h a v e b e e n passed o n to h u m a n s . If this is so, we h a v e to look for a characteristic we h a v e in c o m m o n with cows but not with sheep. A n d that's precisely what we h a v e found', e x p l a i n e d D a v i d K r a k a u e r , director of t h e investigation, yesterday in a telephone interview. (La Vanguardia, 25/04/96) As w e see in the examples above, reporting research verbs are frequendy used: 'demonstrates' ('demuestra' in E x a m p l e 10), 'have discovered' ('han descubierto' in E x a m p l e 11), associated with the collective of researchers a n d with texts. Also, markers such as 'for' ('segun', 'para'), which show a neutral position, are employed to introduce the group of scientists, a n d n o t only to refer to an individ ual authority (as in the second phase). Finally, the popularizxngverh 'to explain' ('explicar') is used in this third p h a s e primarily in relation with authorities ('explained' in E x a m p l e 12). Nevertheless, w h e n scientists are quoted as a generic group, tentativeiepoiting verbs are also assigned, as in the previous phases:
'
DISCOURSE STUDIES
314
(13) L o s cientfficos c r e e n que [los priories] p u e d e n causar graves enfermedades neurologicas cuando pasan de u n a especie a otra e incluso cuando pasan d e u n individuo a otro dentro de la m i s m a especie. (La Vanguardia, 25/04/96) (13) S c i e n t i s t s b e l i e v e that [prions] can cause serious n e u r o b i o l o g i e s illnesses w h e n they pass from one species to another a n d even when they are passed from one individual to another within the same species. (La Vanguardia, 25/04/96) To sum u p , in the third phase, the source is a scientific article o n microbiology published in one of the most prestigious reference journals in science communi cation (Nature). We cannot find any significant discourse function in the way scientific voices are represented, b y individualization, collectivization or objectivization, since all of t h e m are present in the quote frame referring to the same source. T h e reporting verbs used are in the semantic domain of research ('demon strate', 'discover', 'conclude') a n d divulgence ('explain'), following the patterns of science communication. Moreover, markers of neutrality are used frequently to introduce scientific discourse. 7.5 Contrasting Scientific and Non-Scientific Voices As we h a v e already mentioned, the proportion of speech reported from scientists in our selected corpus is only 21.57 p e r c e n t of the total a m o u n t of reported speech; in other words, non-scientific voices are dominant. For this reason it is interesting to contrast the representation of scientific a n d non-scientific voices in order to n o t e differences in the role a n d position of each kind of agent as fea tured in science communication. I n a previous study - reported in Calsamiglia a n d Lopez Ferrero (2001), Calsamiglia a n d Cassany (2001) - the respective contributions of scientific and non-scientific voices in the information o n ' m a d cows' disease w e r e examined. T h e present analysis confirms the results presented in the two articles, in which w e studied use of quotation in the three first days of diffusion of the news in the Spanish newspapers. Table 2 shows the distribution of scientific voices in the present corpus in relation to the representation of identity: Table 2: Presence and identity of scientific voices (96 citations: 21.57 percent) Genericization (16.66%)
Assimilation (25%)
-Determined +Determined "Scientists" Gov Res "Experts" 16 1 14 9
Individualization (43.75%)
Objectivization (14.58%)
-Determined -t-Determined Author WHO 1
29
12
14
I
CALSAMIGLIA A N D LOPEZ FERRERO
ROLE AND POSITION OF SCIENTIFIC VOICES
315
Scientific voices are frequently p u t into a process of genericization, as a com munity without specific reference ('los cientificos', 'los expertos', etc.). W h e n there is a process of assimilationtwo specific groups are represented: government advisors (Gov in Table 2) and collectives of concrete researchers (Res), some of which publish their studies in Nature. As for individualization, this m o d e is presented through a scientific authority (Author) or a World H e a l t h Organization m e m b e r (WHO). I n contrast, Table 3 shows the distribution of non-scientific voices in the corpus: Table 3: Presence and identity of non-scientific voices (349 citations: 78.43 percent) Assimilation (27.79%)
Individualization (70.20%)
Objectivization (2%)
-Determined -(-Determined -Determined -(-Determined Cou Inst Org Cit Mer Org Polit Press Spec 12 17 41 27 5 12 60 134 24 10 7 Abbreviations used in Table 3 refer to the different forms of representations: Forms of assimilation: • Cou: country ('Suiza, Alemania, los alemanes, Reino Unido') • Inst: institution ('el gobierno, la Administration espanola, las autoridades suizas, la U E ' ) • Org: organization ('la O r g a n i z a t i o n M u n d i a l de la Salud, el sector carnico, las organizaciones agrarias, los sindicatos agricolas') Forms of individualization: • Cit: citizen ('dos oficinistas, u n quiosquero, u n ingeniero cincuenton, u n a m a de casa') • Mer: merchant ('un carnicero, u n exportador d e carne, el encargado del restaurante') • Org: organization ('un portavoz de la O M S , el presidente del Sindicato N a t i o n a l d e Ganaderos') • Polit: politician ('el ministro d e la presidencia e n funciones, el lider d e la . oposicion laborista') • Press: press ('Le Monde, The Daily Telegraph, E F E , special correspondents') • Spec: specialist ('sociologo, antropologo'). Forms of objectivization: ('informe, codigo penal') As shown, non-scientific voices are represented m a i n l y as individual voices (70.20 percent), politicians being the social agents most cited. The data tell us how, in this case, politicians are called upon to provide information even on science matters. A s for the reporting verbs used, the press attributes a positive f'assure' in Example 14) and interaction-oriented'position furge', 'call' in Examples 15 and 16) to social agents with apolitical function. I n a critical situation, political
316
DISCOURSE STUDIES
actors are the ones expected to b e active o n their speeches, decision making, rec ommendations. T h e y are pushed to perfom symbolically, involving sometimes a risk to their public face: (14) Tony Blair, lfder d e l a o p o s i c i o n laborista, tambien a s e g u r o que continuant c o m i e n d o carne d e buey, a u n q u e acuso al Gobierno de 'ocultar datos' a la opinion publica, y de 'resistirse a decir toda la verdad' p a r a proteger los intereses d e los ganaderos. (La Vanguardia, 22/3/96) (14) Tony Blair, l e a d e r of the Labour Opposition, also assured the House that h e would continue eating beef, although h e accused the Government of 'hiding facts' from public opinion, a n d of 'refusing to tell the whole truth', to protect the farmers' interests. (La Vanguardia, 22/03/96) (15) Precisamente ayer, su p r e s i d e n t e , J a c q u e s Santer, r e c l a m d a J o h n Major medidas urgentes p a r a acotar la crisis y permitir que la U E pueda levantar el embargo - siquiera parcialmente - que pesa sobre el vacuno britanico. (El Pais, 25/4/96) (15) Precisely yesterday, its president, J a c q u e s Santer, u r g e dJ o h n Major to take urgent measures to b r i n g the crisis to an e n d a n d allow the E U to lift the e m b a r g o (partially) i m p o s e d o n British beef. (La Vanguardia, 25/04/96) (16) U n sector p r e d o m i n a n t e d e la prensa c o m p a r t e esta tendencia. Sir Teddy Taylor, a b a n d e r a d o d e l e u r o e s c e p t i c i s m o , p i d i o ayer una prohibition total de vino y queso frances en el Reino U n i d o en venganza p o r el boicot frances contra la carne britanica. (ABC, 2 3 / 3 / 9 6 ) (16) A major p a r t of t h e press shares this tendency. Sir T e d d y Taylor, a v o w e d Eurosceptic, c a l l e d yesterday for a total b a n o n French wine a n d cheese in t h e U n i t e d Kingdom, in revenge for the French boycott o n British meat. (ABC, 23/03/96) O n the other h a n d , w h e n w e consider the generic a n d collective representation of scientific voices (40 citations) plus their objectivized representation as texts (14 citations), we observe that, unlike what w e saw with non-scientific voices, the p r e d o m i n a n t source is n o t an individual: 'science' is represented b y the scientific c o m m u n i t y as a group, or b y texts. Therefore, the scientific community and texts are perceived b y society as sufficient markers of authority. However, if we leave the 14 textual citations aside a n d focus on t h e 40 citations of personal scientific voices, the proportion of individualized (43.75 percent) a n d generic a n d assimilated (41.66 percent in total) citations are very close. A relevant feature is that b o t h assimilated a n d individualized citations are presented in n o u n phrases (NPs) with determiner. I n the case of individual voices, the determination is related to a personal social identity (name, status, etc.). I n the case of generic and assimilated voices - 'scientist' ('los cientificos'), 'experts' ('los expertos'). 'researchers' ('los investigadores'), etc. - there is generally just a simple determiner preceding a c o m m o n n o u n , having either a n anaphoric or generalizing function;
CALSAMIGJJA A N D LOPEZ FERRERO
ROLE AND POSITION OF SCIENTIFIC VOICES
317
again the linguistic choice appeals to a ' c o m m o n ground' representation of the role of scientific work. As for the reporting verbs, we find scientific and non-scientific voices coincide in their use of positive a n d neutral verbs, such as 'say', 'state', 'assure' ('decir', 'afirmar', 'asegurar'). Nevertheless, there is a relevant difference in the use of specific verbs in each kind of source. As we have noticed, reporting verbs assigned to non-scientific voices can b e described as interaction-oriented: 'claim', 'announce' ('reclamar', 'anunciar'), while those assigned to scientific voices are contentoriented: 'explain', 'believe', 'consider', 'show', 'emphasize' ('explicar', 'creer', 'considerar', 'senalar', 'destacar'), related to mental activities of knowledge, quali fication a n d focalization. S o m e of t h e m reflect a tentative position ('believe', 'consider'), others ('explain', 'show') are characteristic of popularizing discourse. As a summary, in Table 4 w e present the relevant contrasts in the forms of representation of identity a n d in the choice of reporting verbs to introduce scientific a n d non-scientific speech: Table 4: Choice of representation of identity forms and reporting verbs Voice Non-scientific Scientific
Representation of identity Assimilation Individualization Genericization Assimilation Individualization Objectivization
Reporting verbs Positive and neutral Interaction oriented Positive and neutral Content oriented: tentative andpopularizing
8. Discussion At the beginning of this article, w e asked several research questions about the sense, direction a n d effects of the insertion of words from scientific sources into journalist discourse. O u r aim has b e e n to reveal the role of scientific voices o n an issue where they are entitled to talk, a n d to explain the orientation of discourse depending o n the forms of quotation, above all in a situation of crisis such as the 'mad cows' affair has b e e n a n d still continues to b e . A n o t h e r kind of question is related to the use that journalists m a k e of citation: is quoted speech a way of abdicating the responsibility of journalists or a m e a n s of orientating their position o n the topic of reference? As M o i r a n d (2000) points out, journalists compensate for what she calls an 'insecurite discursive' b y bringing to their discourse the very m a n y voices involved in the problem. By doing so they dramatize the conflict. At all events, the a m o u n t of quoted material of nonscientific individuals, mainly politicians, in circulation indicates that responsibility is assigned to socio-political actors, even if the information has its origin in a report b y a g r o u p of researchers. W h y are scientific voices so scarcely represented? O u r p r o p o s e d division of the journal articles studied into different stages m a y help to give an answer to this question. This procedure has b e e n useful in allowing n s rn m ^ J ' a
m
DISCOURSE STUDIES
318
I n our analysis, the first and second phases of the news story are just four days apart, while the third phase comes nearly a m o n t h later. T h e first phase is marked b y the representation of scientists as a collective d e p e n d e n t o n political directives (parliament, ministers, organizations). T h e scientific c o m m u n i t y is presented in the context of social life as a class that is expected to give quality information. T h e first p r o b l e m arises: the d e m a n d s of m e d i a information a n d t h e nature of scientific information collide. T h e y reflect two ways of representing knowledge: science based o n ongoing search, o n the discussion of hypotheses, with no conclusive evidence; the m e d i a representation b a s e d on the idea that scientists are able to provide a n account that will bring a quick solution to the problem. T h e second phase is m a r k e d b y diversification a n d individualization, while written studies are drawn u p o n also as a credible referent. A variety of experts and texts are called u p o n and, according to the state of knowledge, still not established or accepted, different points of view are reported. This is a legitimate position in the context of science, but again the problem arises when it is presented to a public that needs to m a k e urgent decisions, adding to the confusion created in the first phase. T h e third phase is m a r k e d b y reference to a scientific source, the publication of a microbiology research article about the possible infectious agent of the disease. I n this phase the standard way of communicating advances in science is respec ted. T h e progress in scientific information is presented a n d explained within its strict limits through the text a n d the group, only incidentally individualizing the research actors. But t h o u g h it seems that some of the causes of the disease are beginning to b e explained, in the context of social alarm a n d economical conflict, this cannot fulfil the m o r e immediate a n d urgent needs of the general public. As for the evaluative activity of journalists, it seems that the use of reported speech is a m e a n s of orientating their position o n the topic of reference, as well as or even m o r e than a way of abdicating their responsibility to inform objectively. Journalists seem to assume that they reflect public n e e d s a n d fears as well as the loss of control that a health crisis situation generates. If the scientific role is not considered a decisive social role capable of mitigating social alarm on its own. journalists are not able to assign or recognize the limits of scientific knowledge a n d to present such limits as a w o r t h y fact in itself. I n the public context, the consequence is that this knowledge is a source of confusion rather than a contribu tion to a n understanding of the p r o b l e m .
9. C o n c l u s i o n I n conclusion, we can say that the analysis of the 'quoting frame', i n its canonical components of agent identity a n d reporting verbs, gives a representation of what the press considers relevant for that part of information which is based o n words, not o n facts. T h e most outstanding feature is that multiple voices are called u p o n to clarify a situation w h e r e decisions h a v e to b e taken at different levels. I n actual fact, the ' m a d cows' affair is an example of the propagation of ongoing research t h r o u g h political instance. T h e circulation of information generated in
CALSAMIGLIA A N D LOPEZ FERRERO
ROLE AND POSITION OF SCIENTIFIC VOICES
319
a research context, once set in different contexts, m a y create the confusion that we have b e e n able to describe in the analysis of the choice of reporting verbs, showing different attitudes, understanding a n d interests. T h e study of the ' m a d cows' case shows the differing p a c e of science a n d daily life as a gap that is manifested through the way the press, reacting consciously or unconsciously to a situation of crisis, gives rise to different voices. Scientific voices have a limited role in the press, m u c h less weight than those of political actors. Individual scientific authorities (researchers) are presented in a lower position than scientists as a generic class or in a civic/administrative role. As for the communicative position, journalists attribute a neutral and popularizing stance to individual researchers a n d a controversial and tentative position to the scientist in general, b o t h of these contrasting with the positive, interaction-orientedposition attributed to the social agents with a political function. T h e position assigned to scientists is a qualified one b u t n o t efficient enough, managing a kind of knowledge (a) that rests on tentative hypotheses a n d questions under research a n d (b) because the advances of science do n o t give rise to results with immediate answers to social needs. I n fact, the two representations of know ledge show their contradictions w h e n they are placed side b y side: the scientific representation (science domain) a n d the social representation (society domain). A question remains unanswered: whether confidence in scientific work can b e maintained in the m e d i a even if what is c o m m u n i c a t e d is the limited state of scientific knowledge. These conclusions should b e contrasted with the analysis of the role a n d position of scientific voices in the n e w ' m a d cows' episode which peaked at the turn of the century. Approximately five years after the first appearance of the story in the media, a n e w episode of the crisis has developed. Following this rich com municative event should provide ways of accounting for journalists' responsibility in representing science in social life and their contribution to a better public understanding of science.
Notes This is the final version of a study carried out under the auspices of DGES project no. PB96-0292. 'Analisis discursivo de la divulgation cientifica: aspectos pragmaticos, textuales y retoricos' as part of the program of the Direction General de Ensenanza Superior del Ministerio de Education y Cultura (1997-2000) of the Spanish Government, and presented as a paper at the VIIInternational Pragmatics Conference in Budapest (2000). A first version was originally presented as work in progress at the ISimposio International de Andlisis delDiscurso, Madrid, 1998 (see Calsamiglia and Lopez Ferrero, 2001, and Calsamiglia and Cassany, 2001). 1. Dl refers to main discourse by writer 1 (Wl); D2, to quoted discourse by writer 2 (W2). 2. An exhaustive classification of reporting verbs has been proposed by Battaner et al. (2001): assertive communicative ('decir, comunicar, preguntar, asegurar, prometer, manifestar, admitir'), communicative claim ('asegurar, recalcar, insistir, reiterar, precisar, puntualizar'), attentionfocusfobservar, senalar, ver, reflejar, advertir, indicar, destacar').
DISCOURSE STUDIES
320
positive evaluation ('alabar, elogiar, aprobar, dar importancia, destacar, subrayar. recalcar'), negative eca/aatfonfreprochar, desaprobar, criticar, acusar, lamentar, atacar, descartar'), demanding ('pedir, ordenar, mandar, prohibir, exigir, aconsejar, instar a, suplicar, reclamar, solicitar'), forms of speech ('susurrar, tartamudear, suspirar, gritar, exclamar, chillar, carcajearse, reirse, gruflir, murmurar, declamar, tararear, cantar, recitar'), discourse markers ('empezar, comenzar, anadir, seguir, acabar, terminar, continuar, concluir'), discourse action description ('comentar, desarrollar, considerar, plantear, proponer, tratar, sostener, confesar'), interaction description ('intervenir, responder'), metalinguistic markers ('narrar, citar'), opinion verbs ('op'mwc, juzgar, considerar, creer, dar su opinion, estar convencido de, calificar, sugerir, estimar'), counterarguments ('discutir, objetar, rectificar'), analysis ('analizar, distinguir, establecer, hacer una distincion'), comparison ('comparar, ejemplificar'), thinking and observing ('pensar, especular, sentir, ver, oir'). 3. The analysis of citation in the three first days in those newspapers at the moment when it became known that BSE might be transmitted to humans has been studied in Caisamiglia and Lopez Ferrero (2001). 4. In the examples, we stress in bold the linguistic structures (the pre-citations) analysed.
References ABC (1993) Libro de estilo. Barcelona: Ariel. Authier-Revuz.J. (1982) 'Heterogeneite montree et heterogeneite constitutive: elements pour une approche de l'autre dans le discours', DRLAV26: 91-151. Bajtin, M.M. (1979) Estetica de la creation verbal. Mexico City: Siglo XXI. Battaner, M.P. et al. (2001) Aprendery ensendr:La redaccidn de exdmenes. Madrid: Antonio Machado Libros, S.A. Benveniste, E. (1971) Problemas de linguistica generall. Mexico City: Siglo XXI. Benveniste, E. (1974) Problemas de linguistica generalll. Mexico City: Siglo XXI. Caldas-Coulthard, C.R. (1994) 'On Reporting Reporting: The Representation of Speech in Factual and Fictional Narratives', in M. Coulthard (ed.) Advances in Written Text Analysis, pp. 295-308. London: Routledge. Caisamiglia, H. and Cassany, D. (2001) 'Voces y conceptos en la divulgaclon cientifica', Revista Argentina de Linguistica 11(15): 173-208. Caisamiglia, H. and Lopez Ferrero, C. (2001) 'Polifonia en textos periodisticos con informacion cientifica', in J. Garrido Medina (ed.) Lengua, discurso, texto (I Simposio International de Andlisis del Discurso), pp. 2647-64. Madrid: Visor Libros. Charaudeau, P. (1997) Le Discours d'information mediatique: La construction du miroir social Paris: Nathan. Daban, M. and Grau, M. (1996) 'La enfermedad de las vacas locas; analisis de los fenomenos cientificos y mediaticos', Quark 5: 11-18. Donadieu, J. (1997) 'El complejo asunto de las vacas locas. Por una lectura crftica de los raros datos cientificos disponibles', Mundo cientifico 185: 1008-10. Ducrot, O. (1984) 'Esbozo de una teoria polifonica de la enunciation', in El deciry lo dicho. Polifonia de la enunciation, pp. 174-238. Barcelona: Paidos. Giron Alconchel, J.L. (1989) Las formas del discurso referido en el 'Cantar de Mio Cid'. Madrid: Anejos del Boletin de la Real Academia Espafiola, Anejo XLIV. Hyland, K. (2000) 'Academic Attribution: Interaction through Citation', in Disciplinary Discourses. Social Interaction in Academic Writing, pp. 20-40. London: Longman.
CALSAMIGLIA A N D LOPEZ FERRERO
ROLE AND POSITION OF SCIENTIFIC VOICES
321
Informe Quiral (1999) Publicaciones del Observatorio de la Comunicacion Cientifica. Barcelona: Universidat Pompeu Fabra. Janssen, T.AJ. and Van der Wurff, W. (eds) (1996) Reported Speech: Forms and Functions of the Verb. Amsterdam: John Benjamins. Jasanoff, S. (1997) 'Civilization and Madness: The Great BSE Scare of 1996', Public Understanding of Science 6: 221-32. Kerbrat-Orecchioni, C. (1980) L'Enonciation. De la subjectivity dans le langage. Paris: Armand Colin. Krakauer, D.C., Pagel, M., Southwood, T.R.F. and Zanott, P.M. (1996) 'Phylogenesis of Prion Protein', Nature 380: 65-76. Lucy,J. A. (ed.) (1993) Reflexive Language: Reported Speech andMetapragmacity. Cambridge: Cambridge University Press. Maingueneau, D. (1987) 'L'heterogeneite', in Nouvelles tendances en analyse du discours, pp. 53-93. Paris: Hachette. Maingueneau, D. (1991) L'Enonciation en linguistique frangaise: embrayeurs, 'temps', discours rapports. Paris: Hachette. Maingueneau, D. (1993) EUments de Linguistique pour le texte litteraire. Paris: Dunod. Maldonado, C. (1991) Discurso directo y discurso indirecto. Madrid: Taurus. Maldonado, C. (1999) 'Discurso directo y discurso indirecto', in I. Bosque and V. Demonte (eds) Gramdtica descriptiva de la lengua espanola, pp. 3551-95. Madrid: Espasa Calpe. Mendez Garcia de Paredes, E. (1999a) 'La literalidad de la cita en los textos periodfsticos', Revista Espanola de Linguistica 30(1): 147-67. Mendez Garcia de Paredes, E. (1999b) 'Analisis de la reproduction del discurso ajeno en los textos periodisticos', Pragmalinguistica 7: 99-128. Moirand, S. (1997) 'Formes discursives de la diffusion des savoirs dans les medias', Hermes 21 (Sciences et midias): 33-44. Moirand, S. (2000) 'Variations discursives dans deux situations productrices d'un discours sur la science de la presse ordinaire', Carnets du CediscorG: 45-62. Porta, M. (1996) 'Pasando la maroma con John Major a cuestas. Apuntes imaginarios sobre la implausible responsabilidad de una sociedad cientifica ante la inverosfmil crisis de las "vacas locas'", Quark 5: 19-32. Ratzan, S.C. (ed.) (1998) The Mad Cow Crisis: Health and the Public Good. New York: New York University Press. Reyes, G. (1993) Losprocedimientos de cita: estilo directoy estilo indirecto. Madrid: Arco Libros. Reyes, G. (1994) Los procedimientos de cita: citas encubiertasy ecos. Madrid: Arco Libros. Thomas, S. and Hawes, T. (1994) 'Reporting Verbs in Medical Journal Articles', English for Specific Purposes 15(2): 129-48. Van Dijk, T.A. (1988) News as Discourse. Hillsdale, NJ: Lawrence Erlbaum. Van Leeuwen, T. (1996) 'The Representation of Social Actors', in C.R. Caldas-Coulthard and M. Coulthard (eds) Texts and Practice, pp. 32-70. London: Roudedge. Voloshinov, V.N. (1929) ElMarxismoy lafilosofia del lenguaje. Madrid: Alianza Editorial. Waugh, L. (1995) 'Reported Speech inJournalistic Discourse: The Relation of Function and Text', Text 15(1): 129-73. Will, R.G., Ironside, J.W. et al. (1996) 'A New Variant of Creutzfeldt Jakob Disease in UK', The Lancet 347(9006). Zander, U. (1997) Le Journaliste, le scientifique, et la vache folk. Paris: Publications de PUniversite.
13 Relativity and Its Discontents: Language, Gender, and Pragmatics *
-
Deborah Cameron
1. T h e "Social Constructionist Turn" i n L a n g u a g e a n d Gender Research
T
h e approach to language a n d g e n d e r that is p r o b a b l y m o s t familiar to most people (including m a n y scholars w h o are n o t specialists in the field) is one that focuses o n identifying, describing, a n d explaining what are taken to b e characteristic differences in the discourse styles of w o m e n and men. This research tradition (now sometimes referred to variously as "modernist,'' "essentialist," or "second wave," b y contrast with the subsequent "postmodernist," "social constructionist," or "third w a v e " approaches I describe below ) is usually thought to h a v e emerged in the m i d 1970s, with Robin Lakoff s pioneering essay Language and Woman's Place (1975). Lakoff's account of male-female differences exemplified what came to b e called the " d o m i n a n c e " current, which related the Unguistic behaviour of m e n and w o m e n to their unequal social positioning. Later, an alternative "[cultural] difference" a p p r o a c h emerged, in which less emphasis was given to inequalities of power: rather, m e n and w o m e n (and, m o r e especially, boys a n d girls) were seen as m e m b e r s of distinct (sub)cultures within which they acquired differing n o r m s for everyday communication (Maltz and Borker 1982). Since the interactional socio-linguist D e b o r a h Tannen popularized a version of this thesis in h e r best-selling b o o k You Just Don't Understand (1990), a host of pop-psychology a n d self-help writers (notably J o h n " M a r s and Venus" G r a y [1992]), h a v e reinforced its status as folklinguistic c o m m o n sense. 1
Yet most language and gender research today is informed neither b y this view n o r b y the alternative " d o m i n a n c e " view. At the same time that Tannen's work
Source: Intercultural Pragmatics vol. 2, no. 3, 2005, pp. 321-334.
CAMERON
RELATIVITY AND ITS DISCONTENTS
323
was prompting criticism from some of h e r colleagues (e.g., Freed 1992; TroemelPloetz 1991; Uchida 1992), m a n y researchers were questioning some of the funda mental theoretical assumptions on which b o t h " d o m i n a n c e " and "difference" approaches relied, suggesting that b o t h were flawed b y their "essentialist" view of gender. T h e popularity of Tannen's w o r k on male-female misunderstanding, and of the reductive a n d stereotypical "Mars and Venus" accounts that followed, gave increased urgency to this questioning process. W h a t emerged over the next few years was a n e w approach which, as I noted in passing above, has b e e n given a n u m b e r of labels, including "postmodernist" (Cameron, in press), "social con structionist" (Weatherall 2002), a n d "third wave" (Baxter 2003). I have given a m o r e detailed account of this shift and its implications elsewhere (Cameron, in press); the table below provides abrief summary of the key theoretical assumptions characteristic of the n e w a p p r o a c h a n d h o w these contrast with the assumptions m a d e in earlier work. Within w h a t I will henceforth follow the majority of commentators b y calling the "social constructionist" paradigm (though with the terminological reservations 2
Table 1: Contrasting approaches to language and gender "ModernistV'second wave"/ "essentialist" approach
"Postmodernist"/"third wave"/ "social constructionist" approach
Views gender as a stable attribute of individuals, a part of their core identity that is normally acquired in the course of their early socialization. Languageusing reflects or marks the speaker's (pre-existing) gender identity. Focuses on binary gender difference: presupposes the existence of two internally homogeneous groups, "men" and "women," and looks for generalizations about differences between them. Most studies deal with "mainstream" gendered subjects - often white, straight, middle-class, and monolingual.
Views gender as ongoingly "performed" or constructed through engagement in particular local practices; gender is not something you "have" or "acquire" once and for all early in life but something you do continually. Language-using is part of "doing/performing" gender and thus is constitutive of gender. Focuses on gender diversity: assumes that there are multiple masculinities and femininities, inflecting or inflected by other dimensions of social identity (e.g., class, ethnicity, generation); intra-group differences and inter-group similarities are as significant as differences between gender groups. Is also interested in "liminal" cases which challenge the binary logic of gender (e.g., transgendered or "queer" speakers). Favours "looking locally" for explanations of gender-related patterns: masculinities and femininities are produced in specific contexts or "communities of practice," and it cannot be assumed that the same patterns will be found in all or most cases.
Favours global generalizations and "big stories": linguistic gender differences are explained in terms of overarching social structures, e.g., male dominance or separate gendered subcultures.
324
DISCOURSE STUDIES
expressed in n o t e 1), it is possible to distinguish several sub-currents. Some researchers, most of t h e m socio-, applied or "critical" linguists, are influenced b y either J u d i t h Butler's formulation of gender as "performative" (Buder 1990) or the "communities of practice" m o d e l drawn from Lave and Wenger (1991) and most influentially disseminated to language a n d g e n d e r scholars through the work of Penelope Eckert a n d Sally McConnell-Ginet (1992,1999) - or, in some cases, b y b o t h of these. O t h e r s espouse a form of social constructionism that is theoretically rooted in the ideas of H a r o l d Garfinkel a n d Erving Goffman (and feminist elaborations thereof, such as Kessler a n d M c K e n n a 1978; West and Z i m m e r m a n 1987), a n d their preferred approach to the analysis of discourse is c o n v e r s a t i o n analysis (CA), originally a n offshoot of Garfinkelian ethnomethodology. M a n y researchers in the latter group h a v e a disciplinary affiliation to social psychology; s o m e are adherents of "discursive psychology" (cf. Potter and Wetherell 1987) w h o c o m b i n e C A m e t h o d s with poststructuralist theoretical ideas about subjectivity, identity, and "discourse" in the Foucauldian sense (e.g., Edley a n d Wetherell 1997; Wetherell 1998).
2. D e b a t e s w i t h i n Social Constructionism A m o n g social constructionist language a n d g e n d e r researchers, a n d especially the CA-oriented group, there has b e e n vigorous debate during the past few years on the issue of "realism versus relativism" (e.g., Speer 2000). M u c h of this has turned on the methodological question of whether it is legitimate for analysts of discourse to m a k e g e n d e r relevant in analysing data w h e r e the participants themselves do not orient to it explicitly as a concern (see e.g., Kitzinger 2000; Stokoe and Smithson 2 0 0 1 ; Weatherall 2 0 0 0 , 2 0 0 2 ) . I n "classical" C A (see e.g., Schegloff 1997,1998) it is axiomatic that one does n o t go "beyond the data" nor impose analysts' categories in place of participants' o w n (a position that follows the logic of ethnomefhodology, with its insistence o n viewing social order as a local " m e m b e r s ' accomplishment"). For m a n y feminists, o n the other hand, the "classical" approach is o p e n to the criticism that it denies the relevance of gender differentiation/inequality as an overarching principle of social organization, and so prevents analysts from taking proper account of its effect on all social interaction. This debate is sometimes presented as a n argument between m o r e and less radically constructionist positions. G e n d e r "realists" hold that even if gender is diverse a n d locally variable in its actual forms, differentiation/inequaUty between w o m e n a n d m e n is nevertheless a "real" social fact that in some form can be assumed to h a v e "global" relevance. G e n d e r "relativists" d e n y this, arguing thai gender is radically local: it is only "there" w h e n participants in a n interaction explicidy make it relevant, a n d an analytically satisfying account of their behaviour in that case should n o t n e e d to i m p o r t from outside the data m o r e general or "global" assumptions about gender as a macro-social category. S o m e researchers w h o take this position would see the use of global assumptions about gender as coverdy reintroducing a form of essentialism, a n d so vitiating the analysts' claim to b e working within a social constructionist framework. Yet some of their
CAMERON
325
RELATIVITY AND ITS DISCONTENTS
opponents would retort that the kind and degree of essentialism gender "realism" represents is a strategic necessity if discourse analysis is to serve feminist ends. T h e debate, then, is n o t only technical a n d methodological, a question of h o w to go about the practical business of analysing discourse data, b u t also political, a question of what feminist discourse analysis is or should b e centrally about. For "realists," o n e key goal of feminist analysis is to show h o w unequal g e n d e r relations are ( r e p r o d u c e d in various kinds of talk a n d text. "Relativists," while they d o not necessarily reject that goal, tend to define a feminist analysis as one that eschews a priori assumptions about gender, seeking instead to reveal the constructed a n d contingent nature of g e n d e r categories themselves. This debate m a y look quite different from the earlier " d o m i n a n c e versus difference" debate, b u t while there are indeed some significant differences, there are also similarities. I n the "dominance versus difference" case, too, there was a dispute a b o u t the relationship of gender to p o w e r (whether it is legitimate for analysts to postulate that one always implicates the other), which was played out in arguments about analytic p r o c e d u r e . I n her p a p e r " T h e relativity of linguistic strategies," D e b o r a h T a n n e n (1993) w a r n e d against what she regarded as the typical " d o m i n a n c e " analytic p r o c e d u r e of reading p o w e r relations off from the presence in discourse of particular linguistic forms or discourse strategies - for instance, equating the act of interrupting someone with asserting dominance over t h e m . S h e a r g u e d that since the m e a n i n g of a n y form or strategy is contextually "relative," this way of proceeding analytically risks finding meanings in discourse that were n o t salient for the participants, a n d attributing to the participants intentions they did not have. (For example, someone w h o interrupted an interlocutor m a y h a v e intended not to assert d o m i n a n c e over t h e m but to signal solidarity a n d support for what they were saying.) T h o u g h the argument is framed differendy, there are surely resemblances between Tannen's concerns about " d o m i n a n c e " analysts imposing their o w n interpretation a n d / o r political a g e n d a o n discourse data a n d the concerns of contemporary analysts favouring the "classical" C A approach. M o r e generally, it seems feminist discourse analysts a d e c a d e o n are still arguing about s o m e of the same questions the " d o m i n a n c e versus difference" debate turned on, albeit within a different theoretical framework. W h e r e , as analysts, should we look for the workings of g e n d e r in discourse? O n w h a t grounds is it legitimate to argue for the relevance of gender, and of p o w e r relations between g e n d e r groups, in analysing a particular piece of data? T h o u g h the answers proposed have changed, these questions h a v e evidentiy n o t b e e n resolved or transcended b y the turn to social constructionism. A m o n g today's social constructionists, it is generally accepted that discourse strategies do h a v e the property Tannen calls "relativity": the m a p p i n g from linguistic form to discourse function is typically many-to-many rather than oneto-one. But as I pointed out in a response to Tannen's 1993 article (Cameron 1998), accepting the postulate of relativity does not in itself resolve all disputes. Relativity, after all, is not only a n issue for discourse analysts; it is also a practical problem for participants in discourse. If interaction is to g o o n in the ways w e know it does, participants in any specific case must ha *• 1
32(i
DISCOURSE STUDIES
multiplicity of interpretations that might in principle b e licensed b y an utterance or turn-at-talk, and decide what, in that context, would constitute the "best" or most plausible interpretation. T h e question of h o w they do that - a n d h o w the process might b e affected b y g e n d e r a n d / o r p o w e r differences - is not, I would argue, satisfactorily answered in either Tannen's account or the m o r e "radical" social constructionist accounts that h a v e succeeded it. Tannen's argument was that recipients in practice resolve ambiguities by unconsciously invoking gender-specific assumptions about the function and p r o p e r conduct of social interaction. M e n place utterances addressed to t h e m in a "status" frame, whereas w o m e n place the same utterances in a "connection" frame: the result is often misunderstanding. T h u s Tannen (1993) analyses a vignette in which two co-workers, a w o m a n a n d a m a n , are walking between buildings o n a cold day. T h e w o m a n asks the m a n w h e r e his coat is (i.e., remarks o n the fact that h e is n o t wearing one), a n d the m a n responds, "thanks, m o m . " According to Tannen, the w o m a n ' s r e m a r k was intended to convey friendly concern, b u t since the m a n placed it in a status-oriented frame, h e took it as a put-down and responded by putting the w o m a n down - "thanks, mom," addressed b y an adult to s o m e o n e other t h a n their m o t h e r , implies that the ad dressee is inappropriately bossy a n d interfering, a n d is subordinating the other b y treating t h e m as a helpless or irresponsible child. I n the analytic framework favoured b y those o n the "relativist" side of the current debate, there is n o place for the kind of global assumptions about gender (i.e., that m e n as a group seek status a n d w o m e n connection) that are invoked in Tannen's analysis of the "coat" vignette. However, the relevance of g e n d e r to this exchange would b e acknowledged in a "classic" C A analysis m a k i n g use of the "next turn proof procedure." It is a general principle in C A that a recipient's understanding of the prior turn will b e displayed in the response t h e y produce in their own next turn: the analyst can therefore warrant claims about what an utterance m e a n t b y looking to the n e x t turn. I n the "coat" case, the m a n ' s use of the gender-specific appellation " m o m " is evidence that h e takes g e n d e r to b e relevant. If, o n the other h a n d , h e h a d said something like [sarcastically] "I ap preciate your concern," it would still b e possible for the analyst to assert that the m a n interpreted the w o m a n ' s r e m a r k as inappropriate or uncalled for, b u t there would b e nothing to warrant the further claim that this h a d anything to do with gender. So whereas in Tannen's a p p r o a c h gender is globally relevant, a n d b y implication factors into every act of interpretation, in the C A a p p r o a c h its relevance is limited to just those cases w h e r e participants explicitly display their orientation to it b y mentioning it. This is where m y o w n argument begins; for in m y view, both these approaches are unsatisfactory: one does n o t sufficiently restrict the claims a n analyst can m a k e about the interpretive behaviour of participants in discourse, while the other, conversely, is too restrictive, suggesting that nothing can b e relevant for analysis unless its relevance for the participants is overtly displayed in the words they actually utter. Between these extremes, I would suggest, is another possibility: making use of the insights of Gricean pragmatics (Grice 1975), and the elaboration of those ideas in relevance theory (Sperber a n d Wilson 1986).
CAMERON
RELATIVITY AND ITS DISCONTENTS
327
3 . A "Pragmatic Turn"? T h o u g h there are some i m p o r t a n t differences b e t w e e n the Gricean m o d e l of communication a n d relevance theory (which limitations of space p r e v e n t m e from pursuing here), the k e y insight these pragmatic approaches share is that m e a n i n g is inferred rather t h a n decoded: derived not simply b y processing the manifest content of a n utterance or text, b u t b y putting that together with pro positions which are retrievable from context or from background knowledge in order to generate inferences about what is meant or intended. Rather than explicifly providing everything necessary for a n adequate interpretation, utterances/texts will provide evidence that cues the hearer to look for other propositions (i.e., ones that are n o t explicitly encoded) that m a y b e relevant for interpretation. T h e r e is, of course, n o guarantee that the inferences generated b y hearers will lead t h e m to the precise m e a n i n g a speaker intended to communicate. (Meaning is, in that sense, "relative" or indeterminate.) But the inference-based account does imply that in the n o r m a l case the hearer will m a k e the effort necessary to generate a m e a n i n g : s/he will n o t simply conclude that since all strategies are multifunctional a n d polysemous, what this speaker o n this occasion m e a n s b y X is undecidable. A n y inference-based account of m e a n i n g must immediately call into question the idea that something - for example, g e n d e r - has to b e m e n t i o n e d overfly in discourse to count as relevant for a n analysis of what was m e a n t or h o w it was interpreted. I n an inferential account of communication it is assumed that con versationalists in their active efforts to m a k e sense of discourse h a v e far m o r e to bring to b e a r o n that discourse t h a n what is explicidy and overtly e n c o d e d in it; furthermore, it is routinely expected that they will "go b e y o n d the data" i n this way. O n that basis, it is possible to argue for the potential relevance of gender (or rather, of conversationalists' background knowledge and beliefs about gender) even in cases w h e r e gender is n o t explicitly at issue in discourse. For e x a m p l e , the title of m y article "Is there any ketchup, Vera?" (Cameron 1998) quotes a n utterance which a friend of m i n e recalled h e r father addressing regularly to h e r m o t h e r at the family dinner table (the condiment varied - it might b e mustard or pickle rather than ketchup - but the form remained constant). I chose to e x a m i n e this particular case because it provided a counter-example to a claim m a d e in the literature o n male-female misunderstanding, that m e n h a v e difficulty in understanding indirect requests. "Is there any ketchup?" was clearly issued (in this case b y a man) a n d interpreted (by his wife Vera) as an indirect request, that is, b o t h of t h e m understood that the intention was n o t to request information about the availability or otherwise of ketchup, but to request (indirecdy, since the utterance is linguistically coded as a request for information) that Vera go a n d get the ketchup. M y friend also reported that whereas h e r father's use of this strategy invariably elicited the desired response (the action of getting ketchup), if she, the daughter, used the same strategy, her mother typically treated it as the request for information it a p p e a r e d to be, responding with an utterance such as, "Yes, it's o n the kitchen counter."
328
DISCOURSE STUDIES
H e r e we might ask, is it plausible to suggest - as a n analyst using the next turn proof procedure might b e led to suggest - that Vera understands the same utterance as a n indirect request w h e n it comes from her husband, but as a request for information w h e n it comes from her daughter? T h e r e is certainly a difference in h e r overt uptake of the utterance in the two cases, b u t it surely does not follow that Vera must believe the underlying communicative intention to be different. I n a pragmatic framework her contrasting responses are m o s t readily explained as a consequence of her deploying certain beliefs (which in this familial context she takes to b e b a c k g r o u n d knowledge for all parties present) about the nature of h e r obligations to other family m e m b e r s : specifically, that a husband w h o wants ketchup is entitled to h a v e it fetched b y his wife, whereas a daughter w h o wants ketchup is n o t entitled to the same service from her mother. I n that case, beliefs about g e n d e r (more exacdy, about the g e n d e r e d roles of wife/ mother, h u s b a n d , a n d daughter) are relevant for the analysis of Vera's discourse, even t h o u g h gender is n o t explicitly m e n t i o n e d anywhere in the exchanges u n d e r discussion. In m y article, I m a d e the m o r e general suggestion that m a n y cases of socalled "male-female misunderstanding" do not involve any misunderstanding (in the sense of failure to retrieve the speaker's intended meaning or some plausible "best guess" at it), b u t are rather instances of conflict b e t w e e n m e n a n d women w h o are either bringing to b e a r different background assumptions about their roles, rights and obligations, or strategically resisting w h a t they infer to b e the other's assumptions (as Vera is doing w h e n she pretends n o t to hear her daughter's "is there any ketchup" as a request whose i n t e n d e d m e a n i n g is identical to her h u s b a n d ' s w h e n h e utters the same formula). It is for this reason, I argued, that the alleged p r o b l e m of male-female misunderstanding appears to b e c o m m o n in societies w h e r e the influence of feminism has m a d e g e n d e r relations a matter of contestation, whereas it does not appear to prompt the same concern in societies w h e r e m o r e e x t r e m e g e n d e r segregation a n d m a l e d o m i n a n c e r e m a i n the unquestioned norm. If the root cause of the problem were what Tannen claimed that males a n d females acquire their habitual ways of using language in separate and highly divergent subcultures - o n e would surely expect to find exacdy the opposite. T h e kind of conversational g e n d e r trouble Tannen struck such a chord by describing is in m y view related, then, n o t to differences in m e n ' s a n d women's ways of communicating, but to the extent to which they diverge in their assumptions about gender itself. Vera a n d h e r h u s b a n d understand one another perfecdy a n d there is n o conflict between them, because although their social roles are differentiated along conventional gender lines, they b o t h accept this differenti ation as appropriate: they do not disagree o n what each is entitied to expect from the other a n d obligated to provide for the other. For m a n y less traditional couples, this is m o r e a matter of contention and the resulting conflicts are played out in discourse - a m o n g other things, through b o t h real a n d strategic "misunder standings." T h e arguments just outiined (Cameron 1998) w e r e developed in response to Tannen's "cultural difference" approach, b u t they seem to m e n o less pertinent
CAMERON
RELATIVITY AND ITS DISCONTENTS
329
to the debates going o n n o w a m o n g social constructionists. Arguably, even the strictest "relativists" coverdy d e p e n d o n mobilizing the kinds of b a c k g r o u n d assumptions about gender I invoked in the above discussion of the "ketchup" example. Even where the relevance of gender is not disputed, because it is explicitly m a d e relevant - as in the utterance of "thanks, m o m " in Tannen's "coat" vignette the analyst in effect has to "go b e y o n d the data" in order to say anything a b o u t h o w or w h y gender is being m a d e relevant. To u n d e r s t a n d w h y "thanks, m o m " can b e h e a r d in this context as taking t h e prior turn to b e i n a p p r o p r i a t e , j u d g m e n t a l , a n d u n w e l c o m e t o the recipient, we n e e d to access not m e r e l y the dictionary m e a n i n g of " m o m " b u t a cluster of assumptions a n d beliefs about social roles and behaviours to which that appellation is likely to direct the hearer's attention, thus enabling h e r to infer that her previous utterance was unwelcome. A conversation analyst m i g h t reply at this p o i n t that C A is simply n o t interested in what might b e going on in conversationalists' m i n d s - for instance, their intentions a n d their reasoning about others' intentions: what matters is to account for the observable orderliness of their conversational behaviour. As a microanalytic a p p r o a c h which looks at the organization of discourse turn b y turn, CA's great strength is its ability to provide answers to the question "why that n o w ? " or, in other words, give a n account of what is b e i n g accomplished locally b y the production of a particular turn at a particular m o m e n t in the unfolding interaction. But as a n u m b e r of commentators h a v e pointed out, there are clearly supra-local patterns in the way gender is enacted a n d talked about, which cannot b e accounted for using a purely "local" analytic approach. To answer the question "why that n o w ? " it is often necessary to presuppose an answer to the prior question "why that ever?" For example, analysts have repeatedly found people doing extensive identity work in order to claim certain identities/attitudes relating to sex and g e n d e r while disclaiming others: they are at pains to establish themselves as g e n d e r e d in ways that avoid the extremes of either sexism or militant feminism, a n d as heterosexual but n o t h o m o p h o b i c (Edley and Wetherell 1997; Speer 2 0 0 2 ; Speer and Potter 2000). I n their accounts of this behaviour, analysts implicidy m a k e use, expect their readers to m a k e use, and for that matter m u s t suppose that participants themselves were m a k i n g use, of such common-sense propositions as "a n o r m a l person is unambiguously either a w o m a n or a m a n , " "homosexual identities are stigmatized," a n d "being openly prejudiced against certain groups is socially disapproved behaviour." T h e fact that participants i n talk recurrendy take certain issues as requiring careful m a n a g e m e n t makes sense only in relation to background assumptions like those just listed, which are "global" rather than "local" in the sense that they pre-exist the particular conversation being analysed. Strictiy speaking, it is "going b e y o n d the data" to invoke t h e m : b u t if you do n o t invoke them, y o u are arguably missing an important dimension of participants' own behaviour - the use they routinely m a k e of knowledge drawn from sources other than the ongoing interaction itself. With regard to g e n d e r (and other socio-demographic variables such as class, age, and race/ethnicity), we ignore or d o w n p l a y this dimension of participants' behaviour at our peril, since one kind of knowledsrp that < •
330
DISCOURSE STUDIES
to bear on inferring others' probable communicative intentions is their knowledge about the person and/or the type of person they are dealing with. Though the theoretical developments discussed above in connection with the social constructionist turn have led to a strong focus on the speaker's performance of gender, it should not be forgotten that gender is not only performed b y speakers but also ascribed to them b y the recipients of their discourse. A s Weatherall (2000,2002) has pointed out, there are few situations in which an interlocutor's gender is not manifested - visually, vocally, or both in the case of spoken inter action, and through the explicit naming of authors which is typical, though not invariable, in written texts - and also few situations where it is not regarded by most people as salient information. Whenever I speak, the information that I am a woman is available (conveyed b y the pitch and timbre of m y voice, as well as m y bodily appearance in face-to-face contexts) to be used by m y interlocutors in their interpretations; even if m y discourse does not evidence any intention to make m y gender relevant, interlocutors can always choose to treat it as such. One manifestation of sexism as a "global'' social force is the propensity to treat gender, particularly the gender of w o m e n as the marked and subordinated category, as omnirelevant, and, on that basis, to generate inferences which are not warranted by anything in an individual's actual behaviour, but are rather based on propositions about the social category that individual belongs to. Like m a n y w o m e n working in highly traditional male-dominated institutions (in my case, Oxford University), I am made aware of this propensity on a daily basis. I pick up the phone and the person who is calling assumes I must be the professor's secretary rather than the professor. I enter a college or university building and the porter assumes I am a student rather than a member of the faculty. Sometimes, the gendered basis for the inference is indeed displayed in what the caller or the porter says (e.g., the porter addresses m e as "my love," which he would not do to someone he knew to be a senior academic, nor to a male of any social/ professional status); but e v e n w h e n m y interlocutor's judgement of me is displayed through a gender non-specific kind of disrespect (e.g., "hey, where d'you think you're going?"), I know very well, from experience and observation of similar scenarios involving men and w o m e n over time, that I am being treated in this way because of the meaning m y interlocutor attaches to m y gender. Similarly, people of colour learn to identify certain kinds of disrespect as connected to their race/ethnicity, though in this case it is (nowadays) even less likely than in the gender case that the interlocutor will make explicit reference to category membership in performing the act of disrespect. In these examples, there is a temporal or historical dimension: it is the repetition of the same pattern on many different occasions which leads the recipient of disrespectful behaviour to infer, perhaps on the basis of what might otherwise seem scanty or equivocal evidence, that gender or race/ethnicity must also be relevant on this particular occasion. The local instance, in other words, is interpreted in relation to a whole history of comparable instances, as part of a more "global" pattern. So another problem with "relativist" approaches which insist on confining analysis strictly to what can be observed in the local instance is that they downplay or even deny the relevance of participants' histories, and
CAMERON
RELATIVITY AND ITS DISCONTENTS
331
the cumulative knowledge they h a v e built u p during those histories, to their behaviour at a n y given m o m e n t . It is p e r h a p s worth pointing out that J u d i t h Butler's m o s t famous statement about t h e socially constructed a n d performative n a t u r e of g e n d e r actually emphasizes repetition a n d temporality: " G e n d e r is the repeated stylizaiion of the body, a set of repeated acts within a highly rigid regulatory frame, which congeal over time to p r o d u c e t h e appearance of substance, of a 'natural' kind of being [emphasis m i n e ] " (Butler 1990: 33). T h e g e n d e r "realist" Margaret Wetherell somewhat similarly suggests that interpretative strategies "congeal over time": that rather than coming to every n e w utterance or exchange with n o preconcep tions, communicators m a k e use of what she calls "interpretive repertoires ... culturally familiar a n d habitual line(s) of argument comprised of recognizable themes, commonplaces, a n d tropes" (Wetherell 1998:400). She goes o n to suggest that in declining to entertain such non-local constructs o n the principle of n o t going b e y o n d t h e d a t a a n d n o t i m p o s i n g a n a l y s t s ' c o n c e r n s i n p l a c e of participants' o w n , p r o p o n e n t s of classical C A are, i n fact, imposing their o w n very n a r r o w concept of what is relevant for participants locally: It is the conversation analyst in selecting for analysis part of a conversation or continuing interaction w h o defines this relevance for the participant. I n restricting t h e analyst's gaze to this fragment, previous conversations, even previous turns in t h e same conversation, b e c o m e irrelevant for t h e analyst, b u t also b y diktat for t h e participants. We do n o t seem to h a v e escaped, therefore, from t h e imposition of theorists' categories a n d concerns. (Wetherell 1998: 402) I agree with Wetherell's point h e r e as far as it goes; b u t what I h a v e b e e n arguing in this discussion is that t h e p r o b l e m s she is pointing to might usefully b e addressed b y adopting a n a p p r o a c h to t h e analysis of discourse that is informed by the insights of relevance theory. T h o u g h this theory is often thought of as an asocial, cognitivist approach, it is arguably capable of b e i n g applied to p h e n o m e n a which a r e of interest to feminist a n d other "critical" discourse analysts: o n e could focus o n t h e social a n d ideological p r o p o s i t i o n s t h a t conversationalists h a v e access t o , a n d that they u s e in m a k i n g inferences. This approach might enable analysts to give m o r e substance a n d detail to concepts like "interpretive repertoire" a n d to describe h o w in specific instances people mobilize particular interpretive repertoires. It might also enable analysts to ask questions about t h e social distribution of those repertoires a n d t h e assumptions associated with t h e m (i.e., whether there a r e consequential social a n d cultural differences at t h e level of w h a t people take to b e " c o m m o n knowledge" or i n Wetherell's phrase "culturally familiar a n d habitual lines of argument"), without falling into the kind of cultural essentialism for which Tannen's version of inter actional sociolinguistics h a s b e e n criticized. P e r h a p s t h e m o s t i m p o r t a n t insight p r a g m a t i c s h a s t o offer to social constructionist feminist discourse analysis, though, is a general- rh#»nrAhVai
DISCOURSE STUDIES
that in "going b e y o n d the data," or bringing "global" assumptions to bear o n local instances, analysts n e e d n o t b e automatically thought of as imposing their own procedures in place of the participants'. I n a n y inferential approach to communication, it is axiomatic that participants are using other information besides what is actually e n c o d e d in discourse to derive m e a n i n g from what is said. By seeking to m o d e l this process in a principled way, w e might h o p e to arrive at a better understanding of - a n d p e r h a p s a greater degree of consensus about - t h e relevance of gender for communicational behaviour.
Notes 1. There are problems with all these terminologies. The "second/third wave" distinction is somewhat obscure (it alludes to "second wave" feminism, i.e., the post-1968 liberation movement as opposed to the "first wave" nineteenth and early twentieth century suf fragist movement, and posits a more recent "third-wave" feminism whose putative date of onset and distinctive characteristics are, however, disputed). The "essentialist/ social constructionist" distinction is highly value-loaded ("essentialist" has become a smear-term rather than a descriptive label), and is also misleading inasmuch as all feminist theory since Simone de Beauvoir ("one is not born but becomes a woman") could apdy be described as "social constructionist": the whole concept of gender, as distinct from sex, is inherently social constructionist. "Modern/postmodern" is my own preferred terminology, but I concede it has the (considerable) disadvantage of invoking one of the most variable and contested terms in current theoretical discourse (i.e., "postmodernism]"). 2. This table presents the two approaches in an "ideal-typical" form that has the effect of maximizing the degree of difference between them: it should be borne in mind, as always with such oppositions, that the reality is more complicated. It should also be noted that the shift from "second" to "third wave" assumptions is not a matter of simple chronology (that implication is another reason why the "wave" terminology is problem atic). There was no historical moment before which everyone subscribed to the as sumptions set out in the first column, and after which everyone subscribed to the assumptions set out in the second column. Rather there has been a shift in the overall balance between two historically overlapping tendencies, such that "third wave* assumptions are now dominant (which does not mean "second wave" assumptions have vanished altogether).
1
References
Baxter, Judith. 2003. Positioning Gender in Discourse: Feminist Poststructuralist Discourse Analysis. Houndmills, UK: Palgrave. Butler, Judith. 1990. Gender Trouble: Feminism and the Subversion of Identity. New York: Roudedge. Cameron, Deborah. 1998. Is there any ketchup, Vera? Gender, power, and pragmatics. Discourse & Society 9:437-55. . (In press). Language, gender, and sexuality: Current issues and future directions. Applied Linguistics. 4
CAMERON
RELATIVITY AND ITS DISCONTENTS
333
Eckert, Penelope and Sally McConnell-Ginet. 1992. Think practically and look locally: Language and gender as community-based practice. Annual Review ofAnthropohgy 12: 461-90. . 1999. New generalizations and explanations in language and gender research. Language in Society 28: 185-201. Edley, Nigel and Margaret Wetherell. 1997. Jockeying for position: The construction of masculine identities. Discourse & Society 8: 203-17. Freed, Alice. 1992. We understand perfectly: A critique of Tannen's view of cross-sex communication. In Kira Hall, Mary Bucholtz, and Birch Moonwomon (eds.), Locating Power. Proceedings of the Second Berkeley Women and Language Conference. Vol. 1. Berkeley, CA: Berkeley Women and Language Group. 144—152. Gray, John. 1992. Men Are from Mars, Women Are from Venus. New York: HarperCollins. Grice, H. Paul. 1975. Logic and conversation. In Peter Cole and Jerry L. Morgan (eds.), Syntax and Semantics: Speech Acts. Vol. 3. New York: Academic Press. 41-58. Kessler, SuzanneJ. and Wendy McKenna. 1978. Gender: An EthnomethodohgicalApproach. New York: Wiley. Kitzinger, Celia. 2000. Doing feminist conversation analysis. Feminism & Psychology 10: 163-93. Lakoff, Robin. 1975. Language and Woman's Place. New York: Harper and Row. Lavejean and Etienne Wenger. 1991. Situated Learning: Legitimate Peripheral Participation. Cambridge, UK: Cambridge University Press. Maltz, Daniel and Ruth Borker. 1982. A cultural approach to male-female misunder standing. In John J. Gumperz (ed.), Language and Social Identity. Cambridge, UK: Cambridge University Press. 196-216. Potter, Jonathan and Margaret Wetherell. 1987. Discourse and Social Psychology: Beyond Attitudes and Behaviour. London: Sage. Schegloff, Emmanuel. 1997. Whose text? Whose context? Discourse & Society 8: 165-87. . 1998. Reply to Wetherell. Discourse & Society 9:413-16. Speer, Susan. 2000. Let's get real? Feminism, constructionism, and the realism/relativism debate. Feminism ^Psychology 10:539-50. . 2002. Sexist talk: Gender categories, participants' orientations and irony. Journal of Sociolinguistics 6: 347-77. Speer, Susan and Jonathan Potter. 2000. The management of heterosexist talk: Conversational resources and prejudiced claims. Discourse & Society 11: 543-72. Sperber, Dan and Deirdre Wilson. 1986. Relevance. Oxford, UK: Blackwell. Stokoe, Elizabeth and Janet Smithson. 2001. Making gender relevant: Conversation analysis and gender categories in interaction. Discourse & Society 12: 217-44. Tannen, Deborah. 1990. You Just Don't Understand: Men and Women in Conversation. New York: Morrow. . 1993. The relativity of linguistic strategies. In Deborah Tannen (ed.), Gender and Conversational Interaction. New York: Oxford University Press. 165-185. Troemel-Ploetz, Senta. 1991. Selling the apolitical. Discourse & Society 2: 489-502. Uchida, Aki. 1992. When "difference" is "dominance": A critique of the "anti-powerbased" tendency in language and gender. Language in Society 21: 547-68. Weatherall, Ann. 2000. Gender relevance in talk-in-interaction and discourse. Discourse & Society 11: 290-292. . 2002. Gender, Language and Discourse. London: Routledge. West, Candace and Don Zimmerman. 1987. Doing gender. Gender and Society 1 : 125-151. Wetherell, Margaret. 1998. Positioning and interpretive repertoires: d n n v u r c a K ™ - — > — - -
14 The Analysis of Discourse Flow Wallace Chafe
Introduction
L
anguage is a dynamic process. It is easy to forget that fact w h e n one is working with language that has b e e n frozen o n p a p e r or a computer screen, w h e r e it has b e e n t u r n e d into something that can b e examined as if it w e r e a fixed object. So m u c h of linguistic analysis has dealt with language in written form that there is a temptation to think of language itself as having the same static quality (cf. Linell 1982). But language in action is better captured with the m e t a p h o r of a flowing stream. T h e r e are, in fact, two streams, one a stream of thoughts, the other of sounds. T h e two h a v e very different qualities. It is instructive to c o m p a r e the experience of listening to a familiar language with listening to a language o n e does not know. I n the former case it is the thoughts, n o t the sounds, of which one is con scious, b u t in the latter case only the sounds. Sounds are easier for a n analyst to deal with, simply because they are publicly observable. Thoughts are experienced within the m i n d , a n d for that reason are less tractable to objective research. On the other h a n d thoughts enjoy a priority over sounds in the sense that the organ ization a n d communication of thoughts is what language is all about. T h e sounds exist in the service of the thoughts, a n d follow wherever the thoughts m a y take them. It is the thoughts that drive language forward. A basic challenge for dis course analysis is to identify the forces that give direction to the flow of thoughts.
1 Topics I m p o r t a n t a m o n g these forces are w h a t I will b e calling topics. This word has b e e n used in different ways, a n d I should m a k e it clear that I a m n o t using it to Source: Deborah Schiffrin, Deborah Tannen and Heidi E. Hamilton (eds), The Handbook
CHAFE THE ANALYSIS OF DISCOURSE FLOW
335
apply to a constituent of a sentence, as w h e n one speaks of a sentence having a "topic and c o m m e n t " (e.g. Hockett 1958:201), or of "topic-prominent" languages (Li a n d T h o m p s o n 1976), or of "topicalization" or "topic continuity" (e.g. G i v d n 1983). Rather, I a m using it to refer to w h a t is sometimes called a "discourse topic" (Brown a n d Yule 1983: 71), as in "the topic of this paragraph." A topic in this sense is a coherent aggregate of thoughts introduced b y s o m e participant in a conversation, developed either b y that participant or another or b y several participants j o i n d y , a n d t h e n either explicitly closed or allowed to peter out. Topics typically h a v e clear beginnings, although that is n o t always the case (cf. Tannen 1984: 41-3), a n d their endings are sometimes well defined, sometimes not. A s l o n g as a topic remains open, participants in a conversation experience a drive to develop it. I b e g a n chapter 10 of Chafe (1994) with a quote from William J a m e s that nicely captures this drive: I n all our voluntary thinking there is s o m e topic or subject about which all the m e m b e r s of the thought revolve. Half the time this topic is a problem, a gap w e cannot yet fill with a definite picture, word, or phrase, b u t which . . . influences us in a n intensely active a n d determinate psychic way. W h a t e v e r m a y b e the images a n d phrases that pass before us, we feel their relation to this aching gap. To fill it u p is our thought's destiny. Some b r i n g us n e a r e r to that consummation. S o m e the g a p negates as quite irrelevant. Each swims in a felt fringe of relations of which the aforesaid g a p is the term. ( J a m e s 1890, vol. 1: 259) Sensitivity to the topic structure of talk m a y b e a trait that varies with individ uals. Casual observation suggests that p e o p l e are constrained to varying degrees by the n e e d to develop a topic fully before the conversation m o v e s o n to another, and that there is variable recognition of the social right to topic development. O n e wonders if such differences in conversational style can b e traced to differ ences in the degree to which a p e r s o n experiences J a m e s ' s aching g a p a n d the need to fill it. A first step in discourse analysis can b e to listen to a recording of a conver sation with the goal of identifying topics, segments of discourse during which one or m o r e of the speakers talk about "the s a m e thing." Topics are identifiable above all from their content, b u t there are likely to b e phonetic cues as well: sometimes, t h o u g h certainly not always, a longer-than-normal pause before a new topic is introduced; sometimes heightened pitch, loudness, acceleration, or a new voice quality at the outset; sometimes a tapering off in these same prosodic features at the end. O n e m a y find topics varying greatly in length. T h e r e m a y be occasional stretches of discourse during which there appears to b e n o topic at all. But most parts of most conversations lend themselves well to analysis into units of this kind. There appears to b e a basic level of topic-hood, with topics at that level typically included within m o r e inclusive supertopics. T h e latter also have identifiable begin nings a n d endings, but they lack the internal structure that characterizes basiclevel topics a n d d o n o t generate the same drivp f o r r l n c " « > T » — — » —
DISCOURSE STUDIES
336
Each time a basic-level topic is concluded, a n y participant in a conversation has the option of a b a n d o n i n g the current supertopic and, b y introducing a n e w basic-level topic, introducing a n e w supertopic at the same time. With n o internal structure of their own, supertopics can b e a b a n d o n e d whenever a n y included basic-level topic has b e e n completed. After a particular basic-level topic, or some sequence of them, has b e e n chosen for further study, the next step can b e to r e d u c e the flow of language to some written form. T h e w o r d reduce is appropriate. T h e r e is n o way in which the rich ness of natural speech, with all its prosodic complexity, its accompanying gestures and shifts of gaze, and ultimately the entire physical, social, and cognitive context in which it took place - n o way in which all these factors can b e captured in any presendy conceivable written form. T h u s , any attempt to transcribe spoken lan guage o n p a p e r inevitably leaves m u c h out. T h e transcriber needs a system that is m o r e or less adequate for the questions addressed, b u t needs always to keep in m i n d that any system only selects from the totality of observations that might b e relevant (Du Bois et al. 1993; Chafe 1993, 1995). It is useful in this process to identify a unit of transcription that reflects another level of organization. I n addition to basic-level topics, language gives evidence of the organization of thoughts from m o m e n t to m o m e n t into a focus a n d a per iphery: a limited area of fully active consciousness surrounded b y a p e n u m b r a of ideas in a semiactive state. Each focus is expressed in sound with a brief prosodic phrase, typically one to two seconds long, whose properties include o n e or more of the following: a distinctive terminal intonation contour, a n initial resetting of the pitch baseline, the presence of silence before a n d after, a change of t e m p o at the beginning or end, a n d b o u n d a r y changes in voice quality such as whispering or creaky voice. Intonation units are a pervasive feature of natural speech. Not only do they provide a useful w a y of segmenting speech, they are profitably viewed as expressing constantiy changing foci of consciousness, a n d h e n c e their relevance to understanding the flow of thought (Chafe 1994: 53-81). i " <' 1
2 Topic Navigation
I n this perspective a topic can b e seen as a conceptual unit that is too large to be a c c o m m o d a t e d within the limited capacity of fully active consciousness. A topic as a whole can thus b e present only i n a semiactive state. O n c e a topic has been introduced, the m o r e limited focus of active consciousness navigates through it. activating first one included idea a n d then another until the topic is judged to have b e e n adequately covered and closure is judged appropriate. This navigation process is often guided b y a schema, some familiar pattern that provides a path for a speaker to follow (e.g. Bartlett 1932; Chafe 1986). It m a y also b e driven, alternatively or simultaneously, b y the less predictable interaction between con versational participants (Chafe 1994: 120-36). I will illustrate this process with an excerpt from a long conversation in the course of which three w o m e n , w h o s e n a m e s will b e given h e r e as Kathy. Salh. a n d Chris, were discussing teaching practices in a n elementary school classroom
CHAFE
337
THE ANALYSIS OF DISCOURSE FLOW
K a t h y was a n experienced teacher, Sally was a less experienced teacher, a n d Chris was a less involved onlooker. We can take u p this conversation at a point where its forward m o v e m e n t was momentarily at a standstill. T h e previous topic h a d just b e e n closed, a n d if the conversation was to continue s o m e o n e h a d to choose a n d introduce a n e w topic. T h e preceding topics h a d fallen within the d o m a i n of a supertopic I will label Classroom Experiences. T h e default option during such a lull in a conversation is for any of the par ticipants to o p e n a new basic-level topic that remains within the current supertopic, in this case to talk about another classroom experience. T h a t choice would leave the supertopic Classroom Experiences open, a situation that can b e represented with an o p e n parenthesis. T h e introduction of a n e w basic-level topic would then create its own o p e n parenthesis, included within the other. This situation can b e represented as follows, w h e r e the supertopic is shown in italics: (ClassroomExperiences (Classroom Experience 1) (Classroom Experience 2) (Classroom Experience 3 T h e r e are two o p e n parentheses that d e m a n d eventually to b e closed. A n alternative would b e for a n y of the participants in the conversation to i n t r o d u c e a basic-level topic t h a t w o u l d close t h e current supertopic with Classroom Experience 2 a n d establish a n e w a n d different one. Imagine, for example, that s o m e o n e b e g a n talking n o w about a m o v i e she h a d just seen, introducing a n e w basic-level topic that w o u l d simultaneously o p e n a n e w supertopic that could b e labeled C u r r e n t Movies. T h e effect would b e : (Classroom Experiences (Classroom Experience l) (Classroom Experience 2)) (Current Movies (A movie just seen by X A s it h a p p e n e d , Sally chose the default option, opening another topic that remained within the Classroom Experiences supertopic. W h a t she said was: 1
(1) Sally (0.5) What I was gonna tell you about that really frustrates me is that, No one but Sally k n e w where this topic would lead, and for the m o m e n t we can give it the label Something Frustrating. Later we will see h o w the flow of the conversation would m a k e a different label appropriate. T h e words what I was gonna tell you about suggest that Sally h a d p l a n n e d to introduce this topic earlier. Examination of the larger context reveals that she had tried earlier to d o just that. She was unsuccessful in that first attempt be cause Chris interrupted h e r with a different topic. W h a t she said earlier was (numbered (0) because it lay outside the excerpt with which we will be princi pally concerned): (0) Sally.. Meanwhile in the principal's office they're telling me, Two other topics intervened before Sally returned to wh*t ^ Ill
ifll.
a fnni^
_
• •
U~J
DISCOURSE STUDIES
338
while the other topics w e r e being developed. It was thus easily available to b e reintroduced in (1), which was followed b y a second intonation unit whose word ing closely resembled that of (0), as w e will see.
3 Navigation by Schema We can n o w follow this conversation as it unfolded for those engaged in it. At the e n d w e can view a transcript of the conversation as a whole, at the same time considering what, exacfly, such a transcript represents. T h e Something Frustrating topic was at first developed b y Sally as a monologue. T h e r e is a ubiquitous schema for narrative topic development whose m a x i m u m c o m p o n e n t s can be listed as follows (cf. Chafe 1994: 120-36): • summary • initial state • complication • climax • denouement • final state • coda.
'
1
L a b o v a n d Waletzky (1967) suggested a similar schema, b u t inexplicably omitted the climax. A n o p e n i n g s u m m a r y m a y or m a y n o t b e present. Closer to being obligatory is the presentation of an initial state that gives the topic a spatiotemporal a n d / o r epistemic orientation. T h e complication section disturbs the initial state with events that lead to a climax, an unexpected event that constitutes the point of the topic, the reason for its telling. T h e d e n o u e m e n t then provides a relaxation toward a final state in which n e w knowledge provided b y the climax has b e e n incorporated. T h e r e m a y or m a y n o t b e a coda, a m e t a c o m m e n t on the topic as a whole. Sally's statement in (1) summarized the content of what would follow b y saying that it would entail something frustrating. N o t only did she o p e n a n e w topic and assume the floor, b u t at the same time b y using the word frustrates she fore shadowed its organization, creating a n expectation that it would involve some thing desirable followed b y an explanation of w h y that desirable outcome could n o t b e realized. D e c i d i n g just h o w to p r o c e e d r e q u i r e d additional mental processing time o n Sally's part, a n interval during which she uttered a prolonged hesitation sound, followed b y 1.3 seconds of silence a n d t h e n an audible breath before she continued: (2) (3)
Sally
(0.2) uh=, (1.3) (breath) the (0.1) the p e o p . . the principal and stuff they say to me,
I n (3) she repeated, with only partially different words, h e r earlier attempt to introduce this topic, shown above as intonation unit (0). Early in (3) she decided
CHAFE
THE ANALYSIS OF DISCOURSE FLOW
339
to mention the people w h o h a d given her advice. H e r truncated thepeop was an attempt at categorizing that idea, b u t she quickly found a better categorization and p r o d u c e d the interestingly h e d g e d phrase the principal and stuff, followed b y the quote-introducer they say to me. Looking back at (1), w e can see that Sally's consciousness was then operating in what I h a v e called the immediate mode (Chafe 1994: 195-223). T h a t is, Sally was talking about what was still frustrating h e r at the v e r y time she was talking. With (3), however, she m o v e d into the displaced mode b y shifting to things that had b e e n said to h e r at one or m o r e times in the past, displaced from the here and n o w of this conversation. Furthermore, the choice of the generic mode (they say to me, without reference to any particular event) anticipated that the quote to follow would b e generic as well. She was not talking about a particular act of advicegiving, b u t of events less locally specified. (The context makes it clear that she was not using the historical present here.) Sally t h e n b e g a n the quote, shifting h e r voice iconically to a higher pitch that lay noticeably above h e r n o r m a l range. T h e first element in the quote established an affective stance o n the part of the principal a n d the others toward w h a t they were telling h e r : (4)
Sally
(0.9) (tsk) (breath) (begin higher pitch) 6h well,
T h e alveolar click (tsk) as well as the prosody and wording oh well conveyed something of the lack of concern Sally h a d perceived in the advice: the principal and stuff felt that coping with the third-graders was n o big deal. T h e next focus established a frame for the r e c o m m e n d e d action: the idea that Sally should do something specific: (5)
Sally
. . what you dd with those third-graders,
•
1
With this utterance Sally created a second level of displacement. H a v i n g b e g u n in the i m m e d i a t e m o d e in (1) (experiencing her current frustration), she used (3) to shift into the displaced world in which she was given advice, a n d n o w with (5) she m o v e d into the further displaced world of the r e c o m m e n d e d action, a m o r e hypothetical world that might b e realized at some future time. T h u s the sequence of (1), (3), a n d (5) established a setting that was increasingly displaced from the immediacy of the present conversation: what frustrates me the principal and stuff say to me what you do with those third-graders
(immediate) (past and generic) (future and generic)
With this orientation in place, Sally arrived at a point where she could begin expressing the advice that h a d b e e n given her. Putting it all together and deciding how to express it took a little m o r e time, some of which she filled with two intonation units that shed light o n still other aspects of discourse flow:
340
DISCOURSE STUDIES
(6) (7)
Sally
you know, is you just like,
T h e r e are two problems that confront a n y o n e engaged in talk. T h e y are created b y two kinds of unconformity, to b o r r o w a term from geology, w h e r e it refers to a discontinuity in rock strata. I use it h e r e to refer to disparate aspects of h u m a n experience that must s o m e h o w b e b r o u g h t into approximate (but only approximate) conformity if one is to interact with one's fellow h u m a n s . First, there is the inevitable unconformity between an individual's experiences perceptions, actions, a n d evaluations that are either immediate, r e m e m b e r e d , or imagined - a n d the limited resources a language provides for verbalizing them. Second, there is the unconformity that inevitably exists between one mind a n d another. T h e r e is, in short, both a verbalization p r o b l e m a n d an interaction problem. T h e language people p r o d u c e often gives indications that a speaker recognizes both, a n d (6) a n d (7) are examples. So far as the verbalization p r o b l e m is concerned, language c a n n o t fully or adequately express an inner experience. T h e verbalization process allows a speaker to get a useful handle o n the experience and share it to some degree with others, b u t the linguistic organization of ideas is n o t the same as the experience itself. T h e ubiquitous like, found h e r e in (7), is one w a y a speaker can show rec ognition of the unconformity between ideas a n d their verbal expression - a small a n d passing way in which Sally showed h e r recognition that what she was about to say would b e only a roughly satisfactory representation of what she was thinking. So far as the interaction problem is concerned, one m i n d can never fully know what another m i n d is experiencing, a n d language can only imperfecdy bridge the gap. S o m e o n e engaged in a conversation needs b o t h to clothe an inner ex perience in language that will m o r e or less adequately express it, and at the same time find language that will m o r e or less satisfactorily take account of what is believed to b e present in other minds, to the extent that that is possible. The equally ubiquitous you know, the sole content of (6), is o n e w a y a speaker can show recognition of the unconformity between his or her own m i n d a n d the mind of another, in this case signaling that what she was about to say was, to some degree at least, what h e r listeners might have expected a n d n o t something that would b e totally surprising to t h e m . (It can b e n o t e d that (6) a n d (7) were attributed to the people characterized as the principal and stuff, not to Sally herself, b u t of course there is n o w a y to k n o w whether they were anything the principal or a n y o n e else h a d actually said.) It was time n o w to m o v e o n to the complication section of the narrative schema, in this case the actions r e c o m m e n d e d b y the principal a n d stuff: (8) Sally (9) (10)
(0.8) take them=, and put them=, you know with= one of the smarter fourth-graders who's very [verjbal and,
CHAFE THE ANALYSIS OF DISCOURSE FLOW
(11) Chris (12) Sally (13) (14)
341
[Uhhuh,] (0.1) and well-behaved. (0.5) And you . . have them work as a team you know; so that the (0.4) (breath) f6urth-grader can help the thfrd-grader.
At the e n d of (14) the p r o s o d y conveyed a definitive closure of this section. T h e climax then c a m e with a bang, its impact heightened b y the nearly two seconds of silence that p r e c e d e d it as well as b y the forceful wording: (15) Sally
(1.7) (loud) B u t . , that's bullshit.
T h e immediately following d e n o u e m e n t served to justify this evaluation: (16) Sally (17) (18) (19) (20)
(0.1) Because, . (0.5) that just teaches the tmrd-grader=, with the lesser intelligence that, , t• (0.9) that he's worthless; . . you know that he can't learn [sumpm on his ow^n.]
.
i
4 N a v i g a t i o n b y Interaction With (20), Sally completed h e r own d e v e l o p m e n t of the topic she h a d o p e n e d in (1). C a n we say that the conversation h a d n o w r e t u r n e d to a state w h e r e it would h a v e b e e n appropriate for any of the participants to introduce a different topic, either staying within the Classroom Experiences supertopic or introducing a n e w supertopic? T h e question is whether (20) qualifies as a topic ending. We can only speculate o n Sally's goal in opening her topic in t h e first place, b u t we might suppose that she was using (l)-(20) as a way of eliciting s o m e reaction, perhaps s y m p a t h y a n d advice, from her interlocutors. I n any case K a t h y reacted in a w a y that m a y n o t h a v e b e e n what Sally was h o p i n g for. W h a t she said overlapped the e n d of (20): (21) Kathy [No it's not; (22) no it's] not, (23) you can put them in teams like that; With these three intonation units Kathy succeeded in reorganizing the struc ture of the ongoing topic. Until n o w Sally's topic h a d b e e n organized a r o u n d the idea that teams do not work, the idea I labeled Something Frustrating. K a t h y now introduced the idea that teams ao work, thereby organizing the topic into a bipartite structure of thesis and antithesis: into the subtopics Teams D o Not Work and Teams D o Work. Thus, the topic w e are studying could n o w b e relabeled as Using Teams. But what followed took a p a t h that n o one could h a v e anticipated. K a t h y b e g a n b y justifying h e r statement in (23) by trying to modify Sally's conception of the make-up of the teams:
DISCOURSE STUDIES
342
(24) Kathy but you don't put one with one; (25) you put like tw6 fourth-graders with-
i I •
..
Before she finished (25), however, she decided that h e r intent would c o m e across m o r e clearly if she could establish the relative numbers of third- and fourth-graders in Sally's class. After nearly a second of silence she briefly thought in (26) of asking for raw n u m b e r s , b u t truncated that attempt also a n d quickly replaced it with a request for a ratio instead: (26) Kathy (0.8) How many third-graders d(27) What's the . . fl ratio of third- 1] [2 graders to fourth-graders. 2] I n the middle of (27) there occurred o n e of those conversational moments w h e n people talk at cross-purposes, a turbulence in the stream of interactive thought. Sally did not immediately h e a r Kathy's question about the ratio of thirdgraders to fourth-graders, a n d not only Sally b u t also Chris b e g a n to pursue directions of their own, overlapping most of (27): (28) Sally (29) Chris
[1 But they're not 1] [2 You m e a n so they don't feel singled 2] [3 out or w h a t 3]
But Sally quickly a b a n d o n e d whatever she h a d b e g u n in (28) a n d responded to Kathy's question in (27) with some precise information: (30) Sally (31)
[3 Now I have 3] like five third-graders. I have like (0.3) twenty-two kids.
T h e s e two statements elicited the first of the misunderstandings that drove the remainder of this topic. Sally's answer invited some hasty arithmetic that should h a v e yielded the correct n u m b e r of fourth-graders, but Kathy m a d e an error: (32) Kathy (0.2) Okay, have fifteen fourth-graders and five third-graders? (33) so you have We can only speculate on w h y Kathy said fifteen, b u t the subsequent conversation suggests that she h a d b e e n h o p i n g for a whole n u m b e r ratio like fifteen to five, so that each team could h a v e contained three fourth-graders a n d one thirdgrader. T h e question in (33) was a confirmative one, anticipating a positive answer, b u t of course Sally r e s p o n d e d with a correction: (34) Sally (35) (36) (37)
(0.6) No; (0.9) uh=no. (0.1) I have \like (0.2) seven (noise) fourth-graders. (0.1) (sotto vvoce) And five third-graders.
D u r i n g (36) there was a n extraneous background noise that masked the lasi „.,II„KI« tha i*r<-ifrl cmienteen en that Karhv hpard nnlv seven. O n the basis
CHAFE
THE ANALYSIS OF DISCOURSE FLOW
343
of ordinary expectations regarding class siie she responded with surprise, communicated especially by her prosody: ' (38) Kathy You have twelve kids? N o w it was Sally's turn to be surprised. Thinking she had just explained that the correct numbers were 17 fourth-graders and five third-graders, Kathy's question made no sense: (39) Sally
(0.5) What?
But Kathy could only repeat it: (40) Kathy (0.1) You only have twelve kids? Sally repeated her previous answer, this time free of the noise: (41) Sally
(0.4) No.
(42)
(0.3) Seventeen;
Kathy stood corrected: (43) Kathy (0.2) Oh okay, Sally wanted to make certain that Kathy knew that 17 was not the total number in the class, but only the size of the subset on which she had focused: (44) Sally
fourth-gra[ders,]
Amid all this confusion Kathy abandoned her plan to be precise about the numerical composition of the teams. If she had hoped to specify that each team would be composed of three fourth-graders and one third-grader, she n o w found it pointless to insist on such exactitude and fell back on a less precise recommendation: (45) Kathy [so] then what you do is you sprinkle the fifth-graders out evenly. (46) (0.6) And you make . . [the fourth-graders] (0.1) take the responsibility for teaching them. In (45) she made another error, saying fifth-graders instead of third-graders, probably because Kathy herself had taught a fifth-sixth-grade combination in which it was the fifth-graders w h o were the less advanced. Sally corrected her with a questioning intonation while Kathy was uttering fourth-graders in (46): (47) Sally
[Thfrd-graders?]
Kathy then went on to supplement what she had said in (46): (48) Kathy And you engrain in them, (49)
that it's their responsibility to help those little kids.
DISCOURSE STUDIES
344
She added a coda that would drive h o m e the success of the recommended procedure. Sandwiched between her final two intonation units was a protest by Sally, evidently to the effect that she herself had done the same: (50) Kathy That's what I did, (51) Sally I have been. (52) Kathy [and it works.] Even before Kathy finished (52), Chris overlapped with a question whose effect was to open a new, though closely related topic: (53) Chris (54)
[But then you] can you say it's a [part of your] grade?
There followed a lengthy discussion of whether and h o w one should grade the fourth-graders for their mentoring activities. The situation created by (54) was thus as follows: (Classroom Experiences (Classroom Experience 1) (Classroom Experience 2) (Using Teams) (Assigning Grades My intention with this extended example has been to illustrate h o w the stream of language is propelled forward b y the opening of a topic and the creation of a drive for the topic's development until closure is judged appropriate. I have discussed a basic-level topic, ultimately called Using Teams, as an example of the highest level of topic-hood at which there is a coherent trajectory of develop ment. Once open, a topic may be kept moving along a path provided b y a schema, or b y the interaction of separate minds engaged in the conversation, or b y some combination of both. Interactive topic development may be driven b y an inter locutor's desire to agree with or contradict something said by another, or to request needed information the other may possess. This example shows especially well h o w forward movement may be driven by momentary misunderstandings.
5 T h e Text By stringing together all the intonation units that were introduced piecemeal above, one can produce a transcript of this entire segment of the conversation. This kind of object is often called a text, and it is the traditional object of discourse study: 1 2 3 4 5 6
Sally
(0.5) What I was gonna tell you about that really frustrates me is that, (0.2) uh=, (1.3) (breath) the (0.1) the p e o p . . the principal and stuff they say to me, (0.9) (tsk) (breath) (begin higher pitch) oh well, . . what you do with those third-graders, you know, 1
345
CHAFE THE ANALYSIS OF DISCOURSE FLOW
7 8
is yon just like, (0.8) take them=,
9
10 '.
11 12 , 13 14 • 15 16 , 17 18 19 20 21 22 23 :
••' '•
i
i •"
and put them=,
you know with= one of the smarter fourth-graders who's v^ry [ver]bal and, Chris [Uhhuh,] Sally (0.1) and well-behaved. (0.5) And you . . have them work as a team you know; so that the (0.4) (breath) fourth-grader can help the third-grader. (1.7) B u t . . that's bullshit. (0.1) Because, (0.5) that just teaches the thfrd-grader=, with the lesser intelligence that, (0.9) that he's worthless; . . you know that he can't learn [sumpm on his 6w=n.] » r Kathy [No it's not; . _ no it's] not, you can put them in teams like that; 1
24
but you don't put one with one;
25 26 27 28 29 30 31
you put like two fourth-graders with"* (0.8) How many third-graders d What's the . . [1 ratio of third-1] [2 graders to fourth-graders. 2] [1 But they're not 1] [2 You mean so they don't feel singled 2] [3 out or what 3] [3 Now I have 3] like five third-graders. I have like (0.3) twenty-two kids.
32
33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54
Sally Chris Sally
Kathy (0.2) Okay,
Sally
Kathy Sally Kathy Sally Kathy Sally Kathy
Sally Kathy
Sally Kathy Chris
so you have fifteen fourth-graders and five third-graders? (0.6) No; (0.9)uh=n6. (0.1) I have like (0.2) seven (noise) fourth-graders. (0.1) (sotto voce) And five third-graders. You have twelve kids? (0.5) What? (0.1) You only have twelve kids? (0.4) No. (0.3) Seventeen; (0.2) O h okay, fourth-gra[ders,] [so] then what you do is you sprinkle the fifth-graders out evenly. (0.6) And you make . . [the fourth-graders] (0.1) take the responsibility for teaching them. [Third-graders?] And you engrain in them, that it's their responsibility to help those little kids. That's what I did, I have been. [and it works.] [But then you] can you say it's a [part of yourl errade? 1
346
DISCOURSE STUDIES
W h a t k i n d of thing is this? D o e s it h a v e a n y validity b e y o n d being a visual representation of a concatenation of utterances that w e r e p r o d u c e d in sequence as the conversation unfolded through time? O n e possibility, easily discardable, is that it represents something in the m i n d s of one or m o r e of the participants before these things were said. But of course n o one could have planned the above, or h a v e predicted that the conversation would proceed in this way. Is it, then, something that r e m a i n e d in the m i n d s of the participants afterwards? Again the answer must b e n o , t h o u g h perhaps this time n o t quite so unqualified a no. Al though s o m e of the ideas expressed h e r e w e r e p r o b a b l y retained in some form, varying from one participant to another, for at least a while, the details of h o w these thoughts were activated and verbalized during the conversation were surely quickly lost. T h e participants m a y h a v e r e m e m b e r e d for a time that they talked about using teams in the classroom, that Sally did n o t like the idea, that Kathy did like it, a n d so on. But the particular sequence of ideas a n d exactiy h o w they w e r e expressed was surely ephemeral. It is worth noting that spontaneous conversations differ from "oral literature" in this respect. A person m a y r e m e m b e r a ritual or story or j o k e a n d repeat it later in another setting, though with language a n d content seldom if ever identical. But p e o p l e d o n o t repeat casual conversations in the same way. S o m e o n e might say, "That was a good conversation," but n o one would b e likely to exclaim, "Let's say the whole thing again tomorrow." If people d o r e m a r k occasionally, "I think w e ' v e h a d this conversation before," t h e y are hardly thinking of a verbatim repetition. It is worth reflecting o n the fact that the collection a n d study of texts has in the past b e e n slanted toward narratives and rituals whose value lies in something closer to (though seldom identical with) verbatim repetition. Discourse of that kind is m o r e persistent in m e m o r y , a n d in that respect is a little m o r e like written language. I n other words, earlier discourse studies h a v e t e n d e d to favor material that has b e e n closer in nature to written text (Chafe 1981). I d o n o t m e a n to suggest that a text like the above has n o use. W h a t it gives u s is a lasting record of evanescent happenings that w e can examine visually at our leisure. As a kind of time machine, it is a resource that allows us as analysts to view all at once the dynamic processes b y which a sequence of linguistic events was produced. It is a useful tool that can further our understanding of h o w minds and language proceed through time. By freezing temporal events it helps us iden tify the forces responsible for creating them. M y point is that we should not be misled into interpreting this artificial aid to understanding as something that possesses a transcendent reality. O n e m a y sometimes h e a r the view that participants in a conversation are engaged in the joint construction of a text. I suggest that it is better to think of a conversation as a uniquely h u m a n a n d extraordinarily important w a y b y which separate m i n d s are able to influence a n d b e influenced b y each other, managing to some extent, a n d always imperfectly, to bridge the gap between them, not b y constructing a n y kind of lasting object but t h r o u g h a constant interplay of constantly changing ideas. T h e example that has b e e n discussed h e r e suggests a few of the ways in which that can h a p p e n .
CHAFE
THE ANALYSIS OF DISCOURSE FLOW
347
Note 1. Conventions followed in this and the following transcriptions of speech include the following. The numbers in parentheses are measurements (to tenths of a second) of periods of silence. The acute and grave accents mark the nuclei of syllables with primary and secondary accents respectively. Periods show a decisively falling pitch contour, often accompanied by creaky voice, whereas semicolons show a less decisive fall. Commas show any other terminal contour, except that the high rising pitch associated with a yes-no question is shown by a question mark. The equals sign shows a prolongation of the preceding sound. Square brackets show overlapping speech, sometimes indexed with numbers when there might be ambiguity. That is, a segment enclosed in [ 1 . . . 1] overlaps with another segment indexed in the same way, etc.
References
Bartlett, Frederic C. 1932. Remembering: A Study in Experimental and Social Psychology. Cambridge: Cambridge University Press. Brown, Gillian, and George Yule. 1983. Discourse Analysis. Cambridge: Cambridge University Press. Chafe, Wallace. 1981. Differences between colloquial and ritual in Seneca, or how oral literature is literary. In Wallace Chafe, Alice Schlichter, and Leanne Hinton (eds), Reportsfrom Survey ofCalifornia and Other Indian Languages, No. 1, pp. 131-45. Berkeley: University of California, Department of Linguistics. Chafe, Wallace. 1986. Beyond Bartlett: Narratives and remembering. In Elisabeth Giilich and Uta M. Quasthoff (eds), Narrative Analysis: An Interdisciplinary Dialogue. Special issue of Poetics, 15, 139-51. Chafe, Wallace. 1993. Prosodic and functional units of language. In Jane A. Edwards and Martin D. Lampert (eds), lalkingData: Transcription and Coding in Discourse Research, pp. 33-43. Hillsdale, NJ: Lawrence Erlbaum. Chafe, Wallace. 1994. Discourse, Consciousness, and Time: The Flow and Displacement of Conscious Experience in Speaking and Writing. Chicago: University of Chicago Press. Chafe, Wallace. 1995. Adequacy, user-friendliness, and practicality in transcribing. In Geoffrey Leech, Greg Myers, and Jenny Thomas (eds), Spoken English on Computer: Transcription, Mark-up, and Application, pp. 54-61. New York: Longman. Du Bois, John W, Stephan Schuetze-Coburn, Susanna Cumming, and Danae Paolino. 1993. Oudine of discourse transcription. InJane A. Edwards and Martin D. Lampert (eds), TalkingData: Transcription and Coding in Discourse Research, pp. 45-89. Hillsdale, NJ: Lawrence Erlbaum. Givdn, T. 1983. Topic Continuity in Discourse: A Quantitative Cross-Language Study. Amsterdam and Philadelphia: John Benjamins. Hockett, Charles F. 1958. A Course in Modern Linguistics. New York: Macmillan. James, William. 1890. The Principles of Psychology. 2 vols. New York: Henry Holt. Reprinted 1950, Dover: New York. Labov, William, andJoshua Waletzky. 1967. Narrative analysis: Oral versions of personal experience. In June Helm (ed.), Essays on the Verbal and Visual Arts: Proceedings of the 1966 Annual Spring Meeting of the American Ethnological Society, pp. 12-44. Seattle: University of Washington Press.
348
DISCOURSE STUDIES
Li, Charles N., and Sandra Thompson. 1976. Subject andtopic:A new typology of language. In Charles N. Li (ed.), Subject and Topic, pp. 457-89. New York: Academic Press. Linell, Per. 1982. The Written Language Bias in Linguistics. Linkoping, Sweden: University of Linkoping Press. Tannen, Deborah. 1984. Conversational Style: Analyzing Talk Among Friends. Norwood, NJ: Ablex.
r
I
i
i
i
i
i
15 Missing Links in Mainstream CDA: Modules, Blends and the Critical Instinct Paul Chilton
1. T h e Critical Stance
L
et us first attempt a crude m a p of the terrain. Critical approaches to text a n d talk in E u r o p e a n scholarship over the past twenty years or so h a v e i n d e e d b e e n interdisciplinary, b u t this observation has little significance unless one specifies the sources and tributaries. This introductory section is a n admittedly partial attempt to sketch the landscape. 'Critical T h e o r y ' , associated with the Frankfurt School and especially with Jiirgen H a b e r m a s provides o n e source in social theory. This source found itself channelled into a sociological variety which did not claim any technical connection with language studies or linguistics (notably, Stuart Hall, Birmingham Contem p o r a r y Studies). It was also d r a w n u p o n b y scholars whose orientation was towards language use, though these scholars were somewhat eclectic in the social theory that they m a d e use of (Fowler et al. 1979; Fairclough 1989; Wodak 1996; Wodak & M e y e r 2001). A guiding t h e m e was the notion that language can b e used for self-interested ends b y p o w e r groups. 1
A further source was the w o r k of Michel Foucault, whose claims w e r e not always viewed as consonant with those of the H a b e r m a s i a n tendency. Foucault's w o r k has given rise to at least two tendencies. O n e tendency has its context in sociology, political science a n d to some extent in literary studies, and like Foucault himself, does n o t analyse language. This tendency has given rise to m a n y studies, too numerous to list here (but cf. for example Shapiro 1984 for examples). Amongst writers in this vein it is c o m m o n to find the assumption, or claim, that 'discourse' includes the non-linguistic as well as the linguistic. At this point an additional
Source: Ruth Wodak and Paul Chilton (eds), A New Agenda in (Critical) Discourse Analysis, 2005. Dt>. 1.0-51
350
DISCOURSE STUDIES
influence n e e d s to b e m e n t i o n e d , n a m e l y the post-modernist current that was a confluence of ideas from philosophy, literary studies a n d sociology (Barthes, Derrida, a n d others). This kind of work h a d n o contact with formal linguistics b u t m a d e assertions about the nature of m e a n i n g ('free floating signifiers'), epistemology a n d ontology, attacking alleged essentialism, p r o m o t i n g notions such as 'aporia'. This form of relativism provided for the possibility of indefinitely extended critique that runs the risk of being ethically and politically indiscriminate. It is n o t possible h e r e to address it directiy a n d will n o t b e referred to again in this chapter. To return to Foucault, the second Foucauldian influence is to b e found a m o n g those whose declared preoccupation is, b y contrast, with language as such (notably Fairclough 1 9 8 9 , 1 9 9 2 , 1 9 9 5 , 2 0 0 3 ; to s o m e extent Wodak 1996). M a n y of these authors would, however, accept that discourse, understood as language-use, is b u t one (if perhaps the most salient) manifestation of social action. T h e insistence that language is a form of social-action is probably the most important tenet amongst this group (see Fairclough & Wodak 1997:278-279 for a formulation). Those working in this tradition h a v e often h a d some degree of formation in linguistics and h a v e to a greater or lesser extent p r o c e e d e d o n the assumption that the nature of social action can b e elucidated, even u n m a s k e d , b y various kinds of linguistic analysis. O n e can distinguish h e r e several sub-tendencies a m o n g the linguistically oriented. Historically the first was Critical Linguistics ('the East Anglia school'), whose early work, drew on George Orwell for inspiration, Bakhtin and to a lesser extent H a b e r m a s a n d to a n even lesser extent Foucault for its social theory. For its linguistic theory it drew at first o n Chomsky's early versions of transformational g r a m m a r (Hodge & Kress 1993 [1979]). This choice was later r e p l a c e d b y Halliday's so-called systemic-functional g r a m m a r (Fairclough 1989: 13-14; Fowler 1996: 11). T h e second language-oriented trend, chronologically speaking, is Critical Discourse Analysis, most c o m m o n l y associated with Fairclough, Wodak and van Dijk (cf. v a n Dijk 1993; Fairclough 1995; Fairclough & Wodak 1997; CaldasCoulthard & Coulthard 1996). Fairclough, in particular, is influenced b y Foucault, especially in his use of the notions of'order of discourse' a n d 'discourse formation'. Wodak's a p p r o a c h to the analysis of language in use ('discourse' is understood b y all of these authors to b e language) comes from various strands of sociolinguistics a n d ethnography (cf. Reisigl & Wodak 2 0 0 1 : Chapter 2). Van Dijk's work, emerging from relatively formal text linguistics (van Dijk 1977, 1980, 1998; v a n Dijk & Kintsch 1983), has always h a d a cognitive orientation, being concerned with mental schemas that represent the social a n d give rise to stereo types, that in turn give rise to various ideologies, a m o n g which are exclusionary discourses, especially racism. For m a n y in this group of scholars (especially Fowler, Kress, H o d g e , Thibault, v a n Leeuwen, Fairclough and others, b u t not v a n Dijk), Halliday's systemic-functional g r a m m a r has supposedly p r o v i d e d the toolkit for deconstructing the socially-constructed (thus linguistically constructed) 2
CHILTON
MISSING LINKS IN MAINSTREAM CDA
351
Yet after twenty years of this form of scholarship, there r e m a i n serious p r o b lems in the entire endeavour. I will n o t attempt to deal with the postmodernist w o r k that does n o t address issues of language per se, since this would take us too far afield philosophically. I will focus o n C D A understood in a b r o a d sense, occasionally including Fowlerian Critical Linguistics (Fairclough & Wodak 1997; Fowler 1996).
'
2
.
C l a i m s a n d A i m s of C D A
A m o n g the serious problems that, to my-mind, bedevil C D A , are the following. 1. C D A claims, sometimes implicidy, sometimes explicidy, that its practice provides demystifying a n d emancipatory effects (see for example, Fowler 1996: 55ff.; Wodak & Meyer 2 0 0 1 : 10; Weiss & W o d a k 2 0 0 3 : 14f.). T h e r e is of course a n argument about whether a supposedly scientific endeavour should allow itself social a n d political motivation in the first place. O n e m a y argue for objectivity h e r e on the grounds that acknowledging one's interests is per se a form of scientific objectivity, an argument that m a y well b e crucial for all social sciences. This is not, however, m y m a i n point. Rather, I want to pose the question whether C D A has a n y credible efficacy, on its own terms, as an instrument of social justice. A n d if not, or even if the answer to that question is just a bit in doubt, d o w e n e e d it? 2. This question is connected to its m e t h o d s a n d linguistic underpinning. C D A has t e n d e d to draw, as we h a v e seen, o n social theory of a particular type a n d o n linguistics of a particular type. Despite some limited use of work in psychology and cognitive science (e.g. van Dijk 1991,1993a, 1993b, 1998,2004, forthcoming; v a n Dijk & Kintsch 1983; Wodak 1986; Reisigl & Wodak 2001), it appears to b e fair to say that C D A has generally neglected developments in these fields. It has eschewed not only generative linguistics b u t also cognitive linguistics. (The latter at least does n o t figure in any seriously c o m p r e h e n d e d way in C D A literature.) 3 . 1 h a v e just questioned whether C D A has h a d genuine social effects. It is also questionable whether C D A has any theoretically interesting yield for the social sciences, and m o r e especially for linguistics. With regard to the latter, apart from the early attempts b y serious linguists working a r o u n d Fowler, Kress a n d H o d g e , C D A in its later manifestation has m a d e , so far as one can see, n o contri bution to our scientific understanding of the language capacity in h o m o sapiens. This is strange in view of the C D A insistence that language plays such a n e n o r m o u s role in social a n d political life.
3 . W h a t is M i s s i n g from C D A T h e r e is a c o m m o n element underlying these three criticisms: C D A despite some interest in (mental) representations has b y a n d large n o t paid any attention
DISCOURSE STUDIES
352
to the h u m a n m i n d . T h e r e are two consequences of this observation, around which this p a p e r is constructed. T h e y are the following: 1. T h e lack of attention to m i n d is a n important theoretical lack in its own terms. But there is worse. 2. It is possible that taking stock of recent research in the cognitive sciences leads us to the conclusion that w e d o not actually n e e d C D A . If there is anything in this argument, t h e n we n e e d to radically reassess what scholars are doing w h e n they research into the relationship between the socio political d o m a i n a n d the linguistic d o m a i n , a n d the p a p e r concludes with some conclusions o n the matter.
4. W h y C D A N e e d s t o C o n s i d e r M i n d C D A has since its emergence b e e n interdisciplinary, b u t selectively so. T h e aim of this section is to bring to the attention of C D Analysts the existence of a b o d y of empirical research a n d speculative literature which they have tended to ignore. It is a b o d y of w o r k that has clear implications for a possible cognitive approach to the analysis of discourse in social a n d political contexts. T h e r e is a simple a r g u m e n t u n d e r l y i n g the claim that C D A should attend to this cognitive dimension. It starts with the central tenets of C D A itself (cf. Fairclough & Wodak 1997: 258-284). T h e r e are three such tenets: 3
1. discourse is social action (or 'social practice') 2. social action constructs social reality (objects, situations, identities, social relations...) 3. discourse is the use of language. It follows that discourse constructs social reality. This must imply some sort of causal relation between language use a n d social action. T h e r e are a n u m b e r of philosophical questions that seem to arise h e r e , which we can only adumbrate. O n e is the following. It would appear to b e the case that C D A is claiming that discourse stands in a causal relationship to social action (by which we might understand social relationships, group m e m b e r s h i p , the formation of social and political institutions a n d the like). But discourse is said to b e social action, in which cause w e seem to have discourse causing itself. M a n y discourse analysts, particularly those in the post-modernist tradition, would have little problem with this, since they are prepared to adopt the logical conclusion, namely that everything is discourse. C D A authors on the other h a n d are inclined to resolve the tension b y using the term 'constitute' for cause: 'discourse is socially constitutive as well as socially shaped' (Fairclough & Wodak 1997: 258). So far as it goes, this is a useful concept that at least points to a complex dialectally causal relation. However, it is p e r h a p s time to start probing its nature in m o r e detail. To proceed beyond
CHILTON
MISSING LINKS IN MAINSTREAM CDA
353
that particular formula, o n e needs to notice that it is entirely stated with respect to a social d o m a i n that appears a u t o n o m o u s : social action qua discourse has effects o n social discourse that is either (a) also discourse or (b) social action b u t n o t (linguistic) discourse. T h e way out of the l o o p is to introduce the cognitive dimension (as v a n Dijk (e.g. 1980,1998) has done, and does also in the present volume). T h e argument is straightforward. Discourse, that is to say language in use, is p r o d u c e d a n d inter preted b y h u m a n individuals interacting with o n e another. Language can only b e produced a n d interpreted b y h u m a n brains (and vocal apparatus, etc.). If lan guage is produced and interpreted in h u m a n brains, then it interacts on any account of language with other cognitive capacities (as well as m o t o r systems). I n particu lar, if language use (discourse) is, as the tenets of C D A assert, connected to the 'construction' of knowledge about social objects, identities, processes, etc., then that construction can only b e taking place in the minds of (interacting) individuals. This argument seems to indicate that if C D A is to b e a research enterprise, which I take to m e a n an enterprise that enhances h u m a n understanding and know ledge, then what goes o n inside people's h e a d s must b e c o m e a p r i m e concern. T h e r e are m a n y ways in which such a research goal might b e pursued. I n the present chapter I will merely suggest h o w some of the existing research in certain areas of contemporary cognitive science and psychology, particularly those vari eties that take a n evolutionary perspective, might b e used to enrich and explore in n e w ways the sorts of social action a n d social construction that C D A has classically b e e n preoccupied with. Let us, then, consider the specific area of racism or x e n o p h o b i a (not of course necessarily the same thing). C D A is good at showing h o w particular language users establish exclusionary attitudes a n d m a y b e practices b y recurrently a n d selectively asserting certain attributes (i.e. social roles, behavioural characteristics, physical appearance, etc.) of social a n d ethnic groups (e.g. v a n Dijk 1991,1993a, 1993b; Wodak 2000; Reisigl & Wodak 2001). I n other words, C D A draws attention to the existence of stereotyped categorisations in daily talk, elite talk a n d texts. C D A also shows h o w language users categorise behaviour, actions and attributes all of which m a y b e observable facts - in ways that express attitudes towards such facts. For example, immigration, which denotationally refers to spatial move ments of people in relation to political boundaries, is defined as 'invasion' or 'flood'. It is also g o o d at identifying interactive verbal devices for the implication or presupposition of assertions about social and ethnic groups, and in pin-pointing mitigation devices such as 'apparent denial'. Underlying all these features that are found in text a n d talk referring to social a n d ethnic groups is the conceptual process of category formation. O n the actual cognitive processes C D A seems to have little n e w to say, despite the fact that there is an abundant psychological and social psychological literature o n the formation of concepts (but cf. summaries in van Dijk 1991,1993a, 1998 and Reisigl & Wodak 2001). I n general, C D A has done a fine descriptive j o b . W h e t h e r the description has altered public awareness significanuy enough to achieve emancipatory effects is perhaps not the main point here, though it is obviously not irrelevant. T h e disturbing fact is that racism, xenophobia
DISCOURSE STUDIED
354
and other kinds of exclusionary behaviour continue to appear. This suggests that we m a y n e e d to delve deeper into why this kind of category formation is so persistent a factor in social behaviour, and why t h e language forms associated with it are so potent. These are difficult questions: I am not going to give the answers, but merely propose that anyone seriously interested in these issues needs to address such questions. This m e a n s taking a n explanatory stance rather than a m e r e l y descriptive o n e a n d it also m e a n s , I a m suggesting, taking account of ideas developed in cognitive a n d evolutionary psychology.
5. S o m e A s p e c t s of C o g n i t i v e a n d Evolutionary P s y c h o l o g y R e l e v a n t to C D A I n this section I review s o m e recent ideas emerging from cognitive science and cognitive psychology. Since an evolutionary perspective has proved a fruitful thinking device in these a n d other areas, a n d since it can provide an explanatory framework for e x p a n d i n g the discussion of issues that C D A is concerned with. I will bring that too into the picture. But I m a k e n o pretence of completeness in regard to this field. 5.1 Modularity of Mind C h o m s k y postulated b a c k in the 1960s that the h u m a n language faculty was an autonomous m o d u l e of m i n d . I n part, the claim was based o n analogy with other h u m a n cognitive a n d perceptual abilities that w e r e regarded on empirical grounds as modular, in particular vision a n d facial recognition. T h e notion o* modularity was m o r e extensively theorised b y Fodor (1983), whose ideas have h a d a wide impact o n psychology, theoretical a n d applied, o n cognitive science and o n philosophy. Fodor's theory corresponds roughly to Figure 1. For Fodor. the m i n d has two tiers: a collection of input m o d u l e s that are autonomous and h a v e a specific j o b to d o (like the blades of a Swiss a r m y knife) and a central processor that corresponds to w h a t people generally think of as intelligence, creative thinking. T h e former are hard-wired, localised a n d provide inpul to the latter, which is n o t localised a n d is characterised b y 'its non-encapsulation, its creativity, its holism a n d its passion for the analogical' (Fodor 1985: 4). The last point is w o r t h noting a n d w e shall return to it. T h e r e is m u c h controversy surrounding the two-tier proposal, but the mos( significant development of the m o d u l a r concept has b e e n its incorporation into evolutionary psychology, notably b y Cosmides a n d Tooby (1992), who, in broad terms, argue that the m i n d of h u m a n s as we n o w k n o w t h e m has emerged not b y chance or solely b y the impact of particular social a n d physical environments b u t b y the pressures of evolutionary selection. T h e m e n t a l modules evolved for purposes of survival, just like any other organ of the body. Rather than assuming the existence of some generalised intelligence a n d learning ability, the tendency is to regard the m i n d as consisting of m a n y specialised modules, interacting among
Figure 1: Information from perceptual organs to input systems (Modules) (Fodor) one another. O n e type of evidence that is adduced is similar to the familiar Chomskyan argument for the innateness of language: that children develop very m a n y kinds of specific knowledge very rapidly a n d o n the basis of scant input. Cosmides and Tooby think there is a very large n u m b e r of such modules: A face recognition module, a spatial relations module, a rigid objects mechanics m o d u l e , a tool-use m o d u l e , a fear module, a social exchange module, an emotion-perception, kin-oriented motivation, an effort alloca tion and recalibration m o d u l e , a child care m o d u l e , a social inference module, a friendship m o d u l e , a semantic-inference m o d u l e , a g r a m m a r acquisition m o d u l e , a communication-pragmatics m o d u l e , a theory of mind m o d u l e , a n d so on... (1992: 113) Amongst these modules, some of which are empirically attested, some rather more speculative, which should b e of interest to researchers interested in the interaction of language a n d society? Clearly language is one of them, but we will focus o n three other possibly m o d u l a r areas. 5.2 Intuitive Psychology ('Theory of Mind') Somewhat before Fodor, H u m p h r e y h a d begun to propose a specific m o d u l e of social intelligence combined with what is k n o w n as a 'theory of m i n d ' module, which refers to what appears to b e the innate ability of h u m a n s to guess and m a k e pragmatic predictions about the intentions a n d likely future actions of their fellows ( H u m p h r e y 1976). T h e idea has b e e n picked u p a n d further investigated by Leslie (1991) and B a r o n - C o h e n (2001) in n o r m a l and abnormal psychology, and
hv
W h i t P n (VA
1 Q Q 1 \ ~ - : —
-
"
-
-
DISCOURSE STUDIES
35fi
that h u m a n s h a v e a n evolved ability to look for the motives and plans of other"-, an ability that can clearly b e used cooperatively or counter-cooperatively. In an event, applied to language behaviour, it suggests that h u m a n s naturally am automatically tend to look for interpretations of discourse acts in terms of wlu: speakers are intending to convey a n d to d o . 1
5.3 Machiavellian Intelligence Similar claims to those outlined in the preceding p a r a g r a p h have been made also for primates. Byrne a n d W h i t e n (1988) a n d Byrne (1995), for instance, ha*. m a d e empirical studies of chimpanzees that they believe demonstrate clearh that these creatures engage instinctively in social behaviour that display ih<* following: • •
the formation a n d observation of group a n d subgroup boundaries the formation of coalitions in competitive environments, signalled by, inter alia, g r o o m i n g behaviour • the formation a n d observance of d o m i n a n c e hierarchies • the practice of tactical deception motivated b y individual wants.
It is the latter, tactical deception, that has b e e n d u b b e d 'machiavellian intelli gence'. Its workings follow from the ability to r e a d the intentions of others, an(i to recognise the ability of others to r e a d one's o w n intentions, leading to dw ability to mask one's intentions (Whiten 1991). If a distinct machiavellian moduf exists in apes, then, so the evolutionary argument goes, it must exist as a distint t ability also in their descendants, h o m o sapiens. I n the case of language-usine creatures, the ability to verbally mislead, deceive a n d to lie is clearly closeh related. C D A researchers should not rush to dismiss the argument that human h a v e a n evolved instinctive a n d specialised ability to exploit one another upon the basis of abilities developed for cooperative social survival purposes. And there is a further very good reason for C D A to look closely at these ideas. It will have b e e n apparent from the preceding arguments that if an individual A has an innate ability to read the minds of B and use deception for his or her own ends, then conversely, B must h a v e the same ability a n d could in principle use it to read the intention to deceive. This is a highly significant notion for CDA to which we return at the end of this chapter. 5.4 Intuitive Biology We h a v e looked at the claims for innate social a n d psychological modules. Other modules of m i n d have b e e n researched that have to d o with the h u m a n mind^ relation to the non-social world. O n e of these is 'intuitive physics': h u m a n undei standing of gravity, solidity, orientation, force, and so o n appear to b e hard-wirerl, that is, b e present at birth as some sort of neurological template that enables th child to deal with the surrounding environment without having to learn it in tiu way that an empiricist psychology would suggest the h u m a n mental tabula rasa J
CHILTON
MISSING LINKS IN MAINSTREAM CDA
357
has to learn about everything (Spelke 1991). M o r e relevant to the concerns of this paper is 'intuitive biology', which works o n the same general principles. What the innate biology m o d u l e works o n is living entities. T h e research in this field (cf. Atran 1990 a n d 1994; K e n n e d y 1992; Keil 1994) indicates that h u m a n children automatically m a k e assumptions about the t a x o n o m y of'natural kinds'. For instance, they seem to take for granted the difference between animate a n d inanimate entities. Moreover, there is evidence that all h u m a n cultures share a basic framework for classifying the natural world (Atran 1990). T h e s e folk classi fications are highly complex: a reasonable explanation, it is argued, is that there is a c o m m o n genetic blueprint for it. A n analogous argument is often m a d e for the universal innateness of the language instinct. T h e evolutionary argument would b e that h u m a n evolution n e e d e d to select for the ability to classify the living world because of the survival advantages conferred b y such an ability. O n e feature of the research o n intuitive biology that is of interest to C D A is the claim that this m o d u l e includes 'essentialism'. T h a t is to say, h u m a n s seem intuitively impelled to attribute 'essences' to different kinds of things in their taxonomies of natural kinds. A predisposition to essentialism would provide a n important explanatory framework for m a n y social p h e n o m e n a that interest C D A researchers, a n d w e return to it later in this chapter.
6. B e y o n d M o d u l e s : C o g n i t i v e Fluidity As was pointed out above, while the notion of modularity is widely accepted among psychologists a n d cognitive scientists, the precise form a n d significance of the modularity is controversial. Karmiloff-Smith (1992), for example, building on her earlier developmental w o r k with Piaget, argues that m o d u l e s are m e r e l y the basis for development that is subject to variable cultural shaping. But the interpretations of modularity that will concern us here are those of Mithen (1996) and those of Sperber (1994) a n d Cosmides and Tooby (2000). Mithen, whose background is in archaeology, seeks an explanation for the apparent 'explosion' of h u m a n d e v e l o p m e n t that occurred b e t w e e n 60,000 a n d 30,000 years ago. Mithen's hypothesis is that the m i n d is i n d e e d m o d u l a r b u t does not, as in Fodor's account, h a v e a n i n d e p e n d e n t general processor. R a t h e r what h a p p e n e d at the critical m o m e n t in h u m a n evolution was that the auto nomous modules b e g a n to intercommunicate: M i t h e n calls this the emergence of 'cognitive fluidity' (Mithen 1996). Sperber also accepts the m o d u l a r story, b u t accounts for h u m a n creativity by postulating (with plenty of evidence from language) a further module, the "module of metarepresentation', a n e x p a n d e d version of the theory of m i n d module a n d closely tied to language a n d communication. W h a t this m o d u l e must be doing, in Sperber's view, is handling concepts of concepts a n d represen tations of representations. It will also b e giving space to concepts from several otherwise e n c a p s u l a t e d a n d n o n - c o m m u n i c a t i n g m o d u l e s , a n d a l l o w i n g recombinations a n d transformations of the concepts formed in the modules.
DISCOURSE STUDIED
M i t h e n seizes o n Fodor's c o m m e n t about general intelligence having a 'pas sion for the analogical' a n d proposes that the m e c h a n i s m b y which the modular d o m a i n s started to interconnect was m e t a p h o r . Specifically, h e claims that language evolved in the first place for social purposes a n d was 'metaphorically extended' as it was 'invaded' b y concepts coming from the major domains (Mithen 1996:214). A l o n g with Sperber h e retains the metarepresentational module as a key feature of h u m a n intelligence, linked to communication, but sees it as an e x p a n d e d form of social intelligence. This metarepresentational, communicational a n d social m o d u l e takes in non-social concepts from other domains and can process t h e m creatively, p e r h a p s using m e t a p h o r or some other form 01 mental transformation. W h a t M i t h e n a n d Sperber d o n o t do is tell us m u c h about is the nature of these processes, particularly where language, which they regard as a key element is concerned. This is w h e r e the work d o n e in cognitive linguistics over the lasi few decades b e c o m e s signally relevant. I n particular, there are well developed theories of metaphor and conceptual blending that could b e linked with the MithenSperber hypothesis. I n Section 8 below w e explore the possible application of some recent theorising in cognitive linguistics on conceptual blends (Fauconnier & Turner 2002).
7. T h e Point of A l l This for C D A W h y should the b o d y of w o r k i n cognitive science a n d evolutionary psychology, of which I h a v e sketched b u t a small fragment, b e of interest to CDA? In brief the reasons are as follows. i. Machiavellian intelligence is of course political behaviour, whether conceived as institutionalised politics or informal social m o v e m e n t politics or day-to-day interpersonal m a n o e u v r i n g a n d manipulation. T h e work summarised above proposes that the basic framework for such behaviour is innate in humans, not primarily derived from socialisation, socioeconomic forces and the like, a* p r o p o s e d in the existing heavily socialised accounts of C D A . It is quite likely (as M i t h e n a n d Sperber seem to argue) that a n innate social intelligence is in some way fundamental and operates in ways that m a y b e innate. But this is ven different from CDA's existing assumptions about the p r i m a c y of the social a n d the political. If the field is to take account of all relevant science, then h seems inevitable that it has to confront the question of h o w the h u m a n mind works w h e n engaged in social a n d political action, which is largely, for humans, verbal action. ii. Research o n social intelligence has focused inter alia o n the formation ol social categories, specifically racial categories, a n d 'essentialism'. This is familiar territory for C D A practitioners. However, the literature outlined above is scarcely if ever d r a w n o n in these studies. I n Section 8 below, I introduce some more details that might suggest links of research building o n the elements of a cognitive approach already found in C D A work o n racism and ideology.
CHILTON
MISSING LINKS IN MAINSTREAM CDA
359
iii. We h a v e seen that w h e n a perspective is adopted that confronts the question of w h o we are as a species a n d what we are like, there arises the funda mental question of h o w we got to b e so uniquely creative (often unpleasandy so by our own h u m a n values). We h a v e seen the vague b u t recurrent suggestion in the literature, from Fodor to Mithen, that some metaphorical ability seems to b e the explanation of h o w m o d u l a r thought is integrated. W h a t is lacking in the evolutionary-cognitive literature is a detailed theory of metaphor. It is also, as it happens, lacking in a lot of C D A work, w h e r e it is often simply treated as a 'persusasive' rhetorical device of some sort. However, cognitive linguistics (CL) has two closely related theories of cognitive activity corresponding to Mithen's 'cogni tive fluidity'. T h e C L theories on metaphor are those developed b y Lakoff,Johnson and Turner (Lakoff StJohnson 1980; Lakoff 1987; Lakoff & Turner 1989; Lakoff & J o h n s o n 1999) a n d e x p l o r e d e m p i r i c a l l y b y , for e x a m p l e , G i b b s (1994), Boroditsky (1997) a n d Glucksberg (2000). T h e second relevant C L theory I a m alluding to is conceptual blending developed b y Fauconnier a n d Turner, which is of importance to C D A because it is specifically a theory of w h a t conceptual activity goes o n during discourse processing (Fauconnier & Turner 2002). 4
iv. C D A as an academic a n d pedagogical enterprise might not b e necessary at all. This starding inference could b e d r a w n from the claim - indeed the evidence - that h u m a n s h a v e in any case an innate 'theory of m i n d ' and a metarepresentational m o d u l e . If individual h u m a n s are innately machiavellian, they are also innately able to counter one another's machinations. If language is crucial to this ability and associated activity, then they should h a v e a n innate ability not just to use language in machiavellian ways b u t to detect and counter one another's machiavellian use of language. I shall return to this argument, which is put forward b y Sperber a n d others, towards the e n d of this paper. I n the meantime, the question is, given t h e foregoing remarks: W h a t is C D A for if people can do it anyway? Having oudined some of the issues that C D A will h a v e to confront if it is going to b e genuinely interdisciplinary, I n o w m o v e o n to what this m e a n s for research into a specific area of social enquiry that that has b e e n the traditional stamping ground of C D A . T h e area is racism. Section 8 summarises the work largely ignored b y C D A accounts of this area. A n d Section 9 outiines a C L ap proach that explains h o w using cognitive linguistics can provide a n account of what the h u m a n m i n d is doing w h e n it does racist discourse. This approach is not entirely neglected in critical approaches to language use: see for instance Lakoff (1991,1996,2003), Chilton (1996a and 1996b), Chilton and Lakoff (1999), Rohrer (1995), Dirven (2001), Santa A n a (2002) a n d O ' H a l l o r a n (2003).
8. A C o g n i t i v e A p p r o a c h t o R a c i s m Most discourse-oriented approaches to racism also m a k e reference to cognitive issues or to the n e e d for an explanatory framework or both. Van Dijk's approach, for instance, is b r o a d l y sociocognitive and is constructed a r o u n d a standard
DISCOURSE STUDIES
360
psychological m o d e l of m e m o r y , linking semantic m e m o r y to stable social con structs, episodic m e m o r y as a storehouse for previously encountered narratives, a n d working m e m o r y , in which context a n d ongoing discourse are processed (van Dijk 1984, 1991, 1993a, etc.). Reisigl a n d W o d a k (2001: 10-14) review some social-psychological explanations, some psychoanalytic explanations and some cognitive psychology research into categorisation a n d stereotyping bur d o n o t incorporate t h e m directly into their discourse analysis approach. No C D A work, so far as I a m aware, draws o n the evolutionary cognitive science research oudined in Section 5 of the present paper. Essentially, the explanation of racism that has b e e n discussed in the cognitive science literature draws on the theory of mental modules a n d on the postulated existence of mechanisms of transfer between autonomous modules. It is worth considering it further because of its potential for linking cognition, language a n d social behaviour in an explanatory framework. M i t h e n (1996: 224-225) argues that racism arise from a particular use of 'cognitive fluidity' acting o n particular specialised m e n t a l modules, namely technical intelligence (tool making, manipulation of objects) a n d social intelligent (thinking about people, together with ingredients such as theory of mind that we h a v e m e n t i o n e d above). T h e general idea is pictured in Figure 2. Mithen's notion is that the area above the line represents the m o d e r n ti.e. Pleistocene a n d post-Pleistocene) m i n d , while the area below the line represents the separate modules of the early h u m a n m i n d . A better interpretation might b e to say that the area above the line stands for cognitive effects that can come about in the m o d e r n h u m a n m i n d u n d e r certain conditions (maybe the ones that C D A is or ought to b e able to specify). Another element in Mithen's frame work is that intuitive biology (the natural kinds module) accounts for human readiness to accept the existence of different social groups (and label them races, etc.). But it is cognitive fluidity that accounts for our readiness to treat some
Technical intelligence module Figure 2: After Mithen (1996:225)
Social intelligence module
CHILTON
MISSING LINKS IN MAINSTREAM CDA
361
groups as inferior. This can come about w h e n the thinking-about-people m o d u l e merges with the thinking-about-tools module. O r w h e n the natural kinds module merges its h u m a n s a n d animals categories for certain groups of the former. W h a t happens is that people can b e thought of as either animals (thus inferior) or as objects (tools) or both. A slightly different, though related account, comes from the w o r k of Atran (1990,1994), Boyer (1990), German (1994) and Hirshfeld (1994). German's position is that the idea of different h u m a n races comes from essentialist beliefs, which are part of the intuitive biology m o d u l e . As Atran argued, intuitive biology pre disposes h u m a n s to set u p taxonomies of living kinds for animals (a schema obviously adaptive for hunter-gatherers in the relevant environment for early humans). German a n d h e r colleagues (e.g. German, Coley & Gottfried 1994) have done empirical developmental work with children, producing evidence that children spontaneously attribute 'essences' to living kinds. Their findings show that children consequently also take it for granted that living kinds h a v e their own invisible internal workings, that these internal workings cause certain behaviours, that attributed essential characteristics inevitably develop or grow, and that the attributed identity is permanent, despite changes, because something 'inside' is regarded as fixed. So children and adults develop racial categories, assumed to h a v e fixed essences, b y transfer from the living kinds m o d u l e to the people category in the social intelligence m o d u l e . However, Hirschfeld's (1994) experiments suggest that children h a v e a n i n d e p e n d e n t tendency to construct racial categories a n d thus that it is not a matter of merging elements from the essentialist natural t a x o n o m y m o d u l e with elements in the people module. T h e latter position has serious implications. Does it m e a n there is a racist module, a domain-specific competence for racial classification'? If such a m o d u l e included the assumption that some races were inferior to others, does this m e a n racism itself is innate? Sperber (1994) is worried b y this implication and proposes an alternative story within the m o d u l a r framework. I n Sperber's view ' p r o p e r ' domains evolve for survival u n d e r specific environmental conditions, but con ditions m a y change, and the m o d u l e can take o n a different role. This would include the case where such a m o d u l e interacts with components of a later culture (which could of course include linguistically mediated cognitive components). Consequently, for Sperber, a m o d u l e can develop a cultural domain. T h e example Sperber gives is the following. Let us assume, as seems plausible, that a zoological d o m a i n (i.e. biology module) evolved in early humans whose j o b was the conceptualisation of animals. Such a m o d u l e would b e adaptive for dealing with local fauna. T h e n suppose the fauna u p o n which the original m o d u l e was based ceased to exist or b e n o t visible (maybe the c o m m u n i t y m o v e d to another location). It is likely, Sperber argues that the m o d u l e r e m a i n e d active a n d could (perhaps, w e m a y interject, in the way M i t h e n suggests) take input from culture (or as we m a y say, from discourse). Sperber's speculation is that any module of this kind might b e biologic ally inherited in the m o d e r n h u m a n m i n d , b u t b e a rather general metatemplate.
DISCOURSE STUDIES
362
Further, h e speculates, such metatemplates can b e 'initialised' b y cultural input. To quote Sperber directly: A cognitive m o d u l e stimulates in every culture the p r o d u c t i o n and distribution of a wide array of information that meets its input conditions. This information, being artifactually produced or organized b y the people themselves, is from the start conceptualized a n d therefore belongs to conceptual domains that I propose to call the module's cultural domain(s). (Sperber 1994: 55; Sperber's italics) W h a t h a p p e n s then is that in the process of cultural transmission (one may warn to call it 'discourse') causes a mimicking of the m o d u l e ' s actual domain based o n information coming from another domain. It follows that racial classification might result from an ad h o c template derived from the hving-kinds metatemplate, through a n initiaJisation triggered by a cultural input. (ibid.: 58) As to the precise nature of this cultural trigger: the m e r e encounter with a n o m i n a l label used to designate a living thing is e n o u g h to tilt the child's categorization of that thing towards an essentialist construal It is quite possible that being presented with nominal labels for otherwise unidentified a n d undescribed h u m a n s is enough (given a n appropriate context) to activate the initialisation of the ad hoc template. (Sperber 1994: 59) This is of course reminiscent of what discourse analysts h a v e long b e e n saying. H o w e v e r , they h a v e b e e n saying it in a theoretical v a c u u m . W h a t is crucial is that the point being m a d e b y Sperber in the last quotation is based on experi m e n t a l w o r k with children, and is also coherent with the theoretical cognitive evolutionary framework. So w h e r e does this leave the suggestion of a domain-specific racism made b y Hirschfeld? W h a t Sperber says o n this score, somewhat disturbingly, is this: T h e r e is, as Hirschfield suggested, a genetically specified competence [my italics] that determines racial classification without importing its models from another concrete domain. (Sperber 1994: 58) H o w e v e r , a n d the point is extremely important, Sperber's approach seems to p r o p o s e that this competence m a y itself b e derived n o t from a competence thai has racial classification as its p r o p e r domain at all but m o r e indirecdy from a different genetic m o d u l a r competence:
CHILTON
MISSING LINKS IN MAINSTREAM CDA
363
Racial classification m a y b e a m e r e cultural domain, based o n a n under lying competence that does not h a v e any p r o p e r domain. (ibid.) His hypothesis (he stresses that it is only a hypothesis) is in fact that there is a higher level metatemplate whose j o b is to generate ad h o c templates for actual zoology and b o t a n y domains (living kinds). This metatemplate is 'initialised' for racial classification. T h e intitialisation of an ad hoc template for racial classification could well b e the effect of parasitic, cultural input o n the higher-level learning m o d u l e , the function of which is to generate ad h o c templates for genuine living-kind domains such as zoology a n d botany. (ibid.) The upshot in relation to the implications of Hirschfeld is that no racist disposition has b e e n selected for [...] in h u m a n s . However, the dispositions that h a v e b e e n selected for m a k e h u m a n s all too easily susceptible to racism given minimal, innocuous-looking cultural input. (ibid.; Sperber's italics) xVow if we ask what 'cultural input' might b e , it is clear that in view of the genetic endowment for language a n d the fact that h u m a n culture depends critically o n linguistic communication, the cultural input initialising racial classification m u s t be talk and text, in other words discourse. We can sum all this u p as in Figure 3.
Figure 3 : Sperber's imualisation-of-template theory of modules i
364
DISCOURSE STUDIES
T h e value of this kind of approach lies in its potential explanatory power. T h a t is to say, it offers an explanatory account of how what Sperber calls 'cultural input' a n d C D A practitioners m a y call 'discourse' has any effect in the human mind, how it takes hold and has the potential to spread from one mind to another. It also vindicates the w o r k d o n e in C D A , which in effect shows something of h o w the cultural input is achieved. T h e gap a Sperberian account fills is the explanation of w h y the discoursal i n p u t should h a v e any effect and b e so persistent in h u m a n societies. However, there is o n e area that neither the CDA account n o r the Sperberian one tells us about in detail (though Sperber and the other cognitive psychologists a n d archaeologists certainly c o m e close to it). This area concerns the actual workings of the m i n d that give rise to what Fodor calls 'passion for the analogical', what Mithen calls 'cognitive fluidity' and what Sperber variously calls 'cultural input', 'initialisation', 'the effect of parasitic information'. I n the next section I show h o w one C L theory, namely blending theory, can give a m o r e detailed account of these processes than is currendy available either in the cognitive evolutionary literature we h a v e b e e n surveying or in the standard C D A literature.
9. U s i n g C o g n i t i v e M o d e l s of D i s c o u r s e T h e r e are in fact rather few models of discourse that h a v e a distinct cognitive flavour. T h e most sophisticated formal theory, which also claims a cognitive element, is D R T (Discourse Representation T h e o r y : see for example Kamp & Reyle 1993 a n d for developments Asher & Lascarides 2003). Another related type of discourse analysis is Werth's (1999) 'text worlds' approach. However, while C D A cannot seriously claim to m a k e a contribution to discourse analysis without taking account of these highly sophisticated models, I shall leave their further exploration to another occasion. For there is in fact another recent theor etical development which is equally indispensable a n d which h a p p e n s to have a direct bearing o n the matter of cognitive fluidity, a n d thus also b y hypothesis, on the question of racial categorisation and racism as formulated b y the researchers discussed in Section 8. Fauconnier a n d Turner (2002) present a theory of cognitive blending de veloped in a series of earlier papers (e.g. Fauconnier & Turner 1996; Fauconnier 1997). Blending theory is a n attempt to explain o n e of m a n y tricks performed b y the h u m a n mind, specifically in the processing of discourse. T h e starting point is the notion of 'mental space', which m a y b e a structure called u p from longt e r m m e m o r y or a structure o n hold in working m e m o r y : Mental spaces are small conceptual packets constructed as w e think and talk, for purposes of local understanding a n d action. T h e y are inter connected, a n d can b e modified as thought a n d discourse unfold.... In blending, structures from two input spaces is projected to a separate space, the 'blend'. T h e b l e n d inherits partial structure from the input spaces, and has emergent structure of its own.
CHILTON MISSING LINKS IN MAINSTREAM CDA
365=
... input spaces u n d e r construction recruit structure from m o r e stable, elaborate, and conventional conceptual structures that m a y have conventional structures of various sorts: shared frame roles, connections of identity or transformation or representation, metaphoric connections. T h e s e c o n v e n t i o n a l c o n n e c t i o n s a r e fully available to t h e w o r k of blending. Blending m a y exploit, simultaneously, m o r e t h a n o n e kind of c o u n t e r p a r t c o n n e c t i o n (e.g. f r a m e - r o l e c o n n e c t i o n a n d i d e n t i t y connection), through entrenchment, blending can influence conventional structures a n d their conventional connections. Blends themselves b e c o m e conventional. (Fauconnier & Turner 1996: 116) These summaries are almost sufficient as a framework for preliminary exposition. In addition, w e n e e d to note t h e characteristic diagrammatic notation developed by Fauconnier a n d Turner, used b e l o w in Figure 4. Its rationale is b a s e d in set theory a n d t h e mathematical notion of m a p p i n g from one d o m a i n to another. For present purposes all we need to note is that mappings place elements in different domains in c o r r e s p o n d e n c e with o n e another. A n example will illustrate t h e kind of cognitive p h e n o m e n o n that is at issue. Consider this kind of discourse, offered b y Fauconnier and Turner: I claim that reason is a self-developing capacity. Kant disagrees with this o n one point. H e says it's innate, b u t I answer that that's begging the question, to which h e counters, etc. etc. (ibid.) This is a familiar academic genre. H o w it works is b y using two mental spaces, one in which the c o n t e m p o r a r y speaker is m a k i n g assertions a n d another in The anti-semite blend
Living kinds template
Perceptual input 'cutture/discourse
+essentit
Blend space
''Conceptual compression elaboration: disease, medicine frames inferencing
DISCOURSE STUDIES
366
which K a n t is speaking. T h e two spaces are separate. But they do have something in c o m m o n (the concepts of thinking, debating, writing) which is connected to b o t h the first two spaces. It is called the 'generic' space. T h e b l e n d occurs in a fourth m e n t a l space which brings together inputs from the m o d e r n philosopher space a n d inputs from the K a n t space. T h e b l e n d space can also draw on other cognitive frames: in this example, it appears to draw o n the shared 'debate' frame. O n c e the b l e n d has emerged, the m i n d can w o r k in that space, elaborating it. drawing inferences from it, etc. T h e b l e n d e d material can feed back into the two input spaces. It should b e already obvious that a n explicitiy worked out theory along these lines does m u c h to fill in the notion of cognitive fluidity (as suggested also by Turner 2001), t h e transfer a m o n g basic m e n t a l m o d u l e s , postulated by the evolutionary cognitive psychologists a n d archaeologists. I n making that point, 1 a m of course using blending. For I a m merging the Fauconnier-Turner model with the m o d u l a r model. In fact, the suggestion is that it is precisely the modules discussed that can (not always, b u t can) provide the input spaces in the blending system described above. Blending crops u p in all m a n n e r of discourse, however: jokes, poetry, oratory, advertising, ideology, to suggest a few examples. C D A researchers will b e inter ested in ideology. T h e t h e o r y of blending, together with evolutionary cognitive science, could enable us to construct a general cognitive theory of ideology, which h a s b e e n lacking in C D A . W h a t such a theory would give us, it should b e noted, is n o t a m e a n s of combating directiy those with whose ideologies we might dis agree, b u t a m e a n s of deepening our understanding of w h a t it could mean to give a n explanation of h o w a n d p e r h a p s w h y h u m a n m i n d s produce such struc tures. H e r e we can d o n o m o r e than sketch out h o w this task might b e broached. Let us take a simple (because extreme) case, which m a n y people in the world today (but n o t all) would categorically reject: antisemitic nazi propaganda, as found classically in Hitler's Mem KampfS' At t h e core of nazi antisemitic p r o p a g a n d a was a crucial cognitive blend that spread back into m a n y aspects of discourse. As is well k n o w n , Mein Kampf uses, inter alia, the following argument: 5
X is a parasite (virus, bacillus, etc.) in the body of Y bodies can be cured by expelling or destroying parasites doctors expel/destroy parasites, etc. Therefore, Y can be cured if doctors expel X from Y W h a t are the cognitive processes involved in a n argument of this kind? Following blending theory we can see that we h a v e (i) an input space or source domain, which includes the concept 'parasites'. N o w 'parasites' is part of a preexisting structured cognitive frame, which necessarily includes concepts such as 'host' organism. I n fact, it looks u n c o m m o n l y like w h a t Atran, Sperber, Mithen a n d others regard as a mo d u l e of m i n d that has to d o with the taxonomy of living kinds, or naive zoology. Recall that this m o d u l e , according to several accounts, incorporates essentialism: the ascription of inherent a n d constant properties of
367
CHILTON MISSING LINKS IN MAINSTREAM CDA
'kinds'. We also h a v e (ii) a second input space, a target domain, which h e r e is a social group that has contingently b e c o m e distinct in t h e culture w h e r e the dis course (and the formation of concepts in m a n y m i n d s in that culture) is taking place. Clearly, this space is largely formed b y m a n y c o m p l e x inputs from social discourse, including the kinds of linguistic communication ('discourse') studied by van Dijk (1991,1993a), Reisigl a n d W o d a k (2001) and others, not to m e n t i o n the large b o d y of explanatory attempts b y the Frankfurt School and studies of nazi language by, e.g. Ehlich (1989). I n this instance the social group in question isjews. It could also include other contingent social entities, say nation states, or geographical regions. It is highly relevant that larger groups can contain, or b e conceived as containing, smaller groups. I n addition, there is a cross d o m a i n m a p p i n g w h i c h establishes a n a l o g i e s b e t w e e n h o s t - p a r a s i t e a n d n a t i o n subgroup (Jews). In terms of mental spaces, w e have (hi) a c o m m o n (or 'generic') conceptual space. In this instance, the relevant one appears to b e the image schema containercontained (on image schemas, a n d specifically, container, see J o h n s o n 1987 a n d j o h n s o n & Lakoff 1999 a n d o n the importance of this schema in large-scale systems of political belief Chilton 1996a). Finally, w e h a v e (iv) the b l e n d space, where the input spaces are merged. H e r e , a n d this is the crucial point, is w h e r e further inferencing and elaboration is effected. According to Fauconnier and Turner, it is in the blend space that a cognitive p h e n o m e n o n that they term 'con ceptual compression' occurs (Fauconnier & Turner 2 0 0 2 : 9 2 - 9 3 ) . A t the same time, elaboration is a kind of expansion, which in the case of Mein Kampf, a n d antisemitic discourse in general, recruits further cognitive frames, specifically knowledge frames that structure basic knowledge about disease, healers, healing, and curative m e t h o d s . Also recruited in ideological discourse of this type are cognitive frames centred o n the b o d y which draw on the container image schema. Figure 4 represents a simplified picture of the way the m i n d ' s 'cognitive fluidity' operates to creatively p r o d u c e concepts about social groups. Such m o d e l s offer C D A detailed hypotheses about h o w discourse affects social cognition. Notice that they are dynamic, in the sense that they can enter into inference chains. I n the present e x a m p l e , the b l e n d 'nations are hosts, J e w s are parasites' engages the cognitive frame of biology a n d disease. I n turn, the disease frame permits inferences of the kind: 7
(i) (ii)
If a is a parasite in b, a causes disease c If cis a disease, b can b e cured b y rfwith « a n d / o r b y d o i n g / W h e r e d\s (say) a doctor, *for instance is a medicine that (e) kills a. Therefore, to cure b, kill a.
Such inferences are n o t stricdy logical n o r empirical, but based on the b l e n d of one d o m a i n with another. Simply labelling a g r o u p 'parasites' activates a blend, which in turn activates an inferential chain. T h e conclusion 'kill d m a y or m a y not b e implanted a n d result in real exterminations, p o g r o m s a n d genocides. This is a hypothesis a b o u t h o w discourse ('cultural input') can affect ('infect', if one wants to use t h e m e t a p h o r ) h u m a n m i n d s , and lead rn arts V w Y i n m * ™
368
DISCOURSE STUDIES
against other h u m a n s . This link is missing from current C D A . T h e cognitive ac count tells you the mechanisms. W h a t it does not do is tell you why such inferences should lead to certain historical events at a certain time a n d place. If you want answers to this particular why, there are several routes to pursue. O n e is CDA. which might b e able to describe the characteristics of the 'cultural input' (the discourse) that might h a v e activated the blends ('initialised the template' in Sperber's terminology discussed above), facilitated the inferences and promoted the conclusion. It might tell us about the social a n d political conditions that led to certain kinds of communication network that m a d e such input possible or effective. Typically, with the partial exception of the 'discourse-historical ap proach' espoused b y Wodak a n d colleagues, C D A does n o t take this latter step, but sticks at the level of verbal detail. To understand why particular cognitive effects arise a n d why they result in action at a certain time or place is not a ques tion cognitive scientists or linguists address. It is not evident that C D A is equipped to address this question either, without massive collaboration from, at least, the disciplines of political e c o n o m y a n d history. T h e cognitive account oudined h e r e attempts to fill in some of the missing links that are implicit, or ought to b e implicit, in the C D A project - namely, exacdy h o w (i.e. b y what mechanisms) can h u m a n utterances conceivably lead to appalling h u m a n actions?
10. Passive R e c e p t o r s or Critical Instinct? D o We N e e d CDA? To sum u p the argument so far, to understand discourse requires a cognitive perspective. We h a v e reviewed some of the recent w o r k in evolutionary psych ology that appears to h a v e considerable relevance for C D A . We have also re viewed o n e corner of cognitive linguistics, for the same reason. I n Section 9, we h a v e shown two things: one, h o w blending theory can a d d detail to some key claims in cognitive psychology a n d two, h o w a blend of b l e n d i n g theory and cognitive evolutionary psychology can a d d detail ('depth') to our understanding of ideology, a classic topic in C D A . I n this final section w e attempt to draw out some of the implications for C D A of taking this route. To accomplish this we can m a k e further use of recent work in cognitive science. Two kinds of question will guide the discussion. First, does the evolutionary a n d cognitive perspective imply that the h u m a n m i n d is easily manipulated because of the effects due to innate modules and mechanisms like blending? Second, if C D A took into account the cognitive perspective we h a v e outlined, what sort of research would CDA b e c o m e a n d could it still claim emancipatory benefits? T h e question of h o w easily or n o t the h u m a n m i n d can b e tricked, deceived or manipulated through the use of language is a core empirical a n d theoretical question that C D A has never addressed as such. T h e r e are well-known psycholo gical studies o n such p h e n o m e n a as 'groupthink' and o n the framing of narratives b u t littie detailed discussion of the very possibility of verbal manipulation or h o w or to what extent it might work in detail (but cf. Allott forthcoming; Chilton 2 0 0 2 ; a n d Saussure forthcoming). Indeed, from Fowler and others on, in the
CHILTON
MISSING LINKS IN MAINSTREAM CDA
369
wake of Orwell's n e w s p e a k n i g h t m a r e , t h e r e h a s simply t e n d e d to b e a n assumption that people are easily h o o d w i n k e d b y powerful verbalisers. W h a t follows is an attempt to o p e n u p this question to closer scrutiny. We return to recent work in cognitive evolutionary psychology for clues a n d cues. As was noted earlier, Sperber has postulated a genetically evolved metarepresentation m o d u l e which h e regards as central to a n u m b e r of characteristically h u m a n mental and other behaviours, including creativity a n d consciousness. Metarepresentation, according to this hypothesis is a second-order m o d u l e with the ability to form concepts of concepts a n d mental representations of mental representations, one's o w n a n d those of others. T h e evolutionary perspective prompts a form of argumentation that goes as follows. W h y w o u l d h o m o sapiens have developed such a n ability? Because, we m a y speculate, h u m a n s n e e d to k n o w w h a t others are thinking (cf. 'theory of mind', 'machiavellian intelligence'). T h a t is to say, such knowledge has survival value, because it is important to k n o w what others inside and outside one's group are wanting, believing, intending, etc. Further, metarepresentation must b e integral to h u m a n language/communication, for in processing a speaker's (S's) vocalisa tion the hearer (H) forms a mental representation of what S is (probably) rep resenting in their o w n m i n d . I n d e e d , communication occurs w h e n H recognises S's intention that H construct some mental representation corresponding with some representation in the m i n d of S. C o m p r e h e n s i o n is metarepresentation. This line of argument has some possible implications for C D A , at least for that part of it that is concerned with strategic deception through language use. H u m a n s m a y h a v e innate mechanisms for communication a n d comprehension, but these precisely facilitate tactical deception, that is, machiavellian behaviour. It would m a k e n o sense to assume that some h a d a machiavellian module a n d some didn't (which doesn't m e a n that some d o not use it a n d cultivate it m o r e than others, as w o u l d b e argued for any of the other postulated modules). But if everyone has it, it is also necessary to assume that all individuals h a v e the ability to counter it in others. T h e hypothesis is, then, that h u m a n s need to b e able to detect intentions to engage in deceptive action. Since language use is also a form of action, we must ask whether h u m a n s are not also naturally e n d o w e d with the ability to detect deceptive language use. Cosmides (1989), o n w h o m Sperber is drawing, regards some such ability as a specialised mental m o d u l e which she calls the 'cheater detector'. Sperber elaborates o n this in a fashion that should attract the attention of C D A : T h e h u m a n reliance on communication is so great, the risks of deception and manipulation so ubiquitous, that it is reasonable to speculate that all available cost-effective m o d e s of defense are likely to h a v e evolved. (Sperber 2000: 135) Building o n Cosmides, Sperber goes further than Cosmides in grouping together possible c o m p o n e n t modules of this innate ability. First, there is an in-built facility
370
DISCOURSE STUDIES
for checking the reliability of the source of information, in line with the idea of 'source tagging' (Cosmides & Tooby 2000). T h e recursive syntax of h u m a n lan guages permits e m b e d d i n g of propositions as complements of reportative verbs, evidential a n d attitudinal verbs, facilitating the metarepresentation of sources (amongst other things). Second, theory of m i n d supplies the ability to 'be sensitive to subtle signs of deceptive intent [and] read the relevant attitudinal a n d emotional signs' (Sperber 2000: 135). Third, Sperber proposes that there exists a 'consistency filter', by m e a n s of which the h u m a n m i n d is innately inclined to check the logical consist ency of language input, o n the one h a n d with respect to the internal logical consistency of a stretch of talk (or text), a n d o n the other h a n d with respect to external consistency with what the receiver believes to b e true and with alter native informational input. Because a speaker S will recognise a hearer's intention to counter S's deceptive intents, Sperber argues that S will try to penetrate H ' s defences, resulting in "a persuasion counter-persuasion arms race' (Sperber 2000: 136). This point is in one sense obvious, b u t it is important to stress the claim that the modular mental basis of such a spiral is an innate ability. Even m o r e specifically, Sperber proposes that this ability will h a v e evolved in such a say as to give rise to a further innate m o d u l e , a 'logico-rhetorical' m o d u l e . This is a 'persuader' m o d u l e that seeks to penetrate H ' s cheater detector and source tagger. It can display authority and simulate consistency. Again, o n e m u s t assume that this m o d u l e too works both ways: that is, it works for S's goals b u t also for H's defences as a counter-persuader ability. This last point is rather interesting for CDA, since it postulates. ...the development of the ability to scrutinise these argumentative displays and to find fault with them. (Sperber 2000:136) I n other words, h u m a n s m a y already have a critical instinct, even perhaps some thing like a m o d u l e for C D A . Naturally, people m a y n o t use it. But, firsdy, to say that people d o n o t use it, makes a claim that needs substantiating. D o they not? T h e question is not simple, because recognising failure to deploy critical ability requires criteria that would certainly b e highly disputable. Secondly, if people d o not, or do not all, or do not all of the time, this would not demonstrate an innate critical potential does n o t exist. Thirdly, if the claim is that people have the critical potential but d o not use it, then what we n e e d to know is under which conditions is this the case? Historians, political scientists and other disciplines can do this j o b as well as, or better than C D analysts. O n e of CDA's claims seems to b e that it is primarily something about the discourse (utterances) that hampers the use of the h u m a n critical potential a n d p r o m o t e s nefarious notions. This claim has n o t b e e n substantiated. Nor has C D A developed any detailed theory of h o w cognitive influence might occur as result of discourse input to a human brain. T h e cognitive theory outlined in this section is a very modest step in addressing this deficit in C D A .
CHILTON
MISSING LINKS IN MAINSTREAM CDA
371
11. Implications for C D A Whether the proposals of Sperber, Cosmides a n d others are valid can of course only rest o n further empirical a n d theoretical work. But sufficient work already exists for the claims to b e worthy of consideration. In this p a p e r we h a v e (a) argued that C D A cannot neglect the cognitive aspect of communication, (b) rapidly summarised some recent work in the field of evolutionary cognitive psychology, (c) rapidly summarised one recent approach in cognitive linguistics, a n d (d) applied a b l e n d of (b) a n d (c) to o n e aspect of an extreme form of racist discourse. This review a n d application yields two m a i n points. First, the combined cognitive framework can go beyond description and put forward suggestive explanatory stories. Second, the c o m b i n e d cognitive framework raises questions about the status a n d direction of C D A itself. W h a t then are the issues for the status a n d direction of C D A ? i. T h e combined framework of cognitive evolutionary psychology a n d C L yields insights into h u m a n nature a n d h u m a n societies that C D A itself has n o t provided. It m a y b e that C D A , given its adopted m e t h o d s and goals a n d the lin guistic theories it has predorninantly drawn on, m a y b e incapable of going beyond description. I a m thinking here primarily of the kinds of C D A which take their theoretical frames predominantly from social theory, without the incorporation of cognitive perspectives, and which also draw in rather limited ways o n heavily socialised linguistic models. I exclude those less p r o m i n e n t currents that place cognition at the core a n d which d o draw, to varying degrees, o n psychology and on cognitive linguistics. ii. Even if we accept only very tentatively the m o d u l a r m o d e l of mind, and the particular m o d u l e s for which there is evidence or at least plausible deductive arguments, the consequences for C D A as a socially oriented, possibly emancipa tory enterprise, h a v e to b e re-examined. Put blundy, if people have a natural ability to treat verbal input critically, in what sense can C D A either reveal in discourse what people can (by the hypothesis) already detect for themselves or educate t h e m to detect it for themselves? T h e r e are obvious responses to this question. O n e might b e that, granted a 'critical discourse' module, such a module still needs to b e activated a n d developed. T h e r e are problems with such a m o v e , however. O n e is that it places C D A in a position of responsibility and p e r h a p s power, if an educating or developing role is granted to C D A practitioners. Another is the problem of knowing what sort of technical analysis C D A could conceivably provide that could give what is required for s o m e o n e to develop their critical discourse module. Stock responses to C D A include the charge that it is stating the obvious a n d that n o technical apparatus is n e e d e d anyway to point out what is not p e r h a p s obvious. Such responses might b e taken to suggest that people are indeed aware of the existence of their own critical faculties. Moreover, taking the point that technical apparatus is n o t needed, this m e a n s (a) that C D A has n o core scientific p r o g r a m m e of discovery or elaboration a n d then (b) also that its ability to bring strategic and tactical language use to awareness m a y b e n o greater than that possessed b y the non-expert a n d that it is merely engaging in ordinary
DISCOURSE STUDIES
372
iii. T h e last set of points could b e taken to imply that the focal research question might b e formulated as the question: W h a t prevents people using their innate cheater-detecting logico-rhetorical modules to protect their own interests? I a m suggesting that the answer is n o t that they cannot analyse the verbal input a n d n e e d help in spotting, for example, nominalisations or agent-less passives, or intertextual cross-overs. Next to nothing is k n o w n about whether such things do really h a v e any cognitive effects a n d thus social action effects. W h a t is clear, however, is that u n d e r certain social, economic a n d political conditions people m a y n o t b e able, or m a y n o t b e willing, to respond critically. O n e can imagine m a n y such conditions, but it is doubtful that any of t h e m can b e elucidated by purely linguistic or discourse-analytical m e a n s . For they would seem to have to do with economic forces or socio-political institutions that restrict freedom of expression and freedom of access to information, without which y o u cannot use your source-tagging abilities, your cheat detector, your consistency checker, etc. This is not to d e n y the central importance of linguistic communication, but it does cast d o u b t o n whether technical or theoretical research or description in C D A has any special social role. I n fact, if the preceding argument is correct that it is institutional or economic restrictions o n communicative freedom that is crucial, then what is n e e d e d is historical, social, economic and political analysis not the analysis of language itself. And, interestingly enough, it is Chomsky, whose linguistic models have n o social or political critique, w h o has provided an analysis of the restrictions o n expression a n d access in m o d e r n capitalist societies ('A p r o p a g a n d a m o d e l ' in H e r m a n & C h o m s k y 1988: 1-35). iv. T h e last p a r a g r a p h argues that CDA's would-be theoretical a n d technical analyses of text a n d talk m a y have n o direct bearing o n social a n d political con ditions. It m a y b e that it could carry a pedagogical role, b u t in that case it may b e doing n o m o r e than any politically aware person. If it does set research goals, within the normal understanding of research in the western scholarly and scientific tradition, then it m a y necessarily separate itself from social a n d political goals. T h e route that has b e e n indicated in the bulk of the present p a p e r goes in the direction of deductively elaborating empirical a n d theoretical research, and also exploring p h e n o m e n a like racism from within that paradigm. W h a t I have suggested about racism in Section 9 is of n o obvious use in combating racism. If it has a n y value it is of a scientific kind, in so far as it sketches a hypothesis as to h o w certain conceptual structures take hold in the mind and spread. There remain questions about the socio-economic a n d political conditions u n d e r which certain cognitive effects take hold of whole populations b y way of verbal communications (and p e r h a p s other systems), but these are not specifically linguistic matters on which language experts can pronounce. There remain, too, profound philosophical questions about volition, agency a n d ethical responsibility. T h e upshot of these ruminations is that C D A m a y n o t b e the field within which the most perplexing questions can b e pursued that concern the nature of the h u m a n m i n d , of h u m a n language, of h u m a n language use a n d of human society. 8
9
CHILTON
MISSING LINKS IN MAINSTREAM CDA
373
Notes 1. I am most grateful to Patricia Chilton, Bill Downes, Ron Scollon and Ruth Wodak for critical discussion to do with ideas in this paper. This does not mean that they would agree with anything I say in it. 2. See below on cognitive science and cognitive linguistics in a socio-critical context. It is important also to mention the important work inspired by pragmatics produced by Verschueren, Blommaert, Bulcaen (Blommaert & Verschueren 1998; Blommaert & Bulcaen 1997), who have tended to keep their distance both from mainstream CDA and from Hallidayan linguistics (see Verschueren 2001). Downes (2002) points out that Hallidayan systemic-functional grammar fails to take account of inferencing and world knowledge in human language understanding. He also argues that CDA fails to resolve the tension between CDA's claim to reveal social functions of discourse and speakers' conscious intentions and rationalisations. These points are somewhat related to O'Halloran's (2003) arguments that the presence of items and structures in text is not necessarily related to the cognitions established by the reader/hearer on the basis of such text features, as well as to his general application of the distinction between low-effort shallow reading and high-effort interest-oriented reading of texts. 3. The more cognitively oriented linguists concerned with social issues do, I think, assume something like this argument; but these are not the authors typically associated with the central tendency of CDA. 4. The theories in cognitive linguistics referred to have developed quite independendy of the theories of modularity discussed earlier. 5. Interestingly, such an approach takes us back to the origins of the term ideologic as used by Destutt de Tracy, Elements d'idiohgie (1817-1818). 6. Cf. Chilton, 'Manipulation, memes and metaphors: The case of Mein Kampf in L. de Saussure, ed. (forthcoming). 7. The mechanisms by which 'recruiting' occurs are not specified by Fauconnier and Turner, but it seems likely that there is overlap between modular knowledge frames. For example, there is at least a culture-based naive medicine frame that intersects with the naive zoology module. Relevance principles may also apply. 8. It has been pointed out to me by Ruth Wodak that in certain highly structured social spheres, e.g. doctor-patient communication, educational institutions; discourse behaviour has been changed as a result of CDA-type activity. This may be correct. However, there is a risk in such claims of drawing CDA's own charge of the 'technologisation of discourse' (cf. Fairclough 1989, 1995, 2003). 9. 'Epidemiologically', Sperber would say (Sperber 1994); cf. also Dawkins on 'memes' (1999) and Chilton (forthcoming).
References Allott, N. (forthcoming). "The role of misused concepts". In L. de Saussure (Ed.), New Perspectives on Manipulation and Ideologies: Theoretical Aspects. Amsterdam: John Benjamins. Asher, N. & Lascarides, A. (2003). Logics ofConversation. Cambridge: Cambridge University Press. Atran, S. (1990). Cognitive Foundations of Natural History. Cambridge: Cambridge Univer sity Press.
374
DISCOURSE STUDIES
Atran, S. (1994). "Core domains versus scientific theories: Evidence from systematics and Itza-Maya folklbiology". In L. A. Hirschfeld & S. A. Gelman (Eds.), Mapping the Mind. Cambridge: Cambridge University Press. Barkow, J. H., Cosmides, L., & Tooby, J. (1992). The Adapted Mind. New York: Oxford University Press. Baron-Cohen, S. (2001). Mindblindness: An Essay on Autism and Theory of Mind. Cambridge, MA: MIT Press. Blommaert,J. &Bulcaen, C. (Eds.). (1997). Political Linguistics. Amsterdam:John Benjamins. Blommaert, J. & Verschueren, J. (1998). Debating Diversity. London: Roudedge. Boroditsky, L. (1997). "Evidence for Metaphoric representation: perspective in space and time''. In M. G. Shafto & P. Langley (Eds.), Proceedings of the Nineteenth Annual Conference of the Cognitive Science Society. Mahwah, NJ: Erlbaum. Boyer, P. (1990). Tradition as Truth and Communication. Cambridge: Cambridge University Press. Byrne, R. (1995). Thinking Ape. Evolutionary Origins oflntelligence. Oxford: Oxford Univer sity Press. Byrne, R. & Whiten, A. (Eds.). (1988). Machiavellian Intelligence. Social Expertise and the Evolution of Intellect in Monkeys, Apes and Humans. Oxford: Clarendon Press. Caldas-Coulthard, C.-R. & Coulthard, M. (1996). Texts and Practices: Readings in Critical Discourse Analysis. London: Roudedge. Chomsky, N. (1988). Knowledge ofLanguage: Its Nature, Origin and Use. Cambridge, MA: MIT Press. Chilton, P. (1996a). Security Metaphors. New York: Peter Lang. Chilton, P. (1996b). "The meaning of security". In F. A. Beer & R. Hariman (Eds.), PostRealism: The Rhetorical Turn in International Relations. East Lansing: Michigan State University Press. Chilton, P. (2002). "Manipulation". I n j . Verschueren,J.-O. Ostman.J. Blommaert, & C. Bulcaen (Eds.), Handbook ofPragmatics. Amsterdam/Philadelphia:John Benjamins. Chilton, P. (forthcoming). "Manipulation, memes and metaphors: The case of Mein Kampf. In L. de Saussure (Ed.), New Perspectives on Manipulation and Ideologies: Theoretical Aspects. Amsterdam: John Benjamins. Chilton, P. & Lakoff, G. (1999 [1995]). "Foreign policy by metaphor". In C. Schaffner & A. Wenden (Eds.), Language and Peace. Hawood: Academic Publishers. Cosmides, L. (1989). "The logic of social exchange: Has natural selecdon shaped how humans reason? Studies with the Wason Selection Task". Cognition, 31, 187-276. Cosmides, L. & Tooby,J. (1992). "Cognitive adaptations for social exchange." I n j . Barkow, L. Cosmides, &J. Tooby (Eds.), The Adapted Mind. New York: Oxford Univer sity Press. Cosmides, L. & Tooby,J. (2000). "Consider the source: The evolution of adaptations for decoupling and metarepresentations". In D. Sperber (Ed.), Metarepresentations. Oxford: Oxford University Press. Dawkins, R. (1999 [1976]). The Selfish Gene. Oxford: Oxford University Press. Dirven, R., Frank, R., & Hie, C. (Eds.). (2001). Language and Ideology. Volume II: Descriptive Cognitive Approaches. Amsterdam: John Benjamins. Discourse and Society (2000). Volume 11, number 1, ed. R. Wodak, special issue on discourse and racism. Downes, W. (2002). "Linguistic criticism and functional explanation", [unpublished ms] Ehlich, K. (Ed.). (1989). Sprache im Faschismus. Frankfurt: Suhrkamp. Fairclough, N. (1989). Language and Power. London: Longman. H a i r r l n n o r l v NT (FA.). (19921. Critical Language Awareness. London: Longman.
CHILTON
MISSING LINKS IN MAINSTREAM CDA
375
Fairclough, N. (1995). Critical Discourse analysis: The Critical Study ofLanguage. London: Longman. Fairclough, N. (2003). Analyzing Discourse: Textual Analysis for Social Research. London: Routiedge. Fairclough, N. & Wodak, R. (1997). "Critical discourse analysis". In T. van Dijk (Ed.), Discourse as Social Interaction (pp. 258-284). London: Sage. Fauconnier, G. (1997). Mappings in Thought and Language. Cambridge: Cambridge University Press. Fauconnier, G. & Turner, M. (1996). "Blending as a central process of grammar". In A. E. Goldberg (Ed.), Conceptual Structure, Discourse and Language. Stanford, CA: CSLI Publications. Fauconnier, G. & Turner, M. (2002). The Way We Think: Conceptual Blending and the Mind's Hidden Complexities. New York: Basic Books. Fodor.J. (1983). The Modularity of Mind. Cambridge, MA: MIT Press. Fodor, J. (1985). "Precis of "The Modularity of Mind'". Behavioral and Brain Sciences, 8, 1-42. Fowler, R. (1996). Linguistic Criticism. Oxford: Oxford University Press. Fowler, R. (1996). "On critical linguistics''. In C. R. Caldas-Coulthard & M. Coulthard (Eds.), Texts and Practices. Readings in Critical Discourse Analysis (pp. 3-14). London: Routiedge. Fowler, R., Hodge, G., & Kress, G. (1979). Language and Control. London: Roudedge and Kegan Paul. Gelman, S. A., ColeyJ. D., & Gottfried, G. M. (1994). "Essentialist beliefs in children: the acquisition of beliefs and theories". In L. A. Hirschfeld & S. A. Gelman (Eds.), Mapping the Mind. Cambridge: Cambridge University Press. Gibbs, R. (1994). The Poetics of Mind: Figurative thought, Language and Understanding. Cambridge: Cambridge University Press. Herman, E. S. & Chomsky, N. (1988). Manufacturing Consent: the Political Economy of the Mass Media. London: Vintage Books. Hirschfeld, L. A. & Gelman, S. A. (Eds.). (1994). Mapping the Mind: Domain Specificity in Cognition and Culture. Cambridge: Cambridge University Press. Hirschfeld, L. A. (1994). "Is the acquisition of social categories based on domain-specific competence or on knowledge transfer?" In L. A. Hirschfeld & S. A. Gelman (Eds.), Mapping the Mind. Cambridge: Cambridge University Press. Hodge, R. & Kress, G. (1993 [1979]). Language as Ideology. London: Roudedge. Humphrey, N. (1976). "The social function of intellect". In P. P. G. Bateson & R. A. Hinde (Eds.), Growing Points in Ethology. Cambridge: Cambridge University Press. Johnson, M. (1987). Hie Body in the Mind. Chicago: Chicago University Press. Kamp, H. & Reyle, U. (1993). From Discourse to Logic. Dordrecht: Kluwer. Karmiloff-Smith, A. (1992). Beyond Modularity. Cambridge, MA: MIT Press. Keil, F. (1994). "The birth and nurturance of concepts by domains: the origins of concepts of living things". In L. A. Hirschfeld & S. A. Gelman (Eds.), Mapping the Mind. Cambridge: Cambridge University Press. Kennedy,J. S. (1992). The New Anthropomorphism. Cambridge: Cambridge University Press. Keysar, B., Yeshayahu, S., Glucksberg, S., & Horton, W. S. (2000). "Conventional language: how metaphorical is it?". Journal of Memory and Language, 43,576-593. Lakoff, G. (1987). Women, Fire and Dangerous Things. Chicago: Chicago University Press. Lakoff, G. (1991). "Metaphor in Politics". An open letter to the Internet from George Lakoff, http://philosophy.uoregon.edu/metaphor/lakoff-l.htm. accessed 2 9 . 0 3 0 4
DISCOURSE STUDIES
Lakoff, G. (1996). Moral Politics. What Conservatives Know that Liberals Don't Chicago: Chicago University Press. Lakoff, G. (1999). Philosophy in the Flesh. London: Basic Books. Lakoff, G. (2003). "Metaphor and war, again", http://www.alternet.org/story.html? StoryID=15414, accessed 29.03.04. Lakoff, G. &Johnson, M. (1980). Metaphors We Live By. Chicago: Chicago University Press. Lakoff, G. & Johnson, M. (1999). Philosophy in the Flesh. Lakoff, G. & Turner, M. (1989). More than Cool Reason: A Field Guide to Poetic Metaphors. Chicago: University of Chicago Press. Leslie, A. (1991). "The theory of mind impairment in autism: evidence for a modular mechanism of development". In A. Whiten (Ed.), Natural Theories of Mind: Evolution and Development and Simulation of Everyday Mindreading. Oxford: Blackwell. Mithen, S. (1996). The Prehistory of the Mind: A Search for the Origins ofArt, Religion and Science. London: Thames and Hudson. O'Halloran, K. (2003). Critical Discourse Analysis and Language Cognition. Edinburgh: Edinburgh University Press. Reisigl, M. & Wodak, R. (2001). Discourse and Discrimination. London: Roudedge. Rohrer, T. (1995). "The metaphorical logic of (political) rape: The new wor(l)d order". Metaphor and Symbolic Activity, 70(2), Spring 1995. Santa Ana, O. (2002). Brown tide Rising: Metaphors ofLatinos in Contemporary American Public Discourse. Austin, TX: University of Texas press. Saussure, L. de (forthcoming). New Perspectives on Manipulation and Ideologies: Theoretical Aspects. Amsterdam: John Benjamins. Shapiro, M. (Ed.). (1984). Language and Politics. Oxford: Blackwell. Spelke (1991). "Physical knowledge in infancy: reflections on Piaget's theory". In S. Carey & R. Gelman (Eds.), Epigenesis of Mind: Studies in Biology and Culture. Hilldale. NJ: Erlbaum. Sperber, D. (1994). "The modularity of thought and the epidemiology of representation". In L. A. Hirschfeld & S. A. Gelman (Eds.), Mapping the Mind (pp. 39-67). Cambridge: Cambridge University Press. Sperber, D. (2000). "Metarepresentation in an evolutionary perspective". In D. Sperber (Ed.), Metarepresentations, A multidisciplinary perspective. Oxford: Oxford University Press. Sperber, D. (Ed.). (2000). Metarepresentations. A Multidisciplinary Perspective. Oxford: Oxford University Press. Thibault, P. (1986). "Text, discourse and context: towards a social semiotic perspective". Toronto Semiotic Circle Monographs, 3. Turner, M. (2001). Cognitive Dimensions of Social Science. Oxford: Oxford University Press, van Dijk, T. (1977). Text and Context: Explorations in the Semantics and Pragmatics of Discours New York: Longman. van Dijk, T. (1980). Macrostructures: An Interdisciplinary Study ofGlobal Structures in Discours Interaction, and Cognition. Mahwah, NJ: Erlbaum. van Dijk, T. (1984). Prejudice in Discourse. Amsterdam: Benjamins, van Dijk, T. (1991). Racism and the Press: Critical Studies in Racism and Migration. London: Routledge. van Dijk, T. (1993a). Elite Discourse and Racism. Newbury Park, CA: Sage, van Dijk, T. (Ed.). (1993b). Critical Discourse Analysis, Special Issue. Discourse and Society, 4(2). van Dijk, T. (1998). Ideology. A Multidisciplinary Approach. London: Sage.
CHILTON
MISSING LINKS IN MAINSTREAM CDA'
377
van Dijk, T. (1998). "Context models in discourse processing". In H. Oostendorp & S. Goldman (Eds.), The Construction of Mental Models during Reading (pp. 123-148). Hilsdale, NJ: Erlbaum. van Dijk, T. (2004). "Text and context of parliamentary debates". In P. Bayley (Ed.), Cross-Cultural Perspectives on Parliamentary Discourse. Amsterdam: Benjamins, van Dijk, T. (forthcoming). "Contextual knowledge management in discourse production: A CDA perspective". In R. Wodak & P. Chilton (Eds.), A New Research Agenda in CDA: Theory and Multidisciplinarity. Amsterdam: Benjamins, van Dijk, T. & Kintsch, W. (1983). Strategies of Discourse Comprehension. New York: Academic Press. van Leeuwen, T. (1996). "The representation of social actors". In C. R. Caldas-Coulthard (Ed.), Texts and Practices. London: Routledge. Verschueren, J. (2001). "Predicaments of criticism". Critique of Anthropology, 2/(1), 59-81. Waal, F. de (1982). Chimpanzee Politics. London: Jonathan Cape. Weiss, G. & Wodak, R (Eds.). (2003). Critical Discourse Analysis: Theory and Interdisciplinary. Basingstocke: Palgrave Macmillan. Werth, P. (1999). Text Worlds: Representing Conceptual Space inDiscourse. London: Longman. Whiten, A. (Ed.). (1991). Natural Theories of Mind: Evolution, Development and Simulation of Everyday Mindreading. Oxford: Blackwell. Wodak, R. (1986). "Normal and deviant texts. The sociopsychological theory of textplanning." In Y. Tobin (Ed.), From Sign to Text. Amsterdam: Benjamins. Wodak, R (Ed.). (1989). Language, Power and Ideology. Amsterdam: Benjamins. Wodak, R. (1996). Disorders of Discourse. London: Longman. Wodak, R. (2000). "The rise of racism: An Austrian or a European problem". Discourse and Society, 71(1), 5-6. Wodak, R. & Meyer, M. (Eds.). (2001). Methods of Critical Discourse Analysis. London: Sage.
SAGE BENCHMARKS IN DISCOURSE STUDIES
DISCOURSE STUDIES VOLUME II
EDITED BY TEUN A . VAN DIJK
Introduction and editorial arrangement © Teun A . v a n Dijk 2 0 0 7 First published 2007 Reprinted 2008 Apart from any fair dealing for the purposes of research or private study, or criticism or review, as permitted under the Copyright, Designs and Patents Act, 1988, this publication m a y be reproduced, stored or transmitted in any form, or b y any means, only with the prior permission in writing of the publishers, or in the case of reprographic reproduction, in accordance with the terms of licences issued b y the Copyright Licensing Agency. Enquiries concerning reproduction outside those terms should be sent to the publishers. Every effort has b e e n made to trace and acknowledge all the copyright owners of the material reprinted herein. However, if any copyright owners have n o t b e e n located and contacted at the time of publication, the publishers will be pleased to make the necessary arrangements at the first opportunity. •
S A G E Publications Ltd
:
.
•
f
1
i
-
>.
.
^ | '
1 Oliver's Yard 5 5 City R o a d L o n d o n E C 1 Y ISP S A G E Publications Inc. 2 4 5 5 Teller Road Thousand Oaks, California 91320 S A G E Publications India Pvt Ltd B 1 / 1 1 Mohan Cooperative Industrial Mathura R o a d N e w Delhi 110 0 4 4
Area
S A G E Publications Asia-Pacific Pte Ltd 33 Pekin Street # 0 2 - 0 1 Far East Square Singapore 048763 British L i b r a r y C a t a l o g u i n g i n P u b l i c a t i o n d a t a A catalogue record for this book is available from the British Library I S B N : 978-1-4129-3617-0 (set of five volumes) L i b r a r y of C o n g r e s s C o n t r o l N u m b e r : 2006936728 Typeset by Star Compugraphics Private Limited, Delhi Printed o n paper from sustainable resources Printed and b o u n d in Great Britain b y TJI Digital, Padstow, Cornwall
CONTENTS
V O L U M E II
16. T h e Interpenetration of Communicative Contexts: E x a m p l e s from Medical Encounters Aaron V. Cicourel 17. Answers and Evasions Steven E. Clayman 18. Author Identification, Idiolect, and linguistic Uniqueness Malcolm Coulthard 19. Bodies, Beaches a n d Burn-Times: 'Environmentalism' a n d Its Discursive Competitors Nikolas Coupland andJustine Coupland 20. Figures of Speech: Figurative Expressions a n d the M a n a g e m e n t of Topic Transition in Conversation Paul Drew and Elizabeth Holt 21. Universal a n d Culture-Specific Properties of Greetings Alessandro Duranti 2 2 . Moaning, W h i n g i n g a n d Laughing: T h e Subjective Side of Complaints Derek Edwards 2 3 . Trial Discourse a n d Judicial Decision-Making: Constraining the Boundaries of G e n d e r e d Identities Susan Ehrlich 24. T h e Occasioning a n d Structure of Conversational Stories Susan M. Ervin-Tripp andAylin Kiintay 2 5 . Discourse, Social Theory, a n d Social Research: T h e Discourse of Welfare Reform Norman Fairclough 2 6 . C o h e r e n c e in Text vs. Coherence in M i n d T. Givon 27. Action and E m b o d i m e n t within Situated H u m a n Interaction Charles Goodwin 2 8 . Constructing Inferences during Narrative Text C o m p r e h e n s i o n Arthur C. Graesser, Murray Singer and Tom Trabasso 29. Space: Irrealis Objects in Technology Policy a n d Their Role in a N e w Political E c o n o m y Philip Graham
1 20 62 78 96 126 159 183 200 229 258 304 338 390
16 The Interpenetration of Communicative Contexts: Examples from Medical Encounters Aar&nV. Cicourel
Editors' Introduction aron Cicourel is Professor of Cognitive Science a n d Sociology at the University of California, San Diego, w h e r e h e is also e m p l o y e d b y J L JL.the Medical School. For over three decades, Cicourel has worked o n the ways in which knowledge is accessed or manipulated in a variety of contexts, including courts (Cicourel 1968), educational settings (Cicourel a n d Kitsuse 1963), a n d medical settings (Cicourel 1980). His concern for the structuring of everyday experience through social interaction and his critique of hidden proced ures within institutional settings puts his work close to two other contemporary influential figures in the analysis of everyday interaction, Erving Goffman a n d Harold Garfinkel. With the former, Cicourel shares the focus o n situated inter pretation - Goffman's famous leading question "What's going o n h e r e ? " can b e repeatedly h e a r d through Cicourel's work - a n d with the latter h e shares the c o n c e r n for the participants' unconscious inferential processes or implicit principles that characterize everyday thinking. Cicourel's unique approach to sociology, which h e termed "cognitive sociology" (Cicourel 1973), is a n original blend of E u r o p e a n scholarship o n the m u n d a n e basis of scientific thinking Husserl's thought as filtered through Schutz's work was particularly influential in Cicourel's early research - a n d contemporary development within cognitive science. His long-term association with the Center for H u m a n Information Processing a n d his regular exchanges over the last several years with such scholars
2
DISCOURSE STUDIES
as David R u m e l h a r d , George Mandler, Michael Cole, and R o y D ' A n d r a d e has further informed his concern for the integration of traditional micro-sociology with current thinking within psychology a n d anthropology o n the origin of contextspecific knowledge a n d m e m o r y . I n this chapter, Cicourel returns to one of the m a i n concerns of his intellectual career, n a m e l y the question of the definition of context in the analysis of verbal interaction within institutional settings. W h e r e a s formal g r a m m a r i a n s a n d structuralist anthropologists are interested in the discovery of underlying, abstract, often unconscious structures a n d processes that give m e a n i n g to linguistic or ritual forms, Cicourel is interested in the h i d d e n aspects of reasoning a n d infor mation processing that characterize researchers as well as m e m b e r s of a given culture. " H o w d o y o u k n o w that?" is a question that Cicourel often asks his col leagues w h e n they present the results of their research projects. T h e replies (or attempts to reply) h e receives in these exchanges constitute for h i m the starting point for methodological a n d theoretical reflections o n h o w knowledge is stored a n d retrieved a n d issues created or ignored out of a myriad of details which are often used but not talked about b y analysts. Cicourel's ideal model of the relation ship between language a n d context is one in which the participants, before the analysts, h a v e simultaneous access to several levels of analysis, including the local organization of turn-taking a n d their internal syntactic structure, lexical choices and semantic networks, as well as the broader context of the institutional settings where the interaction takes place and within which talk acquires meaning for the participants qua social actors (see Corsaro 1985). H e distinguishes between two senses of context, which h e names "narrow" a n d "broad," and proposes their integration for the analysis of everyday talk. Methodology is one of Cicourel's m a i n concerns. I n particular, h e stresses the importance of ethnographic fieldwork a n d the ways in which it can help situate a stretch of talk within its larger institutional context. I n this chapter, his methodological concerns are integrated with a detailed analysis of particular verbal exchanges within a medical facility. By using a technique similar to w h a t film critics call "slow disclosure" (Sharff 1982), Cicourel slowly uncovers further layers of information about the par ticipants, their institutional roles, their previous exchanges, a n d their shared knowledge. I n so doing, h e is able to directly address the issue of h o w m u c h contextual information should b e m a d e available in the study of spontaneous everyday interaction. I n Cicourel's chapter, we are shown that even the use of initials in a transcript hides u n s p o k e n principles and selective strategies. Until we are told that I D A a n d M R stand for Infectious Disease Attending a n d Medical Resident respectively, we could take t h e m to b e conventional m n e m o n i c devices void of social meaning. But labelling is important for participants (when they talk about each other) a n d researchers alike. W h a t does it m e a n to give a transcript with n a m e s of speakers, with their occupational roles, or with initials? W h a t do we evoke w h e n we say that a certain conversation is between a doctor a n d a patient, or b e t w e e n A a n d B? O n the other h a n d , h o w m u c h information should b e offered? Shouldn't we avoid the possibility of an infinite regress, whereby the
CICOUREL
EXAMPLES FROM MEDICAL ENCOUNTERS
3
whereby only the information that the participants themselves make available in the talk should be evoked b y the researcher. However, Cicourel points out some possible limitations of this program when applied without a critical attitude. The omission of apparent extratextual information can be problematic to the extent to which it obscures information that was at some point relevant to the researcher during the collection of analysis of the material under discussion. The ideal researcher in Cicourel's model of discourse analysis is thus one who does not hide his or her sources of information and research choices but makes them into a c o m m o n resource to be shared with the readers in an attempt to unveil the hidden processes of the selection of information which guides participants and analysts alike in the course of their daily lives.
References Cicourel, Aaron V. 1968. The Social Organization of JuvenileJustice. New York: Wiley. Cicourel, Aaron V. 1973. Cognitive Sociology. Harmondsworth: Penguin. Cicourel, Aaron. 1980. Three Models of Discourse Analysis: The Role of Social Structure: Discourse Processes 3: 101-32. Cicourel, Aaron V., and J. I. Kitsuse. 1963. The Educational Decision Makers. New York: Bobbs Merrill. Corsaro, William A. 1985. Sociological Approaches to Discourse Analysis, in Handbook of Discourse Analysis. Vol. 1: Disciplines of Discourse, ed. Teun A. Van Dijk, pp. 167-92. New York: Academic Press. Sharif, Stefan. 1982. The Elements of Cinema: Toward a Theory of Cinesthetic Impact. New York: Columbia University Press.
T h e i n t e r p e n e t r a t i o n of c o m m u n i c a t i v e c o n t e x t s : e x a m p l e s from m e d i c a l e n c o u n t e r s 1 Introduction A researcher's decision to tape-record conversation or discourse creates a context ual frame that limits what is to b e identified as relevant data, their organization, and the kinds of analysis and inferences to which these data will be subjected. Research on conversation and discourse varies considerably in the extent to which an investigator will describe the circumstances of taping a conversation and her or his involvement with those taped. Similar conditions exist in other types of research on social interaction. For example, in field research, a participant observer must be sensitive to the fact that elicitation procedures can force inform ants to become aware of issues or conditions about which they normally are not aware or may be aware of only in the local context of everyday settings. Silverstein (1981) notes that much if not most of what is of interest to researchers is beyond the informant's ability to articulate meanings in ways we would like ~~
1 TU.. -
.„!
i.:
1-
1 f — ™ — t\,„t
—*-
4
DISCOURSE STUDIES
laboratory experiments of social interaction w h e r e status a n d role relations are foci of attention. T h e investigator's a n d subjects' conceptions of status a n d role relations are contingent on implicit a n d formally defined notions of "context" created for the experimental occasion. We seldom k n o w the extent to which simulated social relations reflect or are contrary to the kinds of experiences sub jects h a v e in their everyday encounters. While m a n y researchers will agree that talk a n d some notion of "context" shape each other as part of a n emergent process that changes through time a n d space, not all students of language use will concede that ethnographic material, participant attributes, and patterns of social organization that are constitutive of talk n e e d to b e included in studies of the structure of conversation or dis course. For example, conversation analysts (Sacks, Schegloff, andjefferson 1974; Schegloff, in press) prefer to focus o n such things as sequential organization, and h o w that organization creates local contextual elements in conversation. T h e researcher's use of social categories in any analysis must b e linked to par ticipants' recognition of the categories as integral to the events being studied. This research attitude can result in a v e r y formal type of analysis. Linguistic approaches to discourse analysis (Brown a n d Yule 1983) also minimize the role of ethnographic and organizational conditions that can also result in a formal analysis of utterance sequences. W h a t remains unclear is the extent to which the decision to tape or use particular materials includes or excludes explicit a n d tacit k n o w l e d g e a b o u t r e p o r t e d a n d u n r e p o r t e d e t h n o g r a p h i c conditions, participant attributes, a n d patterns of social organization that can selectively shape subsequent analysis. T h e researcher can exercise considerable discretion in what the reader will b e shown or told about "context." T h e content of conversation or discourse material can b e m a d e to appear rather transparent w h e n we use brief, formal or informal exchanges among people we do not know, w h o interact in settings described in a limited way, or w h e n w e use casual, m u n d a n e everyday conversations a m o n g friends during informal exchanges. W h e n the research analyst is working in h e r or his own society, a n d the reader is expected to b e from the same society, it is especially convenient to use brief, formal or informal m u n d a n e conversations. T h e investigator's ability to c o m p r e h e n d these exchanges is assumed to b e self-evident and is seldom if ever an aspect of the analysis. But if a fuller analysis of participants' conversation a n d ethnographic understandings about activities, objects, a n d ideas is desired, and that understanding presupposes prior social experience, a n d / o r technical, scientific, or professional training, then other strategies besides a completely local analysis must b e employed. Verbal interaction is related to the task at h a n d . Language and other social practices are interdependent. Knowing something about the ethnographic setting, the perception of a n d characteristics attributed to others, a n d b r o a d e r a n d local social organizational conditions b e c o m e s imperative for an understanding of linguistic a n d non-linguistic aspects of communicative events. In the present chapter, I discuss several settings within a larger organizational
CICOUREL
5
EXAMPLES FROM MEDICAL ENCOUNTERS
"context" includes a n institutionalized framing of activities. In traditional social science theorizing, it is customary to speak of group-derived prescriptive n o r m s that pressure a n d / o r channel p e o p l e with designated tides, p r e s u m e d com petencies, duties or responsibilities into certain physical spaces at certain times in order to engage in a finite n u m b e r of specifiable activities. Within this insti tutionalized context or framing of activities, emergent processes of talk a p p e a r that create a m o r e n a r r o w view of "context" in the sense of locally organized and negotiated interaction. T h e extent to which the investigator knows about the institutionalized a n d everyday events that participants a n d analysts categorize b y the use of specific terms varies. For example, the analyst of particular settings m a y or m a y not b e present w h e n speech events are recorded. Within this m o r e n a r r o w sense of "context," therefore, the analysis of a tape a n d transcript m a y or m a y not begin with the investigator's direct experience with the ethnographic setting in which the speech events w e r e recorded. W h e n the investigator knows little or nothing about the speech event and the event is rather short-lived a n d apparently n o t viewed as part of or influenced b y existing institutionalized or bureaucratic activities, what gets identified as context can b e m o r e easily located in the talk said to b e attended b y the par ticipants. T h e investigator can begin h e r or his analysis b y examining the way the conversation is initiated, the w a y intonation and stress are used, the content of lexical items a n d phrases, the occurrence of pauses or hesitations, turn-taking moves, a n d the w a y topics are introduced, sustained, and altered or changed. T h e focus of attention can b e c o m e the sequential organization of conversation material that presupposes that participants, investigator, a n d reader all share a c o m m o n knowledge base that is largely tacit or unstated b u t which b e c o m e s partially articulated b y the w a y the investigator begins to m a k e claims about the categories to which the participants are said to b e attending. T h e investigator and reader presumably can each examine the same data a n d m a k e claims a n d counter-claims about the clarity a n d substance of the analysis. T h e methodological strategy of using local talk as the source of information in the narrow sense of context can b e self-serving b y the w a y the researcher n o t only ignores prior a n d current organizational or institutional experiences of participants, b u t b y the kinds of data that are presented for analysis. For m a n y students of language a n d social interaction, therefore, the notion of context n e e d not include references to the participants' and researcher's personal, kin, and org anizational relationships and other aspects of complex or institutionalized settings. Casual, fleeting speech events, however, are often constrained and guided b y normative institutionalized features that we associate with encounters in public places described in general b u t vivid terms b y Goffrnan (1959,1963,1971) a m o n g others. These brief exchanges can also carry considerable cultural and inter personal "baggage" for participants because of long-term social relationships u n k n o w n to or u n a t t e n d e d b y the investigator. T h e investigator w h o examines a tUAmfn-n
V»n
J
1
t„
It-
1
~C
J.
i
DISCOURSE STUDIES
(i
T h e present chapter is identified explicitiy with the view that b o t h a b r o a d a n d local sense of context are n e e d e d for the study of language use. T h e focus o n discourse in the pages that follow should n o t b e construed as being relevant only to conversational material, but also applies to the study of single utterances. I examine a complex environmental setting in order to underscore the import ance of context at different levels of analysis. T h e following pages begin with a conversation in which three people emerge as speakers. Subsequent sections provide the reader with a brief overview of the organizational settings a n d the ethnographic circumstances in which the data were obtained. After investigating additional speech events, I close the chapter b y illustrating h o w b o t h immediate a n d other aspects of context must b e taken into account if we are to understand language a n d social interaction in everyday life. T h e focus of this chapter, there fore, differs from b u t also builds o n conversation a n d discourse theorists con cerned primarily with the structural features of conversation such as turn-taking, side sequences, topicalization, coherence, a n d related notions. It is important to locate the analysis of language and social interaction in a wide variety of social activities that are implicitly a n d explicitly k n o w n to the participants a n d investigator.
2 A Routine Conversation T h e speech event presented in Example (1) can b e viewed as a routine conver sation between three unidentified people w h o will b e identified below. Example 1:A conversation between three physicians 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
PA IDA PA IDA PA IDA PA IDA MR IDA MR:
IDA MR IDA MR
(?) (low voice level) Is this the same one;(we?) (ya?) did yesterday? No. This is the eye lady. (?)
Cellulitis Oh. With group A strep..in shock In shock. (Slight rise in voice level) How about that I[t?] was gonna be more interesting |if she didn't |I'm(?) have bacteremia but (laughing and voice level increasing) now she's had (bacteremia so |There's a little, there's little (voice level increases) problem with that that I'll, will go into more as far Yeah. how much shock she really |was in, |was in, right compared to what [abrupt shift] she's a liver lady. u—'t i n ™ w h a t hor hlnnrl nressure sits
CICOUREL
23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41
EXAMPLES FROM MEDICAL ENCOUNTERS
PA MR PA MR PA MR
PA MR PA MR IDA
7
So she didn't have peripheral, evidence of shock really? |Just a low blood pressure. |No, she wasn't, she wasn't ever clamp, you know Uhhuh clamped down or flushed or anything OK I can..and, and she doesn't [abrupt shift] one thing that argues against a lot of neuropathy, you know, from diabetes, for one, she's only had it for three years, but two, you know, her neuro exam an' her an' her peripheral vascular exams is really normal, is normal sensory, good pulses distally, and stuff, and I just have a hard time, Yeah there'd be a lot of sy-sympathetic, you know Right phone calls. Have they got sinus films on yer yet?
Each dot between words = one second. The solid vertical line indicates points of overlap in the exchange. Telling the reader that three physicians are involved is an initial attempt at penetrating what is occurring in this encounter. Lines 1-12 strongly suggest that the participants are engaged in a discussion regarding the medical condition of some other party. T h e investigator's control over w h e n and what material will b e shown the reader can, therefore, vary considerably a n d can create different interpretive frames. Saying that three physicians (they also could b e other health care providers) are talking implies that they are p r o b a b l y in a medical facility of some kind, b u t there is n o indication of precisely what kind or part of a health care facility they are in. Nor can w e b e sure if the patient is nearby, is a n inpatient or an outpatient. We k n o w nothing about the physicians' gender or personal characteristics nor their specialities. Nor do we know if this information is relevant to the analysis. T h e opening lines (1-8) of E x a m p l e (1) d o not mention the term "physicians" nor is there any indication in lines 1-3 that a "patient" is involved. H a v i n g iden tified the three participants as "physicians," we can also say that they are talking about a "patient" w h o has b e e n d u b b e d the "eye lady," for some unstated reason, or a patient w h o is experiencing difficulty with her eye. Although the content of lines 1-3 of E x a m p l e (1) ("Is this the same one [we?] [ya?] did yesterday? N o . This is the eye lady.") does n o t suggest a health care facility, physicians, and a patient with eye problems, the term "cellulitis" in line 5 might b e recognizable enough to send us to a medical dictionary as we begin our analysis. Alternatively, we might seek help from informants or e x a m i n e further sequences of conversation in the h o p e that other lexical items, phrases,
8
DISCOURSE STUDIES
b y I D A in line 7 ("group A strep..in shock"), which could refer to a n infection. T h e c o m m e n t or observation b y PA in line 8 ("In shock. H o w about that") could b e interpreted as expressing interest in the categories supplied b y I D A a n d p e r h a p s mild surprise. T h e subsequent remarks b y I D A in lines 9 , 1 1 and 12 of E x a m p l e (1) ("I[t?] was g o n n a be m o r e interesting if she didn't h a v e bacteremia but n o w she's h a d bacteremia so") suggest that PA a n d I D A are b o t h familiar with the terms "cellulitis," "group A strep," "in shock," a n d "bacteremia." T h e significance the analyst attributes to particular utterances or to their sequential order, and claims about the participants' perception and understanding of a speech event, are often the result of examining subsequent utterances (Sacks, Schegloff, a n d Jefferson 1974). If we do not invoke institutional a n d local sociocultural details with which to identify the participants of conversation, the analysis of m e a n i n g is nearly impossible. I will n o w abruptly shift m y m o d e of analysis b y telling the reader something about the participants in E x a m p l e (1) a n d about h o w the data w e r e obtained. T h e setting is generally k n o w n to the reader and investigator alike because of a commonsense understanding of categories such as "patient," "health care facility," "medical center," "hospital," and "medical school." Stating that the talk is in a medical setting does n o t tell the reader that there is medical talk taking place n o r who is doing the talking a n d with what credentials. But claiming the par ticipants in a given speech event are physicians talking a b o u t a particular (technical sense of) "patient" orients us to categories we associate with particular m e m b e r s of a medical setting a n d a basis for m a k i n g claims about the use of these categories in their talk. T h e t e r m "patient," for example, is used in at least two ways in this chapter. For the reader of this chapter, the term "patient" p r o b a b l y implies a com monsense notion of s o m e o n e w h o is n o t feeling well or is "sick." For the health care expert, however, the t e r m "patient" can activate thoughts about s o m e o n e with neurophysiological, biochemical, a n d psychopathological s y m p t o m s and conditions that are part of a n extensive technical compositional semantics and clinical as well as m u n d a n e experiences. For the most part, the t e r m "patient" will b e used in the second sense just noted. Thus, even m u n d a n e terms such as "eye lady" (line 3 , E x a m p l e 1) a n d "she's a liver lady" (line 19, E x a m p l e 1) imply an activation of technical knowledge a n d experiences that are part of complex socially organized activities.
3 T h e R e s e a r c h S e t t i n g a n d Types of Participants T h e case material I present in this chapter represents a small portion of data from field research in two teaching hospitals. T h e material is p a r t of routine institutionalized bureaucratic activities that are typically associated with health care delivery in Western countries. M y observations a n d tape-recording of nrKvitffts m o s d v involved physician-patient, physician-physician, or physician-
CICOUREL
EXAMPLES FROM MEDICAL ENCOUNTERS
9
T h e medical settings observed are typical of teaching hospitals b u t are n o t necessarily representative of other clinic or hospital settings in which there is an absence of H o u s e Staff (interns, residents, a n d training fellows). T h e inclusion of House Staff adds hierarchical relationships a n d h e n c e additional bureaucratic conditions that differentiate teaching from non-teaching hospitals. Additional details about organizational conditions, the influence of status a n d role relation ships, a n d the e x p e r t i s e or b a c k g r o u n d k n o w l e d g e associated w i t h s u c h designations will b e p r o v i d e d as the chapter unfolds. T h e range of speech events i n these settings can b e diverse and can include a quick exchange of greetings, brief exchanges of gossip or rumors about patients, or staff, a n d discussion about the local baseball or football team's fortunes. T h e r e also can b e brief and lengthy exchanges between physician a n d patient, technical discussions a m o n g physicians a n d between physicians and nurses or technicians, and somber or emotionally charged exchanges between family m e m b e r s a n d health care personnel. T h e exchanges can occur in a variety of settings such as the patient's r o o m , in a clinic examining r o o m , a hall or corridor, an e m p t y elevator, the cafeteria, at the x-ray facility, the nurse's station, or a laboratory. O v e r a period of m o n t h s , I observed a n d then began to tape-record various clinical exchanges. These exchanges included outpatient clinics where patients m a y b e seen initially or as a follow-up from prior visits or inpatient service. I also accompanied the resident or attending physician o n w a r d rounds, or visits to the x-ray facility, to daily laboratory rounds, a n d to weekly grand or teaching rounds where cases of unusual interest are presented before attendings (one or m o r e physicians qualified to supervise residents) from the county and residents from several hospitals. T h e ethnographic activities alluded to a b o v e obviously provided m a n y opportunities to express m y ignorance about medicine a n d to ask questions of different experts. T h e field research was facilitated b y m y official connection with a school of medicine. K n o w i n g some of m y informants independently of the research setting because of committee activities in the School of Medicine m a d e participation in the research setting easier and enabled m e to ask m a n y " d u m b " questions as a non-medical specialist. But familiarity with the settings can also create p r o b l e m s . For example, m y colleagues sometimes forgot that I a m n o t a physician a n d spoke to m e as if I w e r e fully capable of understanding technical matters that were discussed. I often pretended to seem informed in order to not disturb the speech event, but later would h a v e to ask for help. My ethnographic experiences, consultations with people called "physicians," and occasional uses of a medical dictionary enable m e to take a n u m b e r of lib erties in describing the material in E x a m p l e (1). I can tell the reader that there is a female inpatient located in a medical teaching hospital with an eye p r o b l e m as well as additional difficulties associated with the technical terms "group A strep" (line 7), "shock" (line 7) a n d "bacteremia" (line 11). I rely o n the reader's commonsense understanding a n d m y use of expert informants to claim that the nut
cnmanna
10
DISCOURSE STUDIES
Less transparent h e r e is the r e m a r k b y I D A (lines 9-12) that the case would h a v e b e e n m o r e interesting if the patient did n o t have bacteremia.
4 R e e x a m i n i n g t h e C o n v e r s a t i o n i n E x a m p l e (1) T h e exchange in E x a m p l e (1) does not state that the three participants are at tached to a health care facility of some kind n o r is there any mention of a patient. T h e conversation begins in line 1 with the introduction of a topic ("Is this the same one . . . " ) and a reference to some object or s o m e o n e that h a d something d o n e to it the day before. I n Section 2 above, I alluded to the possible content of lines 1-7 of Example (1). A n analyst might choose to ignore further discussion of content at this point and instead could focus o n w h o speaks first and h o w s o m e o n e responds, iden tifying turns, the selection of n e x t speaker, topicalization, side sequences, ad jacency pairs, registers used, topic and comment, deictic, anaphoric, metaphoric, a n d metanymic expressions, to mention some familiar aspects of the study of language in social interaction. T h e opening question b y PA ("Is this the same one [we?] [ya?] did yesterday?") in line 1, E x a m p l e (1), can b e seen as topicalizing an unidentified b u t particular type of object that other participants presumably recognize as fitting into a specific subset of a general category. We can subsequentiy categorize this object as a "patient" in the technical sense b y assuming the participants are health care pro viders of some kind. T h e response b y I D A in line 3 , E x a m p l e (1), rejects a parti cular subset of the unidentified hypothesized category b u t does allude to another subset b y stating "This is the eye lady." I n lines 5, 7, a n d 12 of E x a m p l e (1), I D A seems to clarify the hypothesis about the category "patient" b y describing par ticulars such as "cellulitis," "group A strep," "in shock," and "bacteremia." T h e opening question b y P A in line 1, E x a m p l e (1), might also b e loosely interpreted as one part of a n adjacency pair (e.g. question-answer sequence) in which the second part consists of the " N o " of line 3 b y I D A . Perhaps this type of analysis might b e stretched even m o r e if we ignore the " N o " of line 3 and the " O h " of line 6 and say that the second part of the adjacency pair consists of fragments from lines 3 , 5, 7, and possibly 12 of E x a m p l e (1). To paraphrase the exchange, PA's question in line 1, "Is this the same one (we?) (ya?) did yesterday?" is answered b y I D A in lines 3 , 5, 7, a n d possibly 12 ("No. This is the eye lady," "Cellulitis," "With group A strep...in shock," and " n o w she's h a d bacteremia"). If we go b e y o n d this loose illustration of sequential analysis, we can ask h o w m u c h local context is the analyst required to construct in order to convince herself or himself a n d the reader that the interpretations being claimed are adequate for some sequence of text?
5 W h a t Counts- as R e l e v a n t E t h n o g r a p h i c Details?
CICOUREL
EXAMPLES FROM MEDICAL ENCOUNTERS
11
to I D A w h e n posing the question about the "same o n e " in line 1. T h e speech event can b e further clarified b y noting that PA refers to a Pathology Attendant w h o is the Chief of the Microbiology Laboratory, a n d I D A is the Infectious Dis ease Attending. M R is the infectious disease Medical Resident for the rotation period during which the event occurred. T h e exchange reported in E x a m p l e (1) occurred after M R a n d I D A h a d both interviewed the patient a n d h a d discussed the case. Their participation in the encounter reported in E x a m p l e (1) h a d already b e e n anticipated because of the expected earner exchange. PA is the expert h e r e and methodically examines a series of small slips of p a p e r each m o r n i n g at 11:00 a.m. in which n e w or add itional information about a patient's cultures is noted. I D A is also a n expert about microbiological aspects of each patient, particularly the clinical ramifi cations of the case. By addressing I D A , PA asks the responsible clinical expert about the case in order to contextualize the laboratory findings. T h e r e are other people in the corner of a large r o o m (which is actually one part of the suite in which the microbiology labs are located). T h e r e often is a fourth-year medical student doing a m o n t h ' s rotation in infectious disease cases, a pathology resident who assists PA with laboratory details including the cultures for each patient, slides, and petrie dishes to b e examined that morning, an infectious disease resident from pediatrics, a n d o n e or two teaching fellows. In a teaching hospital, I D A m a y begin with a few comments o n the case b u t normally turns it over to M R for the details. After this, PA first examines the slide or cultures a n d either direcdy describes the m o r p h o l o g y of the organism(s) or invites the medical student or fellow or a resident to tell the group what organism(s) is/are involved. P A then invariably gives the group additional infor mation about the nature of the organism, the kinds of disease a n d / o r symptoms associated with the laboratory findings, and the patient's likely prognosis with different types of treatment. T h e opening line of E x a m p l e (1), therefore, assumes coherence b y reference to a. presumed activity d o n e with thus far unidentified others the day before. As noted above, the unidentified others in the present case are the M R a n d I D A . T h e response b y I D A in line 3 shifts the topic to s o m e o n e called "the eye lady." T h e inaudible c o m m e n t b y PA in line 4 of E x a m p l e (1) could have b e e n a re quest for details about the case. I D A provides one detail ("cellulitis"). F r o m m y watching his facial expression a n d hearing the intonation of his voice ("Oh" in line 6), I infer that the term is of interest to PA. W h e n I D A provides additional information about the patient, PA repeats the last phrase ("in shock") a n d with a slighdy rising voice level seems to b e expressing mild surprise ("How about that"). T h e n I D A , in lines 9, 11, and 12, refers to h o w m u c h m o r e interesting the case would h a v e b e e n if the patient h a d n o t h a d "bacteremia" ("I[t?] was gonna b e m o r e interesting if she didn't h a v e bacteremia but n o w she's h a d bacteremia so"). M R , however, in lines 10 a n d 13-17 ("There's a litde, there's little p r o b l e m with that that I'll, will go into m o r e as far h o w m u c h shock she really was in") seems to b e questioning the patient's "shock" a n d these doubts are related to the issue of bacteremia with or without shock.
12
DISCOURSE STUDIES
I D A and M R w h e n the two discussed the case and k n e w that there were doubts expressed (in two separate medical histories) b y other physicians about the patient's "shock." M R ' s completion of his r e m a r k from line 17 (stated as "com pared to what" in line 19) was followed b y an abrupt shift in the topic. Presumably the patient's "shock" was to b e contrasted with an unstated something else ("compared to what") when M R decided to abrupdy shift the topic to the patient's liver ("she's a liver lady, you know"). So in addition to being "the eye lady" of line 3, the metaphor shifts to that of "liver lady." T h e reader's technical knowledge m a y b e strained b y the phrase "a liver lady." M y participation in the setting tells m e w e h a v e a patient with a current eye p r o b l e m a n d a prior liver p r o b l e m attributed to a history of alcohol abuse. O n the basis of previous conversations with informants, I can report that one clinical sign of shock is very low blood pressure. M R ' s r e m a r k in line 21 ("It m a y n o t b e real high to start with") refers to the time of admission for the patient w h e n h e r blood pressure was observed to b e low b u t w h e r e n o o n e (in cluding the patient) was aware of h e r n o r m a l blood pressure. T h e inference of possible shock, therefore, can b e problematic d e p e n d i n g o n the patient's n o r m a l b l o o d pressure. I D A h a d contributed to M R ' s remarks through their previous conversations the day before, a n d the "Right" uttered twice b y I D A in line 22 could signal agreement with the present observation b y M R a n d / o r IDA's prior contribution. After reading the transcript, however, I D A reported that the intent of saying "Right" twice was to hurry M R into expediting his delivery a n d remarks. I n lines 2 3 - 4 , P A provides some clarification of the patient's alleged shock b y his "declarative-interrogative" that the patient a p p a r e n c y "didn't h a v e periph eral, evidence of shock really? J u s t a low b l o o d pressure." Although appearing to b e a question, PA's remarks are spoken authoritatively through the use of a somewhat "matter of fact" or "self-confident" intonation a n d the mildly assertive nature of the content. T h e r e is also the ethnographic element that h e is the primary expert h e r e a n d can b e indirectiy asserting this expertise b y his remarks in lines 2 3 - 4 . M R ' s rambling remarks in lines 2 5 - 3 6 of Example (1) address several possible topics that presuppose background in clinical medicine. For a n attending, M R ' s remarks can b e c o m e the basis for a n organizational assessment of the resident or novice in infectious diseases. As non-medical specialists, students of language use in social interaction, we can only superficially assess the import of the material in Example (1) unless we seek the help of informants and have access to the kind of ethnographic information provided thus far.
6 Expanding the Ethnographic Context I h a v e presented a partial description of the w a y the selection of different ethno graphic or organizational settings a n d types of speech events or social interaction
CICOUREL
13
EXAMPLES FROM MEDICAL ENCOUNTERS
technical matters that affect h o w patients are admitted into a medical setting, are diagnosed, a n d treated b y different health care delivery personnel. T h e present case involves a 48-year-old white female whose chief complaint was recorded in h e r admission history a n d physical examination as "swelling of the left eye." Additional details can b e found in Cicourel (1987). Of interest here is h e r p r i m a r y diagnosis of "periorbital cellulitis" or what was described as a "right-upper-lid abscess with periorbital cellulitis." M y informants h e l p e d m e translate this last phrase into a n infected, swollen-shut left eye, that seemed highly inflamed to m e , with purplish coloring of the skin a r o u n d the eye that was apparently caused b y b l o o d seeping into the skin. T h e patient's left side of her face was swollen as well as h e r eye a n d ear. T h e r e was pus oozing out of the area of the eyelids. I accompanied M R w h e n h e first went to the medical w a r d to see the patient. Before interviewing the patient, M R went to the nurse's station to review the patient's chart. A s is often the case with infectious disease patients, an attending physician other than I D A usually will see a patient admitted to the medical service. I n the present case, two attendings h a d interviewed the patient a n d h a d ordered laboratory tests before asking for a n infectious disease consultation. T h e results of these tests were reported in the chart. M R ' s infectious-diseaseoriented interview of the patient, therefore, was in part influenced b y his having had access to the patient's medical chart. M R ' s questions, therefore, reflect aspects of these prior experiences. In addition, this interview is influenced b y his prior medical school train ing. T h e medical student's knowledge of infectious diseases can b e traced to one or two courses (depending o n the quarter of semester basis of the medical school) in microbiology. I attended some of the lectures in microbiology a n d also interviewed the director of the course, w h o is also PA in the conversation of Example (1). With the help of informants, it is possible to trace aspects of the basic science concepts a n d clinical medicine experience medical students are exposed to before they enter an intern or residency program. Knowing something about the physician's prior experiences in medical school and as a H o u s e Officer can provide the analyst a n d reader with possible aspects of the background knowledge tacitiy or explicitly assumed or employed during a speech event. I do not have time to present all of the content of m y interview with PA in his role as the microbiology course director, but instead will provide the reader with a few highlights that can b e linked to the conversation in E x a m p l e (1). Trying to trace the kinds of basic a n d clinical science concepts that medi cal students and H o u s e Staff are exposed to before they attain the status of "attendings" is n o t m e a n t to imply that such codifiable knowledge is the primary source of background necessary for the diagnostic reasoning a n d treatment dir ected at patients. I n m y work o n medical communication a n d diagnostic reason ing, I have stressed the importance of tacit clinical experience a n d the everyday understanding of m u n d a n e social interaction as aspects of a w o r l d p r e s u m e d to be known in c o m m o n but n o t necessarily e x a m i n e d reflexively b y participants : „ j-l
„
3; .1
. „il
.
.
_ r i : f . Ir~i:
._l m o o
inn/?
irvnrrt T
n >1
DISCOURSE STUDIES
14
codified knowledge (textbooks, detailed lectures, laboratory practices, apprentice like experiences) into our understanding of bureaucratic and non-bureaucratic language use and social interaction. Inasmuch as sociolinguistics tends to address various kinds of everyday events in which the analyst can "pass" as an expert because of her or his background in language use and social interaction, the analysis of content is often viewed as fairly straightforward. By calling the reader's attention to the role of multiple ethnographic and/or organizational settings and informants, I underscore the complexity of the term "context" or "local" and more abstract senses of culture or social organization for understanding language use and social interaction. According to the microbiology course director during an interview obtained early in m y research and noted above, students should remember the following: Example 2 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
"...what are the important bacteria, and, in terms of human disease, and what are their unique microbiologic characteristics...The streptococci can be differentiated into the hemolytic ones and the non-hemolytic ones and green ones. Those that are alpha-hemolytic make green colonies. And among the beta-hemolytic ones we have the group A strep, which is really the only accepted cause of bacterial pharyngitis...The streptococci are in chains, the pseudomonas and the enterobacteriaceae are gram negative rods, and they look a lot alike on the gram stain... And if you look at the streptococci for a minute, the group A strep, is able to cause sore throats primarily because of the M protein... ...what are the unique microbiologic characteristics that i permit me to recognize this organism, let's say right out of the blood culture bottle like you've seen us do up there?....Did this patient have cellulitis, or lymphangitis? Yes. Well, then it's probably a group A strep...
The technical remarks b y the pathologist made during the interview conducted before the completely independent laboratory conversation of Example (1) refer to the clinical consequences medical students are expected to remember from courses in microbiology. The references to "streptococci" (line 3), "beta-hemolytic ones" (fine 6), and "group A strep" (line 7) in Example (2) are direcdy related to the "sore throats" of line 12 and are a significant aspect of the diagnostic reasoning employed in the present case. Notice the reference to "cellulitis" and its asso ciation with "group A strep" in lines 17-18. The pathologist's remarks provide clear evidence of prior training on the part of physicians that plays a role in jobrelated speech events. Such training may be important not only to participants, but also to the analyst's understanding of medical communication and semantic aspects of diagnostic reasoning in the case under review and in medical settines generally.
C1COUREL
EXAMPLES FROM MEDICAL ENCOUNTERS
IS
of a medical student's laboratory experiences with the microorganisms n o t e d in Example (2). Without this background knowledge based o n studying textbooks, listening to lectures, a n d working in a laboratory setting, it would b e rather dif ficult for a participant to display h e r or his group membership in the conversation of Example (1). I n the case of a medical resident, there will h a v e b e e n prior ex periences as an intern in which aspects of the material in E x a m p l e (1) will h a v e b e e n discussed. Another source of background knowledge that can influence speech events in medical settings is the w a y that novice physicians are exposed to a n initial lecture (also given to t h e m as a written handout) w h e n they begin their training. T h e following lines are from lecture notes developed b y an infectious disease attending physician. I quote selectively from the written h a n d o u t given to the House Staff. Example 3 1 "I. Septicemia - microbial agents in the bloodstream... 2 B. One is led to the diagnosis of bloodstream 3 infection by a sudden change in clinical 4 state... 5 All sorts of organisms can cause hypotension, but, 6 in the absence of hypoxia (pneumonia) or heart 7 failure, gram-negative bacteremia with endotoxemia 8 is by far the most common cause of sustained 9 hypotension... 10 D. Other causes of infectious shock 11 3. Streptococcal shock (group A strep), often in 12 the absence of bacteremia, associated with soft 13 tissue infections (usually severe)..." Of interest are references to "microbial agents in the bloodstream" (line 1), "the diagnosis of bloodstream infection b y a s u d d e n c h a n g e in clinical state" (lines 2-4), a n d "organisms [that] can cause hypotension" such as "gram-negative bacteremia with endotoxemia" (lines 5-7). A sudden change in clinical state can b e a drop in b l o o d pressure (hypotension), which i n turn can b e linked to the release of bacteria or endotoxins in the bloodstream that can lead to shock. T h e most succinct r e m a r k can b e found in lines 10-13 of E x a m p l e (3) because it comes fairly close to the case described in E x a m p l e (1). T h e patient was found to have group A strep a n d bacteremia as noted b y I D A in Example (1). T h e case revealed bacteremia but M R claimed the patient m a y not have b e e n in shock. T h e presence of shock was never confirmed. T h e remarks contained in E x a m p l e (3) and given to H o u s e Staff at the two hospitals in which I conducted m y research direcdy associate this background material to specific aspects of the exchange in Example (1). I will close this section b y quoting a brief excerpt from M R ' s original interview with the natient. Thp intprvipw nrrnrrprl nrinr tn fhp snpprh pvpnt rpnnri-prl in
DISCOURSE STUDIES
16
Example 4 1 2 3 4 5 6 7 8
I wanna take (?) your blood. (?) blood pressure (?) ( ) (mumbling) A lot of people get, bacteria in their blood, and ya get uh shock. [Low monotone intonation] When your blood pressure goes down ' because it means (?) the bacteria releases certain toxins, depending on (?) area (?) [raised intonation] Has your, blood pressure always ran kind of on the low side, ma'am?
•
Before M R a n d I entered the w a r d to see the patient, M R h a d reviewed the patient's chart and was aware of the two prior medical histories therein a n d hence was also aware of the suspected diagnosis of bacteremic shock. M R was also aware of m a n y other medical facts a n d hypotheses about the patient's condition. M R ' s explanation (only partiy shown here) to the patient about h e r medical problems (Unking a drop in b l o o d pressure to a prior release of "certain toxins'') can b e linked direcdy to the prior medical histories that existed in the patient's chart, the infectious disease attending's written h a n d o u t , a n d the pathologist's description of what medical students should r e m e m b e r of their course work in microbiology. I n turn, all of these matters, plus M R ' s interview with the patient, h a d an effect o n the remarks b y PA, I D A , a n d M R in E x a m p l e (1).
7 Concluding Remarks T h e chapter has presented a few of the large n u m b e r of interpenetrating medical events that form the local context for language use a n d social interaction in any given encounter. I have sought to discuss the topic of local context in terms of those cultural a n d organizational constraints, normative expectations, a n d im mediate conditions that surround local speech events as they unfold. T h e notion of interpenetrating medical communicative contexts seeks to place the local mutual shaping of talk and context into a framework that incorpor ates structural a n d processual aspects of social organization a n d reasoning during social interaction. T h e medical setting I chose as the point of departure for the present chapter was the source of a conversation b e t w e e n three physicians. T h e material presented in E x a m p l e (1) can b e subjected to different types of inter pretation, contingent o n the analyst's orientation to language use a n d social interaction a n d the extent to which the notions of background a n d context are employed. T h e local interaction in E x a m p l e (1) is intended to call the reader's attention to outside organizational experiences or institutionally guided a n d constrained social interaction a n d / o r speech events that are direcdy relevant to the way local exchanges are likely to occur. For example, the interaction described in E x a m p l e (1) can b e linked to the microbiological aspects of medical education (Example 2) a n d the lecture o n Septicemia (Example 3) given to the H o u s e Staff at the beerinnine of the medical academic year. All social interaction a n d / o r
CICOUREL
17
EXAMPLES FROM MEDICAL ENCOUNTERS
T h e analyst's decision to describe the materials in Example (1) as an exchange between three health care delivery workers or professionals creates a cultural context for the analyst a n d reader a n d simultaneously calls attention to the use of social categories that hopefully can b e attributed to participants' recognition of their relevance for the speech event being examined. T h e analyst's decision to focus o n particular sociolinguistic notions (topicalization, turn-taking, speech act categories, coherence, types of debds, frozen expressions, preferred sequences, etc.) rather than the interrelationship b e t w e e n discourse a n d culture or the interrelationship b e t w e e n discourse a n d linguistic theory can trigger different expectations o n the part of the reader. Telling the reader that the three participants in E x a m p l e (1) are physicians, that the exchange took place in a hospital microbiology laboratory while they were discussing a patient with suspected bacteremic shock, and preparatory to examining the patient's b l o o d cultures, creates a cultural context within which the analyst a n d reader can understand what transpires. O u r understanding deepens when the physicians are identified as a pathology attending, an infectious disease attending, a n d a n infectious disease medical resident. Not only are the expressive style a n d the content observed contingent o n organizationally con strained status a n d role relationships that can influence w h o speaks first, whose views will b e influential or will prevail vis-a-vis action to b e taken, but the analyst's and reader's ability to c o m p r e h e n d the significance of b r o a d e r a n d local sociocultural issues m a y require extensive consultation with medical experts. T h e present case actually b e g a n with M R a n d the researcher going to the nurse's station of a university medical center medicine w a r d to examine the medical chart of a patient suspected of having bacteremic shock. M R was per forming an organizational obligation of initiating a consultation requested of the Division of Infectious Diseases. T h e m a t e r i a l from t h e m i c r o b i o l o g y c o u r s e p r e s e n t e d a b o v e reflects information medical students are expected to know, a n d the excerpts from the lecture given to n e w H o u s e Staff reflect the kinds of training experiences a novice like an intern or resident physician is likely to possess w h e n reading a patient's chart and t h e n interviewing the patient. By the time the case reached the microbiology laboratory rounds depicted in Example (1), the patient h a d also b e e n interviewed b y I D A : M R a n d I D A had discussed the case in some detail; I D A h a d interviewed the patient; a n d IDA had consulted the patient's chart, added to it, and h a d consulted two standard textbooks on microbiology. T h e analysis of material in E x a m p l e (1), the excerpts o n the microbiology course (Example 2), the infectious disease attending's lecture notes for n e w H o u s e Staff (Example 3), a n d the fragment from M R ' s interview with the patient (Example 4) represent a small part of a complex set of interpenetrating contexts. In the present chapter, I h a v e d o c u m e n t e d only minimally the way informants helped m e understand organizational conditions, local speech events and social interaction, the m e a n i n g of c o m p l e x concepts, and procedures from basic a n d r»ltntr«ol mnA + fn]
n/«*an.r
~ —
—T1
1
1
1
DISCOURSE STUDIES
18
I have characterized as unavoidable aspects of organizational and local constraints a n d processes that are integral to rethinking "context." A nagging issue that u n d o u b t e d l y remains for m a n y readers is the familiar o n e that a n infinite regress c a n occur w h e r e b y t h e observer presumably must describe "everything" about a context. Such a d e m a n d is of course impossible to satisfy because n o o n e could claim to h a v e specified all of the local a n d larger sociocultural aspects of a context. Observers or analysts, like participants of speech events, must continually face practical circumstances that are a n integral part of all research or everyday living. A s researchers, w e obviously privilege some aspects of a context while minimizing or ignoring other conditions. T h e observer is obligated to justify w h a t h a s b e e n included a n d w h a t h a s b e e n e x c l u d e d a c c o r d i n g to stated theoretical goals, m e t h o d o l o g i c a l strategies employed, a n d the consistency a n d convincingness of a n argument or analysis. M e a n i n g a n d understanding in everyday life are contingent o n cognitive a n d linguistic activities, i.e., a knowledge of h o w to use language structures that w e call m e t o n y m s , metaphors, a n d other imaginative constructions associated with particular lexical items, phrases or utterances, a n d personal experiences. T h e r e fore, some specification of t h e environmental conditions i n which these lan guage practices emerge, are e m b e d d e d , a n d evolve should guide the researcher's depiction of context. I h a v e tried to indicate some of the conditions that c a n b e identified i n a medical setting to illustrate t h e w a y different senses of t h e term "context" c a n b e pursued conceptually a n d empirically.
Acknowledgments Parts of this chapter were initially presented at the Invited Session "Rethinking 'Context': Language as an Interactive Phenomenon, Part I" (Society for Linguistic Anthropology and General Anthropology Division), American Anthropological Association annual meeting, Philadelphia, PA, December 1986. A version of the paper appeared in the Social Psychology Quarterly. The present version has been expanded slightly and further revised. I am grateful to Charles Davis and Elizabeth Ziegler for their invaluable support and advice, and to Sandro Duranti, Charles Goodwin, Douglas Maynard, Hugh Menan, James Wertsch, and two anonymous referees for their helpful substantive suggestions on earlier drafts. Douglas Maynard also provided very useful editorial suggestions.
References Brown, Gillian, and George Yule. 1983. Discourse Analysis. Cambridge: 'Cambridge University Press. Cicourel, Aaron. 1982. Language and Belief in a Medical Setting, in Contemporary Per ceptions of Language: Interdisciplinary
Dimensions, ed. Heidie Burnes, pp. 48-78.
Georgetown: Georgetown University Press. . 1986. The Reproduction of Objective Knowledge: Common Sense Reasoning in i. i r , . . ! . : — n
CICOUREL
EXAMPLES FROM MEDICAL ENCOUNTERS
19
Cicourel, Aaron. 1987. Cognitive and Organizational Aspects of Medical Diagnostic Reasoning. Discourse Processes 10: 346-67. Goffman, Erving. 1959. The Presentation of Self in Everyday Life. Garden City, New York: Doubleday Anchor Books. . 1963. Behavior in Public Places. New York: The Free Press. . 1971. Relations in Public. New York: Basic Books. Sacks, Harvey, Emanuel A. Schegloff, and Gail Jefferson. 1974. A Simplest Systematics for the Organization of Turn-Taking for Conversation. Language 50: 696-735. Schegloff, Emanuel A. In press. Between Macro and Micro: Contexts and Other Con nections, in The Micro-Macro Link, ed. J. Alexander, B. Giesen, R. Munch, and N. Smelser. Berkeley and Los Angeles: University of California Press. Silverstein, Michael. 1981. The Limits of Awareness, in Sociolinguistic Working Paper M, pp. 1-30. Austin, Texas.
i
17 Answers and Evasions Steven E. dayman
W
h e n Albert Gore was Bill Clinton's vice presidential running m a t e in 1992, Gore's position o n abortion b e c a m e the focus of controversy. As a legislator, Gore h a d opposed federal funding for most abortions, b u t n o w h e was expressing support for it as part of Clinton's health care reform plan. I n a n aggressive interview conducted b y S a m Donaldson, Gore received a barrage of tough questions exposing this apparent contradiction. H e was m o mentarily rescued b y a commercial break, at which point h e was urged b y his m e d i a advisor to sidestep questions of this sort: " D o n ' t b e afraid to turn their questions. If they ask y o u about [abortion], just say . . . 'I want to talk today about the n e w direction that Governor Clinton and I w a n t to take the country."" There is a widespread perception that politicians are frequently evasive u n d e r questioning from m e m b e r s of the news media, a n d this perception is n o t without merit (Bavelas et al. 1988; Bull 1994,1998; Bull & M a y e r 1993; C l a y m a n 1993; Greatbatch 1986b; Harris 1991). T h e impetus to resist a line of questioning is understandable, given the adversarial character of contemporary journalism. In news interviews - as well as press conferences a n d other forms of interrogation journalists are drawn to questions that are unflattering, incriminating, or other wise hostile in character. If answered straightforwardly, these can inflict damage on a politician's policy objectives, career prospects, and personal reputation. W h e n 1996 presidential candidate B o b Dole was questioned about the addictiveness of tobacco, his equivocal response was roundly attacked b y the Clinton campaign and received weeks of unfavorable m e d i a coverage. O n the other side of the Atlantic, a question about the Labour Party's opposition to nuclear weapons p r o m p t e d party leader Neil Kinnock to give a cavalier response that was subse quently exploited b y the Tories and h e l p e d ensure Thatcher's reelection in 1987. A n d it is n o t only politicians w h o face such dangers: n o interviewee is i m m u n e . I n response to a question about the absence of African Americans at the managerial
CLAYMAN
21
ANSWERS AND EVASIONS
level in professional baseball, Al C a m p a n i s , vice president of the Los Angeles Dodgers, m a d e s o m e racially insensitive remarks which caused such a n u p r o a r that h e was fired the next day. To avoid consequences like these, interviewees m a y b e motivated to b e less than forthcoming in the face of hostile questioning. However, evasiveness has a downside. Answering questions is treated as a basic moral obligation, n o t only for public figures in journalistic interviews b u t also for interactional participants m o r e generally (Schegloff 1968, Heritage 1984:245-53, R a y m o n d 1998). But while interactants generally expect each other to b e properly responsive to questions, the responsive conduct of politicians is perhaps m o r e closely scrutinized, so that attempts to resist, sidestep, or evade can b e cosdy in a variety of ways. Interviewers themselves monitor for evasiveness, and such maneuvers can b e met with probing follow-up questions and negative sanctions (Greatbatch 1986a). W h e n Clinton administration Treasury secretary R o b e r t Rubin sidestepped a difficult question about gasoline taxes and instead offered a rosy analysis of the overall economy, h e was r e p r i m a n d e d for his slipperiness: " W h e n we ask people like you a simple question . . . y o u do the light fantastic instead of giving us a direct answer." Sanctions like these, in addition to u p p i n g the ante for a genuine answer, also constitute the previous response as i n d e e d evasive, thereby expos ing and highlighting that quality for the broadcast audience. Interviewers have not always b e e n quite so aggressive, but the impetus to monitor for a n d r e s p o n d to evasiveness is n o w rooted in the adversarial culture of contemporary American journalism (Clayman and Heritage in press a).Journalist-interviewers gain profes sional status on the basis of aggressive questioning, a n d they pride themselves on the skill with which they can pursue a n d pin d o w n recalcitrant interviewees. Monitoring b y journalists can also extend b e y o n d the occasion of the inter view itself. Subsequent news stories about interviews a n d press conferences often contain excerpts (in the form of quotations a n d sound bites) that show public figures to b e refusing to answer questions, or initially resisting questions, or ans wering only after being repeatedly pressed to do so (Clayman 1990). W h e n Texas governor George W. Bush, early in his 2000 campaign for the presidency, gave less than forthcoming answers to questions about his cocaine use, there followed a virtual feeding frenzy of stories dissecting the ramifications of his evasiveness. A similar fate befell President Bill Clinton following his artful responses to questions about M o n i c a Lewinsky. A n act of evasion can thus b e c o m e a news worthy event in its own right, a n d a persistent source of unfavorable publicity. Even w h e n journalists allow such maneuvers to pass without comment, there may still b e consequences for the broadcast audience a n d in public opinion. Insofar as a resistant response is noticed b y audience m e m b e r s , they will seek to account for the b r e a c h of conduct via inferences that are apt to b e unflattering or mcriminating. Audience m e m b e r s m a y infer that the interviewee has some ulterior motive for avoiding the question, or that h e or she has something to hide. American citizens m a y h a v e a constitutionally protected right to remain silent in the face of police questioning, so that silence cannot b e treated as incriminating in rmirrsnf l a w hur rmhliV fimiroc liaira n«-> cn/4i r>«-/-i(-a/.fJ™i in t-U^ ^.mi-i f „,.U1:„ 2
n
22
DISCOURSE STUDIES
I n a nutshell, then, interviewees face a dilemma. T h e r e are various pressures, b o t h from journalists and from t h e audience, from within t h e interview a n d in subsequent m e d i a coverage, to "just answer the question." But when the question is adversarial, there a r e cross-cutting pressures to take precisely the opposite course of action. For those contemplating a resistant response, the p r o b l e m is h o w to r e a p the benefits while minimizing the various costs associated with that risky course of action. T h e m a n a g e m e n t of interactional resistance, and the construction of answers m o r e generally, is a neglected topic in studies of broadcast news interviews. Although there is lively interest in evasiveness as a p h e n o m e n o n , prior research focuses mainly o n distributional matters - o n which politicians tend to sidestep questions disproportionately, a n d o n the circumstances u n d e r which they do so (Bavelas et al. 1988, Harris 1991, Bull & Mayer 1993, Bull et al. 1996; Bull 1998). Less attention has b e e n paid to m o r e fundamental issues, such as what distinguishes an "evasion" from a genuine "answer," t h e elementary forms that resistance to a question can take, and h o w acts of resistance are m a n a g e d through specific discursive practices (Greatbatch 1986b, C l a y m a n 1993, Bull 1994, C l a y m a n & Heritage i n press b). This article provides a n overview of the dynamics of answering a n d resisting questions in broadcast news interviews. A n exhaustive treatment of this complex subject is, of course, b e y o n d the scope of a single essay. T h e present objective is to delineate some of the m a i n issues involved a n d some of the m a i n practices mobilized to m a n a g e responses that resist the agenda of a question. T h e primary database consists of m o r e than 100 news interviews broadcast in the United States a n d England over t h e past two decades. A second a n d m u c h smaller data set consists of journalistic questioning in other contexts, including U.S. presidential press conferences a n d campaign debates.
C o n c e p t u a l Preliminaries Evasiveness is an elusive p h e n o m e n o n , and its analysis is fraught with conceptual pitfalls. T h e concept is, in the first instance, a familiar part of interactants' ordinary language for characterizing a n d sanctioning conduct, b u t it is also deployed b y social scientists in technical analyses of such conduct. This raises numerous difficulties, including the question of the perspective from which evasiveness is to b e assessed. O n e a p p r o a c h treats the ANALYST'S PERSPECTIVE as primary. H e r e , the bound ary between "answering" a n d "evading" is something that the analyst determines b y formulating a clear-cut operational definition, which can then serve as a benchmark from which to assess particular responses. This approach can generate informative results (e.g., Harris 1991, Bull & M a y e r 1993, Bull 1994), b u t it becomes problematic w h e n t h e analyst's assessment diverges from that of the participants themselves. It is, after all, t h e participants' own understandings of •*
•
1
' '*
il
i.1
nohmllir
Jmmlnra
CLAYMAN
ANSWERS A N D EVASIONS
23
tradition that informs this essay, h e r e the PARTICIPANTS' PERSPECTIVE is treated as of primary importance in the analysis of responses. T h u s , every effort is m a d e to ground analyses in the understandings a n d orientations of interviewers and interviewees as these b e c o m e manifest in the interaction itself. This ideal can b e difficult to achieve in practice. O n e p r o b l e m is that the participants m a y n o t necessarily agree on the i m p o r t of a particular response. While an interviewer m a y treat a given response as improperly evasive, the interviewee w h o p r o d u c e d it m a y treat it as an essentially valid way of dealing with a difficult and p e r h a p s flawed question. I n this connection, the very terms used b y the analyst to characterize responses can b e c o m e problematic. "Evasion" connotes moral impropriety a n d thus m a y b e seen as e m b o d y i n g a contestable perspective on the action u n d e r analysis. A more fundamental difficulty is the fact the participants' understandings are not always transparent, and they m a y at times b e designedly opaque. Consider that when an interviewee sidesteps a question, h e or she m a y strive to conceal that fact in an effort to avoid various negative consequences that might otherwise follow (e.g. hostile follow-up questions, negative inferences). Correspondingly, even if the interviewer recognizes that the question has b e e n sidestepped, h e or she m a y decide to "let it pass" in the interest of m o v i n g the interview along. It is thus possible that a n act of evasion m a y occur that is fully apparent to b o t h participants, yet neither party registers that fact in any demonstrable way. Accordingly, well-grounded analytic judgments must draw n o t only on re sources internal to the particular instance under examination, but also on patterns of conduct that cut across numerous cases. Moreover, to maintain analytic clarity, I will adhere to the following terminological convention: I will reserve the term ""evade/evasive" for actions that are treated as inadequately responsive b y the interview participants; other terms - e.g. resist, sidestep, agenda-shift - will b e used m o r e broadly to encompass responses that depart from the agenda of the question, b u t which the participants m a y n o t necessarily treat as inadequate on that occasion.
Doing "Answering" Any analysis of this subject must begin b y considering the fundamental nature of ANSWERING as a type of social action. I n the organization of interactional sequences, answers are responsive actions that b e c o m e relevant only o n the completion of a question, and questions in turn set an agenda of topics and tasks to be dealt with in subsequent talk (Heritage in press). Moreover, after a question has been produced, interactants monitor the ensuing talk to determine h o w it embodies an answer to the question (Schegloff 1968, 1972). W h a t , then, consti tutes an answer in this context? This m a y at first seem obvious: an answer is an action that addresses the agenda of topics and tasks posed b y a previous question. What is less obvious is precisely h o w such responsiveness is displayed b y inter1
i.U
TT7-\
J
-
1
1
.
.
.
»
-
-
—
•
24
DISCOURSE STUDIES
fact that there is n o single p r i m a r y indicator or marker of "answering" - unlike "questioning," which is typically m a r k e d b y interrogative syntax. H o w , then, do IEs indicate that they are indeed being responsive to the question at h a n d ? I n other words, h o w do they A C C O M P L I S H O R D O " A N S W E R I N G " ? As a point of departure, consider that there are various paths or trajectories that answers m a y follow. Some answers take a R O U N D A B O U T TRAJECTORY; they begin with a unit of talk which c a n n o t in itself b e construed as a possible answer, but which is part of a larger stretch of talk that can b e seen in its entirety as answering. For instance, w h e n a conservative politician is asked (ex. [1], lines 1-2) about the attractions of a n e w workfare proposal - which would require those receiving u n e m p l o y m e n t benefits to work for t h e m - his initial remarks (lines 3-10) do not, b y themselves, answer the question. Instead of talking about the advantages of workfare, h e begins b y attacking the current u n e m p l o y m e n t p r o g r a m as "ludicrous" (lines 3-7), a n d h e then goes on to say that reducing benefits is n o t a viable solution (lines 7-10). O n l y after h e has in effect ruled out these other courses of action, taking three full sentences to do so, does h e speak direcdy to the issue of the advantages of workfare (lines 11-13). (1) UK BBC Today: Social Security Cuts 1 IR: . . . Mister Howell what are the attractions as you see them: 2 uh- of this workfare idea? 3 RH: .hh Well (.) hh it seems to me to be ludicrous that we 4 are spending according to the government more than 5 eight bjllion pounds: in support of the unemployed on 6 condition that they do nothing whatsoever .hhh to(r) 7 help society, .hh And I believe the time has come 8 when- when we've got to recognize: that (.) par::ing 9 down benefits is not the an:swer. That isn't how 10 savings can be made. 11 .hhh Savings ku- huge savings could be made: if ahm 12 (.) one the unemployed people were offered the right 13 to work and given an opportunity to work.... Although the initial remarks could not b y themselves stand as an answer, they are n o t irrelevant to the question at h a n d . Indeed, the initial assessment of alter natives to workfare is a relevant prelude to rendering a comparative assessment of workfare itself. T h u s , considered holistically, this entire turn can b e seen as occupied with the task of answering the question. A n d I R s generally treat such r o u n d a b o u t answers, once completed, as adequate. However, I R s do n o t in the first instance encounter C O M P L E T E D turns; they monitor a n d evaluate turns INCREMENTALLY, while they are unfolding. From that in-progress vantage point, the subsequent trajectory of a response m a y b e decidedly uncertain. I n the previous example, the I E could simply h a v e attacked the alternatives without ever advancing a n affirmative argument for workfare. To
CLAYMAN ANSWERS AND EVASIONS
25
This is a leap that IRs are not always willing to make. Thus, roundabout answers are initially vulnerable to being h e a r d as evasive a n d are subject to countermeasures from the I R . Consider ex. (2), an excerpt from a 1985 interview with Pat Buchanan, shordy after h e b e g a n serving as President Reagan's second-term director of communications. T h e I R rather delicately makes the point (lines 2-6) that other administration officials d o n ' t seem to like Buchanan very m u c h a n d have leaked that view to the press, a n d h e goes o n to ask Buchanan (line 7) h o w that can h a p p e n . B u c h a n a n responds (lines 9-11) b y noting that there was a lotta that in the first term, a n d h e begins to explain why. T h i s could b e the first component of a r o u n d a b o u t answer that will eventually deal with the current situation, but it could also b e an effort b y Buchanan to deflect the discussion away from himself. T h e I R takes the skeptical view, analyzing it as an incipient evasion; he interjects at line 12 (arrowed), pointing out that you weren't in in the first term, thereby treating Buchanan's turn-thus-far as irrelevant a n d unresponsive. (2) US, 3 June 1985, Nightline: Patrick Buchanan 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
IR:
Continuing our: conversation now with Pat Buchanan. Pat- uh:- (0.2) to pat it as gently as I ca:n there're some people: fairly high up in this administration who seem to be able to contain their enthusiasm for you, .hhh And every once in a while stories crop up in thuh press that one can only assume come from some o'those folk. (0.2) How does that sort of thing happen, in an administration. (.) PB: tlk .hhh Well I think 'ere was a lotta that in the first terim Ted, an' = uh: >I flunk one o'thuh rgasons< was you had-= IR: —> = Well you weren't in in the first [term. PB: [Right, but you had three chiefs of- (.) ehsta:ff virtually, you had Baker (.) Deaver and Meese .hhh An' there was it seemed to me from the ojslsr.de an awful lot of leaking on one er against one er another, from secondary an' tertiary personnel .hhhhh A:n' since Don Regan came in we've been goin' through a bit of a transition, there was some o'that I think back in April, .hhhh but since thuh transition's been complete I haven't seen any of it an' we don't expect to see as much in the:: uh .hhh in the second t e r m . . .
it is possible that the design of the final question (line 7) unintentionally encour ages Buchanan to begin his response as h e does. Perhaps to soften what is plainly a face-threatening subject, the I R depersonalizes the question somewhat b y asking how that sort of thing can h a p p e n , a n d this m a y license Buchanan's decision to talk initially about the larger history of Reagan administration leaks. But whatever may have prompted this initial response, a n d whatever direction Buchanan m a y i
i -
i
.i
.i
«
•
DISCOURSE STUDIES
2fi
A more common type of trajectory establishes the "answering'' character of the talk early on. This trajectory may be termed M I N I M A L A N S W E R P L U S ELABORA T I O N ; it begins with a first unit of talk in response that provides the information targeted by the question, albeit in a minimal way, followed b y subsequent talk that clarifies and elaborates. For instance, a yes/no question can prompt an initial one-sentence expression of affirmation or negation before that answer is elaborated. A n explicit "yes" or "no" m a y be included in the initial response, which is the normative way of answering a yes/no question (Raymond 1998). For example: (3) US, 22 Feb. 1985, IR:
AB: ->
Nightline.
South African State of Emergency
tch .hh Are you willing (.) personally to renounce the violence (.) in that country. (0.6) .hh Yes I will. I mean I have said so on Saturday I was on a platform . . .
Similarly, a wh-type question (how longbelow) can prompt an initial one-sentence provision of the requested information (arrowed) prior to further elaboration: (4)
UK, IR: ZM:
Newsnight
China 2
And how long how long will that take and how long has he got to prove he can do it? -> .hhhh Ah:: (0.2) it tt-(0.2) maybe it take uh —> one or two years (.) to to do that. (0.7) i And I think ah t h a t . . .
Like a standard written paragraph, this trajectory of answering begins with an initial unit of talk that minimally fills the information gap targeted by the question, while subsequent talk furnishes details that clarify, support, or elaborate. Furthermore, the initial remark's relevance to the question is often marked by various surface features of the remark's design. Perhaps the most obvious way of marking question-relevance is to incorporate some of the wording of the question into the initial response (Roth 1996, Schegloff 1998). The repetition may involve a single key word - confrontation in ex. (5): (5) U K
Newsnight.
Tienanmen Square Uprisings
IR
Jonathan first (.) let me ask you:, uh what is the latest situation are we any nearer: the actual (.) straight —> confrontation between the troops and the students (.) in the square. JM: => Well I think we've already had this confrontation. The: uh citizens of Peking .hhhh and of course . . . A lareer Dhrase m a y also be repeated, such as s i d e effects in ex. (6):
CLAYMAN
ANSWERS A N D EVASIONS
27
(6) US, 8 Dec. 1985, Face the Nation: Cancer Treatment IR:
SR:
—» .hhh Now tell us about uh = the side effects, .hhh Is it as toxic (.) a:s chemotherapy, < Is it as poisonous: an- (.) to the system and what=are the other —> side effects. => The side effects could be quite severe, but they're =* somewhat different than the kinds of side effects that one sees with- uh with chemotherapy, .hhh The major => side effect is a buildup of fluid in the body . . . .
An I E m a y go still further, incorporating the entire framework of the question into the initial response a n d t h e r e b y matching his response word for word to the question at h a n d : (7) US, 13 March 1979, World at One: Mineworkers Strike IR: —> And what do you think the result of the ballot will be. AS: => 1::: think that the result of the ballot will most probably be acceptance of the deal:, (.) but it could be: er closer than most people expect. This latter m o d e of response can s e e m exaggerated or hypercorrect, a n d it m a y indicate an undercurrent of resistance or hostility to the question. But the general import of repeating lexical items from the question remains m u c h the same across these examples. T h r o u g h this practice, IEs can p r o p o s e that they are attending to the question in detail a n d are thus properly responsive to the issues that it raises. Most of the practices e x a m i n e d thus far involve similarities in phrasing, b u t other practices for doing "answering" work quite differentiy. Certain I N D E X I C A L EXPRESSIONS, n a m e l y those involving A N A P H O R I C R E F E R E N C E , h a v e m e a n i n g s that are inextricably linked to the prior question. T h e simplest type involves the use of a p r o n o u n that refers to the issue raised b y the question, such as that in ex. (8) (arrowed): 3
(8) US, 22 July 1985, MacNeiVLehrer. South Africa IR:
.hhhhh Reports today: are: of course that the violence has continued uh what have you heard tha- whether er not the state of emergency is. in fact working. (0.4) HB: tch .hhhhh It is perhaps too soon:: to make —» make a judgement on that
Here the I R asks whether the South African government's imposition of a state of emergency is working to stop the violence there. T h e I E ' s response - It is perhaps too soon to make a judgment on that- ends with a p r o n o u n that acquires its meaning b y referring to the matter raised b y the previous question.
DISCOURSE STUDIES
28
It is not only pronouns that have this back-referencing character. For example, certain verbs are also heavily context-dependent for their meaning, such as was in ex. (9) (arrowed): (9) UK, 5 Nov. 1979, ATV Today. Innocent Man Int:
TS:
.hh Were you surprised when you: w- went to court, an- and indeed went dawn, (0.2) —> hhhh .hhh I was mos-1-1 certainly was, in fact I-1 .hh all the way up to the- to the court...
I n the context of the prior question (Were you surprised), the response (I certainly was) can readily be understood as m e a n i n g "was surprised." Units of talk that are shorter than a sentence also tend to b e parasitic o n the question for their meaning. For example, the initial phrasal response below Child support offices - can b e understood in context as advancing a claim that "child support offices will b e the j u d g e of good cause." (10) UK, Today. Child Support IR: RH:
.hh But who's going to be the judge of that, the judge of good cause:. —> .hhhh Child support offices. Ahm in the local offices and in our regional centers .hh ah we'll look at £ach case very very carefully —
Finally, certain turn-initial discourse markers also refer to the previous question. Answers to why-type questions m a y b e prefaced b y because, which identifies what follows as an explanation fitted to the question. For example: (11) US, 18 Sept. 1992, MacNeil/Lehrer. Perot IR: RP:
->
So why don't you go ahead and (.) say: I'm (.) a candidate for president? [Because that's not (.) where the organization is now. Our organization (.) is to:tally focused on try:ing to get both parties to do the job. (0.7) That's why.
i.
I n all these cases, the sense of the initial r e m a r k is, b y itself, indeterminate at least to some extent. Observers m u s t refer to the question to disambiguate the remark, and in so doing its m e a n i n g is "filled in" in a way that is thoroughly question-relevant. I n analyzing these various practices for doing "answering," it m a y seem that we are e x p e n d i n g a great deal of energy for a m o d e s t payoff. H o w e v e r , these practices are far m o r e significant than the analysis thus far suggests. A s w e shall
CLAYMAN ANSWERS AND EVASIONS
29
see, not only do they figure in genuine efforts to answer the question; in addition, lEs can use t h e m subversively in m a n e u v e r s that are substantively resistant.
D i m e n s i o n s of Resistance Resisting a question is, like answering, a complex p h e n o m e n o n . We can begin to dissect this p h e n o m e n o n b y drawing a basic conceptual distinction between two dimensions or aspects of resistance. The Negative Dimension The NEGATIVE aspect is manifest to the degree that the IE's response falls short of a n adequate answer to the question. I n the strongest variation o n this t h e m e , the I E declines to provide any information at all that bears o n the question. For example, w h e n a m e m b e r of the L a b o u r Party is asked about his willingness to serve in a cabinet committed to unilateral nuclear disarmament, h e flatly refuses to answer (arrowed): 4
(12) UK, Greatbatch 1986b:451: Unilateral Nuclear Disarmament IR:
You wouldn't serve in a Cabinet committed to lulinilateral nuclear disarmament of Britian would you Mister Shore? PS: .hh What I fjo believe:: er: Mister Day (which) —> I will not a:nswer that question, I'm not (.) -» deliberately answering that question. What I do believe is thi:s. I do actually genuinely believe la:ng believe: (d) .hhh that unilateral initiatives: (.) can assist (.) multilateral disarmament. In a less extreme form of negative resistance, the I E provides a n answer that is partial or incomplete. For instance, in the face of a two-part question, an I E may address one part while leaving the other unanswered. T h u s , in a N i x o n press conference h e l d during t h e Watergate p e r i o d , a journalist first asks (beginning at arrow 1 in ex. [13]) whether N i x o n is personally investigating char ges that his campaign funds were mishandled, a n d he then asks (arrow 2) whether the charges will h u r t his b i d for reelection. 5
(13) US, 29 August 1972: Nixon Press Conference JRN: 1 —> Mr. President, are you personally investigating • the mishandling of some of your campaign funds, 2 —> and do you agree with Secretary Connolly that these charges are harmful to your reelection? RN: Well, I commented upon this on other occasions, and I will repeat my position now.
DISCOURSE STUDIES
30
3 —¥ With regard to the matter of the handling of campaign funds, we have a new law here in which technical violations have occurred and are occurring, apparendy, on both sides. As far as we are concerned, we have in charge, in Secretary Stans, a man who is an honest man and one who is very meticulous - as I have learned from having him as my treasurer and finance chairman in two previous campaigns - in the handling of matters of this sort. Whatever technical violations have occurred, certainly he will correct them and will thoroughly comply with the law. He is conducting any investigation on this matter, and conducting it very, very thoroughly, because he doesn't want any evidence at all to be outstanding, indicating that we have not complied with the law. N i x o n targets only the first question for response (see arrow 3), a n d h e never gets a r o u n d to addressing the second question about political implications. A n o t h e r type of incomplete answer is a simple yes-or-no reply to a yes/no question. I n the news interview context, IEs normally are expected to produce elaborated answers (Heritage & Greatbatch 1991), so that minimal yes-or-no responses (arrows 1 in ex. [14]) can b e seen as inadequate. T h e y indicate tacit resistance to the b r o a d e r agenda of the question, a n d they regularly lead IRs to p r o b e for further elaboration (arrows 2): (14) US, Meet the Press, 24 Oct. 1993 IR:
. . . .hh Madam Attorney General you've testified this week- u- in front of Congress abou:t .h violence and television, .hhh And said that if the TV industry didn't in effect clean itself up, clean its act up, .hhh there may be government intervention. Government regulation. (0.4) Thuh New York Ti:mes in an editorial said that (.) you embarked on a quote (0.3) Didju? IE: 1 -> No. (0.2) IR: 2 —> .hhhhWha:t kind of government intervention are you thinking about? Would you ban: programs like NYPD: Law and Order, would you [uh: IE: 1 -> [No. (•)
IR: 2 IE:
—» W- Wh:at are we talking about. We're talking about (.) asking the media to slop talking (.) about what it promises to do, and do i t
CLAYMAN ANSWERS AND EVASIONS
31
The Positive Dimension Resistance has a positive dimension as well. This is manifest to the degree that in IE moves b e y o n d the parameters of the question, saying and doing things that were not specifically called for. These departures v a r y greatly in magnitude and in kind. T h e most dramatic form of departure involves a substantial change of topic. For instance, following a question about nuclear waste disposal (lines 1 - 3 in ex. [15]), an expert n o t only refuses to answer (lines 5-6); she then goes on (lines 7-10) to refute allegations m a d e m u c h earlier in the p r o g r a m regarding the Three Mile Island accident a n d its health consequences. T h e issue of p o w e r plant accidents is rather far r e m o v e d from that of routine waste disposal efforts. (.15) US, 6June 1985, Nightline: Nuclear Waste 1 2 3 4
5 6 7 8 9 10
IR:
Continuing our conversation now with Doctor Rosalyn Yalow. Doctor Yalow uh- ehh lejnme put it in very simple terms. If it's doable, if it is: easily disposable, why don't we. (1.0) RY: Well frankly I cannot- (.) ANswer all these scientific questions in one minute given to me. On the other hand there was one horrible thing that happened tonight that you have- .h in addition extended, .hh And that is thuh NOtion that there is an increased incidence of cancer associated with the Three Mile Island accident.
Here the IE veers sharply away from the topic of the question a n d toward a .substantially different area of discussion. Alternatively, a response m a y lie within the question's topical parameters but perform a task or action other than what was specifically requested b y the question. Thus, in ex. (16), w h e n Senator B o b Dole is asked whether h e would support the reappointment of the Federal Reserve Board chairman (lines 1-3), Dole offers a generally favorable assessment of the chairman's past performance (line 4 ) . but h e does n o t specifically endorse his reappointment (presumably in deference to the political i n d e p e n d e n c e traditionally granted to the Fed). This evasive maneuver does n o t escape the notice of the I R , w h o presses again for an explicit endorsement (line 5): (16) US. This Week: Senator Bob Dole (from Donaldson 1987) ( IR: 2 3 4 BD: 5 IR:
Talking about money, what about Paul Volcker, whose term is up next year? Would you like to see him reappointed to the Fed? I, I think he's been very effective. Well, would you like to see him reappointed?
Here, the response is, b r o a d l y speaking, within the topical parameters of the question — it is "about" Paul \folcker—hut it nan hp rpmrAfA a c mraemo nnnathcAa^
DISCOURSE STUDIES
32
T h e m o s t subtle form of evasion is e m b o d i e d in those responses that alter the terms of the question ever so slighdy. Consider ex. (17), from a n interview with the deputy secretary of defense, J o h n Deutch, o n the subject of Gulf War syndrome. Across this excerpt, the I R tries to get D e u t c h either to confirm or to d e n y reports that U.S. troops were exposed to chemical weapons during the Gulf War. H o w e v e r , D e u t c h will not b e p i n n e d down; although h e repeatedly offers denials, all are in some w a y qualified. 6
(17) US, 12 March 1995 60 Minutes: Gulf War Syndrome 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
JD:
IR: JD:
IR:
JD: IR: JD: IR: JD:
hh Our most th:prough (0.2) and careful efforts to determine (.) whether chemical agents were us:ed in the Gulf, (.) .hh lead us to conclude that there was no: (.) widespread use of chemicals against U.S. troops. = =Was there any use. = Forget w [idespread. Was there any use.] [I- I do not belie:ve ] I do not believe there was any: o:ffensive use of chemical agents b_y.: .hh uh- Iraqi: (0.2) uh military: (.) troops. Ther [e was not[Was there any- any accidental use. Were our troops exposed in any way:. (0.4) .hhh Uh-1 do not believe that our troops were: expo:sed in any widespread way to: u [h: chemical [In any narrow way. = In any way. hh .hh The defense science board did an independent study of this matter: .hh [and foji:nd in their judgement* [(.) =that there was BQ:: confirmation .hh of chemical: (0.2) weapon (0.2) widespread use: in the Gulf.
T h e deputy secretary first denies that our troops were exposed to any widespread use of chemical weapons (lines 1-4). T h e I R seeks to disallow this qualifying adjective b y asking about any use whatsoever (line 5), b u t the I E again qualifies his response b u t in a different w a y (6-8), this time denying only that there was offensive use. W h e n the I R tries to disallow that qualification (10-11), the I E switches back to his previous qualifying adjective, restricting his denial to widespread exposure (13-14). This m a n e u v e r is countered m o r e aggressively b y the IR, w h o interjects his next pursuit just after the qualifying adjective is introduced a n d before the I E has reached a possible completion point (14-15). H e makes o n e last attempt to elicit a blanket denial, b u t the I E holds fast to his m o r e cau tious position (17-21). T h e deputy secretary's caution is not difficult to understand: a n y confirmation w o u l d p r e s u m a b l y e x p o s e the g o v e r n m e n t to n u m e r o u s lawsuits, while a flat denial m a y b e unsupportable a n d m a y even contradict cur rent or future evidence. H e m a n a g e s to avoid b o t h alternatives repeatedly, and in a rather subtie way. H e presents himself as if h e were being dutifully responsive,
CLAYMAN
33
ANSWERS A N D EVASIONS
Finally, even responses that do address the agenda of the question, but contain additional turn components that shift away from that agenda (as in exx. [18-20]), are treated as problematic in t h e n e w s interview context (Greatbatch 1986b). This is a product of t h e distinctive turn-taking system that organizes news inter view talk, which obligates IEs to restrict themselves to the action of answering IRs' questions (Greatbatch 1988, Heritage & Greatbatch 1991). Given this nor mative constraint on I E s ' talk, any attempt to p r o d u c e something other than an answer - even as a supplement to a n otherwise responsive answer - m a y b e re garded as an i m p r o p e r diversion from the agenda at h a n d . By changing the topic of a question, the task that it poses, or the specific terms in which it is framed, I E s can loosen the strictures inherent in being o n the receiving end of a n interrogation. A n y such m a n e u v e r is resistant in the affirmative sense a n d constitutes an effort to SHIFT T H E A G E N D A of the question.
O v e r t Practices Given the various dimensions of resistance, h o w are such m a n e u v e r s actually managed b y IEs? IEs deploy a wide range of discursive practices in such contexts, and these can b e u n d e r s t o o d in terms of the way they reduce the risks associated with a resistant course of action. O n e set of practices is used w h e n IEs choose to b e u p front and explicit about what is taking place. T h e strategy of resisting a question OVERTLY has, from the IE's point of view, an obvious disadvantage: It renders the resistance conspicuous, and hence m o r e likely to b e noticed b y the I R and the m e d i a audience. However, this disadvantage is offset b y a n equally important advantage: H a v i n g admitted the evasion, an I E can take steps to minimize the d a m a g e that it might otherwise cause. T h r e e forms of d a m a g e control will b e e x a m i n e d as they figure in efforts to shift away from the a g e n d a of t h e question. Deference to the IR IEs often preface their agenda shifts with remarks that display s o m e degree of DEFERENCE T O T H E IR. Perhaps the greatest deference is conveyed w h e n the I E actually REQUESTS P E R M I S S I O N F R O M T H E IR to shift the agenda. In ex. (18), for example, a C h i n a expert first answers a question about whether recent civil unrest will strengthen t h e position of reform-oriented officials in that country (lines 4-13), a n d h e then goes o n to talk about various other matters 116-25), including the generational shift in Chinese leadership and p r o b l e m s of corruption. However, h e does not raise these other matters without first requesting (arrow 1) and receiving (arrow 2) permission from t h e I R to do so. 7
(18) UK, Newsnight: Civil Unrest in China 1 IR:
Well what do you think do you think this strengthen:s
o
/1 n\ - —— * i
i
. i t
i
•-'
-
-
DISCOURSE STUDIES
34
4 DH: I think that (0.2) Jao Ze Young just as he was 5 responsible for bringing (.) China out of the turbulence. 6 which followed the .hhh uh resignation of Hu Yao Bung as 7 General Secretary in=uh January nineteen eighty seven. 8 .Hhh just as he (.) brought China out of that turbulence 9 he will bring Chi:na out of this turbulence .hhh and I 10 think his stature has already been increased (.) by 11 recent events (.) .h and ah (.) I'll go out on a limb 12 and say: I think it's likely to be increased further 13 .hh by future events 14 1 —> but I would like to make two very quick points.= 15 IR: 2 —» =Yerv. quickly if you would. 16 DH: There's a genera/.tional thing he:re. .Hhh U:um (0.4) ih 17 Deng Zhao Peng is going to be ekghty fi:ve on the twenty 18 second of August this yea:r. .Hh he joi:ned the 19 Communist Pa:rty (0.3) .h in nineteen twenty fou:r. .Hh 20 When Mister Baldwin had become prime minister for the 21 first time in this coun:try:. Just. .Hhh (0.3) Secondly 22 (0.3) corruption. .Hh A Jot of (.) what is: (.) ca:lled 23 corruption .hh is in fact the by:produ:ct (0.1) of a 24 system of multiple pricing .hhh which I think is going to 25 have to be rela:xed. A similar request appears in ex. (19), an excerpt from a debate interview con cerning health care reform. T h e I E - a health insurance industry executive who opposes President Clinton's health care reform plan - is asked whether antireform T V ads disclose the fact that they were paid for b y the insurance industry (lines 1-4). She answers this question in the affirmative (5-8), b u t she then goes o n (arrow 1) to ask the I R for permission to c o m m e n t o n an issue raised earlier in the p r o g r a m b y a reform p r o p o n e n t (Ron). W h e n the I R grants permission (arrow 2), she proceeds to address this other issue (which has to do with whether the so-called Coalition for H e a l t h Insurance Choices is actually an association of insurance companies masquerading as a grass-roots public interest group). (19) US, 21 Oct. 1993, MacNeil/Lehrer. Health Care Ad War 1 2 3 4 5 6 7 8 9 10 11 12
IR:
LJ:
IR: IR:
.hhh Well Miss Jenckes he raises an interesting question.= Again just as a matter of strategy your ad doesn't say:: that it's sponsored by the heal:th (.) insurance companies] [Margaret that's absolutely incorrect, .hh Our a:ds (.) whether they're on TV, .h our print a-advertisements. that appear in newspapers .hh even radio spots indicate that we have paid for it.= 1 —> Let me may- just make one commfent in [terms of wha [t Ron: says. [.hhh [wh[ih2 -> Al [lright.]
CLAYMAN
ANSWERS AND
15 16 17
EVASIONS
35
you always start with like minded people. But whether .. you're an agent or a broker, .hh you have legitimate health care concerns yousejf....
Requests for permission openly acknowledge that a shift of the agenda is in the works. In ex. (19), the I E specifically indicates (lines 9-10) that she wishes to respond not to the IR's question, b u t to a point m a d e earlier b y another I E . At the same time, however, such requests defer to the I R as the one w h o is properly in charge of the discussion agenda. T h e I E m a y also offer w h a t Greatbatch 1988 has t e r m e d a T O K E N R E Q U E S T FOR PERMISSION to shift the agenda; this resembles an actual request but is not treated as requiring a response from the IR. T h u s , in ex. (20). a discussion of newly proposed legislation to restrict access to abortion, an anti-abortion advocate answers a legalistic question about the wording of legislation, but she then goes on to argue that current law is too permissive. She prefaces this agenda shift with a request-like object, can I also point out (arrowed). (20) UK, Afternoon Plus: Abortion 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
IR: JK: IR: JK: JK: (): JK:
Jill Knight may I ask you how far that's going to be put into practice and [what- who: is going to de£i:de what= ['Ye-s' =i[s serious and what [is a substantial['Ye:s' [.hhuh Well of course the doctor:: and u::h thuh[in other a'.reas wh(h)ere medical- th'medical profession= [( ) =is practiced .hhh doctors've been quite capable of deciding what's serious. (.) and what substantial means, —> .hhh And can I also point out, .hh that u::h Professor Huntingford whom you had on .hh your program in Decgrnber:: .hh supporting the abortion act .hhh u::h eh said (.) really (.) again quite recently there's no do(h)ubt abo(h)ut it=we have got abortion on request, .hhh and this is what parliament did NOT ask for . . . A
.Notice that the I R makes n o attempt either to grant or to refuse permission, and the IE does not seem to expect a response. Indeed, she actively discourages a response b y building her token request as an incomplete clausal unit (can I also point out that...), which projects further talk to come. Because they provide so little opportunity for response, token requests are somewhat less deferential than their full-fledged counterparts. Nevertheless, they do show the I E to b e "going through the motions" of seeking permission, thereby continuing at least to acknow ledge the principle that it is the I R w h o normally sets the agenda. Whether they are "genuine" or "pro forma," requests for permission openly acknowledge the fact that a n a g e n d a shift is being contemplated. At the same time, however, these practices alleviate some of the interpersonal damage that
DISCOURSE STUDIES
36
to the I R ' s control over the course of the discussion. A n y effort to seek permission mitigates that threat b y deferring to the I R a n d sustaining a sense in which the I R remains at least formally i n charge. Minimizing the Divergence A second form of d a m a g e control involves downplaying the agenda shift b y portraying it as insignificant, a m i n o r digression from the agenda established b y the question. Requests for permission often contain M I N I M I Z I N G CHARACTERIZA T I O N S , such as reference to "a very quick" or "just o n e " c o m m e n t . (21) UK, Newsnight Civil Unrest in China DH: But I would like to make two very quick points
'
(22) US, 21 Oct. 1993, MacNeil/Lehrer. Health Care Ad War LJ:
Let me may-just make one comment in terms of what Ron says . . .
I n addition to temporal a n d numerical minimizers, the inclusion of the adverb "just" further downgrades what is about to b e said, as in the preceding example and again in the following: (23) UK, Today: Child Support RH: Can I say just to (set) the context... I n each of these ways, the divergence is cast as a slight digression from the framework of the question. Justifying the Shift I E s m a y also strive to EXPLAIN A N D JUSTIFY their efforts to divert the discussion. Justifications m a y b e e m b e d d e d within requests for permission. For example, in ex. (24), a discussion of the 1992 vice presidential debate, a Republican stra tegist first responds to a question about the performance of Ross Perot's vice presidential candidate, Admiral Stockdale, but h e t h e n shifts the agenda (lines 14-16) to defend George Bush's flip-flop o n abortion, a n d h e prefaces this shift with a token request for permission (arrowed): (24) US, 1 2 3 4 5 6 7 o
13 Oct. 1992, Nightline: Presidential Debate IR: .hhhh Uh Bill Kristol, does: S_tockdale's performance tonigh:t take some of the air:: (.) out of the the Pe [rot balloo ]n:: just as it was getting bhown up again. WK: [pHHHhhh] (0.6) WK: Uh:: I'm not sure about that Chris. I think the: ah: two things were remember- we'll remember about Admirable o*„^l. i f t m i r a i Qir.Mrrlalp tnnio4it ah! are his: verv strong
CLAYMAN
ANSWERS AND EVASIONS
10 11 12 13 14 15 16
37
and his statement about the important of cah- 'portance of character, .h to leadership. Both of those statements will: (reboun:d) to the benefit of President Bush. .hh Ah and on that last point —> if I could just speak to Molly's point: uh before the break, uhm President Bush changed his mind about abortion an:d said s o . . . .
This token request contains justificatory elements, and these are embodied in how the requested action is characterized. Instead of asking generically to "say one more thing'' or "make an additional point,'' this IE asks specifically to address Molly's point before the break. This way of characterizing what he wants to do is not technically necessary for the token request to be intelligible; it is a choice that provides an implicit rationale for the agenda shift. In the context of a debate interview involving partisan IEs, portraying the shift as a response to a point made earlier by an opposing IE tacitly justifies the shift on the grounds of fairness and the principle that partisan accusations should not be permitted to stand unanswered. In addition, the shift is characterized in relation to an impending commercial break, further justifying the maneuver by explaining why it is being launched at this particular point in time. Justifications can also appear outside of permission requests, where they tend to be more explicit and elaborate. A n example is highlighted in ex. (25). The IR asks whether corporate mergers are creating monopolistic entities, and the IE briefly addresses this issue, but he then raises other concerns about mergers. Before doing so, however, he justifies this shift (arrowed) on the basis that it will concern the most important problem with corporate mergers, one that has not yet been addressed in the interview. (25) US, 5 June 1985, Nightline: Corporate Mergers IR:
HM:
.hhhh Senator Metzenbaum take me hack to the- to that difference: that uh Mister Forbes made a moment ago, between monopolies and what we have today:, which it seems in- in some instances is moving .hh at least (0.2) gradually in the direction of a monopoly, is it not? (0.3) Well I think that gome mergers (.) don't have any element of monopoly in them at a:ll. .hh (.) Uh for example General Motors buying Hughes Aircraft (I'm-) not at all certain that there's any monopoly (.) issues there. (0.5) —» On the other hand I think the real concern tha hiasn't —> been addressed (.) previously (.) in this program (0.7) HAS to do with the fact that... ((parenthetical comment omitted)) . . . when you have a major merger of this kind, (0.2) of the KINd that we've been talking about on this program, (.hh) you haff to worry A does it eliminate (.) competition
DISCOURSE STUDIES
38
secondly you have to be concerned as to the impact (.) on the shareholders, (0.4) and third but certainly not least of the three, (.) is the impact upon the community . . . T h e rationales offered for agenda shifts tend, not surprisingly, to exclude naked self-interest as a motivating factor. Instead, such rationales fall into one of two basic categories. O n e argument, c o m m o n in p a n e l interviews involving partisan IEs, is based o n an implicit principle of fairness a n d the n e e d to respond to points raised b y opposing I E s . Ex. (24) typifies this rationale, b u t other examples are c o m m o n p l a c e : (26) US, 3 Feb. 1992, MacNeil/Lehrer. Haitian Refugee Repatriation BA: Ahm: let me just respond to a few things that (.) Congressman Rangel said . . . (27) US, 21 Oct. 1993, MacNeil/Lehrer. Health Care Ad War LJ: Let me may- just make one comment in terms of what Ron: says Alternatively, unsolicited material m a y b e justified o n the basis that it has a significant bearing o n the overarching subject at h a n d . This type of rationale is illustrated in ex. (25); other examples include the following: (28) UK. Today: Child Support RH: Can I say just to (set) the context...
j
(29) US, 5 May 1996, This Week: Gas Tax Repeal RR:
But I think there's really a- a mu:ch b:igger (0.4) this is part of a much bigger picture . . .
I n either case, the import of this practice remains m u c h the same. Such accounts acknowledge the fact that a shift of the agenda is in progress, and they even grant that this constitutes a b r e a c h of interview etiquette. But b y providing a justification g r o u n d e d in principles of fairness or relevance to the discussion agenda. I E s present that b r e a c h i n a favorable light. The Special Case of Refusing to Answer Justificatory accounts b e c o m e particularly elaborate and strenuous w h e n the IE overtly refuses to answer the question altogether. This follows from the fact that such refusals constitute a particularly strong b r e a c h of etiquette. It is one thing to m a k e some effort to answer the question before proceeding to shift the agenda; it is quite another to decline to answer altogether (Greatbatch 1986b). Accordingly, justificatory accounts are crucial in this context. Various rationales m a y b e offered to account for a refusal to answer, but they *
-
1
:
—
1 T " > " Anflart
rosnnnsihilitv a w a V frOID
CLAYMAN
ANSWERS A N D E V A S I O N S
39
is to claim that the information necessary to answer the question is unavailable. Thus, in ex. (30), when a medical researcher is asked whether a new cancer treatment may have other applications, he suggests that the information is not yet available because the relevant research has not yet been done: (30) US, 8 Dec. 1985, Face the Nation: Cancer Treatment IR:
Uh two final questions. Doctor Rosenberg, d'you see this having application for other diseases, like multiple sclerosis or even A:IDS, (0.4) SR: We haven't yet begun: to explore that, although I think possibilities exist 'at need to be investigated 'n I think other:: scientists will be looking at those questions. Notice that the IE does not merely assert that he D O E S N ' T K N O W the answer, which might taint his reputation as a medical expert. In general, D O N ' T K N O W accounts in news interviews usually contain some further explanation for the lE's lack of information. Indeed, when a simple don't know account is offered without any further explanation (arrow 1 in ex. [31]), it is pursued b y the IR yarrow 2) in a way that strongly implies that the IE ought to know the answer in some form. 131) US, 22 July 1985, News Hour: South Africa IR: HB: 1 [R: 2
HB: IR:
Who are these people. (0.7) —> I do not know.= —> = You don't know the naych-1 don't mean their names obviously but I mean what kind of pepphe are fallingfare falling- (0.4) into the category uh those = [I would ha=thetch- need to be arrested.
Alternatively, the IE may suggest that he or she knows the answer but is unable to provide it under current circumstances. The temporal limitations of the broadcast interview are often cited - as in ex. (32), when a nuclear physicist asserts that she cannot answer all these scientific questions in one minute given tome (arrowed). (32) US. 6 June 1985, Nightline: Nuclear Waste IR:
Continuing our conversation now with Doctor Rosalyn Yalow. Doctor Yalow uh- ehh lemme put it in very simple terms. If it's doable, if it is: easily disposable, why don't we. (1.0) RY: —> Well frankly I cannot- (.) answer all these scientific questions in one minute given to m e . . . .
40
D I S C O U R S E STUDIES
delicacies of official negotiations to deflect questions. In ex. (33), from a discussion of the federal budget, Senate majority leader Bob Dole is asked whether it will be necessary to cut social programs, raise taxes, or reduce defense spending in an effort to reduce the deficit, but Dole declines to answer (arrowed), arguing that to do so would be premature in advance of formal negotiations. (33) US. 8 Dec. 1985, Meet the Press: Bob Dole AH:
You can't have it both ways sither.>On this program< you have said that you don't think, .hhh that you'll eliminate thirty to fifty programs, [an'] Senator Packwood= [()] =says ya have to, .hh Number two you say you hope you will not have a tax increase, [.hhhh And]= [But I do.] =number- and number three you say ya h:or±e you can have ad[m o s t] three percent on: .hhh on: on=
BD: AH: BD: AH: 0:
[(
AH:
BD:
—> —¥ —> —»
)]
=defg:nse, .hh And yet you hafta cut fifty billion next year. Now which o'those three is gonna give Senator, (0.4) I think that's going to happen sometime next year when these of as:: uh in- leadership positions=set=down with the President and make the hard choice. I don't think I'd make it today: .hhhhh ih=in December of 1985.
Notice that this account, by proposing that T O A N S W E R would be inappropriate, can also be taken to imply that T H E Q U E S T I O N soliciting this answer was inappro priate, although this is an unstated imputation of an account that remains focused primarily on the inappropriateness of answering. Occasionally, however, an IE will go one step further b y asserting outright that the question is improper and hence unworthy of an answer- in effect, deflec ting the question by attacking it. In ex. (34), for example, when a Serbian spokes person is asked if recent prisoners of war are being beaten (lines 1-2), he suggests that the line of questioning is unnecessarily provocative and biased (5-10): (34) US, 15 July 1995, NPR All Things Considered: Serbia 1 IR: Are they being beaten? Or will you be: are you treating 2 them (u-) humanely according to international conventions. 3 IE: [hhh! 4 (.) 5 IE: Well I mean your line of questioning ieally suggests that 6 we are the most awful creatures on garth. That we a:re 7 heating the prisoners, raping jvomen, and so on and so forth. 8 .hh Please I think I have been very: uh: uh correct in my 9 answers, an' I would expect you to: .hh be more correct in 10 your line of question=because it's extremely provocative....
CLAYMAN
41
ANSWERS A N D EVASIONS
By attacking the question in this way, the I E b o t h justifies his failure to provide an answer and deflects the discussion away from the substance of the question and toward the m a n n e r in which it was raised. It is rare for an I E to refuse a question flady, without providing a rationale of some sort. W h e n that does h a p p e n , it can c o m e across as a n extremely hostile gesture (arrowed in ex. [35]). Consider h o w a Labour politician flatly refuses a question concerning his willingness to serve in a cabinet committed to unilateral nuclear disarmament: (35) UK, Greatbatch 1986b:451: Unilateral Nuclear Disarmament] IR:
:
PS:
You wouldn't serve in a Cabinet committed to lujmilateral nuclear disarmament of Britian would you Mister Shore? .hh What I do. believe:: er: Mister Day (which) —> I will not a:nswer that question, I'm not (.) —» deliberately answering that question. What I dp believe is thi:s. I do actually genuinely believe lo:ng believe: (d) .hhh that unilateral initiatives: (.) can assist (.) multilateral disarmament....
Here the I E not only declines to offer a justification; h e casts his refusal to answer as a deliberate, willful choice. This can b e heard as a powerful - albeit implicit atuck on the legitimacy of the question (and, b y implication, o n the j u d g m e n t of the IR. who asked it), which is treated as so transparently u n w o r t h y that its rehisal requires n o justification. At the opposite extreme are refusal accounts that are nonhostile in character and inflict minimal d a m a g e to the interpersonal relationship between I R a n d IE. The most notably b e n i g n practice involves refusing to answer AS A MATTER Of GENERAL POLICY; I E S m a y assert, in effect, that they never answer questions of thai sort. In ex. (36), for example, w h e n Arthur Scargill is asked if he is planning to ran for the presidency of the National U n i o n of Mineworkers (lines 1-2), he refuses to say then a n d there, pointing out that h e ' s b e e n similarly unresponsive to every other pressman over the past forty-eight hours (3-5): 0 6 ) UK, 13 March 1979, World at One: National Union of Mineworkers 1 IR: Muster Scargill will you run for the presidency of the National 2 Union of Mineworkers. 3 AS: .hhh er Mister Day: I must give you the same answer that I've i been giving every other pressman over the past forty-eight 5 hours, .hhh If and when Mister Gormley officially (.) hands in 6 his resignation and that's by no means certain .hhhh cr during: 7 this year or at any time during the next three years .hh then I 8 will give (.) serious consideration to the matter . . . Treasury Secretary Robert R u b i n does something very similar w h e n asked about thp AWaftirvn
f,W
i- —
1
-
T
T
1
D I S C O U R S E STUDIES
42
three-and-a-half-year-old policy of n o t c o m m e n t i n g o n the future course of the financial markets (lines 2-4): (37) US, 5 May 1996, ABC This Week with David Brinkley: Robert Rubin 1 IR: But which way are they going now?= 2 RR: =>For three and a half years.< .hhh gaim. I have had for 3 three and a half years a policy of >not commenting on 4 what markets are gonna d o . . . . As a form of d a m a g e control, the general-policy account has advantages that extend b e y o n d its justificatory import. By emphasizing that there is a prin cipled rationale underlying the refusal, this type of account also has the effect of D E P E R S O N A L I Z I N G the refusal. It is presented not as a n idiosyncratic response to a particular question from a particular I R , but as a general policy applied to all questions of that sort. This helps to ensure that the refusal will not b e taken as an act of defiance against the I R p e r se. Furthermore, insofar as this practice implies that any further efforts to elicit a n answer will prove fruidess, it also FINALIZES the refusal. It thereby inhibits follow-up questions a n d represents a strong bid to close d o w n the entire line of inquiry. 8
C o v e r t Practices O v e r t practices h a v e their counterpart in strategies for resisting a question COVERTLY. Covert practices are used mainly in the context of positive resistance, or talk that departs from the agenda of the question. W h a t m a k e s t h e m covert is that the IEs avoid any explicit acknowledgment of the fact that a shift is in pro gress, a n d they m a y also take steps to conceal that fact. For the I E , the obvious advantage of a surreptitious strategy is the possibility of "getting away with it"; if d o n e with enough subtiety, it m a y escape the notice of the I R a n d m a n y audience m e m b e r s . O n the other h a n d , if the m a n e u v e r is noticed, it can b e particularly cosdy for the I E . Those w h o sidestep questions while pretending to answer them risk b e i n g seen as devious a n d manipulative, a n d this is over a n d above the negative inferences generated b y the resistance itself. Furthermore, such inferen ces cannot b e forestalled via forms of " d a m a g e control,'' because the coven nature of the practice precludes any explicit remedial effort. IEs can, however, reduce the likelihood that the resistance will b e noticed b y taking steps to render it less conspicuous. Subversive Word Repeats and Anaphoric Pronouns Earlier w e considered h o w a stretch of talk becomes recognizable as a n "answer," and we described a variety of practices - including lexical repetitions and anaphork p r o n o u n s - that are implicated in processes of doing "answering." However.
CLAYMAN
ANSWERS AND EVASIONS
43
A recognizable answer can b e constructed without any of these practices; conversely, when such practices are used, they provide n o guarantee that a fullfledged answer has b e e n given. I n d e e d , I E s can use these s a m e practices subversively to provide a kind of surface camouflage for m a n e u v e r s that are substantively resistant. A straightforward example of this sort appears in ex. (38), a n interview with Arthur Scargill of Britain's National U n i o n of Mineworkers. T h e interview took place just as the mineworkers w e r e preparing to elect a n e w president, and Scargill was discussed as a likely candidate. I n a question seeking to distinguish the candidates on the left, the I R (lines 1-2) asks Scargill to explain the difference between your Marxism and Mr. McGahey's Communism. Scargill launches his response (line 3) with a repeat of a k e y w o r d from the question: The difference is. . . . By virtue of this literal repetition, h e appears to b e m o v i n g straightforwardly to answer the question: (38) UK, 13 March 1979, World at One: Striking Mineworkers 1 IR: 2 3 AS: 4 5
.hhh cr What's the difference between your Marxism and Mister McGahey's Communism. —> er The difference is that it's the press that £onstandy call me Ma:rxist when I do not. (.) and never have (.) er er given that description of myself....
Bui appearances can b e deceiving: Scargill uses the word difference to m e a n something quite other than what it m e a n t in the IR's original question.. I n that question, the difference refers to a distinction between two candidates, Scargill vs. McGahey, and their ideologies. I n the response, the difference refets to a distinction between two interpretations - b y the press vs. b y Scargill - of Scargill's ideology in particular. This semantic shift i n the m e a n i n g a n d reference of the differences part and parcel of a m o r e encompassing shift in the agenda. Scargill veers away from the question p e r se in order to counter a presupposition that was e m b e d d e d within it - that h e is in fact a Marxist. To b e sure, this is a relatively mild diver gence, in that Scargill does n o t change the subject so m u c h as propose that die question is inapposite. Nevertheless, h e does not, stricdy speaking, answer the question in the way in which it was framed. A n d yet h e presents himself as if he were being dutifully responsive. By repeating a key lexical item from the question (What's the difference... —> The difference is...), h e packages his response as if' it were filling the information gap targeted b y the question. Like word repeats, anaphoric p r o n o u n s can b e used subversively. Consider ex. :39), an exchange with a spokesperson for presidential candidate Ross Perot. The (R prefaces his question with a c o m m e n t o n the a m o u n t of m o n e y Perot is planning to spend o n T V advertising during the final weeks of the campaign (lines 1 -3), and h e goes o n to ask w h e t h e r Perot is actually gonna get out and meet with the voters... (4-5). T h e I E begins to respond b y saying Let's talk about this (6), using a p r o n o u n that refers to the questioning turn a n d thus seeming to promise a b o n a fide a n s w w
DISCOURSE STUDIES
44
(39) US, 15 Oct. 1992, MacNeil/Lehrer. Presidential Debate 1 IR: .hh Mister Milfor:d ah r- your maa Ross Perot is gonna 2 spen:d at lea::st ten millr.ion dollars .hh in TV advertising 3 >in the final two and a half weeks of this campaign.< 4 .hhh Is he actually gonna get out and- and meet with 5 voters, campaign like the other candidates? 6 CM: Yeh. well 1- let's: (.) talk about this: for a second. 7 Ah- the- the other two: (.) candidates recei:ve over 8 fifty five point two million dollars directly from the 9 taxpayers, .hh Mister Perot is spending his own money. 10 In addition to the fifty five million dollars that they 11 get from the federal government, .hh they get over a 12 hundred million dollars in so::ft money. The whole way 13 the (.) political process is financed is something we 14 objec:t to::, .hhhh I-1 imagine he will spend ten 15 million dollars on media 'cause what we see is something 16 very extraordinary in American politics, .hh Usually, 17 with the passage of ti:me independent candidates go 18 dawn: in the polls:, .hh Since we have entered the 19 race according to the polls, we've gone from seven to 20 fifteen percjent.... And we think Perot is gonna win. .((Some 20 lines of transcript omitted)) 21 IR: 22 23 24
. . . You: you gave a good answer.=You never answered my question, how:ever. .hh >Part of thuh political process< is for candidates actually to get out and mee:t with voters. (0.3) >Is he gonna do tha:t?
R a t h e r t h a n answer the question about meeting with the voters, the I E responds instead to the prefatory comment, offering a lengthy justification for Perot's ad vertising expenditures (lines 6-14). H e then goes o n to c o m m e n t o n Perot's rise in the polls and his chances of winning (15-20). This shift is obscured b y the initial back-referencing statement, b u t it is n o t exacdy invisible. T h e I R pursues the matter (21-24), explicitiy sanctioning Milford for having never answered my question. Subversive w o r d repeats and anaphoric p r o n o u n s also can b e used in com bination, as i n ex. (40), from a British debate interview concerning a proposal to m a k e abortions m o r e difficult to obtain. T h e excerpt begins with the I R asking Jill Knight, a n outspoken o p p o n e n t of abortion, about o n e aspect of the legisla tion, which would shorten the time period for legal abortions. A key word in the question is concern; it is central to the final question (line 13), which asks about the IE's level of concern regarding the m o r e restricted time frame, a n d it also appears in the preceding statement (12) and the earlier lead-in (3-4), both of which m a k e reference to widespread public "concern" about the n e w restrictions: (40) UK, Afternoon Plus. Abortion
CLAYMAN
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
ANSWERS A N D EVASIONS
45
same. (.) and indeed (.) have caused great deal of concern. (0.4) But fixst you'll note .hhh is the clause about (.) time limits h in which h abortions can be .h legally= (JK): =*(yes)*= IR: =ha:d. And the time limit h (.) according to the bill has now dropped .h from twenty eight weeks .h (.) to twenty wee[ks. (JK): [Yes.= IR: =Now< a lot of people are very concerned about this. [.hh How concerned are you. (JK): [*yeh* JK: -> .hhh Uh: (.) I think this is right. I think that um: .hh again one's had a lot of e:uh conflicting evidence on this but .hh what has come ou::t h an' —> I think that .h the public have been concerned about —> this, .hhh is that there have been th'most distressing cases, .hhh of (.) live (.) kicking babies who have been destroyed, .hh I've had nurses come to me in great distress (0.2) about this .hh and uh there was undoubtedly (0.1) throughout the whole —> (ambit) of public opinion .hh very great concern .h on this whole question....
In her initial response, the I E appears to b e m o v i n g to answer the question straightforwardly. H e r first r e m a r k (/ think this is right) refers to the IR's prior talk and seems to b e expressing some form of confirmation or a g r e e m e n t A n d when she begins to elaborate, she twice uses that same key word, concern (ar rowed). However, this comes to m e a n something very different h e r e than it did originally. She uses concern to m e a n "concern about late term abortions"; but in die original question it m e a n t "concern about the m o r e restricted time frame" and. by unplication, the m o r e restricted access to abortion that this entails. This semantic shift is intertwined with a m o r e encompassing shift in the topical focus of the response vis-a-vis the original question. It is useful to consider the ramifications of a m o r e overt m o d e of resistance. The I E could have said something like " I ' m n o t the least bit concerned about a -horter time frame; what worries m e is the destruction of live and kicking babies!" But that would place her in direct disagreement with the viewpoint e m b e d d e d in the question, and it would m a k e h e r vulnerable to being seen as insensitive to the plight of those seeking abortion services. H e r actual course of action obscures such implications. She presents herself as if she were straightforwardly answering, and agreeably expressing "concern," while surreptitiously veering away from the question in the way in which it has b e e n framed. T h e cover for this m a n e u v e r is provided b y the back-referencing confirmation a n d the lexical repetition, the latter serving as a kind of pivot between the question's agenda a n d the somewhat different direction pursued subseauentlv.
DISCOURSE STUDIES
46
Operating on the Question A g e n d a shifts can b e obscured in other ways. Before "answering" a given ques tion, a n I E m a y first refer to, characterize, or paraphrase the question at hand. T h e s e various operations can modify the question in a w a y that b o t h facilitates and conceals a shift of the agenda. T h u s , n o t only can IEs adjust the surface form of a response to fit the question, they can also, in effect, adjust the question to fit the response that they intend to give. To illustrate, consider ex. (41), an excerpt from a n interview with a presidential candidate, Senator G a r y Hart. T h e 1988 interview was p r o m p t e d b y media reports suggesting that H a r t h a d a n extramarital affair with a y o u n g woman n a m e d D o n n a Rice. A t one point, h e was asked specifically if h e h a d an affair with Miss Rice (arrow 1). I n t h e course of his answer, H a r t reformulates the question (arrow 2), b r o a d e n i n g it so that it is m a d e to concern his marital fidelity over t h e past 2 9 years, including periods during which h e and his wife were publicly k n o w n to h a v e b e e n separated. U p o n completing this reformulation, h e provides a n "answer" (arrow 3), a n admission of infidelity. But the param eters of his admission h a v e b e e n set n o t b y the original question, but by the reformulation. (41) US, Nightlirur. The Best of Nightline IR:
GH:
Uh- (0.5) I told you::. (0.4) some days ago when we §po:ke, and I told our audience this evening that I would ask you bath questions. I will ask you the first now: just before we tak a biea:k because I think I know what your answer's gonna be.= 1 —> =Did you have an affair with Miss Rice? 2 - » . . . .hhhh Mister Koppel (1.1) if the question: (.) is in the twenty nine y:ear:s of my marriage, including two public separations have I been absolutely and totally faithful: to my wife .hhh 3 —» I regret to say the answer is n o : . . . .
T h e advantages of such a transformation should b e obvious. I t enables Hart to a p p e a r "forthcoming," b u t in response to a question that, b y virtue of its gener ality, is m u c h less pointed. His admission is thus less politically damaging than it might otherwise h a v e been. I n effect, H a r t manages to "steer the question" in a m o r e desirable direction. I n this particular example, Hart seems to acknowledge the fact that the question has b e e n modified. Notice that his reformulation is offered tentatively within a n ^ c l a u s e (Mister Koppel, if the question is. . .). This case is thus comparatively overt in the w a y in which it shifts the agenda. O t h e r question reformulations are asserted m o r e forcefully, as if they were faithfully preserving the essence of the original question. Ex. (42) comes from the 1988 vice presidential debate; although n o t officially labeled a "news inter view," this event was organizationally similar in having the candidates respond *«m-.«oj-;rtno f.-/-vTTn „ ninol «-»f i m i i m n l i c f r o T n fTiia firsf- n i i p s f i n n to Dan Ouavlfi. a n
CLAYMAN
ANSWERS AND EVASIONS
47
critical of Bush's decision t o choose Q u a y l e as his r u n n i n g m a t e (lines 4-16); the journalist then asks Q u a y l e w h y h e h a s n ' t m a d e a more substantial impression on his own Republican colleagues (16-19). Q u a y l e begins his response (21-24) by reformulating the question i n t e r m s of his general qualifications for t h e presidency: (42) US, 5 Oct. 1988, Bentsen-Quayle Debate 1 2 3 4 5 f> 7 8 9 10 11 12 13 14 15 16 17 18 19 20 2) 22 23 24 25 26 '27 2«
JRN:
hhhh Senator you have been criticized as we all know:: for your decision to stay out of the Vietnam war::, (0.3) for your poor academic record, .hhhhhh but mpjre troubling to so::me are some o'thuh comments that've been made by people in your own party, tch .hhh Just last week former Secretary of State Hai::g. .hh said that your pi:ck. (0.2) was thuh dumbest call George Bush could've rna[:dc. AUD: [h-h-hhxhxhx[hxxXXXXXXXXXXXX=] JRN: [Your leader in the Senate] AUD: = XXXXXXXXXXXXXXXXXX[XXXXXXXxxxxxxx(5.8)] JRN: [Your leader in the Senate ] Bob D_p_:le said that a Jaetter qualified person could have been chosen, .hhh Other Republicans have been far more critical in private, .hhhh Why d'you think that you have not made a more substantial impression on some of these people who have been able to observe you up clo:se. (1.5) DOj -> .hhhhhh The question goe::s (1.0) to whether I'm qualified (l.l) to be vice president, (0.8) .hhh and in the case of a:: (.) tragedy whether I'm qualified to be president. (0.6) .hhhh (0.7) Qualifications for:: (0.2) the office of vice president 'r president (1.0) are not age alo:ne. (1.5) you must look at accomplishments: (1.0) and you must look at experience....
This is a substantial transformation. O n o n e level, it m o v e s from subjective impressions of Quayle - which m a y b e difficult to explain o r refute - to his qualifications considered as a n objective matter. T h e r e is also a change i n the presup positional loading of t h e question. T h e original question is p r e s u p positionally negative: In b o t h the preface a n d the wording of the question itself, it presumes that Quayle did not i n fact m a k e a g o o d impression a n d asks w h y this was so. In contrast, the reformulated version is presumptively neutral [whether I'm qualified) and thus facilitates a m o r e u p b e a t response. Despite the magnitude of this transformation, it is asserted affirmatively a n d without qualification (The qutstion goes to . . .). Quayle thus oroDoses that h i s r»fnrnr.iiia*i<-»r, o , , , . „ —
DISCOURSE STUDIES
48
To appreciate the significance of this practice for m a n a g i n g a n agenda shift, it might b e useful to consider what the preceding exchange would look like without a reformulation: (43) [Invented] JRN: . . . Why do you think that you have not made a more substantial impression on some of these people who have been able to observe you up close? DQj Qualifications for the office of vice president or president are not age alone. You must look at accomplishments and you must look at experience. W h e n t h e "answer" is m a d e t o follow t h e question without a n y preparatory work, it is manifesdy disjunctive. Against this b a c k d r o p , the importance of the reformulation is that it affiliates t h e matter-to-be-pursued with t h e matter-thaiwas-inquired-about, thereby minimizing the discrepancy between the two. In effect, the reformulation provides a version of the question that the subsequent response can b e seen as "answering." I n ex. (42), t h e I E operates o n t h e question as a whole object. But IEs also m a y operate o n a c o m p o n e n t of the question - a phrase, a prefatory statement, or o n e p a r t of a multi-part question. I n ex. (44), from a N i x o n press conference during the Watergate period, N i x o n targets t h e first part of a two-part question for reformulation. T h e journalist first asks (beginning at arrow 1) whether Nixon is personally investigating charges that his campaign funds were mishandled, a n d h e then asks (arrow 2) whether the charges will h u r t his bid for reelection. After a prefatory remark, N i x o n produces a reformulation (arrow 3) that high lights the first part of the question: (44) US, 29 August 1972, Nixon Press Conference JRN: 1 —» Mr. President, are you personally investigating the mishandling of some of your campaign funds, 2 —> and do you agree with Secretary Connolly that these charges are harmful to your re-election? RN: Well, I commented upon this on other occasions, and I will repeat my position now. 3 -» With regard to the matter of the handling of campaign funds, we have a new law here in which technical violations have occurred and are occurring, apparently, on both sides. As far as we are concerned, we have in charge, in Secretary Stans, a man who is an honest man and one who is very meticulous - as I have learned from having him as my treasurer and finance chairman in two previous campaigns - in the handling of matters of this sort. Whatever technical violations have occurred, certainly he will correct •i
... j
<.i—-„.,„P.I„ cnmnlv with the law. He
CLAYMAN ANSWERS AND EVASIONS
49
conducting it very, very thoroughly, because he doesn't want any evidence at all to be outstanding, indicating that we have not complied with the law. This reformulation is subversive in two respects. First, it replaces a key term from the question that implied w r o n g d o i n g (mishandling) with a m o r e favorable t e r m (handling). Furthermore, the r e m a i n d e r of the response deals exclusively with the matter of the investigation; N i x o n never gets around to the second p a r t of the question, regarding the consequences for his reelection campaign. This omis sion may not h a v e b e e n accidental. Dealing with the investigation question enables Nixon to present himself as "doing something" about a scandal within his administration, and thus i n d e p e n d e n t of the morally tainted forces which brought ir about. I n contrast, the issue of whether the scandal will hurt his cam paign seems, at least from his standpoint, less advantageous. Although this omission is clear in retrospect, it was not evident at the outset that vime form of evasion was in progress. Nixon could have gone o n to answer the second question. Indeed, it is standard practice for IEs, w h e n "reaching back" to deal with something other t h a n the last c o m p o n e n t of the questioning turn, to indicate as much b y referring to or reformulating that aspect of the question (Clavman 1993). This is because IEs normally address the final c o m p o n e n t of the question (Sacks 1987); conversely, they take steps to w a r n listeners w h e n an atypical response trajectory is in the offing. T h u s , Nixon's operation was initially accountable as an effort to m a n a g e an atypical response trajectory rather t h a n to avoid the second question. There are still more subde variations on the practice of operating on the ques tion. In the cases examined thus far, the operation is exposed within a discrete unit of talk, but it also m a y b e e m b e d d e d within some other activity - assertions of agreement or disagreement, for example. Ex. (45) shows how, in the course of claiming to a g r e e / d i s a g r e e w i t h s o m e aspect of the q u e s t i o n , a n I E c a n embeddedly reformulate that question (arrowed): (45) US. 22 July 1985, MacNeil/Lehrer: South Africa 1 JW: 2 3 A 5 rt FW: —> 7 —» 8 9 10 for I.) 12 13
But isn't this (.) d. declaration of the state of emergency:: (.) an admission that the eh=So_uth African government's policies have not worked, an' in fact that the um- United States (0.2) administration's policy of constructive engagement (.) has not worked. I do not agree with you .hhhh that the approach we have taken (.) toward South Africa is- ay- is an incorrect approach. .hhhhh We. want (0.5) to see that s- system change. We wanta see South Africa end apartheid. We wanta see basicrightsestablished all South Africans, .hhhh We wanta see peace and stability in that country, .hhh An' that's a PERfectly respectable goal. Second. The way we have pursued it .hhh I also believe .hhh is the most SENsible way: in dealing with a
DISCOURSE STUDIES
50
H e r e , the I E first asserts disagreement (/ do not agree with you) and then char acterizes the object of his disagreement (that the approach we have taken...) in a m a n n e r that transforms the terms of the question. T h e transformation is subtie yet advantageous for the IE, who was then a U.S State Department official during the Reagan administration and is h e r e defending Reagan's policy of "constructive e n g a g e m e n t " T h e original question asked (after some preliminary talk) whether the U.S. policy of constructive engagement has not worked. This is reformulated in the statement of disagreement as a question about whether U.S. policy is an incorrect approach. T h e latter version is very m u c h easier for the I E to refute. It is difficult to argue with the original assertion that U.S. policy has not worked, since at the time of the interview apartheid r e m a i n e d intact. But o n e can assert the overall correctness of U.S. policy even in the face of its manifest failure to bring about an e n d to apartheid, a n d this is precisely what the I E does in his subse quent response.
Two Case Studies H a v i n g analyzed various practices for managing resistance as they are employed across a wide range of interview circumstances, it is n o w time to apply these analytic resources to some singular cases. O n e objective of these case studies is to illustrate the p o w e r of the analytic apparatus developed thus far to elucidate just h o w particular noteworthy I E s are able to elude the grasp of an advancing line of questioning. Moreover, while a systematic analysis of audience reactions is b e y o n d the scope of this p a p e r , these case studies are suggestive of the impact t h a t s u c h p r a c t i c e s c a n h a v e o n s u b s e q u e n t m e d i a c o m m e n t a r y and on pubic opinion. Dan Quayle and the Succession Question D u r i n g the 1988 U.S. presidential campaign, the two m a i n vice presidential candidates, senators Lloyd Bentsen and D a n Quayle, squared off in a nationally televised debate. T h e format h a d the candidates responding to questions from.. panel of four journalists, making it n o t unlike a multi-IR interview or small press conference. Opportunities to follow u p a n d pursue evasive answers wert m o r e Hmited t h a n in ordinary news interviews, because h e r e each journalisi could ask only o n e question at a time, a n d the order of questioners was predeter m i n e d . Nevertheless, processes of resistance a n d pursuit were both very much in play. A n extended tug-of-war developed a r o u n d the issue of presidential succr sion. It began w h e n D a n Quayle - a youthful senator a n d George Bush's runrm. m a t e - was asked what h e would do if the president died or b e c a m e incapacitat I for some reason. T h e p u r p o s e of this question was to test Quayle's readiness! : assuming the presidency in a n emergency: W h a t would b e his plan of actioii; J
CLAYMAN ANSWERS AND EVASIONS
51
The first journalist to raise this question was Brit H u m e of A B C News. Noting the apprehensions p e o p l e m i g h t feel about Q u a y l e b e i n g a heartbeat away from the presidency (ex. [46], lines 1-4), h e asks Q u a y l e to describe, in the event of his sudden succession to power, the first steps that you'd take and why (5-10): t46) US, 5 Oct. 1988, Bentsen- Quayle Debate 1 JRN: Senator I wan-1 wanna take you hack if I can 2 to the question Judy as:-asked you about some 3 o'the apprehensions people may feel about your 4 being a heartbeat away from the presidency, .hhhh 5 And let us assume if we can for the sake of this 6 question that you become vice president- an:d the 7 president is incapacitated for one reason or 8 another and you hafta take the reins of power. 9 .hhhh When that moment a::me, w- what would be 10 thuh first steps that you'd take (0.2) and why::. 11 (3.2) 12 DQ: .hh First I'd- first I'd say a prayer (1.1) tch for 13 myself (2.3) and for thuh country I'm about to 14 lead, (2.4) And then I would (1.1) assemble his 15 (1.1) people and talk (0.8) .hhh 16 -> And I think this question keeps going bjjxk to: (1.0) 17 the qualifications and what kind of (1.1) of vice (8 president 'n (0.7) in this hypothetical situation (9 (1.0) if I had to assume:: (0.8) thuh responsibilities 20 of: (0.3) president what I would be. (1.0) .hhh 21 And as I have said (1.2) tch age akv.ne. (0.3) .hh 22 although I can tgll you h.h after the experience of: 23 these last few weeks 'n the campaign I've added 24 ten years to my a[ge, 25 AUD: [x-x-x-x-x-x[-x (1.7) 26 [Age alone. (1.0) 27 is not (0.2) the only (0.5) qualification, .hhhh 28 You've got to l&ok at experience. (.) And you've 29 got to look at accomplishments.... Quayle makes an initial stab at answering t h e question (lines 12-15), b u t it is rather half-hearted and insubstantial. H e says only that h e ' d say a prayer a n d would assemble his people and talk. H e then proceeds to reformulate the question (arrowed, lines 16-20), veering a w a y from the issue of his plan of action for assuming the presidency in an e m e r g e n c y and toward the m o r e general issue of his overall qualifications for the presidency. H e then goes o n to discuss his quali fications at length, ruling out age as a qualification a n d focusing o n experience and accomplishments. In t h e e n d , talk a b o u t qualifications d o m i n a t e s his response, only the first part of w h i c h is r e p r o d u c e d h e r e . Although this shift is m a n a g e d covertly a n d is obscured b y the use of a n fcjUtiii Question r p f n r n r m l o * - ; ^
DISCOURSE STUDIES
52
I n ex. (47), after a full r o u n d of questioning from t h e other panelists, H u m e regains the floor a n d pointedly pursues the question (line 6). Before doing so, however, h e carefully justifies this m o v e b y calling attention to t h e inadequacy of Q u a y l e ' s previous response, summarizing it in a w a y that highlights its feebleness: You said you'd say a prayer, and you said something about a meeting (4-5). H e then presses Q u a y l e to elaborate (6). Notice that some audience m e m b e r s in the hall begin to laugh at this point (8), displaying appreciation of H u m e ' s derisive c o m m e n t a r y a n d aligning with h i m in his pursuit of a n answer. (47) US, 5 Oct. 1988, Bentsen-Quayle Debate 1 JRN: 2 3 4 5 6 7 8 AUD: 9 DQj 10 11 12 -» 13 14 15 16 17
Senator I wanna take you back to the question that I asked you earlier about what would happen if you were to: take over in an emergency and what you would do first and why:: .hhhh You said you'd say a prayer:: and you said something about a meeting, (.) What would you do next. (•) h-hhhhhhhhhhhhhhhhhhhhhhhhhhhhhh[hh-h-((laughter)) [I don't believe that it's (0.6) proper for me:: to: .hh get into the specifies: (0.5) of a hypothetical (.) situation like tha:t (1.2) The situation is: (0.8) that if (0.8) I was called upon (0.7) to senve (0.7) as the president (0.4) of this country, or the responsibilities of the president of this country, (1.0) would I be capable and qualified (0.2) to do that. (0.5) .hh and I've tried (0.4) to list the qualifications. (1.0) of twehve yeans in the United States Congress
But that answer remains elusive, as Quayle again sidesteps the question, although his method of doing so here is rather different. Given that his prior covert maneuver has b e e n exposed b y H u m e ' s pursuit, Quayle n o w chooses a m o r e overt mode of resistance. H e explicitly justifies his failure to provide a m o r e substantia) answer b y characterizing t h e focus of inquiry as a hypothetical situation and suggesting that it would b e improper to answer in specifics (lines 9-11). H e then shifts t h e agenda (12-15) i n precisely the same direction as before - away from his p l a n of action a n d toward his overall qualifications for t h e presidency. Thus, while Quayle's resistance is n o w overt a n d o n record, it is also justified and accounted for. Unfortunately for Quayle, justificatory accounts d o n o t necessarily bring the line of questioning to a halt; such accounts can b e argued with a n d contested This is what the very next questioner o n the panel - Tom Brokaw of N B C News - does, relinquishing whatever question h e h a d p l a n n e d to asked in order to pursue t h e succession question yet again. I n ex. (48), Brokaw begins with a dis claimer to the effect that that h e doesn't m e a n to beat this drum until it has no mon sound left in it (lines 1-2). H e then takes issue with Quayle's account for not ani x
— - » « * r t m f it u after all the reasw
CLAYMAN
ANSWERS AND EVASIONS
53
question but with a pointed assertion (12-16) that surely you must have some plan in mind for assuming the presidency in an emergency, since it has h a p p e n e d to so many vice presidentsin recent years. By rejecting Quayle's previous account for not answering, and b y pressing the issue in a m o r e pointed way, Brokaw has increased the pressure for a genuine response. (48) US, 5 Oct. 1988, Bentsen-Quayle Debate 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34
JRN:
Senator Quayle I don't mean to beat this drum until it has no more sound left in it but to follow up on Brit Hume's question w:hen you said that it was a hypothetical situation, .hhhh it is Sir after: all: the reason that we're here tonight. .hh [h because you are= DOj [Mhm JRN: — [running [not just for vice president,] AUD: [ x x [x-x-x-xxxxxx= ] = xxxxxxxxxxxxxxxxxxxxxx [xx-x-x-x-x (4.4) ] JRN: [And if you cite the] experience that you had in Cjjngress, (0.2) surely you must have some plan in mind about what you would do: if it fell to you to become >piesident of the United States< as it ha:s to so. many vice presidents .hh just in the last twenty five years er so. (0.3) DQj teh .hhh Lemme try to answer the question one more ti:me. I think this is the fourth ti:me, (1.0) [ that I have had this question, .h [and I think= JRN: [(this is-) [Third time DOj =that-.hh three times, (0.8) that I have had this question, and I'll try to answer it again for you. (0.3) as clearly as I can. (0.7) .hh Because 1 —> the ougstion you're asking. (1.3) is what (.) kind (.) of qualifications .hhhhhh does Dan Ouayle have to be president. (1.0) tch 2 —» What kind of qualifications do I have 3 —» and what would I do: (1.0) in this kind of a situation. (0.4) what would I do in this situation, .hh I would (1.9) make sure. (2.1) that the people in the cabinet, (0.9) 'n the people 'n the advisors to the President, (.) are called in, (0.2) an I'll talk to 'em, (0.5) an I'll work with ' e m . . . .
After commenting o n the n u m b e r of times he's h a d this question (lines 19-22), Quayle promises to try to answer it again for you as clearly as I can (23-24). Quayle then does something that is very puzzling o n its face. H e launches into yet another question reformulation (beginning at arrow 1) that begins to reframe the issue once again as a matter of qualifications. However, in the course of this reformu-
DISCOURSE STUDIES
54
which h e subsequently elaborates (30-34). W h y does Q u a y l e start to veer away from the agenda of the question, only to return to it subsequently? T h e solution to this puzzle lies at the nonvocal level. After Q u a y l e launches into his reformulation (at line 3 in ex. [49]) a n d completes the focal word qualifications (4), Brokaw begins shaking his h e a d (6), a n d h e continues to do so until the reformulation reaches a first possible completion point. I n this way, Brokaw nonvocally rejects Quayle's b i d to shift the agenda. (49) US, 5 Oct. 1988, Bentsen-Quayle Debate 1 DOj . . . and I'll try to answer it again for you. 2 (0.3) as clearly as I cj,n. (0.7) .hh Because 3 the auegtion you're asking. (1.3) is what (.) 4 kind (.) of qualifications 5 .hh [hhhh does Dan Quayle have to be president. (0.5)] (0.5) 6 JRN: [((Headshaking)) ] 7 DQ: tch What kind of qualifications do I have 8 and what would I do: (1.0) in this kind of a situation. (0.4)] 9 JRN: [((Nodding))] 10 DQ: And what would I do in this situation, .hh I would (1.9) 11 make sure. (2.1) that the people in the cabinet, (0.9) 12 'n the people 'n the advisors to the President, (.) are 13 called in, (0.2) an I'll talk to ' e m . . . This rejection is consequential; Q u a y l e subsequently a b a n d o n s the incipient agenda shift and returns to the original agenda (lines 7-8). Brokaw n o d s approv ingly (9), and Quayle proceeds to elaborate on his emergency plan. T h u s , while Quayle initially steers the question in a different direction, Brokaw steers hini right back. However, Quayle's return to the original a g e n d a is m a d e to appear as if it is unrelated to what Brokaw has done. Notice that Quayle does not respond imme diately to the headshakes; h e continues to speak through the headshaking until the reformulation is possibly complete (line 5). H e also places s o m e distance between the completion of the headshakes and the start of his continuation, i allowing one full second of silence to elapse, a n d then backtracking a bit when ' h e continues (7). Furthermore, w h e n h e finally gets to the plan of action com p o n e n t of the reformulation, h e links it to the previous c o m p o n e n t with and: a is thus introduced as a supplementary rather than a contrastive matter. By these various means, Q u a y l e constructs his reformulation so that it can b e seen as e single continuous action rather than an "about-face" in response to Brokaw prompting. I n other words, h e presents himself as if h e h a d b e e n headed in thi> direction all along. I n the aftermath of this debate, most observers declared Lloyd Bentsen the I decisive winner, and extensive m e d i a commentary focused o n Quayle's perform- I ance a n d its shortcomings. A c o m m o n criticism was that h e came across as I overly "rehearsed" or "programmed" in his remarks. This widespread impression I 1
- -
il
„L»,uf
I
CLAYMAN
ANSWERS AND EVASIONS
55
experience - as a favored r e s p o n s e to various kinds of questions (see also ex. [42]). Moreover, h e repeatedly uses the same basic practice - which w e have termed "operating o n the question" - to fit that favored response to the question ai hand. This recurrent m o d e of resistance is first m a n a g e d covertly, but it is subsequendy exposed b y persistent follow-up questions from the panel of journal ists, at which point it b e c o m e s transparently manipulative. H e r e , then, patterns of response within the event are congruent with a n d thus appear to h a v e b e e n consequential for subsequent m e d i a commentary. The Affairs of Bill Clinton Perhaps more than for a n y other A m e r i c a n president, Bill Clinton's conduct in answering questions - n o t only in n e w s interviews a n d press conferences, but in courtroom depositions as well - h a s h a d clear effects o n his political fortunes and public image. H e is a notoriously skilled interrogatee, a d e p t at turning questiuns to his advantage while appearing to b e dutifully responsive. However, w h e n these practices have b e e n exposed as strategies of evasion, the negative repercus sions have been substantial. T h e following analysis focuses on questions regarding the delicate subject of extramarital affairs. Early in the 1992 presidential campaign, allegations surfaced about an ex tended affair between Clinton a n d Gennifer Flowers. These allegations emerged just as Clinton was b r e a k i n g from the p a c k of Democratic candidates to b e c o m e th e front-runner in the primary campaign, placing his buoyant candidacy in serious jeopardy. In an effort to confront the issue a n d p u t it to rest, b o t h Bill and Hillary Clinton a p p e a r e d o n the 60Minutes p r o g r a m . T h a t interview has b e e n called one of the great performances in A m e r i c a n presidential politics, a n d it was widely credited with rescuing t h e Clinton candidacy. Many factors u n d o u b t e d l y contributed to this outcome, b u t at least part of the success can be attributed to the m a n n e r in which Governor Clinton dealt with the core questions concerning his relationship with Gennifer Flowers. Although he admits in a general w a y to having h a d " p r o b l e m s " a n d "difficulties" in his marriage, specific questions about the alleged affair with Flowers are m e t with what seem at first glance to b e straightforward denials. However, o n analysis it becomes apparent each response falls at least a hair's-breadth shy of a full-fledged denial. Within the framework developed in this article, Clinton's responses are covertly resistant and extremely subtle in the m a n n e r in which they elude the agenda of the question. Consider the first question of this kind, in which the I R , Steve Kroft, raises Flowers's claim of a twelve-year affair with Clinton (lines 1-3): •Bm (50) US. Jan. 1992 60 Minutes: The Clintons 1 IR: 2 ^^^•v 3
She's alleging (0.2) and has described in some detail in thuh super market tabloid .hh what she calls a twelve year affair with you.
DISCOURSE STUDIES
56
Clinton responds to this allegation (line 5) with a simple assertion to the effect that it is false. This assertion is squarely on-topic, but it is nonetheless only minimally responsive. As demonstrated earlier, given that elaborated answers are the norm in the news-interview context, m i n i m a l one-sentence r e s p o n s e s are tacitiy resistant to the agenda of the question. I n the present case, Clinton's unelaborated denial - That allegation isfalse - is n o t particularly informative about his relation ship with Flowers. Is h e denying any extramarital affair whatsoever? O r is he merely denying an affair that lasted twelve years? By responding minimally, Clinton is able to issue a denial in a way that avoids specificity regarding what, exacdy, is being denied. T h e I R notices the ambiguity in Clinton's denial, a n d h e pursues the question in a w a y that seeks to resolve it (lines 1-4). H e tries to pin Clinton down to an absolute denial, reformulating the prior response as categorically denying that you ever had an affair with Gennifer Flowery. (51) USJan. 1992, 60 Minutes. The Clintons 1 2 3 4 5 6 7 8 9 10 11 12
IR:
I'm assuming from your answer (0.4) that you're (.) categorically denying (.) thafc=you ever had an affair. (1.0) IR: with Gennifer Flowers. BC: .hh I said that before. (.) .hh °uh° An' so has she. HC: (hmhhmhhmh) (.) BC: hh=huh Wh(h)en £these st(h)ories came out£ (0.5) shg: an' thee other people invghved, (0.2) nh: denied them, (.) An' denied them (.) repeatedly, (.) An' she changed her story when she was paid.
O n c e again, Clinton seems at first to b e cooperating with the agenda of tbf question b y confirming this version of his denial. However, h e avoids a straight forward "yes" or "that's right," asserting instead that h e h a d already denied the affair o n some prior occasion (line 5), an occasion which remains unspecified in his response. H e t h e n proceeds to talk at greater length about Flowers's own previous denials of the affair (6-11), a n d h e suggests that the recent change in her story was motivated b y the m o n e y she received from the tabloid that first published it (12). Thus, while h e fosters the impression of having categorically denied a n affair, o n closer analysis it becomes apparent that h e never quite do« so in the here-and-now, o n his o w n accord. H e adopts a n interactional footing in which he is merely relaying denials previously expressed b y himself and othen This interview went over well at the time, a n d Clinton would go on to win both the nomination and the election decisively. But his performance would eventually c o m e back to haunt him. Early in 1998, w h e n Clinton was called to give a deposition in the PaulaJ o n e s lawsuit, h e admitted u n d e r questioning that, he did i n d e e d h a v e a n affair with Flowers. Shortly thereafter, the original I
CLAYMAN ANSWERS A N D EVASIONS
57
Around the same time, allegations about another affair surfaced, this time involving a young W h i t e H o u s e intern n a m e d M o n i c a Lewinsky. Shortly thereiuter, Clinton was questioned about the affair in a n interview o n NewsHour. O n c e again, what initially appear to b e forthright denials of the affair are in fact coverdy resistant. In this case, the specific m o d e of resistance is a subtle shift in the terms of the question, a shift involving the verbal tense in which it is expressed: (52) US, 21 Jan. 1998, NewsHour. Clinton ( 2 3
IR: You had QO sexual relationship with this [young wo [man.] IE: [ml [Th-] IE: There is not a sexual relationship. That is accurate.
The IR asks Clinton to confirm that h e had no sexual relationship with this young woman. Clinton eventually issues a confirmation (That is accurate), b u t only after reformulating the issue from past to present tense (There is not a sexual rela tionship). This response does n o t necessarily rule out a n affair that is over and done with, but to the casual listener it might seem that Clinton has denied an affair altogether. This type of tense shift would later b e c o m e notorious w h e n it was exposed as a strategy e m p l o y e d b y Clinton in the Paula J o n e s deposition. The Clinton case is a powerful illustration of the distinctive attractions a n d risks associated with a covert m o d e of resistance. T h r o u g h o u t these examples, Clinton never owns u p to the fact that h e is not answering the question fully or straightforwardly; the resistance remains unacknowledged and extremely subtle. This approach enabled Clinton to survive various difficult m o m e n t s in his original presidential campaign and his tenure in office. But because some of these responses have been exposed as evasive through subsequent events, h e has paid a price over the long term in the form of damage to his political a n d personal reputation. The indictment against h i m is n o t m e r e l y that h e sidesteps questions - m a n y politicians are guilty of that - but that h e is deceitful about it. T h e so-called "Slick Willie" factor should stand as a cautionary note to public figures contemplating a coven mode of resistance.
Discussion In this age of political cynicism coupled with anxiety about the decline of civility in public life, it is tempting to assume that virtual anomie n o w characterizes the domain of public discourse. Politicians, in such a world, would n o longer b e b o u n d by traditional norms a n d could thus ignore with impunity the questions they receive in journalistic interviews a n d press conferences. T h e practices examined in this article, taken together, reveal a m o r e complex state of affairs. Although it is true that resistant and evasive responses are commonplace, these are managed with considerable care. W h e n resistance is d o n e overfly, interviewees take steps to control the damage that m a y b e caused thereby. W h e n it is done covertly, there are corresnondinc pffnrrs tn m n r a o i *L*> ..„„:..<—• • 1
58
D I S C O U R S E STUDIES
consequences that can follow from the breach of conduct e m b o d i e d in an act of evasion, a n d they demonstrate that such a n act continues to b e regarded AS a breach b y those involved. O n the other h a n d , b y virtue of these practices, politicians a n d other public figures can indeed gain substantial "wiggle r o o m " for pursuing their own agendas even u n d e r the most persistent interrogation. It is possible that such practices have evolved over time in relation to the changing culture of the journalistic profession. I n b o t h England a n d the U n i t e d States, journalistic questioning has become less deferential and more adversarial since the 1950s ( d a y m a n & Heritage in press a, b). This could p r o m p t interviewees to b e c o m e m o r e resistant and/or to develop m o r e sophisticated m e t h o d s for dealing with difficult questions, jusi as t h e y h a v e developed other strategies for managing adversarial encounters with journalists (seejones 1992). Alternatively, increasingly adversarial question ing could h a v e precisely the opposite effect: Insofar as adversarialness include* a greater propensity to ask follow-up questions that pursue evasive response* (Clayman a n d Heritage in press a), it could encourage interviewees to adhere m o r e closely to the question agenda. Whether there actually has b e e n a systematic change in public figures' responsive conduct - in either the propensity toward resistance or in the practices for managing resistance - must await further research. Finally, this analysis has methodological ramifications for efforts to quantif) responsive conduct as a prelude to examining its distribution across individual politicians, journalistic environments, or historical time periods. It should h abundantly clear that responses cannot properly b e dichotomized as either "an> wers" or "evasions," for between these black-and-white categories are numerou shades of gray comprised of varying m o d e s a n d degrees of resistance (cf. Bui 1994). Moreover, these variations m a y h a v e nontrivial consequences for how th public forms j u d g m e n t s about a politician's m o r a l character. To take just on example, a skillfully m a n a g e d covert approach could enable a politician to gr through a hostile interrogation without any negative repercussions. Alternative! as the case of Bill Clinton aptly demonstrates, if covert evasions are exposed |v subsequent events, the politician m a y b e seen in retrospect as devious and man pulative, and this outcome m a y b e worse for the politician than if h e or she hf initially acted overfly with appropriate forms of damage control. According! quantitative and distributional studies would do well to p a y m u c h closer attenti to the specific practices through which interviewees deal with questions.
• .
.
Notes
Portions of this article were presented at the 1996 conference of the American Sociological Association. The author would like to thank John Heritage, Manny Schegloff, and the anonymous reviewers for their critical input. 1. This incident was discussed in a documentary film by Brian Springer entided $1*. 2. This professional skill is often thematized in broadcast journalists' autobiographia
CLAYMAN ANSWERS AND EVASIONS
59
Mr. President! (Donaldson 1987). Both tides are common prefaces to hostile follow-up questions, and they evoke images of an aggressive interviewer doggedly pursuing his elusive prey. 3. This kind of full-form repetition, where the entire framework of the question is preserved in the initial response, appears disproportionately in certain environments. It is often found in exchanges that (i) are highly argumentative, or (ii) involve a re sponse that ultimately departs from the agenda of the question. It may be that the practice of repeating the framework of the question, rather than accepting it and building on it, conveys a modicum of independence from the prior course of ques tioning, and is thus useful in resistant and hostile environments. •t. Notice that the IE refers to the IR by name ("Mister Day") as he launches his resistant response. It turns out that this is a highly recurrent practice across various forms of resistance. It appears to be related to the pragmatics of address term usage in inter action generally: Speakers often address their recipients by name when expressing deeply fell opinions and personal feelings (dayman 1998), particularly when such opinions or feelings are oppositional in character. 5. Such multi-part questions are more common in press conferences than in news interviews. 6. This excerpt is from a taped interview segment on 60 Minutes, and it appears to have l>Ken edited in such a way as to highlight and sharpen the tug-of-war between the IR and Deutch. 7. Similar requests for permission and token requests are used by IEs when speaking out "i turn (Greatbatch 1988). Such practices generally serve to mitigate departures from the normative turn-taking arrangements in the news interview, and agenda shifts represent a specific type of departure that is mitigated in this way. 8. The array of refusal accounts discussed here may not be equally available to all IEs in all circumstances.
References Bavelas, Janet B., et al. (1988) Political equivocation: A situational explanation. Journal of language and Social Psychology 7:137-46. Bull, Peter (1994). On identifying questions, replies, and non-replies in political interviews. Journal ofLanguage and Social Psychology 13:115-31. — (1998). Equivocation theory and news interviews. Journal of Language and Social Psychology 17:36-51. —Judy Elliott, Derrol Palmer, & Libby Walker (1996). Why politicians are three-faced: The face model of political interviews. BritishJournal ofSocial Psychology 35:267-84. — , & Mayer, Kate (1993). How not to answer questions in political interviews. Political Psychology 14:651-66. dayman. Steven E. (1990). From talk to text: Newspaper accounts of reporter-source interactions. Media, Culture and Society 12:79-104. — (1993'.. Reformulating the question: A device for answering/not answering questions in news interviews and press conferences. Text 13:159-88. — (I99H . Some uses of address terms. Paper presented at the annual meeting of the National Communication Association, New York City, November 1998.
60
DISCOURSE STUDIES
Clayman, Steven E., & Heritage,John (in press b). The News Interview:Journalists and public figures on the air. Cambridge: Cambridge University Press. Davidson,Judy (1984). Subsequent versions of invitations, offers, requests, and proposals dealing with potential or actual rejection. In J. M. Atkinson and J. Heritage (eds.), Structures ofSocial Action, 102-28. Cambridge: Cambridge University Press. Day, Robin (1993).... But with respect London: Weidenfeld and Nicolson. Donaldson, Sam (1987). Hold on, Mr. President. New York: Random House. Greatbatch, David (1986a). Some standard uses of supplementary questions in news interviews. In J. Wilson and B. Crow (eds.), Belfast working papers in language and linguistics, Vol. 8, 86-123.Jordanstown: University of Ulster. (1986b). Aspects of topical organisation in news interviews: The use of agenda shifting procedures by interviewees. Media, Culture and Society 8:441-55. (1988). A turn-taking system for British news interviews. Language in Society 17: 401-30. Harris, Sandra (1991). Evasive action: How politicians respond to questions in political interviews. In P. Scannell (ed.), Broadcast Talk 76-99. London: Sage. Heritage, John (1984). Garfinkel and ethnomethodology. Cambridge: Polity. (1985). "Analyzing news interviews: Aspects of the production of talk for an over hearing audience. In T.A. van Dijk (ed.), Handbook ofdiscourse analysis, vol 3, Discoun and dialogue, 95-119. New York: Academic Press. (in press). Designing questions and setting agendas in the news interview. In J. Mandelbaum, P. Glenn, & C. LeBaron (eds.), Unearthing the taken-for-granted: Stud in language and social Interaction. Mahwah, NJ: Erlbaum. , & Greatbatch, David (1991). On the institutional character of institutional talk: The case of news interviews. In D. Boden and D. H. Zimmerman (eds.), Talk and social structure, 93-137. Berkeley: University of California Press. Jones, Bill (1992). Broadcasters, politicians, and the political interview. In BillJones and Lynton Robins (eds.), Two decades in Britishpolitics, 53-77. Manchester, UK: Manchester University Press. Koppel, Ted, & Gibson, Kyle (1996). Nightline: History in the making and the making t; television. New York: Random House. Pomerantz, Anita (1984). Agreeing and disagreeing with assessments: Some features w preferred/dispreferred turn shapes. InJ. M. Atkinson andJ. Heritaage (eds.), Struaum of social action: Studies in conversation analysis, 57-101. Cambridge: Cambridge University Press. Raymond, Geoffrey (1998). The structure of responding: Conforming and nonconforming responses to yes/no type interrogative. Paper presented at annual meeting of the National Communication Association, New York, November 1998. Roth, Andrew (1996). Turn-final word repeats as a device for 'doing answering' in ai institutional setting. Paper presented at annual meeting of the Pacific Sociological Association, Seattle, WA, April 1996. Sacks, Harvey (1987 [1973]). On the preferences for agreement and contiguity in sequen ces in conversation. In G. Button and J.R.E. Lee (eds.), Talk and Social Organisatin. 54-69. Clevedon, England: Multilingual Matters. Schegloff, Emanuel A. (1968). Sequencing in conversational openings. America Anthropologist 70:1075-95. (1972). Notes on a conversational practice: Formulating place. In D. Sudnow (ed.!. Studies in Social Interaction, 75-119. New York: Free Press.
CLAYMAN ANSWERS AND EVASIONS
61
Schegloff, Emanuel A. (1998). Word repeat as a practice for ending. Paper presented at annual meeting of the National Communication Association, New York, November 1998. Whalen, Jack; Zimmerman, Don H.; & Whalen, Marilyn R. (1988). When words fail: A single case analysis. Social Problems 35:335-62.
18 Author Identification, Idiolect, and Linguistic Uniqueness Malcolm Coulthard
Introduction
[
t is n o w over thirty-five years since J a n Svartvik p u b l i s h e d The Evans Statements: A Case for Forensic Linguistics (1968), in which h e demonstrated that mcrinnnating parts of a set of four linked statements - purportedly dictated to police officers b y T i m o t h y Evans a n d incriminating h i m in the killing of his wife a n d b a b y daughter - h a d a grammatical style measurably different from that of uncontested parts of the statements. It was later discovered thai b o t h victims h a d actually b e e n m u r d e r e d b y Evans' landlord, J o h n Christie. This m a r k e d the birth of a n e w discipline - the linguistic investigation of author ship for forensic purposes. Little m o r e h a p p e n e d for a quarter of a century, with the notable exception of w o r k b y R o g e r Shuy in the U S A (1993, 1998), but during the past ten years, there has b e e n a rapid growth in the frequency with which lawyers and courts in a n u m b e r of countries h a v e called u p o n the expertise of linguists in cases of disputed authorship. T h e texts e x a m i n e d range from questioned suicide notes, through anonymous letters, mobile p h o n e text messages a n d contemporaneous police records of b o t h interviews a n d confession state ments, to the essays of students suspected of plagiarism.
I d i o l e c t a n d U n i q u e n e s s of E n c o d i n g T h e linguist approaches the p r o b l e m of questioned authorship from the theor etical position that every native speaker has their o w n distinct a n d individual
COULTHARD AUTHOR IDENTIFICATION, IDIOLECT, AND LINGUISTIC UNIQUENESS
63
version of the language they speak a n d write, their own idiolect, a n d the assump tion that this idiolect will manifest itself through distinctive a n d idiosyncratic choices in texts (see Halliday etal. 1964: 75; A b e r c r o m b i e 1969). Every speaker has a very large active vocabulary built u p over m a n y years, which will differ from the vocabularies others h a v e similarly built u p , not only in terms of actual items but also in preferences for selecting certain items rather t h a n others. T h u s , whereas in principle a n y speaker/writer can use a n y w o r d at a n y time, speakers in fact tend to m a k e typical a n d individuating co-selections of preferred words. This implies that it should b e possible to devise a m e t h o d of linguistic finger printing - in other words that the linguistic 'impressions' created b y a given speaker/writer should b e usable, just like a signature, to identify them. So far, however, practice is a long w a y b e h i n d theory a n d n o o n e has even b e g u n to ••peculate about h o w m u c h a n d what kind of data would b e n e e d e d to uniquely characterize an idiolect, n o r h o w the data, once collected, would b e analysed and stored; indeed work on the very m u c h simpler task of identifying the linguistic characteristics or 'fingerprints' of whole genres is still in its infancy (Biber 1988, 1995; Stubbs 1996). in reality, the concept of the linguistic fingerprint is an unhelpful, if n o t actu ally misleading metaphor, at least w h e n used in the context of forensic inves tigations of authorship, because it leads us to imagine the creation of massive databanks consisting of representative linguistic samples (or s u m m a r y analyses) of millions of idiolects, against which a given text could b e m a t c h e d a n d tested. In fact such an enterprise is, a n d for the foreseeable future will continue to b e , impractical if not impossible. T h e value of the physical fingerprint is that every sample is both identical a n d exhaustive, that is, it contains all the necessary information for identification of an individual, whereas, b y contrast, any linguistic sample, even a very large one, provides only very partial information about its creator's idiolect. This situation is c o m p o u n d e d b y the fact that m a n y of the texu. which the forensic linguist is asked to examine are very short indeed - most suicide notes and threatening letters, for example, are well u n d e r 200 words lon$ and many consist of fewer t h a n 100 words. Nevertheless, the situation is n o t as b a d as it might at first seem, because such texts are usually accompanied b y information or clues which massively restrict the n u m b e r of possible authors. Thus, the task of the Unguistic detective is never one of identifying an author from millions of candidates o n the basis of the unguistic evidence alone, b u t rather of selecting (or of course S e l e c t i n g ) one author from a very small n u m b e r of candidates, usually fewer than a dozen and in many cases only two (Coulthard 1992, 1993, 1994a, 1994b, 1995, 1997; Eagieson. 1994). An early and persuasive e x a m p l e of the forensic significance of idiolectal co-selection was the U n a b o m b e r case. Between 1978 a n d 1995, s o m e o n e living in the USA. who referred to himself as F C , sent a series of b o m b s , o n average once a year, through the post. At first there seemed to b e n o pattern, b u t after several years the FBI noticed that the victims seemed to b e p e o p l e working in UnivfirsiKpfinnrl AJi-U^n^ n « J
] J
'
64
DISCOURSE STUDIES
Industrial Society and its Future, from s o m e o n e claiming to b e the U n a b o m b e r , along with an offer to stop sending b o m b s if the manuscript were published. I n August 1995, the Washington Postpublished the manuscript as a supplement and three m o n t h s later a m a n contacted the F B I with the observation that the d o c u m e n t sounded as if it h a d b e e n written b y his brother, w h o m h e h a d not seen for some ten years. H e cited in particular the use of the phrase 'cool-headed logician' as b e i n g his brother's terminology, or in our terms a n idiolectal pref erence, which h e h a d noticed a n d r e m e m b e r e d . T h e F B I traced a n d arrested the brother, w h o was living in a w o o d e n cabin in M o n t a n a . T h e y found a series of documents there a n d performed a linguistic analysis - one of the documents was a 300-word newspaper article o n the same topic written a decade earlier. T h e F B I analysts claimed major linguistic similarities b e t w e e n the 35,000 and the 300 w o r d documents: they shared a series of lexical a n d grammatical words a n d fixed phrases, which, the F B I argued, provided linguistic evidence of com m o n authorship. T h e defence contracted a linguist, w h o counter-argued that o n e could attach n o significance to these shared items o n the grounds that a n y o n e can use any word at any time and therefore shared vocabulary can h a v e n o diagnostic signifi cance. T h e linguist singled out twelve words and phrases for particular criticism, o n the grounds that they w e r e items that could b e expected to occur in any texi that was arguing a case - at any rate; clearly; gotten; in practice; moreover; more ot less; on the other hand;presumably;propaganda; thereabouts; and words derived from the roots argu*andpropos*. T h e F B I searched the internet, which in those days was a fraction of the size it is today, but even so they discovered some 3 million documents which included one or m o r e of the twelve items. However, when they narrowed the search to documents which included instances of all the twelve items they found a m e r e 69 and, o n closer inspection, every single o n e of these documents p r o v e d to b e a n internet version of the 35,000-word manifesto. This was a massive rejection of the defence expert's view of text creation as purelv o p e n choice, as well as a powerful example of the idiolectal habit of co-selection a n d a n illustration of the c o n s e q u e n t forensic possibilities t h a t idiolectal co-selection affords for authorship attribution. T h e first e x a m p l e will b e taken from the area of plagiarism detection. 1
O n D e f i n i n g a n d D e t e c t i n g Plagiarism
•
At its simplest, plagiarism, or m o r e accurately the type of plagiarism we as lin guists are competent to deal with, is the theft, or unacknowledged use, of texi created b y another. As m y o w n university's website expresses it:
i
Plagiarism is a form of cheating in which the student tries to pass off s o m e o n e else's work as his or h e r own Typically, substantial passages are 'lifted' verbatim from a particular source without p r o p e r ' attribution
COULTHARD
AUTHOR IDENTIFICATION, IDIOLECT, AND LINGUISTIC UNIQUENESS
65
Any investigation of plagiarism is based consciously or unconsciously o n a notion of idiolect. I n other words it is expected that a n y two writers writing o n the same topic, even if intending to express very similar meanings, will choose an over lapping, but b y n o m e a n s identical, set of lexico-grammatical items to d o so. Indeed, and m o r e importantly for some cases I will treat below, linguists from all persuasions subscribe to some version of the 'uniqueness of utterance' principle (Chomsky 1965; Halliday 1975) a n d so would expect that even the same p e r s o n speaking/writing o n the same topic o n different occasions would m a k e a different set of lexico-grammatical choices. It follows from this that, in any comparison of two texts, the m o r e similar the set of items chosen, the greater the likelihood that one of the texts was derived, at least in part, from the other (or, of course, that both were derived from a third text), rather than being composed independendy. In most cases involving students there is little doubt about guilt, as these two examples of essay openings from J o h n s o n (1997: 214) demonstrate - all items which student B 'shares' with student A are highlighted in bold: A. It is essential for all teachers to understand the history of Britain as a multi-racial, multi-cultural nation. Teachers, like anyone else, can be influenced by age old myths and beliefs. However, it is only by having an understanding of the past that we can begin to comprehend the present. B. In order for teachers to competendy acknowledge the ethnic essential to understand the history of Britain as a multi-racial, nation. Teachers are prone to believe popular myths and beliefs; only by understanding and appreciating past theories that we anticipate the present.
minority, it is multi-cultural however, it is can begin to
Even these short extracts provide enough evidence of shared items to question the originality of at least o n e of the essays, or b o t h , of course, if a third text later proves to b e the c o m m o n source. W h e n this level of sharing is also instanced in other parts of the same texts there is n o r o o m for doubt or dispute. T h e case of essay C. however, is not as clear-cut (items which C shares with one or both of essays A and B are highlighted): C. it is very important for us as educators to realise that Britain as a nation has become both multi-racial and multi-cultural. Clearly it is vital for teachers and associate teachers to ensure that popular myths and stereotypes held by the wider community do not influence their teaching. By examining British history this will assist our understanding and in that way be better equipped to deal with the present and the future. Even though there is still quite a lot of shared lexical material h e r e , it is evident that the largest identical sequences are a m e r e three running words long. Even so, one would still want to categorize this degree of lexical overlap, if instanced other parts of the text, as unacknowledged, through m o r e sophisticated hnr-
DISCOURSE STUDIES
66
Birmingham observation that 'Typically, substantial passages are "lifted"'. I will n o t discuss h e r e the important question of whether a significant proportion of student written texts, which technically fall within the textual definition of plagiarism, are n o t the results of deliberate attempts to deceive at all, b u t rather a consequence of what is coming to b e k n o w n as 'patchwriting', that is genuine but flawed attempts b y students, w h o have s o m e h o w failed to acquire the aca demic rules for acknowledging textual borrowing, to incorporate the work of others into their own texts (see Pecorari 2 0 0 2 ; H o w a r d 1999). J o h n s o n ' s (1997) solution to the detection of this k i n d of student plagiarism or collusion, was to m o v e away from using strings or sequences of items as diag nostic features a n d to focus instead o n the percentage of shared individual lexical types a n d tokens as a better measure of derivativeness. Intensive testing has shown that this measure of lexical overlap successfully separates those essays which share c o m m o n vocabulary simply because they are writing o n the same topic, from those which share m u c h m o r e vocabulary because one or m o r e of t h e m is derivative (see Woolls a n d Coulthard 1998). For example, in Johnson's study, whereas essays A, B, a n d C shared 72 different lexical types in their first 500 words, a set of three other essays from the same batch, whose authors had n o t colluded, shared only 13 lexical types, most of which w e r e central to the topic u n d e r discussion. Further w o r k (Woolls 2003) has shown that the most sig nificant evidence is n o t the m e r e quantity of shared lexis, b u t rather the fact that, in the case of some shared items, b o t h texts have b o t h selected them and t h e n only used t h e m once. As such, 'once-only' items are, b y definition, not central to the m a i n concern of the text, otherwise they would h a v e b e e n used m o r e frequentiy. T h e chances of two writers i n d e p e n d e n d y choosing several of the same words for single use are so r e m o t e as to b e discountable. 2
If proof were n e e d e d of the distinctiveness a n d diagnostic p o w e r of words used once-only - hapaxes as they are technically labelled - it comes from suc cessful internet searches in cases of suspected plagiarism. Experience confirms that the most economical m e t h o d to use w h e n checking the internet for suspected plagiarized text is to search using distinctive collocates whose individual items occur only once in the text in question. I will exemplify with the opening of a story written b y a n 11-year old girl: The Soldiers (all spellings as in the original) Down in the country side an old couple husband and wife Brooklyn and Susan When in one afternoon they were having tea they heard a drumming sound that was coming from down the lane. Brooklyn asks, 'What is that glorious sound which so thrills the ear?' when Susan replied in her o sweat voice 'Only the scarlet soldiers, dear,' The soldiers are coming, The soldiers are coming. Brooklyn is confused he doesn't no what is happening. Mr and Mrs Waters were still having their afternoon tea when suddenly a bright light was shinning trough the window.
COULTHARD
AUTHOR IDENTIFICATION, IDIOLECT, AND LINGUISTIC UNIQUENESS
67
The first paragraph is unremarkable, but the second shifts dramatically,' What is that glorious sound which so thrills the ear?'. T h e story then m o v e s back to the opening style, before shifting again t o ' What is that bright light I see flashing so clear over the distance so brightly?;'. It is h a r d to believe that the same author could write in both styles a n d raises the question of whether the other borrowed text(s) might b e available o n the internet. If one takes as search terms three pairs of collocated hapaxes 'thrills - ear', 'flashing - clear' a n d 'distance - brightly' one again sees the distinctiveness of idiolectal co-selection; the single pairing 'flashing - clear' yields over half a million hits on Google, b u t the three pairings together a m e r e 360 hits, of which the first thirteen are all from W. H . A u d e n ' s p o e m ' O W h a t is that sound'. T h e p o e m ' s first line reads ' O w h a t is that sound which so thrills the ear' while the beginning of the second verse is ' O what is that light I see flashing so clear O v e r the dis tance brighdy, brightly?'. If one adds a seventh w o r d and looks for the phrase 'flashing so clear', all of the hits return Auden's p o e m .
Do People Repeat Themselves? Whereas (occasional) identical strings in two texts which are supposed to have different authors can b e indicative of 'borrowing' or theft, it is harder to argue the case when the texts are (supposedly) p r o d u c e d b y the same author but o n different occasions - even w h e n there is n o suggestion that the author h a d sight of the first text when s/he was producing the second. T h e example I want to use is from a famous English m u r d e r case, dating from 1978, w h e r e one piece of strongly contested evidence was a record of a police interview with a suspect. In this case, four m e n were accused, a n d subsequendy convicted, of killing a 13-year old newspaper delivery b o y , Carl Bridgewater, solely on the basis of the confession of one of them, Patrick Molloy - there was n o corroborating forensic evidence a n d Molloy retracted his confession, b u t to n o avail. H e ad mitted that h e did actually say (most of) the words recorded in his confession, but insisted that h e was being told what to say, while h e was dictating the con fession, by a policeman w h o was standing b e h i n d him. H e also claimed that h e had only m a d e the confession after being physically and verbally abused for some considerable time. The police, however, as support for the reliability of the confession, produced a contemporaneous handwritten written record of a n interview which t h e y claimed had taken place immediately before the confession and which contained substantially the same information expressed in the same language as the con fession statement. Molloy denied that the interview h a d ever taken place - in his version of events h e was being subjected to abuse at that time - a n d counterclaimed that the interview record h a d b e e n m a d e u p later o n the basis of the then pre-existing confession. A s is evident from a cursory glance at the two extracts below taken, respectively, from the statement which Molloy admitted maicinDr n n H
*->io
1
-
1
-•
1
1
*
•
DISCOURSE STUDIES
68
the basis of such similarities, that either o n e of the two documents was derived from the other or that b o t h h a d b e e n derived from a third. However, at the time of the original trial, n o linguist was called to give evidence - in fact there were n o forensic linguists practising in Britain at the time - so it was left to the lawyers to evaluate the linguistic significance of the interview and confession. As a re sult, the same p h e n o m e n o n , massive identity in phrasing a n d lexical choice, was argued b y the defence to b e evidence of falsification, a n d b y the prosecution to b e evidence of the authenticity and reUability of both texts, o n the grounds that here was an example of the accused recounting the same events, in essentially the same linguistic encoding, o n two separate occasions. Extract from Molloy's statement (17) I had been drinking and cannot remember the exact tune I was there but whilst I was upstairs I heard someone downstairs say be careful someone is coming. (18) I hid for a while and after a while I heard a bang come from downstairs. (19) I knew that it was a gun being fired. (20) I went downstairs and the three of them were still in the room. (21) They all looked shocked and were shouting at each other. (22) I heard Jimmy say, 'It went off by accident'. (23) I looked and on the settee I saw the body of the boy. (24) He had been shot in the head. (25) I was appalled and felt sick. Extract from disputed interview with Molloy P. (18)
How long were you in there Pat? I had been drinking and cannot remember the exact time that 1 was there, but whilst I was upstairs I heard someone downstairs say 'be careful someone is coming . P. Did you hide? (19) Yes I hid for a while and then I heard the bang I have told you about. P. Carry on Pat? (19a) I ran out. P. What were the others doing? (20) The three of them were still in the room. P. What were they doing? (21) They all looked shocked and were shouting at each other. P. Who said what? (22) I heard Jimmy say 'it went off by accident'. P. Pat, I know this is upsetting but you appreciate that we must get to the bottom of this. Did you see the boy's body? (Molloy hesitated, looked at me intendy, and after a pause said,) (23) Yes sir, he was on the settee. P. Did you see any injury to him? (Molloy stared at me again and said) (24) Yes sir, he had been shot in the head. P. What happened then? (25) I was appalled and felt sick. 1
1
t
COULTHARD AUTHOR IDENTIFICATION, IDIOLECT, AND LINGUISTIC UNIQUENESS
69
this assertion d e p e n d o n two c o m m o n l y held mistaken beliefs: first, that p e o p l e can and d o say the same thing in the same words on different occasions a n d secondly, that people can r e m e m b e r and reproduce verbatim what they and others have said on some earlier occasion. T h e former belief can b e demonstrated to be false either b y recording a person attempting to recount the same set of events on two separate occasions, or b y simply asking a witness to repeat w o r d for word what s/he has just said. T h e second belief used to h a v e some empirical support, at least for short stretches of speech (see K e e n a n et al. 1977 a n d Bates et al. 1980), but was seriously questioned b y Hjelmquist (1984), w h o d e m o n strated that, even after only a short delay, people could r e m e m b e r at best 2 5 per cent of the gist a n d 5 per cent of the actual w o r d i n g of what h a d b e e n said in a five minute two-party conversation in which they h a d participated. Confirmatory evidence about the inability to r e m e m b e r even quite short single utterances verbatim was specially c o m m i s s i o n e d from Professor Brian Clifford a n d presented at the 2003 'Glasgow Ice C r e a m Wars' Appeal. This was used to chal lenge successfully the claim of police officers that they h a d independendy remem bered, some of t h e m for over a n hour, verbatim a n d identically, utterances made b y the accused at the time of arrest. Clifford's experiment tested the ability to remember a short, 24-word utterance and found that most people were able to recall verbatim n o m o r e t h a n 30 to 40 percent of what they h a d h e a r d . By the time of the Bridgewater Appeal in 1997, it was possible to provide extra supporting evidence of two kinds. First, as a direct result of J o h n s o n ' s [ 1^97) work on plagiarism discussed above, which demonstrated the significance of vocabulary overlap, an analysis was done of the shared vocabulary in the two Molloy texts. It b e c a m e evident that the highlighting in the two Molloy extracts presented above actually understates the similarities between the two texts - a closer examination revealed that there was in fact n o t one single word in Molloy's statement, neither lexical n o r grammatical, which did not also occur in the inter view record. I have only seen that degree of overlap o n one other occasion, w h e n two students h a d in fact submitted identical essays for assessment. Ironically, the computer analysis showed the degree of similarity to b e only 97 percent - die 3 percent of different words m a d e u p of spelling errors p r o d u c e d b y one of the two students. In the Bridgewater case t h e r e also was secondary, supporting linguistic evidence, of a different kind, to support the claim that the interview record was both falsified and based o n the statement. If we assume that the police officers had indeed, as Molloy claimed, set out to create a dialogue based on the m o n o logue statement, they would h a v e faced the major p r o b l e m of what questions to invent in order to link forward a n d a p p a r e n d y elicit the pre-existing candidate answers which they h a d derived from the s t a t e m e n t I n this scenario o n e would expect there to b e occasions w h e n a question did n o t fit successfully into the text into which it h a d b e e n e m b e d d e d - a n d i n d e e d there are. In a developing interview, a police question usually links backwards lexically, repeating word(s) from the previous answer. H o w e v e r , in creating a question to 3
fit 3 DXe-exisrinGr a n s w e r
i-hora
ic
o l , . ™ , ™ *.U-
J
ii
. ,1
DISCOURSE STUDIES
70
sequence - (21) ' T h e y all looked shocked a n d were shouting at each other.' (22) 'I h e a r dJ i m m y say, "it went off b y accident"' - which appears w o r d for word in the interview record, except that the two sentences are separated b y the inserted question ' W h o said what?'. However, in this context the w o r d 'said', although it is cataphorically u n r e m a r k a b l e - said links with say - is anaphorically o d d be cause the m e n have just b e e n described as 'shouting'. O n e would therefore h a v e expected an anaphorically cohesive follow-up question to b e either 'What/ W h y were they shouting?' or ' W h o was shouting (what)?'; one would certainly not predict 'who said what?'. T h e choice of 'said' is a most u n e x p e c t e d choice except of course for s o m e o n e w h o knows that the n e x t utterance will b e 'I h e a r d J i m m y say . . . ' - then it has an evident logic. A n example of a grammatical misfit is w h e r e the statement version 'on the settee I saw the b o d y of the boy. H e h a d . . . ' is transformed into ' Q . D i d you see t h e boy's body? Yes sir, h e was o n the settee'. T h e statement version correcdy uses the p r o n o u n ' H e ' because the referent is the ' b o y ' in 'the b o d y of the boy', b u t the reformulated version in the police interview, 'the b o y ' s body', would be m o r e likely to h a v e elicited 'it' as a referent. We also find examples of process misfit: in the exchange r e p r o d u c e d below, the question 'what h a p p e n e d ' re quires a report of a n action or a n event, but in fact the response is a description of two states: P. What happened then? M. I was appalled and felt sick. H a d the reply b e e n 'I vomited', it would, of course, h a v e been cohesive. Similar process misfits are: P. M. P. M.
What were the others doing? The three of them were still in the room. What were they doing? They all looked shocked
It is possible to continue in this vein, but I think these examples are sufficient to show that textual oddities like these support the claim, which was based on the identity of expression, that the interview record was created from the pre-existing statement. Sadly, it was n o t possible to test the acceptability a n d persuasiveness of these arguments in court, as the C r o w n conceded the appeal shortly before the due date, w h e n compelling n e w evidence from d o c u m e n t a n d handwriting analysts emerged to convince the judges of the unsafeness of the conviction.
T h e Evidential V a l u e of S i n g l e Identical Strings I n the Bridgewater Four case there was a whole series of identical strings of i words to support the claim that the interview record was derived from tfc
COULTHARD
AUTHOR IDENTIFICATION, IDIOLECT, AND LINGUISTIC UNIQUENESS
71
due to borrowing rather t h a n identical encoding o n two separate occasions, the claim of fabrication was supported b y other linguistic evidence of a different and independent kind. T h e final questions I will address in this article are h o w much weight can o n e place o n a single identical string and h o w significant is the length of a string w h e n assessing its evidential significance? These questions go to the heart of current thinking about uniqueness in language production. As Sinclair (1991) pointed out, there are two complementary assembly prin ciples in the creation of utterances/sentences; o n e is the long accepted principle that sequences are generated w o r d b y w o r d o n an ' o p e n choice' basis. W h e n strings are created in this way, there is for each successive syntagmatic slot a large n u m b e r of possible, grammatically acceptable, paradigmatic fillers a n d thus one can easily, if n o t effortlessly, generate m e m o r a b l e b u t meaningless se quences like 'colorless green ideas sleep furiously'. T h e other assembly principle proposed m u c h m o r e recently as a result of corpus work (Sinclair 1991), is the 'idiom principle', according to which pre-assembled chunks m a d e u p of frequent collocations and colligations are linked together to create larger units. I n practice, both principles w o r k side b y side, which m e a n s that any given short string might be produced b y either principle a n d therefore might b e either an idiosyncratic combination or a frequently occurring fixed phrase. Nevertheless, the longer a sequence is, the m o r e likely it is that at least some of its components h a v e b e e n created b y the o p e n choice principle a n d consequently, the less likely that the occurrence of this identical sequence in two different texts is a consequence of two speakers/writers coincidentally selecting the same chunk(s) b y chance. The data I will use for exemplificatory purposes c o m e from the A p p e a l of Robert Brown in 2 0 0 3 . As in the Bridgewater Four case, h e r e too there was a disputed statement and a disputed interview record; the difference was that Brown claimed that the statement was in reality a dialogue which h a d b e e n represented as a monologue. H e claimed that a police officer h a d asked ques tions to which Brown said h e simply replied 'Yes' (Judge's Summing-up, p. 95 section E), and that, although the interview did occur, the record of it was m a d e up afterwards - ' n o police officer took any notes' (Judge's Summing-up, p. 93 section E). Below are two sentences from the statement m a t c h e d with items occurring in the (invented) interview record: [i) Statement Interview (ii) Statement Interview
I asked her if I could carry her bags she said 'Yes' I asked her if I could carry her bags and she said 'yes' I picked something up like an ornament I picked something up like an ornament
In what follows I h a v e used examples from Google, rather t h a n from a corpus such as the Bank of English or the British National Corpus, o n the grounds that Google is accessible to the layperson for w h o m the argument is designed. While the above utterances/sentences m a y n o t seem remarkable in themselves, neither of them occurs even once in the h u n d r e d s of millions of texts that Gooele searches
72
DISCOURSE STUDIES
String
Instances
I picked I picked something I picked something up I picked something up like I picked something up like an
"
an ornament like an ornament something like an ornament
' _
1
, 1,060,000 780 362 1 0 73,700 896 2
I asked I asked her I asked her if ' I asked her if I I asked her if I could I asked her if I could carry I asked her if I could cany her I asked her if I could carry her bags
2,170,000 284,000 86,00') 10,400
if I could if I could carry
2,370,000 1,600
7,770
7 4 <>
It is evident that 'if I could' a n d p e r h a p s 'I asked h e r ' h a v e the characteristics of pre-assembled idioms, but even t h e n their co-selection in sequence is rare, ai 7,770 occurrences. T h e m o m e n t o n e adds a 7th w o r d , 'carry', the odds against it occurring b e c o m e enormous, with a Google search yielding only 7 instances I n d e e d rarity scores like these begin to look like the probability scores DNA experts proudly present in court. However, unlike the D N A expert, the linguist expert has the disadvantage that everyone in the courtroom considers themselves to b e a language expert and, a m o n g other things, 'knows' that they can remember what they a n d indeed others said in past conversations a n d feels confident thai they can 'repeat' what was said verbatim at a later date. It will never b e enoug) for the linguist to simply assert the uniqueness of encoding, it will need to b demonstrated in a n accessible way.
A n A t t e m p t at Court Persuasion W h e n faced with the p r o b l e m of having to convince the A p p e a l Court judges in o p e n court in the R vs. Robert Brown A p p e a l I prepared the following presentation, which I h o p e d would b o t h interest and persuade the three judges of the evi dentiary strength of the identical formulations discussed above. I should say that, as in the Bridgewater case, this was n o t the only linguistic evidence to support Brown's claims about the unreliability of the police records. As a first step I used Google to find out something about other cases involving L o r d Justice Rose, w h o was to preside. T h e first three citations for the words
COULTHARD
AUTHOR IDENTIFICATION, IDIOLECT, AND LINGUISTIC UNIQUENESS
73
Guardian Unlimited | Special reports | Archer loses a p p e a l b i d . . . was not present at today's hearing, h a d his application for permission to a p p e a l against the conviction rejected within hours. L o r d J u s t i c e Rose, sitting with . . . (www.guardian.co.uk/archer/article 0,2763,759829,00.html—30k) 1 accessed this citation, part of which is r e p r o d u c e d below as 'Guardian Extract', and from it selected the first p h r a s e quoted from Lord Rose - 'For reasons w e will give later in the day' - which I h a v e highlighted in bold. Given the nature of Appeal Court j u d g m e n t s , which are often released after the decision has been announced, this seemed a n u n r e m a r k a b l e phrase a n d yet a Google search resumed only 7 occurrences of the phrase - all of which, o n closer examination, proved not only to b e attributed to Lord Rose, b u t were all in fact different reportings of the same uttering at the end of the Archer appeal. Even reduc ing the phrase to the apparently less specific 6-word utterance 'For reasons we will give later' only p r o d u c e d two m o r e examples, this time n o t uttered b y Lord Rose. Thus, h e r e was an e x a m p l e of the uniqueness of an a p p a r e n d y ordinary utterance b y L o r d Rose himself. uuardian extract Archer loses appeal bid Lord Justice Rose, sitting with MrJustice Colman and Mr Justice Stanley Burnton in London, told Archer's QC Nicholas Purnell: 'For reasons we wiU give later in the day we are against you in relation to conviction.' At the start of the hearing Nicholas Purnell QC, outlining the grounds of appeal, said: 'The submission that we make on behalf of Lord Archer is that the first and fundamental ground which interconnects with all the other grounds of appeal was that the learned trial judge wrongly exercised his discretion not to sever the trial of Edward Francis.' Mr Purnell said the decision of the judge, MrJustice Potts, not to sever the trial of Francis had an 'unbalancing effect on the equilibrium' of the trial. Counsel argued that Mr Francis was 'in a position effectively as a substitute prosecution witness and a substitute prosecutor'. I then took three shorter phrases quoted in the same article, this time from Nicholas Purnell, Lord Archer's Q C , each of t h e m seeming, at least to this lay reader, to be equally u n r e m a r k a b l e phrases for a lawyer to utter, thefirstand fundamentalground, unbalancing effect on the equilibrium, and a substitute prosecution witness. For these phrases I found 7, 10 a n d 4 instances respectively, but again for every phrase all the instances were reports of the same single occasion of uttering by M r Purnell. Armed with these examples, taken from an audience-relevant text, I h o p e d to convince the judges that uniqueness of utterance was a demonstrable fact. Sadly, at a pre-trial case conference, the defence barrister w h o h a d chosen to call me as an expert witness smiled indulgently a n d described m y intended pre sentation aS Whimsical and dpridprl n o t tn urn it PVM-h,»II(^»1" ~ " - «<-' 1-
DISCOURSE STUDIES
74
A u t h o r Identification i n A m e r i c a n Courts While the analytical techniques and arguments and the derived opinions reported above would b e a n d i n some cases already have b e e n accepted i n British courts, the situation is less clear i n the U S A . I n a recent article entitled ' T h e linguist on the witness stand: forensic linguistics in A m e r i c a n courts', Tiersma and Solan (2002) noted that: although t h e [American] legal system h a s often w e l c o m e d linguistic expertise, there are a n u m b e r of areas in which they are m o r e hesitant to d o so. O n e example is the use of linguistics to identify authors. (Tiersma a n d Solan 2002: 229) Tiersma a n d Solan cited the rigorous d e m a n d s of the A m e r i c a n legal system's Daubert criteria, which, in their opinion, m a n y authorship identification method fail to meet. T h e D a u b e r t criteria w e r e created i n a S u p r e m e Court ruling at tb< e n d of a n appeal in the case of Daubert vs. MerrellDow Pharmaceuticals, Inc. (5(1* U.S. 579 (1993)). Essentially t h e argument was over whether expert evidenr could b e rejected o n the grounds that t h e experts involved h a d n o t publishetheir work. I n their ruling the S u p r e m e C o u r t observed that ' t h e adjecth "scientific" implies a grounding i n the m e t h o d s a n d procedures of science' and then went o n to propose four criteria with which to evaluate the 'scientific-ne^of a m e t h o d : 1
1 2 3 4
whether t h e theory offered has b e e n tested; whether it has b e e n subjected to peer review a n d publication; t h e k n o w n rate of error; a n d whether the theory is generally accepted in t h e scientific communi' (509 U.S. 593) quoted i n Tiersma a n d Solan (2002).
T h e r e is an extensive a n d lucid discussion of the D a u b e r t criteria as applied linguistic evidence in the article b y Solan a n d Tiersma in this issue (pp. 448-* I n o n e sense, Tiersma a n d Solan are raising purely American proble because in the British a n d Australian legal systems it is t h e expert rather tl. the m e t h o d that is recognized, so these courts can a n d d o allow opinion evidti from anyone considered to h a v e : specialised knowledge based o n . . . training, study or experience [pro vided that the opinion is] wholly or substantially based o n that knowledge (Evidence Act 1995 Sec 79) However, knowing that their evidence would also satisfy t h e Daubert critt gives extra confidence to British linguist/experts a n d it is quite conceivable i similar criteria might b e introduced into British Courts at some point in -Ii - r : _ J : . j 4 „ „ l injimiaiits
A Will!.
COULTHARD AUTHOR IDENTIFICATION, IDIOLECT, AND LINGUISTIC UNIQUENESS
75
Northern Ireland, for example, has recendy ruled that forensic phonetic evi dence based solely on auditory analysis, that is with n o acoustic support, is n o longer permitted. Although I await the views of lawyer-linguists Solan and Tiersma with interest if not with some trepidation, I would argue that the methods of author attribution discussed above do meet the four Daubert criteria:
1. Whether the Theory Offered has been Tested Work by many people on a large number of cases has shown that there is no longer any dispute that the occurrence of shared identical items is conclusive evidence that two texts have not been independendy created; what remains to be agreed is how few shared identical items are necessary to support a decision.
2. Whether [the Theory] has been Subjected to Peer Review and Publication Publications like this one and those b y Johnson (1997), Woolls and Coulthard (1998) and Woolls (2003) have b e e n subjected to peer review, there have been many presentations on this kind of authorship assignment at international conferences and to peer audiences in universities worldwide. In addition, the Copycatch Gold collusion detection program is in use internationally in over 50 universities, including the British Open University.
3. The Known Rate of Error This is perhaps the most difficult criterion - in cases of plagiarism it is traditional to err on the side of caution, and so I know of no cases of error. However, in this article, in the final section, I have taken the extreme position that a single and relatively short string can be conclusive evidence; this in itself is a challenge to the academic community to test the error rate and at the same time to fix an acceptable statistical equivalent for 'beyond reasonable doubt'.
Whether the Theory is Generally Accepted in the Scientific Community There is no doubt that the basic tenets of idiolectal variation and the uniqueness of utterance are generally accepted across the whole linguistic community; the disagreements are over how far certainty of assignment depends on the amount and kind of shared vocabulary and on the length of individual sequences and their composition in terms of idiomatic and open choice items.
Conclusion evidence discussed above suggests that the concepts of idiolect and IPnPtC n f
ni*«-»-
1
76
DISCOURSE STUDIES
questions a b o u t a u t h o r s h i p with a h i g h d e g r e e of confidence. A s d e m o n strated w e can use t h e concepts to h e l p us search w h e n w e suspect plagiarism a n d to categorize a n d classify w h e n w e already h a v e texts of various kinds whose authorship is suspect or disputed. T h e r e are still m a n y author identification p r o b l e m s w h e r e the methodology is less developed a n d reliable a n d where Solan a n d Tiersma's cautions are well-heeded, b u t the future for author iden tification is encouraging.
Notes 1. For an accessible version of events, from someone who wrote a report on the language of the manuscript, see Foster (2001). The full text of the Unabomber manuscript is available at: http://www.panix.com/~clays/Una/. 2. An automated version of this analytic method, produced by Woolls (2002), is now available as the computer program Copycatch Gold. 3. http://news.bbc.co.Uk/7/hi/scotland/3494401.stm
References Abercrombie, D. 1969. 'Voice qualities' in N. N. Markel (ed.): Psychalinguistics: An Introduction to the Study of Speech and Personality. London: The Dorsey Press. Bates, E., W. Kintsch, C. R. Fletcher, and V. Giulani. 1980. 'The role of pronominalisation and ellipsis in texts: Some memorisation experiments,'./ouraflZ of Experimental Psychology: Human Learning and Memory 6: 676-91. Biber, D. 1988. Variation across Speech and Writing. Cambridge: Cambridge University Press. Biber, D. 1995. Dimensions of Register Variation: A Cross-linguistic Comparison. Cambridge: Cambridge University Press. Chomsky N. 1965. Aspects of the Theory of Syntax. Cambridge, MIT Press. Clemit, P. and D. Woolls. 2001. 'Two new pamphlets by William Godwin: A case of computer-assisted authorship attribution,' Studies in Bibliography, 54: 265-84. Coulthard, R. M. 1992. 'Forensic discourse analysis' in R. M. Coulthard (ed.): Advances in Spoken Discourse Analysis. London: Routiedge, pp. 242-57. Coulthard, R. M. 1993. 'Beginning the study of forensic texts: Corpus, concordance, col location' in M. P. Hoey (ed.): Data Description Discourse. London: Harper Collins, pp. 86-97. Coulthard, R. M. 1994a. 'Powerful evidence for the defence: An exercise in forensic discourse analysis' in J. Gibbons (ed.): Language and the Law. London: Longman, pp. 414-42. Coulthard, R. M. 1994b. 'On the use of corpora in the analysis of forensic texts,' Forensic Linguistics: the International Journal of Speech, Language and the Law 1/1: 27-43. Coulthard, R. M. 1995. Questioning statements: Forensic applications of linguistics, text of inaugural lecture. Birmingham: English Language Research. Coulthard, R. M. 1997. 'A failed appeal,' Forensic Linguistics: the International Journal o< Sbeedu Language and the Law 4/ii: 287-302.
COULTHARD
AUTHOR IDENTIFICATION, IDIOLECT, AND LINGUISTIC UNIQUENESS
77
Foster, D. 2001. Author Unknown: on the Trail o/Anonymous. London: Macmillan. Halliday, M. A. K. 1975. Learning how to Mean. London: Edward Arnold. Halliday, M. A. K, A. Mcintosh, and P. Strevens. 1964. The Linguistic Sciences and Language Teaching. London: Longman. Hjelmquist, E. 1984. 'Memory for conversations,' Discourse Processes 7: 321-36. Howard, R. M. 1999. 'The new abolitionism comes to plagiarism' in L. Buranen d A. M. Roy (eds): Perspectives on Plagiarism and Intellectual Property in a Postmodern World. Albany: State University of New York Press, pp. 87-95. Johnson, A. 1997. 'Textual kidnapping - a case of plagiarism among three student texts,' Forensic Linguistics: The International Journal of Speech, Language and Law 4/ii: 210-25. Keenan, J. M., B. MacWhinney, and D. Mayhew, 1977. 'Pragmatics in memory: A study of natural conversation,' Journal of Verbal Learning and Verbal Behavior 16: 549-60. McMenamin, G. 2004. 'Disputed authorship in US Law,' InternationalJournal ofSpeech Language and the Law (formerly Forensic Linguistics) 11/1: 73-82. Pecorari, D. E. 2002. Original Reproductions: An Investigation of the Source Use of Postgraduate Second Language Writers. Unpublished Ph.D. thesis, University of Birmingham. Shuy, R. 1993. Language Crimes: The Use and Abuse of Language Evidence in the Courtroom. Cambridge MA: Blackwell. Shuy, R. 1998. The Language of Confession, Interrogation and Deception. London: Sage. Sinclair.J. 1991. Corpus, Concordance, Collocation. Oxford: Oxford University Press. Solan, L. and P. Tiersma. 2004. 'Author identification in American courts,' Applied Linguistics. 25:448-65. Stubbs, M. 1996. Text and Corpus Analysis. Oxford: Blackwell. SvartvikJ. 1968. The Evans Statements: A CaseforForensic Linguistics. Goteborg: University of Gothenburg Press. Tiersma, P. and Solan, L. 2002. 'The linguist on the witness stand: Forensic linguistics in American courts,' Language 78:221-39. Wbolls. D. 2002. Copycatch Gold a computerised plagiarism detection program. Woolls. D. 2003. 'Better tools for the trade and how to use them,' Forensic Linguistics. The Intemationaljoumal ofSpeech, Language and Law 10/i: 102-12. Woolls, D. and Coulthard, R. M. 1998. 'Tools for the trade,' Forensic Linguistics. The International Journal ofSpeech, Language andLaw5/'v. 33-57.
19 Bodies, Beaches and Burn-Times: 'Environmentalism' and Its Discursive Competitors Nikolas Coupland andJustine Coupland
Linguistic Perspectives o n t h e Natural E n v i r o n m e n t
W
hile there is as yet n o consensus on h o w the term 'ecolinguistics" should b e applied (cf. Alexander et a l , 1993; Fairclough, 1993), it would b e a n appropriate t e r m for a critical perspective which concentrated o n the semantics of expressions used to represent the natural environment, environmental practices or policies (cf. Alexander, 1993; Michael, 1991). Alter natively, because language is the m e d i u m through which ecological knowledge is fabricated, developed, transmitted a n d understood, a form of ecolinguistic research needs to investigate h o w relevant knowledge is shared or confined - for example as between scientific a n d m e d i a texts, a n d between experts a n d non-experts. Are non-experts simply informed or actively 'de-skilled' (Giddens, 1991) in the mediation of environmental knowledge? Bell's work o n the reporting and reception of climate change information [1989,1991] gives a n important lead, in this regard. R a t h e r differendy from these approaches, we want to argue for a 'competing discourses' formulation for ecolinguistics in the present paper. W h e n people seek, as activists, critics, theorists or green politicians, to change social priorities, values or practices relating to the global environment, we become aware of environmentalism as a site of struggle. These struggles are generally represented as rather concrete disputes over h u m a n actions (whether or h o w to recycle waste, or to drive diesel cars), or over political decisions with practical consequences (whether to build a n e w by-pass, b u r n orimulsion in a p o w e r station, or limit fishing). But to the extent that such struggles are founded on arguments between political priorities and value systems, established b y different groups with varying degrees
COUPLAND A N D COUPLAND
BODIES, BEACHES AND BURN-TIMES
79
of power and influence, and enacted rhetorically (in conversations or the media), it is appropriate to construe t h e m as competing discourses (cf. Lee; 1992, Candlin and Maley, in press). Discourse in this context is intended in the broadly Foucauldian sense of a particular, internally coherent set of values a n d orientations, which is held to b e normative, persuasive or simply unexceptional within specific groups or m o v e ments, which guides practical action, b u t which is configured around and through a finite set of claims, statements or accounts, a m o u n t i n g to a stance or (partial) world-view. T h a t is, following the b r o a d priorities b y m a n y contributors to this journal, discourses are taken to b e b o t h ideological a n d textual, a n d basically amenable to forms of linguistic or rhetorical or semiotic analysis. 'Envkonmentalism' itself can therefore b e construed as a discourse, or, m o r e appropriately, as an inter linked and evolving set of discourses, established to counter pre-existing discourses. We see this a p p r o a c h as b e i n g able to build quite directiy u p o n the critical theoretic work of Beck (1995) a n d Giddens (1991), in their characterization of 'the [late-modern, contemporary] risk society' a n d of popular involvement with environmental risk in particular. Douglas a n d Wildavsky's (1982) social a n d specifically cultural perspective o n risk similarly requires the empirical support of particular grounded analysis. A discourse analytic tradition can support Nelkin's claims about the viability of a linguistic perspective o n risk a n d the n e e d to expose 'the conflicting values that perpetuate disputes over risk' as a supplement to 'rational' a n d 'objective' approaches (Nelkin, 1985, p . 9). ' H u m a n responses to risk', as Nelkin writes, 'are filled with contradictions' (p. 15), a n d in the environ mental d o m a i n these are occasioned b y the vast disjunction b e t w e e n the hugely global etiology of risk a n d t h e local circumstances of p e o p l e ' s particular responses, a n d b y the conflicts between ideologies of modernity/progress a n d conservatism/restraint. According to Beck, individuals' responses to environmental risk are necessarily located within t h e constraining 'freedom' of consumer choice. C o n t e m p o r a r y technology a n d 'progress' h a v e killed off nature and sought to reconstruct it within the parameters of technology a n d consumerism. Therefore, 'the social "consumption" of nature renders invalid all those concepts a n d theories that conceive of nature as the counter-image of h u m a n activity a n d power, to which it must b e subjugated' (1995: p . 38). Since there is a prevailing technicist confusion between nature a n d society (p. 47), individuals are largely restricted to making their responses to environmental hazards within the d o m a i n of consumerist activities, rather t h a n engaging with natural p r o b l e m s a n d solutions directly. The complexities a n d contradictions of r e s p o n d i n g to acknowledged environ mental problems should therefore b e a p p a r e n t in the relevant social discourses, and amenable to critical commentary.
A n Empirical A p p r o a c h
80
DISCOURSE STUDIES
I n this paper, we develop an account of h o w U K print m e d i a coverage of these issues sustains a series of contradictory moral, ideological and practical positions. O u r intention is to show h o w imperatives and proscriptions that are articulated as key parts of 'environmental' discourses are systematically u n d e r m i n e d by other formations, either within the same or across different texts. T h a t is, we want to illustrate h o w a rather abstract, b u t still textually grounded, discourse analytic a p p r o a c h to 'the environment' can show where a n d h o w 'environmentalism' is constructed in U K print media, and h o w those same m e d i a offer us specific m e a n s of challenging or ignoring 'environmental' imperatives. (The attributive 'environmental' is part of the paper's problematic, a n d should b e r e a d as subject to qualification a n d doubt, whether or n o t w e place it in inverted commas.) We collected a sample of U K print m e d i a during 2 weeks inJ u l y 1994 (that is during the 'so-called British summer', where 'so-called' is so-called). T h e sample was n o t in any w a y pre-planned, except insofar as w e consciously chose a 'high s u m m e r ' period. We surveyed all national newspapers a n d the 2 3 most popular magazines (10 weeklies, three 2-weeklies a n d 10 monthlies), j u d g e d mainly by circulation statistics. I n the first scan of these data, we identified texts making any reference to any aspects of ozone depletion (but n o t 'ground-level ozone'), UVA and U V B problems, sun-tanning or burning, a n d protecting skin from sun expos ure. T h e texts included newspaper a n d magazine reports and features of variable length a n d p r o m i n e n c e , b u t also weather reports a n d product advertisements. 2
Ozone-Melanoma Discourse As a starting point, we might identify a recurring 'environmental' discourse within these data. It is a discursive formation that is relatively recent, b u t seemingly m u c h better established in t h e s o u t h e r n t h a n i n the n o r t h e r n h e m i s p h e r e anglophone cultures. Its k e y c o m p o n e n t s might b e represented as in Figure 1. We refer to this configuration as ' o z o n e - m e l a n o m a ' discourse. Note its pro gression, at least in h o w the keywords are arranged in Figure 1, from causes to effects a n d from p r o b l e m s to solutions. As summarized in the last line of the Figure, the discourse frames this particular environmental issue initially in scientific terms, m o v i n g from a scientific cause t h r o u g h a n interpretation of health consequences (with 'health' still ingrained in a biomedical domain). But through the n e e d that it asserts for us to b e aware of a n d perhaps responsible for our o w n health, a m o r a l dimension comes into play, leading to potential revision of self-presentational ideals - specifically in redefining the desirability of fair skin. T h e m e d i a sample includes a strand of newspaper reports focusing o n a par ticular, n e w scientific discovery, to the general effect that sun-protection creams m a y give poorer protection against skin cancer t h a n was previously thought. Text 1 (Daily Telegraph, 7 J u l y 1994) is one of these reports, and it articulates many of the k e y processes a n d propositions that w e h a v e taken to b e criteria! elements of o z o n e - m e l a n o m a discourse. T h e report is located within the referential d o m a i n of academic science ('a senior scientist said'; 'professor of chemical pathology'; 'Tests in his laboratories'; 3
COUPLAND A N D COUPLAND
BODIES, BEACHES AND BURN-TIMES
Sun creams 'no barrier' to cancer By Peter Pallot, Health Services S t a f f SOME sun creams may not protect against skin cancer, a senior scientist said yesterday. Manufacturers have con centrated on blocking "burn ing" ultraviolet radiation — U V B — and have largely overlooked another form of r a d i a t i o n , U V A , s a i d Dr Joseph Lunec, professor of chemical pathology. Tests in his laboratories at Leicester University showed UVA. could damage genetic material within cells which may predispose to cancer, he said. He had developed chemi cal markers which identified UVA damage to genetic ma terial in skin cells cultured in the laboratory, he said. But the link .had not been demon strated in people. His findings are to be pre sented to a conference in Venice in October and in Australia in November. Cases of skin cancer have grown rapidly in Britain and now exceed 40.000 a year, although around 90 per cent are easily treated. Prof Lunec added: "Com panies are conducting only a little investigation into UVA and there should be more." Sun creams are given sun protection factors according to how much protection they offer to UVB. But that was no
guide to their effectiveness against UVA, he said. There has been concern in the sun cream industry about UVA. Boots, the pharmacy chain, has introduced"a UVA star rating system. Prof Lunec declined to say which commercial brands he had used in tests, but he said: " C r e a m s n e e d to c o n t a i n more anti-oxidant vitamins to offer adequate protection; The people currently produc i n g c r e a m s are not fully aware of the effects of UVA. "In Australia, where expor s u r e to t h e s u n is m u c h greater than here in Britain, t h e y are more c o n c e r n e d about this issue. "We have been experienc ing a spell of hot weather and all the people you s e e coming into work burned by the sun have been exposing thenar selves to potentially danger? ous levels of UVA." Last night Prof Gordon M c V i e . of t h e C a n c e l R e s e a r c h C a m p a i g n said manufacturers had concen trated on protecting against U V B because all research suggested that this, rather than UVA, was the one that did the damage. If t h e s c i e n c e p r o v e s , sound, it would be the first t i m e t h a t UVA had beer implicated.as a cancer agent he added.
81
DISCOURSE STUDIES
82
OZONE DEPLETION I N C R E A S E D H A R M F U L U V R A Y A C T I O N ( U V A and
UVB)
T H R E A T T O H E A L T H A N D LIFE SKIN DAMAGE A N D MELANOMA D E A T H RATES ( I N D I V I D U A L PAIN) AWARENESS BURNING VS TANNING SKIN PROTECTION CHEMICAL BARRIERS EXPOSURE LIMITATION
•
SUN AVOIDANCE
, '
S U N - B A T H I N G IS N A I V E ( O R C R A S S O R G R O T E S Q U E ) FAIR S K I N I S G O O D S C I E N C E - HEALTH - MORALITY - SELF-PRESENTATION
, ,
.
Figure 1: Key elements of ozone-melanoma discourse relates a scientific discovery to a health problem affecting large numbers of people, a n d so to an epidemiological trend ('Cases of skin cancer h a v e grown rapidly in Britain a n d n o w exceed 40,000 a year'), although there are reported to b e differ ences between northern and southern hemispheres in these trends a n d associated risks ('In Australia, where exposure to the sun is m u c h greater than here in Britain'). T h e threat to health is represented as a physical o n e ('could damage genetic material'). Therefore what is at issue is 'burning' not 'tanning' ('burned by the sun'). Limiting d a m a g e is repeatedly (in fact o n four occasions) cast in terms of 'protection' b y means of a physical a n d / o r chemical 'barrier' (headline) of varying 'effectiveness'. T h e text is a prototype of a n explicitly 'risk-formative' discourse It references 'risk' a n d 'threat' directly. It also establishes a hazardous outcome, gives a probability index, and extrapolates to undesirable consequences/problems H o w , then, might a competing discourses perspective enrich the analysis of h o w such m e d i a texts are socially constituted - h o w they stand as textualized versions of cultural discourses, relative to other discourses realized by other texts (actual or potential)? Even to claim, as w e have d o n e implicitiy, that texi< like Text 1 articulate a n 'environmental' discourse suggests that there are other discourses with which they have to compete. A r e there discursive configurations already marshalled as compering discourses to oppose the imputations of scientificmedical o z o n e - m e l a n o m a discourse? First, there is a 'traditional' versus 'contemporary' opposition at work. The original articulation of the scientifically based statement of the ozone hole prob' l e m (about a decade ago) expressed a n e w p h e n o m e n o n , a n e w threat, deriving from n e w technical discoveries. T h e local story that drives Text 1 is based in i report of findings from new research. T h e issue of being aware of the health consequences locates us - as readers a n d as the people potentially at risk from cancers - as m o r e aware than w e were in the past, or as m o r e aware than 'trad- J i o p n n i w rpa c.ts aefainst old 1
COUPLAND AND COUPLAND BODIES, BEACHES AND BURN-TIMES
83
then arise. W h a t discourses articulate 'traditional' orientations to sun-exposure? Do they still have a currency, and if so, where? Also, are there competing contem porary discourses which similarly oppose o z o n e - m e l a n o m a discourse? Which, how and where? Relating to the first of these questions, our media sample suggests there is indeed a well-established, 'traditional' discourse of sun-exposure. Its key elements might b e summarized as in Figure 2, a n d w e might identify this configuration as a discourse of 'sun-is-fun'. LEISURE IS FUN SUN IS PART OF SUMMER LEISURE
is MARKED (relative to work) is RARE (in Britain)
LEISURE SUN
THE MORE SUN THE BETTER SUMMER LEISURE IS BODY-BASED BODY EXPOSURE IS SEXY AND FUN
Figure 2: Key elements of sun-is-fun discourse Sun-is-Fun Discourse The Western Mail (the self-styled 'national newspaper of Wales') of 12 J u l y 1994 carries a story u n d e r the headline 'Doctors issue a roasting on t h e hottest day of the year' (Text 2). T h e text begins b y characterizing the traditional British d a y at the beach ('bumper d a y for ice-cream sellers a n d deck-chair attendants, yesterday's hottest temperature of the year . . . ' ) a n d e m b e d s information about
Doctors issue a roasting on hottest day of year VEDXX IT w w a bojoper By SIMON fCTRY day fat I m n i W awOars sod disek-ehair attendants, day. Mash at Gwestt and rtvurdty'f hottest tem southern Powers hit IPC The reeding in Britain was perature of the year highest brought strong warnings at laWmhfSthi Suffolk, where tram doctors an the WswJfcef f t f f a noffon denge" at* skin cancer. prfcsa — Pou* 2 It kxih. only SO adnata* for Comment — Poos 8 svopk tn walastobora nnder the M K j y a ] son* A ben scald be eat step a a j f c j o T e i s wate treated to a rand, with 80*0 OSD afternoon. Fair lata. Shetland, recorded th* coolest with Just i r C ( 6 m today. The top temperature jost Wets* warned rulneraMe paopl* BO* to go oat usprotec- pipped tile year's •previoos peak of 3M*C <8*T) recorded t*dtothe 0»ns heat at Heathrow Airport on Jane TaBsarmteraa Mating to Tt.rC WTTt in Cardiff jestar- 24.
The "barn time" — or the avenge time a lair-skinned parson could spend in the son unprotected without burning — was 30 adnata* in SoothEast Wales yesterday and jnst flee adnata* loafer naar also sounded poihiuon warning* for high •ulphowiiractd* levels in the north of England and hick aJtMspo-dhnlds levels in •Usadott. Today, Cardiff weathermen forecast that it will remain hot and sonny, but it will start to cool off tomorrow when tem perature* are forecast to fall to 20-tt-C (BS-TOT), the monthly average-
84
DISCOURSE STUDIES
burn-times a n d cancer dangers within the story ('brought strong warnings from doctors o n the dangers of skin cancer'). T h e feature goes o n to endorse sun-is-fun direcdy in places: Temperatures soaring to 27.6°C (80°F) i n Cardiff yesterday M u c h of G w e n t a n d southern Powys hit 26°C. T h e highest reading in Britain was atLakenheath, Suffolk, w h e r e sun-lovers were treated to a 30°C (86°F) afternoon. Temperatures are reported to b e 'soaring' (an ethereal metaphor) and the representation that they have b e e n 'hitting' high n u m b e r s evokes a n annual competition to achieve the highest recorded figure (cf. ' T h e top temperatures just p i p p e d the year's previous peak'). T h e sun is a 'treat' for 'sun-lovers' w h o will r u e the fact that 'it will start to cool off tomorrow'. T h e contexting of burn-time information ('20 minutes i n South-East Wales yesterday') allows, a n d arguably encourages, readers to bracket t h e p r o b l e m status of the health-risk, o r at least builds i n t h e availability of a n alternative stance. Although t h e text clearly references 'burning', 'warnings' a n d 'dangers' of 'going out unprotected' mucb as Text 1 did, it also offers sun-is-rare as a mitigation of sun-as-risk, a n d offer& sun-is-fun as a legitimation of b o d y exposure. This pattern of intermixing discourses is repeated in m a n y different texts in the sample, for example Text 3 (Try Noon' from the Sun newspaper of 12 July 1994).
FRY NOON
You can get skin cancer in the local park as easily as on a Greek beach
*J DAW** HCESOM B A L M \ Britain is putting Mediterranean fiotspois in the shade, with temperatures roday expected to top yesterday's 86° F BUT tt it ALSO CASTING a SHADOW OF {TENTH AVER THOUSANDS NF SUN WOR SHIPPERS M THF FORM <\( SKIN CANCER
th* iiih j * high in i*>* ihy Evca ao. poll* i u f | u : lh»' « n l y a third of Bntofli h*v» c h a n c e d iheiv lunbathtnt habira, with juat flva per WII of women saying they *™&> prefer to b e pale and I wttl*F and a fifth nill nixH'ia *>• t u n b i t h r It " . r* t inportuntty ll akin I«NC«I - • *-flnftii »arly Ui" chancvf o> » rW* • r e good more than t^- p»r Cere ttf iturrerero ^ • riwer than Pve vearr ,
Over t o * latf «wi year* there has b**n a 5fi p a r c r n i . 1 K tn Britain'* 4*«th r a | r from t h e i l s M s r
more on t h e lorsb Around 3,300 people a, j M T i n Britani develop t h e lethal malignant airlanoma and 111,000 th* l*ss ifllV y«*l ll it til J BIKT*eriout but still dangerou* orrtoru. romporfd with lAS In non-melanoma iktn cancel lfl74—a m e OR M? » r r e m Hona MacKie arofpnoi ... The only cancer t o kill more »i Gla*»ow 'Urn*IB of the luftft hut the unn >j dermatology nrilty. say*: Th? mtnaff flostnc fa*i a b a u l protecting vi*rr*i»e» Women who i«nd tr- ATA, fnm the sun'g ay». I« gwiuig .«! in the aun lonarr. cui through biJi peop]* thin) it number male virtim* three (o only e p u l i s wh ie^n, ar shirti o(fv. loSi. p r a t e r ' m m i tokrs
Young: Wifi
(
midday wilhowl a n ; .irtd they'd q u i ' " nghtl,, think you were m a d They don't * w m to Craxp th*i an hour in t h r Jot-el park *( runcht'rne can damage y o u r > k t n *n e w M j the n n e wiy Pun of t h e problem ih* lotiff t l m - )B# Dtiwetn whan thr sklr. U damaged nd " h e n cani**r nevclopa 'it r a n take beiwvei i t u -rtd Ml year* <-IR i h * -fTrrti :o J» seen " r h e Mcl Ofln-r now IRMI*_ v.«ira v l o l « radiation warnin* * during iwrnmc. feoekemis ThV amount oi W]\ pari. .nB itw arauna: i hiiflteM iwlw—"i • Jam "'ft 3p"> -vbe-i1
fha#» m o i l at *rt I*J aalnnad peoplt esp«vta" ihaaa with reef hour anil (M-I lw—Scotland haa « e t i «n* it the Mjgraf >wea in CM** with W o - - ^-o» mo?- ifeanr Fhe aey ..rnptcm* < tor is any chanc* >ur Ut» turraee If • brown pal if pear* tnai waa not ilm* lv !an. at if a molt *vu paler or cfMnfv* M a x ol bleeda * e - > JTINI' Immediately Dr inhn hawk --N*UII»JM (
5
DETMALNLOGIKL at SI {^nrami
TlmpiU), I-ONDUN, u a - UrmVOU WON' (HRNK LL " (J"i*l* «Ufc t/« b - »'«iwr» V.*it r#
BDLLNH' ITL ,i taint** will
|*«H*
'Vh*. , „ ,-JT,,.
COUPLAND A N D COUPLAND
BODIES, BEACHES A N D BURN-TIMES
85
This is ostensibly another report about skin cancers, and the text again expresses the associated health-threat in terms closely reminiscent of the previous ones we have considered. But its opening lines project high temperatures as being, once ;igain, good news ('Balmy Britain'), and in places the text glorifies t h e m a n d marks them as achievements (expected to top yesterday's 86°F'). H i g h tempera tures in Britain are unusual, a n d it is newsworthy that Britain, for once, is out doing the Mediterranean ('putting Mediterranean hotspots in the shade'). I n Text 4, the Daily Star's very similar lead-in to its story of the same day is: 'Stick the Med . . . Britain's real h o t stuff. But the implied competition is over more than temperature itself. I n Text 4, heat is overtly constructed as desirable, s i n c e those 'under a cooling cloud' are 'some unlucky Scots'. Britain's being 'hot sniff evokes a competition between E u r o p e a n regions a n d races to b e recognized as centres of sexual or at least sensual experience. T h e practice of bodily exposure to the sun is represented as 'basking', 'sweltering' and 'broiling', images which m i x ileal, moisture a n d passivity in ways that are as evocative of receptive sex as they are of some m o d e s of cooking ('roasting'). As later examples suggest, the lexicons of tabloid sex a n d of s u m m e r heat (which itself is linked, of course, to bodily exposure) are entirely consonant with one another. T h e U K tabloid semi-nude is as likely to b e characterized as 'sizzling' or as a scorcher' as is a h o t summer's day. At the same time, the p h e n o m e n o n is not purely tabloid. A headline from The Times (also of 12 J u l y , n o t copied here) is: 'Britain to sweat for just one m o r e day', a n d the feature goes o n to c o m m e n t that 'most of the country continues to bask in sunshine'. Our argument this far, then, is that there is a recurrent ambiguity in m a n y media texts' formulation of o z o n e - m e l a n o m a discourse, which, interspersed with reports of health-risks a n d warnings of skin damage, offer a 'traditional' m e a n s of resisting its evidentiality a n d its applicability to individuals, perhaps especially Brits. The Sun's 'Fry N o o n ' piece (Text 3) itself reports the findings from 'polls' that "just five percent of [British] w o m e n . . . [say] they would prefer to b e pale and healthy and a fifth [are] still rushing to sunbathe at every opportunity'. (This is repeated almost verbatim in The Timer H o m e News column of that day.) This gap between risk-information a n d day-to-day practice m a y b e partly explained by the rhetorical alignment w e find in s o m e m e d i a texts against their own imprecations towards safer sun-use. A text such as Text 4 is therefore a selfsubversive text, aligning against k n o w n eco-risks in the pursuit of hedonistic fulfilment and sensual sun-experience. As it suggests, 'Most h u m a n a n i m a l s . . . ' can persist with bodily exposure 'in the h o p e of getting a good, safe tan' [even] 'after the Government issued skin cancer alerts'. T h e r e are k n o w n risks, 'But at the moment hardly a n y o n e cares . . . in the top sun spot of Europe!'. The gendering of sun-is-fun discourse is implicit in the texts w e h a v e so far considered. But several other instances from the sample juxtapose verbal articu lations of ozone - m e l a n o m a discourse with stereotypic visual images of y o u n g w o m e n in bathing suits. For example, the page 15 spread of T h e Daily Express (7July 1994, not copied here) has the full-page headline 'Doctor's pin-up insult'. -
FINE-O? TOO HOT London Weather Centre B H JOHN ItaUftHNCf spokesman confirmed unlucky Scots under a The main problem may cooling cloud — were just be the usual one — jams, For t h e average 86*F heat which applying lotion to their with people.- heading for has u s sweltering will last at least a bodies in the hope of get coast m droves as they day or t w o more A n d it could e v e n ting a good, -afe tan after the did yesterday hit NINETY, say weathermen the Government issued There were tailbacks on As yesterday's top temperature — the skin cancer alerts. routes to Blackpool and besi. for three years — reached 88* at FJFCRNHCATLI IN SUFFOLK, Scarborough And a 10Northolt, Middlesex London was also WFTB 86~. WAS MS broiling mile one bulk up on the warmer than the Mediterranean mm Bombmj mad red-hotter in Essex But while that was good news for THAN ROME, Nicosia. M25 The extra exhaust fumes; almost everyone, the capi ISTANBUL AND BARCELONA. could also cause smog. tal's zoo keepers put a Even rainy Manchester But at the moment damper on our sun lun by basked under clear skies hardly cares in asking us to spare a and 79* . though further the topanyone sun spot of Europe thought for their RHINOS. north the sun definitely didn't have its hat on as FIVE DAILY STAR LIBRARY WORKERS WERE TOWTING THEIR the mercury in the Fair SUCCESS LAST NIGHT ALTER SCOOPING £10,00« IN a RADIO For black species favour Isles stopped at 54* SHOW BREMKFSST BONANXA. ANDY WILLIAMS ENTERED HIM ites Rosie and Josh hawBut m the south, the 46- GOLFER Jeremy T. SELF AND WORKMATES JOHN GRAY, DAVID OODDARD, PAT had co be given rubdowns year-old July record of 93* Brenno. 16, died when he BRIDIES AND MARTIN EDWARDS IN THE TREBOR EXTRA with moisturising cream to could be under threat, slammed his club against a Strang MINTAL CHALLENGE ON VIRGIN \XiS* RUAS AND prevent their thick skins before the temperature bench after a bad shot, and drops Idler in the week •FONO'S BREAKFAST SHOW THEY ALL HADtoTONEIN at THE cracking in the heat' the broken shaft snapped Most human animals, SAMETIMEAND THEN RING- IN TO CLAIM THEIR PRIZE. "It's been very warm but back and pierced his heart, though — apart from somt may «et warmer still ' a at Gloveraville Nev. York
S T I C K t h e M e d . . . Bri t a i n ' s r e a l h o t stuff.
Breakfast toast
Text 4: The Daily Star, 12 July 1994. © Express Newspapers pic.
1
Ullllflllil
O O
G
JO BS
w c D
M C/I
COUPLAND A N D COUPLAND
BODIES, BEACHES AND BURN-TIMES
87
the headline story, u n d e r the sub-head ' W o m e n get a raw deal o n the N H S [National Health Service], claims Minister'; a n d (rightmost) the story of high risks even with the use of suncreams, u n d e r the heading ' S U M M E R S U N T A N S H I D E C A N C E R R I S K ' . Physically dividing (or linking) these stories is a n untitled image of a y o u n g w o m a n wearing a swim-suit and jewellery, a p p a r e n d y in bright sunlight, in a submissive pose, her hands held u p to her hair. T h e image lends itself as potential support for the 'pin-up' headline, since it can b e seen as an instance of an actual 'pin-up' photograph. But it also visually contextualizes the verbal o z o n e - m e l a n o m a text to its right, again offering a subversive reading of that text. T h e image's stereotypic qualities conjure traditional, tabloid inter pretations of female sexuality which are then proxemically forced into interaction with the right-panel's headline ' S U M M E R S U N . . . ' . T h e r e are m a n y such instances, including the Daily Star's (11 J u l y , p . 5) close-up of Kim Basinger, pouting, apparently in illustration of the headline 'Star Kim's lip cancer o p agony'. T h e text contrasts two themes: o n the one h a n d ' H O L L Y W O O D superstar Kim Basinger', 'Beautiful blonde K i m ' (and the photo graphic image); o n the other h a n d , 'her agonising batde against cancer', 'painful laser surgery o n tumours that appeared o n her lips', 'the most painful experience of her life', 'It took three nurses to hold m e down. T h e y stuck a large needle u n d e r m y lip', 'they lasered off the cancerous parts' (and the headline). As in previous instances, o z o n e - m e l a n o m a discourse is here in competition with traditional tabloid visual a n d verbal imagery of female sexuality. W h a t is a d d e d in this case is an implicit element of moral chastisement, using the o z o n e - m e l a n o m a issue apparently to chastise those whose lifestyles are hedonistically linked to the sun. Kim is said to b e 'allergic to the sun', to b e 'married to movie h u n k Alec Baldwin', and to h a v e h a d a 'surfer ex-husband'. H e r e w e get a fleeting glimpse of o z o n e m e l a n o m a discourse as part of a shifting morality of anti-hedonism. Skin protec tion as part of (one form of) eco-awareness can b e constructed as a n e w asceticism, aligned against the traditional h e d o n i s m of sun-is-fun. O n other occasions, however, 'sun-is-fun' gains unqualified approbation. Today newspaper's piece (12 J u l y , p . 11, not copied), h e a d e d '86° Absolutely Sweltering', places its verbal text against a photo of forward-leaningjoanna Lumley, tanned and wearing a low-cut T-shirt:'... hottest day of the year - several degrees more sizzling than the Greek islands enjoyed . . . T h e best of British sunshine . . . In the South of France, Nice failed to top 80 f a h r e n h e i t . . . A n d , Today's weather m a n Philip E d e n warns: "Make the most of it. Things will cool down in most places b y the weekend. T h e n o r t h is suffering a l r e a d y . . . " ' Overall, the sample suggests that the tabloid press construct sun usage primarily as a deserved reward, a treat. This is a pro-fun, anti-political correctness, and in its o w n terms even a libertarian discourse. Its stance is distinctively British working class (since 'the M e d ' has ' m o r e than its fair share', a n d 'working p e o p l e deserve their day out', etc.). Sun-use, through (predominandy female) b o d y ex posure, falls easily into the pattern of 'just a bit of fun' tabloid representations of leisure. It offers a ready counterposition to the serious, scientific, medical, rational discourse of the o z o n e - m e l a n o m a discourse.
88
DISCOURSE STUDIES
Body-Culture Discourse But contemporary discourses of the body, even in Britain, h a v e in some ways transcended these tabloid representations. In late-modern societies, the b o d y is semiotically too important, a n d has too m u c h definitional potential, to function m e r e l y as a solar panel for hedonistic leisure. At the same time, it is too potent a signifier to b e h i d d e n from view, u n d e r the pressures of scientistic priorities. 'Laying out' is not what all late-moderns d o , b u t body-culture m o r e generally is o n e of their focal concerns. A n d b o d y discourses refashion and u n d e r m i n e both pro- a n d anti- 'post-ozone' rhetoric to a significant extent. Bodily exposure to the sun entails a bio-medical health-risk, b u t it also entails a risk to personal and social standing often referred to as 'premature ageing'. Social stereotypes of old age are in m a n y respects based o n physical (including physiological a n d physiognomic) images a n d definitions of ageing, b u t sagging flesh a n d wrinkles are undeniably also the hallmarks of (genuine) d e e p old age. Woodward's (1991) comments about the 'unwatchability' of old age, based o n her observations of responses to fine art portrayals of elderly n u d e s , is apposite. T h e notion of'gerontophobia' is also relevant here, encapsulating m a n y people's experience of reticence and even revulsion w h e n confronted b y others', a n d no doubt the prospect of their own, ageing. T h e r e is a traditional (liberal humanist, a n d sometimes religious) discourse that works against attaching any significance to the b o d y , a n d therefore to the ageing b o d y ('beauty is only skin-deep . . . age cannot w i t h e r . . . ' ) . But in late or high modernity the b o d y has b e e n i m b u e d with remarkable symbolic status, to the extent that, for some, it is the single most potent signifier of selfhood. Not only Giddens, but also Shilling (1993) and Featherstone (1991) are among those w h o h a v e recentiy captured, at the level of social theory, the general char acteristics of high m o d e r n body-culture. F r o m their writings, w e can derive the following elements of another putative discursive formation, which we can refer to as body-culture discourse (see Figure 3). I n relation to ageing, it is v e r y clear that body-culture discourse encodes an ageist, anti-elderly ideology. A s Shilling writes, the body-as-symbol is par excellence the domain of the 'quintessentially m o d e r n individual w h o is y o u n g a n d never dies' (1993:196). C o n s u m e r culture, in Featherstone's account, recycles ageist images, m a n y of which centre o n the youthful b o d y : 'Certain themes, infinitely revisable, infinitely combinable, recur within advertising and consumer culture imagery: youth, beauty, energy, fitness, m o v e m e n t , freedom, r o m a n c e , exotica, luxury, enjoyment, fun' (1991: 174). T h e associated moral agenda is again largely a mass m e d i a imposition: 'Adver tising, feature articles a n d advice columns in magazines a n d newspapers ask individuals to assume self-responsibility for the w a y they look. This becomes important not just in the first flush of adolescence and early adulthood, for notions of "natural'' bodily deterioration and the bodily betrayals that accompany ageing b e c o m e interpreted as signs of m o r a l laxitude' (Featherstone, 1991: 178)
COUPLAND A N D COUPLAND
BODIES, BEACHES AND BURN-TIMES
89
THE BODY EXPRESSES THE INDIVIDUAL WE HAVE A MORAL CONTRACT WITH OUR BODIES HEALTH IS EXTERNAL AS WELL AS INTERNAL RECREATION IS WORK (THE NEW ASCETICISM) THE BODY IS A COMMODITY AND A FORM OF CAPITAL CONSUMERS CAN BE MODELS, BY BODY SHAPING AND COLOURING BODY-CULTURE RESISTS/DENIES AGEING BEAUTIFUL BODIES ARE YOUTHFUL BODIES YOUTHFUL BODIES ARE HEALTHY BODIES HEALTHY BODIES ARE (OFTEN) TANNED BODIES
Figure 3: Key elements of body-culture discourse 'create' ourselves, to an extent including physically, from all manner of role-models offered to us, particularly via the media. I n a consumer culture, w e can c o m m a n d the belief (ultimately illusory, b u t highly seductive a n d involving) that w e can transcend t h e lifespan a n d our o w n ageing (Featherstone, 1991: 177). Against this background, discourses i n the d o m a i n of sun, risk, health a n d tanning take o n n e w meanings. Body consumerism a n d ageism can infiltrate a n d colonize b o t h o z o n e - m e l a n o m a a n d sun-is-fun discourses. I n late modernity, science-based a n d health-based rationales for limiting sunexposure c a n b e supplanted b y fear-of-ageing (gerontophobia), b u t to the same practical effect that is, sun-avoidance. If (young) m o d e r n s d o limit their sun-exposure, it m a y be under pressure from a wholly nonenvironmental discourse, which asserts attract iveness over health, personal a n d social priorities over responsiveness to scientific facts, and perceived youthfulness over actual longevity. Exposure to sun brings 'premature ageing', a n d ageing is what body-culture has b e e n developed to resist. 4
But there is an internal dilemma. For body-culture to b e enacted, bodies n e e d to be seen. Similarly, bodies n e e d to b e 'displayed' a n d 'advertised' if they are to enter the relational market a n d realize their symbolic capital (Bourdieu, 1977). Investment in the b o d y can only see a return if bodies are there to b e traded in the open marketplace. I n fact, there are few fully legitimate public display-sites for bodies other t h a n beaches, a n d bodily disclosure as part of s u m m e r leisure therefore has a particular salience in late-modernity. Scantily clad bodies (espe cially female) m a y p o u r from the mass media, b u t people's desires to role-match, even to b e c o m e the m e d i a images, are severely curtailed. From the consumerist standpoint, the fun element of sun m a y therefore b e overridden b y the sun's role as a facilitator for, a legitimizer of, bodily revelation. S u n declares the market place open. Media texts in our sample again reflect the interpolation of bodyaesthetic, anti-ageing (and ageist) concerns into the o z o n e - m e l a n o m a a n d the traditional sun-is-fun discursive formations, either d i r e c d y or indirectiy. T h e frequent references to 'damaging y o u r skin' (e.g. the 'Fry N o o n ' text, Text 2, earlier) im plicate the risk of so-called ' p r e m a t u r e ageing'. T h e ' P i n u p ' text that we referred to (from the Daily Express, 7 J u l y 1994), like m a n y others, suggests that even UVAraVS 'cause aceinff'. And Vflh thp Slinf-annprl V i n r l v t i n e a n u r c i c t e n t - c o m i n t i c
90
DISCOURSE STUDIES
continues to b e a population (and we h a v e seen that, according to m e d i a reports a n d skin-cancer rates, it is a majority) w h o will forego the benefits of sunavoidance, whatever their ideological bases. O n e possible response is to reassert the hedonistic short-termism of sun-isfun. As Shilling writes: Investment in the b o d y also has its limitations. I n d e e d , in one sense, the effort e x p e n d e d b y individuals on the b o d y is d o o m e d to failure. Bodies age a n d decay, and the inescapable reality of death appears particularly disturbing to m o d e r n people w h o are concerned with a self-identity which has at its centre the body. After all, what could signal to us m o r e effectively the limitations of our concern with the y o u n g a n d fit, ideally feminine or masculine b o d y than the brute facts of its thickening waistline, sagging flesh and inevitable death? (1993: 7) Confronting the inevitability of physical decrement, we might shorten the time-span of our physicalized self-identities, a n d decide to live in rather locally relevant phases ('while I ' m y o u n g e n o u g h to b a r e m y b o d y o n the beach'). But this is a radical stance, a n d consumer culture is ready to step in with a n ideological solution to a lived d i l e m m a w h e r e it can b e turned to economic advantage. It is u n d e r these circumstances that medical conglomerates and their advertisers have promoted the notion of 'healthy tanning'. Healthy tanning enacts the late-modern imperatives of ascetic body-culture, by-passing the dated h e d o n i s m of sun-isfun, a n d pays lipservice to the scientific respectability and personal responsibility required in ozone-melanoma discourse. But above all it legitimizes b o d y exposure to the sun a n d reinvents this practice as an entirely rational lifestyle choice. It is a classically commodified solution to an ideological dilemma, a n d it is hardly surprising, therefore, that the largest category of m e d i a text-types in our sample is advertisements selling anti-wrinkle, safe-tanning products. Consider Text 5, page 2 8 of theJ u l y 1994 issue of Prima magazine (a magazine directed at y o u n g w o m e n ) , the beginning of a feature tided 'It's O K to tan'. T h e text, like its title, is essentially a legitimation, cast as a n e w revelation ('everything has changed'). I n fact, it offers an apparently n e w solution to an im plied dilemma. It reasserts skin 'tanning' over skin 'burning', and glamorizes dark (or 'golden') skin. 'Risk' of skin ' d a m a g e ' is still clearly referenced - in fact the twin risks of m e l a n o m a a n d visible ageing - but t h e y are presented as familiar and controllable ('for the past decade we've b e e n constantly w a r n e d ' , ' . . . thor oughly d r u m m e d into us'). T h e magazine, a n d the technology it introduces to readers, offer the means of control. Sexual allure ('a fabulously sexy golden shade'; ' a glorious golden glow') is to b e earned through a n urgent ('vital') regimen of self-attention ('a properly achieved tan'; ' a sun-tan, carefully gained'; 'what you mustn't d o is burn'). Sun use is reinstated as a legitimate late-modern practice, rescued from the d r a b asceticism of o z o n e - m e l a n o m a discourse which has been superseded (we are asked to believe) b y a n e w asceticism of control within the
COUPLAND A N D COUPLAND
BODIES, BEACHES A N D BURN-TIMES
91
This sinner it is Mflitelyokayto tail- bstwhat yew mustn't ao is burn. Your Prima UuninjcardiiN to vital on holiday, helping you to a fabuioftsiy sexy
Y
y«&, tt & rally ouef How yee am have that gtorioui golden glow without putting yoai 1,-jWiiindjoursldnaliisJt.Foi i v pur decade w e ' w - b m ir-~u*ntly warned about the - 'V-'ofdHMill'lutmwiDlel i > v fhe rides of maeatad -.. yini, ageing awl. mow i'- r"rtan!ly, skin ameer, havt nem Hi .highly drummed into ... fl'ttno* it would *5etDiba» c . i ihinf Ma changed A nroptny a c h l m d tan p r o t e c t s y o u r rtta Nc«revOTlishowilhat«»unMl, careful!' gained using aijh'fAcicir uinacreera, could •ctuiil r pnxect your skin from **, icn>; term damage. Sun burn is the n u k r culprit when it emu to .sin dotage: so U you ran avoid it, ihen you
Text 5: Prima,July 1994. © Gruner &Jahr, U K
Retrospective We began this analysis by identifying one strand of an 'environmental' discourse
92
DISCOURSE STUDIES
exposing skin to t h e sun, linked to ozone depletion in the u p p e r atmosphere. It would not b e true to say that this scientific discourse is eclipsed b y alternative formations. M a n y texts i n the sample d o refer to such risks, a n d build their own meanings in relation to a discourse of ozone-melanoma. To that extent, the textual data suggest the science-led discourse of risk appears to h a v e received a reason able exposure in the U K print media. But t h e m e d i a sample also shows that there at least two different, altern ative ideological a n d rhetorical formations which compete with this particu lar e n v i r o n m e n t a l discourse - w h a t w e called sun-is-fun a n d body-culture formations - often within t h e same m e d i a texts. If w e look back at w h a t w e posited as t h e k e y elements of t h e 'environmental' discourse, w e c a n in fact see h o w its competitors marshal themselves to resist its impact, b o t h b y refutation a n d b y colonization. Sun-is-fun is essentially a refutation, a refusal to recognize one set of priorities a n d t h e imposition of an alternative set: it asserts leisure over science; the persistent rarity of sun (in Britain) over a perceived increase in environmental risk; hedonism over asceticism; a 'bit of fun' over serious scientific debate; the working person's priorities over political correctness; traditional values over n e w science. I n contrast, body-culture discourse largely colonizes environmentalism. It over lays t h e science of skin protection o n t o p of the science of risk; it relocates risk within the individual's use of chemical treatments rather than in large-scale changes to t h e environment; it overlays t h e asceticism of self-protection o n t o p of the asceticism of sun avoidance; consumerist solutions o n top of natural solutions; individual responsibilities o n top of c o m m u n i t y responsibilities; a n apparendy n e w science of safe tanning o n top of the newness of ozone depletion science. But body-culture also colonizes sun-is-fun discourse. It technologizes leisure and makes 'work' out of 'laying out'; it reinvents 'laying out' as a n identity project; it adds a narcissistic dimension to b o d y exposure; it commodities s u m m e r leisure as part of consumer culture. Of the various opposing formations, therefore, it seems that the complex of body-culture priorities h a s the strongest potential to resist environmentalism, at least in t h e specific context of ozone layer depletion. T h e discourse processes w e h a v e examined i n this analysis can in fact b e seen as a coherent set of devices acting to p r o m o t e what Giddens (1991) calls t h e sequestration of nature - latem o d e r n societies' practice of obscuring the facts a n d imperatives of the natural environment through t h e distorting effects of mass-mediated institutions. The most profound potential interpretation is that Beck m a y similarly b e correct in arguing that knowledge of environmental hazards does not, i n itself, constitute an avenue for remediation - a n escape route. Individual j u d g e m e n t about the environment, h e suggests, is n o longer anchored i n individual experience (Beck, 1995:52). T h e print m e d i a which establish environmental risks themselves lead us into commodified 'solutions' which fail to address the underlying environ mental concern. M o r e generally, w e h o p e to h a v e established that the juxtaposition of 'lan„,.„„„' iAie,~m*-Bai ..rUV. «fVio a m H r n n m o n t ' i c not- inhprpntlv 'inannrnnriate' or
COUPLAND A N D COUPLAND
BODIES, BEACHES AND BURN-TIMES
93
constructive for environmental research at a fundamental level. I n this p a p e r , we have adopted social semioticians' (or critical linguists') approach to context. Part of H o d g e a n d Kress's (1988) thesis, for example, is that ' T h e site in which a text occurs typically contains instructions as to h o w it should b e read and what meanings should b e found in it' (p. 68). Context is therefore taken to incorporate sets of obligations a n d predispositions, which act as constraints not only on which messages should b e conveyed a n d h o w they should b e conveyed, but also o n how messages should b e interpreted. This feeds direcdy back into a community's dominant value systems and, ultimately, its social a n d political organization. This perspective has o p e n e d u p the possibility of reading 'environmental' a n d other culturally specific values from m e d i a texts. But it is a relatively small step to reconceptualize 'context' as including 'envir onment', even in the senses of 'natural' and 'global' environment. Ultimately, then, 'environment' is m o r e than just another 'social issue' that can b e addressed by linguistics. Linguistic theories of context m a y help us understand h o w the concept of 'the [natural] environment' has traditionally b e e n interpreted, and how it m a y usefully b e re-interpreted. For example, the relationship between linguistic behaviour and context has traditionally b e e n one of'figure' to 'ground' (Goodwin and Duranti, 1992; Hanks, 1992). T h e analysis of language a n d communicative behaviour has traditionally entailed a focal concern with the structural a n d semantic properties of the be haviour itself (the figure), a n d an almost universal under-attentiveness to context (the ground). This abstracting of behaviour from context has a p p e a r e d natural, for example because the structural analysis of language-as-code is analytically more secure than any form of structural analysis of context. Human action more generally seems similarly to offer itself for analysis independently of context, in that we can characterize our lived fives as a series of internally structured local activities - going to work, doing leisure, having relationships, etc. - each with their own event structure. Yet a current consensus amongst context theorists is in challenging the appro priateness of the figure-ground relationship. Language, that is, can n o longer b e said to exist 'within' a context, from which it is theoretically independent: these are the 'increasingly m o r e interactive a n d dialogically conceived notions' of context that G o o d w i n a n d Duranti refer to (1992: 1). M e a n i n g derives from the interplay between language texts a n d context, taken to include other textualized meanings. This is Volosinov/Bakhtin's principle of 'dialogic' utterance a n d 'multiple voicing' (e.g. Bakhtin, 1981). Texts find their meanings in and in relation to the contextual meanings that comprise societies, cultures and ideologies, just as those contexts are continually b e i n g confirmed, challenged or refined in the generation and reception of n e w texts. It follows that language, a n d in fact all social action, is a c o m p l e x of processes that, inescapably, dialogically implicates and actively valorizes the natural environment. Environmentalism, in any a n d all of its guises, m u s t look to language for its constitutive core. A s linguists, w e 11 J J. . . . . . »
94
DISCOURSE STUDIES
Notes
We acknowledge permissions to print copies of original newspaper and magazine extracts granted by the following: The Telegraph, The Western Mail, The Sun, Daily Star (Express Newspapers), Prima (Gruner &Jahr, UK). 1. The published rationale for the symposium referenced as Alexander et al. (1993) was given as follows: 'Areas of work to be presented and discussed include questions such as: How are ecological problems articulated in texts? What do we learn about ecological problems through various texts? Which syntactic, semantic and pragmatic means are used in communication about ecological matters in different contexts? Which linguistic methods are appropriate for this kind of language research? Which theories of language are most adequate for investigation of the ecological problematic?' Many writers involved in those proceedings acknowledged Michael Halliday's seminal contribution to this developing field in his keynote address at AILA (The International Association of Applied Linguistics) 1990 (Halliday, 1990). 2. We are very grateful to Pam Perkins who collated and annotated the media print sample, and contributed important insights to our analyses. The sample was compiled under the guidance of Allan Bell, who has been involved in all aspects of this phase of the study. Chris Lane and Janet Holmes have also made important formative inputs into this work. 3. An explicit aim of our future work is to contrast northern and southern hemisphere mediated and lay discourses about the environment. 4. Giddens's position is: 'The reflexive project of the self, which consists in the sustaining of coherent, yet continuously revised, biographical narratives, takes place in the con text of multiple choice as filtered through abstract systems. In modern social life, the notion of lifestyle takes on a particular significance. The more tradition loses its hold, and the more daily life is reconstituted in terms of the dialectical interplay of the local and the global, the more individuals are forced to negotiate lifestyle choices among a diversity of options' (1991: 5).
References Alexander, RJ. (1993) 'Introduction to the Aims of the Symposium on Ecolinguistics at AILA 1993, Work so far and Some Ecolinguistic Principles to Pursue', in Papers fir the Symposium Ecolinguistics: Problems, Theories and Methods. Odense University: Research Group for Ecology, Language and Ideology. Alexander, RJ., Bang, J.C., Door, J. and Verhagen, F.C. (1993) 'Symposium on EcoIinguistics: Problems, Theories and Methods', paper presented at the Association International de Linguistique Allique, Xth World Congress of Applied Linguistics. Amsterdam, August. Bakhtin, Mikhael (1981) The Dialogic Imagination, ed. M. Holquist, trans. C. Emerson and M. Holquist. Austin: University of Texas Press. Beck, U. (1995) Ecological Politics in an Age ofRisk. Cambridge: Polity Press. Bell, A. (1989) 'Hot News: Media Reporting and Public Understanding of the Climate Change Issue in New Zealand', Project report to the Department of Scientific and Industrial Research and Ministry for the Environment. Wellington, Victoria University,
COUPLAND A N D COUPLAND
BODIES, BEACHES AND BURN-TIMES
95
Bell, A. (1991) 'Hot Air: Media, Miscommunication and the Climate Change Issue', in N. Coupland, H. Giles andJ. Wiemann (eds) 'Miscommunication'and Problematic Talk, pp. 259-82. Newbury Park, CA: Sage. Bourdieu, P. (1977) Outline ofa Theory of Practice. Cambridge: Cambridge University Press. Candlin, C.N. and Maley, Y. (in press) 'Intertextuality and Interdiscursivity in the Discourse of Alternative Dispute Resolution', in B.L. Gunnarsson, P. Lanell and B. Nordberg (eds) The Construction ofProfessional Discourse. London: Longman. Douglas, M. and Wildavsky, A. (1982) Risk and Culture. Berkeley: University of California Press. Goodwin, C. and Duranti, A. (1992) 'Rethinking Context: An Introduction', in A. Duranti and C. Goodwin (eds) Rethinking Context, pp. 1-42. Cambridge: Cambridge Univer sity Press. Fairclough, Norman (1993) 'Language and Ecology: An Exploratory "Position" Paper', from discussion at a symposium at the University of Lancaster, April. Featherstone, M. (1991) The Body in Consumer Culture', in M. Featherstone, M. Hepworth and B. Turner (eds) The Body: Social Process and Cultural Theory, pp. 170-96. London: Sage. Giddens, A. (1991) Modernity and Self-Identity: Self and Society in the Late Modem Age. Cambridge: Polity Press (in association with Basil Blackwell). Halliday, M.A.K. (1990) 'New Ways of Meaning: A Challenge to Applied linguistics', Journal of Applied Linguistics 6: 7-36. Hanks, W.F. (1992) 'The Indexical Ground of Deictic Reference', in A. Duranti and C. Goodwin (eds) Rethinking Context: Language as an Interactive Phenomenon, pp. 43-76. Cambridge: Cambridge University Press. Hodge, R. and Kress, G. (1988) Social Semiotics. Cambridge: Polity Press. Lee, D. (1992) Competing Discourses: Perspective and Ideology in Language. London: Longman. Michael, M. (1991) 'Discourses of Danger and Dangerous Discourses: Patrolling the Borders of Science, Nature and Society', Discourse & Society 2(1): 5-28. Nelkin, D. (1985) 'Introduction: Analysing Risk', in D. Nelkin, The Language of Risk: Conflicting Perspectives on Occupational Health, pp. 11-24. Beverly Hills, CA: Sage. Shilling, C. (1993) The Body and Social Theory. London: Sage. Woodward, K. (1991) Aging and its Discontents: Freud and Other Fictions. Bloomington: Indiana University Press.
20 Figures of Speech: Figurative Expressions and the Management of Topic Transition in Conversation Paul Drew and Elizpbeth Holt
O
ne feature of conversation that contributes to its colloquial character is that, from time to time, speakers use certain forms of idiomatic expressions which may be regarded as FIGURES O F SPEECH. The following fragment from a telephone conversation illustrates this kind of Unguistic object, the use of which w e shall examine in this article. (1)
[Field:X:(C):l:l:l:6] Lesley: .hh He wz a (.).p a buyer for the hoh- i- the only horse hair fact'ry left in England. Mum: Good gracious, (0.3) Lesley: And he wz their buyer. (•) Mum: Hm::: Lesley: .t Mum: Hm:. —> Lesley: So hg had a good innimgs did [n't he. Mum: [I should say so:
Just before this fragment, Lesley has told her mother about the death, at the age of 79, of a mutual acquaintance. Here, instead of describing this man's long and full life in some literal fashion, she chooses instead the figurative expres sion (arrowed turn) "he had a good innings" - an image derived from the game of cricket.
D R E W A N D HOLT
FIGURES O F SPEECH
97
The matter of a speaker's selecting a figurative expression, in preference to a more literal way of saying or describing something, is particularly visible when the speaker changes what was going to be a literal or non-idiomatic form and replaces it with a figurative expression, as illustrated in the self-repairs (arrowed) in the following extracts. (2) [US:20] Vic: Dey took their business fuh three days en ney —» came back en, took care a' her dey g i - .hh —» showered her with, whatever kinda gifts they wanted. (0.7) Vic: In her mi:nd stuck the showering of gifts. (3) [Field: 86A] (Lesley has called her husband, Skip, at work, to tell him that the police have been round to investigate the burglary of their home the previous night) Lesley: An they said OH YOU'RE VERY PRECISE when I said what ti_:me (.) I thought it was, (0.2) Lesley: BU' FORTunately I'd had a peep at the cjoxk .hh (1.2) Skip: Well I - 1 knew about what time it was I knew it was about (1.1) Skip: Ten- about uh (ten- say) between ten and twenty paiist. (0.2) Skip: (Obviously knew ) Lesley: .hhh Well I said it was about three twenty one I —> think because by the time w e - (.) w e - the penny —> had dropped an' we got up again that was three twenty finve, In 2, it looks as though Vic had been going to say that, as a way of "taking care of her," they "give" her (gifts). However, h e cuts off "gi-(ve)" and replaces it with the figurative "showered her w i t h . . . gifts." In 3, Lesley alters what she had been going to say at "because by the time we-"; judging from the way she con tinues her turn, that projects either "by the time we r e a l i z e d . . . " or "by the time we got up again." In any case, she changes it to "by the time . . . the penny had dropped," thereby either replacing "realized" with "the penny had dropped" or inserting that expression into an otherwise non-idiomatic account. In each case, the speaker starts to produce a literal version of something, but then does a selfrepair which substitutes that version (or possibly, in 3, adds to it) with a figurative description. In such cases, a speaker's SELECTION of a figurative expression is particularly exposed in self-repairs involving the substitution of a figurative for a more literal description. But it is generally the case that the use of figurative expressions in conversation is a property of the D E S I G N of turns at talk. As with any other fea ture or component of turn design - lexical, syntactic, intonational, or prosodic a figurative expression may b e selected as part of the way in which a turn at talk
DISCOURSE STUDIES
98
sequential properties of turns that h a v e t h e selection of a figurative expression as o n e of their design features. Parenthetically, w e should n o t e that w e are taking a somewhat different dir ection of inquiry than that followed b y most research in this field. Very broadly, research into the whole range of idiomatic language (including n o t only figurative expressions b u t also proverbs, oxymora, "syntactic idioms" etc.) h a s b e e n dir ected toward uncovering the following: 1
(a) T h e grammatical characteristics of, a n d structures underlying, idiomatic constructions (Chafe 1968, Weinreich 1969, Fillmore et al. 1988, Abeille 1995, Nicholas 1995, Schenk 1995). (b) T h e semantic interpretative competencies a n d cognitive processing involved in their c o m p r e h e n s i o n (Bobrow& Bell 1973, Gibbs 1980, Estill&Kemper 1982, Gibbs 1987, Popiel & M c R a e 1988, Schraw et al. 1988, Schweigert & Moates 1988, Gibbs & Kearney 1994). (c) T h e connections between syntactic properties of idioms a n d their inter pretation (e.g. Gibbs & Nayak 1989). 2
Across all these fields of research in linguistics a n d (psycho) linguistics, the data or test materials are almost invariably instances of idioms abstracted from any actual dialogic context (indeed, m u c h of the research, especially in the area of idiom comprehension, is experimental). I n short, it appears that there has b e e n Utile systematic investigation into the use of idioms in naturally occurring dialog, at least, in EngUsh. By contrast, w e are inquiring h e r e into the interactional role that figurative expressions play w h e n they are used in conversation. Figurative expressions are a feature of t h e design of turns at talk: T h e y are selected as part of the way turns are constructed so as to "fit" with, or play a coherent interactional role in, the sequences in which those turns are p r o d u c e d . T h u s the c o m p o n e n t s of a turn's construction, at whatever level of tinguistic production, are connected with the activity that the turn is b e i n g designed to perform in the unfolding interactional sequence of which it is a part, a n d to t h e further development of which it contributes. T h a t is, the most proximate context in which a turn is produced, a n d i n which it is recognizably coherent, is its S E Q U E N T I A L context H e n c e w e treat figurative expressions as o n e of those linguistic c o m p o n e n t s of turn design through which speakers m a n a g e , collaboratively, certain sequen tially e m b e d d e d activities. We explore h e r e whether an account c a n b e given of the occasions w h e n speakers employ figurative expressions, in terms of their S E Q U E N T I A L D I S T R I B U T I O N or E N V I R O N M E N T S . Therefore, rather than seeking any kind of "cognitive processing" account for speakers' selections of figurative ex pressions in speech, w e can regard using figurative expressions as a systematic, socially organized practice for producing certain kinds of (verbal) conduct - as, indeed, are all components of the design of turns at talk. J u s t what that organized practice, a n d what activities figurative expressions m a y b e employed to perform in conversation, is the focus of this inquiry. 3
4
r*,r
«R ,MM™ o v t i r o c d n n s " m
w e m p a n those conventionalized complex units
DREW A N D HOLT
FIGURES OF SPEECH
99
that are phrasal (e.g. "stark naked," "by and large," "zero tolerance," "window of opportunity") or even lexical. (For an overview, see the introduction to Everaert et al. 1995.) I n c o m m o n with these other idiomatic complex units, figures of speech are formulaic or relatively "fixed" in composition - syntactically, lexically, and sometimes intonationally. H e n c e , like words, they m a y b e learned separately as single units of the language (Fillmore et al. 1988; but for a different view, see Gibbs & Nayak 1989). Generally, they are semantically opaque; their meanings are not the same as, or cannot b e found from, the meanings of their constituent words (Bolinger 1976, Weinreich 1969). This creates the distinction between their literal and figurative meanings - a distinction which, of course, is the starting point of m u c h of the research literature cited above o n the recognition and comprehension of idioms. At any rate, the conveyed meanings of such expres sions are not literal b u t metaphorical, as is illustrated b y "had a good innings," "shower with gifts," and "the p e n n y h a d dropped," in exx. 1,2, a n d 3 respectively. The metaphorical character of figurative expressions is therefore another pro perty that distinguishes t h e m from some other routinized or relatively fixed expressions (including "automatic language," Van Lancker & Kempler 1987) which are part of the idiolect of a language. O u r primary data are a corpus of telephone calls recorded b y a British family at intervals over a three-year period, consisting of approximately 15 hours of telephone conversations between family m e m b e r s , between m e m b e r s of the family and their friends, and occasionally with colleagues. This corpus was sup plemented b y various other corpora of A m e r i c a n a n d British telephone calls, amounting to about 2 0 hours of recorded a n d transcribed conversations (for the transcription conventions, see Atkinson & Heritage 1984). F r o m these data a sample collection was m a d e of approximately 2 0 0 instances of speakers using figurative expressions. 5
8
7
8
Figurative Expressions i n Topic Termination S e q u e n c e s An initial and fundamental step toward identifying the patterns or organizations associated with a linguistic p h e n o m e n o n is to see whether that p h e n o m e n o n has any discernible distribution in talk. While the sense of, or basis for, "distributionalizing a p h e n o m e n o n " can v a r y according to one's analytic perspective, here we m e a n determining whether the p h e n o m e n o n has any regular or recurrent position in sequences of talk (following the sense of "distributionalize" used b y Sacks 1992:2.422). W h e n w e adopted this analytic way of examining the col lection of figurative expressions in our corpus, a very striking pattern emerged. In this pattern, following a turn in which a speaker produces a figurative expres sion, the co-participants briefly agree with one another, after which o n e or the other introduces a n e w topic of conversation. H e n c e it appears that the use of a figurative expression is associated with topic termination a n d transition to a new/next topic. 9
To illustrate
the*
ni-i«fJt-»al
foot....—
:
1
100
DISCOURSE STUDIES
expression "had a good innings" - though now considering the turn in which that was produced in relation to more of the surrounding talk. Lesley and her mother had been talking, for some minutes before this extract, about the illnesses (and possibly "final" illnesses) of some elderly friends. (4) [Field:X:(C):l:l:l:6] 1 Lesley: I don't know'f you remember Missiz Milbeck th't 2 use to go to chu:rch. 3 (0.8) 4 Mum: (Mi[ssiz) 5 Lesley: [Uh: uh- he. wz the vicar's ward'n anyway he 6 die:d suddenly this week .hhh and he wz jtill 7 wo:rking. 8 (0.3) 9 Mum: ( [ ) 10 Lesley: [He was seventy ni:ne, 11 (0.3) 12 Mum: My.: wo:rd? 13 (0.2) 14 Lesley: Y[e:s he: was um 15 Mum: [(You've got s'm read) workers down theire, 16 heh 17 Lesley: .hhh He wz a p - uh: Ye:s. Indee:d .hh He wz a 18 (0.2) .p a buyer for the hoh- i - the 19 only hgrse hair fact'ry left in England. 20 Mum: Good gracious. 21 (0.3) 22 Lesley: And he wz their buyer, 23 (.) 24 Mum: Hm::: 25 Lesley: .t 26 Mum: Hm:. 27 Lesley: So he. had a good innungs did[n't he. 28 Mum: [I should say. so: 29 Ye.:s. 30 (0.2) 31 Mum: Marvellous 32 Lesley: .tk.hhhh Anyway we had a very good evening o:n 33 Saturdaiy. 34 (.) 35 Mum: Ye;s? 36 Lesley: We went to North Cadbury: an' Gordon came too ...
The kind of topic transition with which the production of a figurative expression is associated in our corpus occurs in 4:27-33. In response to Lesley's producing a figurative version of this acquaintance's having lived a long and full life (4:27), Mum assents very positively to that assessment, "I should gay. so: Ye.:s. (0.2) \* n„,.„» ^ . O Q _ Q ^ Aft<w riiic T J J S I P V introduces what is evidendv a Quite
DREW A N D HOLT
FIGURES OF SPEECH
101
Before considering in more detail some of the features of that topic transition sequence, and the role within it of the use of a figurative expression, we present further instances of this pattern of topic transition. To highlight the pattern we are identifying, the turn in which one speaker produces a figurative expression is marked by arrow 1; the turn(s) in which the recipient briefly agrees or con curs with the idiomatic assessment (sometimes followed by the speaker who produced that assessment) is shown as arrow 2; and then the turn in which one or the other speaker introduces a new topic is shown as arrow 3. (5) [Field:J86:1:4:6] Gwen: You know it wz: so lovely an' everything[y know Lesley: [nYg:s. Gwen: AH character 'n (0.3) beautiful. Lesley: Ye:s. (0.7) - > Gwen: B't I suppose she must V come € the end of 'er (.) tether 'n just walked out the [n. —» Lesley: [Yes. (0.8) -> Lesley: Oh what a shame. (•) —> Gwen: Ye:s [it's a shame —> Lesley: [Anywa:y e - so you don't know any mo:re th'n ... (6) [Field:2:3:9] (The "he" referred to is Robert Maxwell, a well-known British publisher and businessman) Steven: Well he didn't either 'ee had a bad start (when) 'ee had iz (0.3) .t.k .hh father shfit by the £Jazis 'nd iz uh uh .hh mother died in: Auschwitz yih know [so Lesley: [Qh really:?= Steven: =So eez [had the: ( ) Lesley: [Oh'z a Je:w is he. Jeiw? (•)
Steven: Q_h yeah. (•) 1 —» Steven: He's had k- eez a Czechoslovakia Jew so [eez Lesley: [Yes 1 —> Steven: had quite a- checkered career already= Lesley: =eh heh 2 -» Steven: .hhhh [Yeah 2 -> Lesley: [Ye:h. (0.2) 3 —> Lesley: .hhh Alri:ght. Well I'dl get my husband then: to get in touch with the address. = Steven: =Thanks very much. (7) [Kamunsky III: 16] (Talking about a mutual friend's car) Myrn: Qh. What color izzit. =
DISCOURSE STUDIES
Myrn: Alan: Myrn:
£) how gro [ :ss. [That's the color V iz car, He's (hanitel- he p - uh bl-l-luh.) He colored it dark bro [iwn? Alan: [Well he wanted a burnt orange, en it came out shit hlfi:wn. (0.3) Alan: ihhhhhhh[ih .hhh ih-hn-hn ihhee Myrn: [Ooo::::::::::::::. Alan: it's rilly l:loo(h)oo heh it's n h - hnh-hu[hh it'sMyrn: [D'z it lookbaid.?= 1 —> Alan: = Yeheh- heh well it grows o(h)n yeh- heh [.hheh 2 -> Myrn: [Ooo::[:oh:: Alan: [I call it 'iz pudding on wheeds, (•)
3 —> Alan:
.hhh B't ghhhhnywhhay,. .((move to talk about who else will be at the party)) (8) [JGI(S):X15:4-5] P: but u - certain: (.) things wi:ll do that you know they're re- ruthey're BOUND to [.hhh in certain industry. M: [Yah. (0.2) M: Yah, P: Djfferent things'll pick up when it- begins to be spring of the yea:r and everything, M: Yah. 1 -> P: .hhh But I think it'll iron itself out, 2 —> M: I sure hope [so. 3 —> P: [Pll see you Tuesday. (9) [Rah:(18):5-6] (Discussing difficulties cleaning double glazed windows) Jenny: An' the trouble is you see if you tighten th'clips too much they snap. Ann: Yes. Well that's what I do. Breakin'them. Jenny: Ye:s:[:, Ann: [I've been a bit more careful this time but the trouble is I don't get th'm ( ) tih the windows actually. (•) 1 -» Ann: You can't wi:n really. 2 -> Jenny: No::. (•)
2 —» Jenny: 2 -» Ann: 3 —> Ann:
Oh no. (0.2) No. (0.3) Have you been to schogj this morning.
DREW A N D HOLT
FIGURES OF SPEECH
103
(10) [Field:M88:l:5:ll] Robbie: . . . but I spoze we de learn 'n she hasn't been to other sdiools'n I've learned 'n awf'l lot 'n last three years. Lesley: Mm:. (0.2) Lesley: Mm. 1 —> Robbie: B't take this with a dollop'v salt you kno::w I'm-I'm baisic'ly quite happy b't quite relieved it's the sheer organization 'n getting all, everything done in th' da:y. 2 -» Lesley: Ygs: that's ri:ght,= 2 -> Robbie: =Ye[s. 3 -» Lesley: [Yes. Ye[h .tch .hhhhhh. Wuh3 —> Robbie: [Qk a y, W'ddiyou wanna talk f me abou(h)[t 3 -> Lesley: [Eh: WE:11 eh WHAT I RANG up about was . . . . These cases illustrate clearly a sequential pattern in which a figurative expres sion is produced in a turn that results in the termination of a current topic. That termination is managed b y the participants in a brief exchange of agreements, followed by a transition to a new/next topic. From this pattern it appears that figurative expressions are associated with an organization for closing a current topic of talk and moving on to a next topic. Several components of this organ ization for topic termination/transition will be examined here in closer detail: (i) the character of figurative expressions in this position as summaries of the current topic; (ii) the sequence through which topic transition is managed; and (iii) the ways in which speakers construct a next turn as introducing a "new" topic.
Figurative Expressions as S u m m a r i e s Ai the point where each of extracts 4-10 begins, a speaker is telling the coparticipant about something: about the death of a mutual friend in 4, the h o m e )f friends who have just separated in 5, the troubled life of Robert Maxwell in 6, the car of a friend in 7, and so forth. The speakers are reporting empirical facts or details; and in each case, the one doing the reporting is the speaker who pro duces the turn in which a figurative expression is used. Thus, just before ex. 6, Steven and Lesley have been talking about the famous (perhaps infamous) British publisher Robert Maxwell, touching among other things on the rumor that he has a son who is so severely disabled that he lives permanently in hospital, and on Maxwell's risky business dealings (data not shown). Here, at the beginning of the extract, Steven tells Lesley about the death of Maxwell's parents when he was young. He then produces a figurative summary of these tragic circumstances and allegedly nefarious dealings: "so eez had quite a- checkered career already-." The character of this as a summary is perhaps particularly evident, in that this is designed as an upshot of the prior detailing, signaled b y "so . . . " (see also 4:27).
104
DISCOURSE STUDIES
I n 9, A n n has b e e n telling J e n n y about her difficulties cleaning her doubleglazed windows, concluding h e r account of these difficulties with "you can't wi:n really." I n each case, the empirical detailing or reporting is brought to a conclusion with a figurative expression which s o m e h o w summarizes what the speaker has b e e n reporting. Used in this way, these figurative expressions serve as s u m m a r y assessments of what has b e e n reported - assessments being generic forms of summaries (Jefferson 1984:211). T h e y are "assessments" in that these expressions convey a certain positive or negative value to b e attached to the circumstances that the speaker has b e e n describing, including optimism ("it'll iron itself out") or re signation ("you can't win"). I n producing figurative assessments, speakers move away from, or step out of, their report of empirical details. I n ex. 4, w h e n Lesley uses the idiomatic assessment "he h a d a good innings," she is n o t n o w adding to the list of facts about the m a n that m a d e his life, a n d death, notable. T h e figurative expressions do not themselves include further information about the person who died in 4, n o r about the house of the friends w h o h a v e separated in 5, Maxwell's early life in 6, the friend's car in 7, or the economic difficulties facing the P's industry in 8. I n p r o d u c i n g a figurative expression, the speakers are becoming empirically disengaged. T h e y are n o t contributing further details b u t rather are assessing a n d summarizing the previously reported empirical information. Em ploying a figurative expression can b e a device to summarize a topic, and thereby to draw it to a conclusion. This disengagement from empirical details, m a n a g e d through producing a figurative assessment, is highlighted b y the way in which that assessment is not connected particularly or exclusively to its contiguously prior turn. Rather, the expression relates back over m u c h of the prior talk/details, to assess aspects of the topic that h a v e b e e n talked about in earlier turns; indeed, the expression m a y refer back to a n d summarize or assess circumstances talked about over the whole of that topic. I n ex. 4, w h e n Lesley says that h e "had a good innings," she is not merely summarizing the information in her immediately prior turns. T h e idiom plainly does not refer particularly to the immediately prior information that, w h e n the friend died, h e was a b u y e r for the only horse-hair factory in England. Instead, it refers back over the whole topic, including the details in each of Lesley's turns in the extract - that h e was the vicar's warden, that h e was still working w h e n h e died, a n d that h e died aged 79. By connecting back beyond the adjacently prior turn/information, the expression " h a d a good innings'" summarizes the information across the entire prior announcement/telling, and it thereby begins to detach the talk from an item-by-item sequential development of the topic. Although the data are too extensive to explicate this in detail, this is also true of the other examples cited above. I n ex. 5, G w e n a n d Lesley have been discussing the b r e a k u p of the marriage of people they know, in the course of which G w e n m e n t i o n e d h o w lovely and full of character this couple's home was (see Gwen's first two turns). Again, "coming to the e n d of her tether" is a
DREW A N D HOLT
FIGURES OF SPEECH
105
we have already m e n t i o n e d that Lesley a n d Steven were previously discussing Maxwell's business dealings a n d the r u m o r concerning his disabled son; h e n c e , when Steven says that Maxwell has h a d "quite a - checkered career," h e can b e heard as referring back to a n d summarizing all those troubled circumstances, as well as the immediately prior information that b o t h Maxwell's parents were killed b y t h e Nazis. I n 10, Robbie's figurative s u m m a r y "take with a dollop of salt" refers back over m u c h of the very lengthy prior topic, concerning h e r ex periences teaching at a school at which Lesley has also taught. T h u s a figurative summary can serve as a device to close d o w n a topic b y connecting back b e y o n d the immediately prior turn or information, often to t h e beginning of the topic. A figurative expression has the property of being able to summarize a n d assess more than whatever is referred to in its contiguous prior turn; in this w a y it c a n detach the talk from the progressive development of a "next item" of information relevant to an ongoing topic. There are, of course, other m e a n s besides figurative assessments b y which a current topic can b e brought to a close; e.g., repetition is c o m m o n l y associated with terminating a topic. Moreover, there are other kinds of assessments besides figurative expressions with which a speaker c a n summarize the talk thus far. Hence the account w e are developing here is not a general account for the termin ation of topics: T h a t would require a n exploration of the range of techniques b y which speakers c a n s u m m a r i z e a n d disengage from w h a t they h a v e b e e n talking about (cf. M a y n a r d 1980, Button 1990). O u r focus h e r e is to account for a systematic position i n which speakers recurrentiy employ figurative ex pressions - namely, that they d o so in summarizing a topic, a n d thereby d r a w it to a close. It is worth noticing, however, that figurative expressions m a y have a special place a m o n g the techniques available for summarizing a n d concluding a topic. First, it is striking that they are commonly, i n d e e d overwhelmingly, employed by speakers to summarize what they themselves h a v e b e e n telling/reporting, rather than what the other, their co-participant, has b e e n telling. T h e y are, then, "self-summaries": summaries of one's own talk/topic, a n d n o t of the other's talk. Furthermore, their u s e as generic assessments detached from empirical particulars m a y give t h e m a certain "power" in summarizing a n d closing topics e.g., in contrast to t h e use of repetitions or other forms of assessment, which a r e more closely tied to prior empirical details (Pomerantz 1984a). A glimpse of this can be seen in 4:17-27, where Lesley initially repeats one of the details concerning the friend's career. After M u m ' s initial response to the details of this m a n ' s e m ployment, a n d a pause (4:20-21), Lesley repeats that "he wz their buyer" (4:22); following this, there is something of a hiatus - a slight pause - before M u m does repeated ruminative acknowledgements. I n outiine, it appears very m u c h as though the topic termination is i m m i n e n t before Lesley's figurative s u m m a r y in 4:27. It seems that Lesley is looking for a w a y to close t h e topic, tries with a repetition that does n o t quite succeed (in repeating "Hm:::," M u m doesn't find a way to m o v e out of that topic a n d onto a next), a n d then tries again with a 10
"SITOnH'er" r.lnsinor m n t m
Kxr » w ~ J — - •
ft
—
DISCOURSE STUDIES
100
they fail in an initial attempt at topic closure, in ex. 4 through repetition. This suggests that figurative expressions m a y h a v e greater "power" in drawing topics to a conclusion than other closing techniques such as repetition or (non-figurative) assessments; this represents a rather different approach to the "vividness" (Chafe 1968) or "intensity" (Labov 1984) of idiomatic a n d figurative expressions.
T h e Topic Transition S e q u e n c e We oudined in the previous section the properties that are associated with the s u m m a r y status of figurative expressions, a n d which m a k e t h e m objects for concluding a topic. T h e key evidence for their terminal character, however, is that co-participants treat these figurative turns as terminal, as bringing the topics to an end. This is manifest in the brief turns that c o m e after the figurative turn and before the introduction of the n e w topic. In these turns, the co-participants manifesdy decline to develop the topic further. Consider, for example, the se quence in ex. 9, following A n n ' s figurative s u m m a r y of the trouble she is having cleaning double-glazed windows. (from 9) Ann: Jenny: Jenny: Ann: Ann:
You can't wi:n rgally. No,".
M W Oh no. (0.2) No. (0.3) Have you been to school this morning.
T h e recipient's (Jenny's) response to the s u m m a r y "You can't wi:n really" is to agree: "No::. (.) O h n o . " Ann, in turn, does a token reciprocal agreement or con firmation, "No", after which she introduces a n e w topic. T h e recipient h a d the opportunity, in the slot after the figurative turn, to continue talking about the topic of Ann's difficulties cleaning windows; she could have developed or elabor ated that topic in a variety of ways. Instead, she produces a near-minimal unelaborated agreement; in doing so, the recipient declines the opportunity to develop that topic any further. H e r response is topically fitted, but it does not offer a n y further resources to talk m o r e about that topic. H e n c e she begins the disengagement from and closure of the current topic. W h e n in response A n n produces a (nearly) identical token, she likewise de clines to take the opportunity in that slot to say any m o r e about the previous topic. She thereby accepts or confirms the prior m o v e b y J e n n y to disengage from the current topic; as a result, they are n o w collaboratively m o v i n g toward the closure of that topic. T h a t closure is finally sealed b y A n n ' s introduction of ™ .i 1 „ fi ,„Hvp e u m m a r v . followed b y each of m
DREW AND HOLT FIGURES OF SPEECH
107
sequence through which co-participants collaboratively disengage from a current topic and move to a next. Two further illustrations of that sequence follow. (from 6) 1 —> Steven: Lesley: 1 —)• Steven: Lesley: 2 -> Steven: 2 -> Lesley: 3
He's had k- eez a Czechoslovakianjew so[eez [Yes had quite a- checkered career already= =eh heh .hhhh[Yeah JYa:h. (0.2) -» Lesley: .hhh Ahi:ght. Well Fill get my husband then: to get in touch with the address. =
(from 10) 1 —> Robbie: B't take this with a dollop 'v salt you. kno: :w I'm— I'm baisic'ly quite happy b't quite relieved it's the shg.er organization 'n getting all, everything dsne in th' da:y. 2 —» Lesley: Ye.s: that's ri:ght,= 2 -> Robbie: =Ye[s. 3 -» Lesley: [Yes. Ye [h .tch .hhhhhh Wuh3 —> Robbie: [O k a y, W'ddiyou wanna talk t' me abou(h)t In each of these fragments, the turn in which one speaker produces a figurative summary (arrow 1) is followed by reciprocal and nearly identical "agreements" by both speakers (arrow 2); after this, one of them introduces a n e w topic (arrow 3). "Agreement" is meant broadly, to include topically fitted or appropriate responses such as the sympathy tokens ("Oh what a shame") with which Lesley responds and Gwen reciprocates in 5, and the sympathetic or affiliative manner in which M responds to P's report of her troubles in 8. Agreement, sympathy, or affiliation may describe the particular response b y a recipient to the figurative summary in question; in any case, contiguity or alignment between the coparticipants is expressed in some way (i.e. through the minimal character of the agreements, typically "Yes" or some such object) so as to foreclose further talk about that topic and to enable the speakers to move o n to a next topic. 12
Thus the topic transition sequence associated with figurative summaries (which may apply more generally to topic transitions following other forms of topical summary/closure) exhibits a kind of standard form, which can be repre sented schematically as: 1 2 3 4
—> -» —» —>
Speaker A: Speaker B: Speaker A: Speaker A/B:
Figurative summary Agreement (or other expression of contiguity) Agreement/confirmation Introduces next topic
This can be considered a "standard sequence" for topical closure and transition to a next topic, after a toDical summarv - thr> nmA^r-n^ ~r ~ -e ^
"5
108
DISCOURSE STUDIES
elided - e.g., instead of p r o d u c i n g a further agreement/confirmation token in response to recipient's agreement, speaker A (the o n e w h o p r o d u c e d the figurative summary) moves straight to the introduction of a n e w topic, as in ex. 8. (from 8) 1 - » P: .hhh But I think it'll iron itself out, 2 —» M: I sure hope [so. 4 -» P: [I'll see you Tuesday. I n other cases, speaker B (the recipient of the figurative summary) m a y pro d u c e a minimal agreement token a n d then o p e n a n e w topic right away, without waiting for a n y further token from speaker A. I n either case, the production of a figurative s u m m a r y (arrow 1) is responded to b y a minimal form of concurrence b y the recipient (arrow 2), which is followed b y the introduction of a n e w topic (arrow 4). This results in a n attenuated form of the standard sequence, through the elision of step 3 in the schematic m o d e l above. This is evidence that the re cipient's topically disengaged m i n i m a l concurrence with the prior speaker's production of the figurative s u m m a r y is sufficient to give either speaker the op portunity to e m b a r k o n a n e w topic. T h e standard transition sequence that we h a v e described represents the sequence in which co-participants collaborate to close o n e topic of conversation a n d begin a n e x t topic. T h e significance of this sequence is that it reflects - or, m o r e properly, is the product of - the fact that co-participants regularly a n d sys tematically orient to the s u m m a r y a n d closing implications of the figurative expression in the prior turn. T h a t orientation is displayed in the way that both speakers systematically withhold any further topical development or elaboration. By producing minimal agreements (or other contiguous actions), recipients dis play their recognition that the idiomatic turn is s o m e h o w sufficient to termin ate that topic; their responses are therefore designed not to say any m o r e about that topic, b u t rather to exit from it. Subsequently, speaker A collaborates in the closure, either t h r o u g h further minimal agreements (frequently repetitions of speaker B's response), or b y eliding the sequence and going straight to a next topic. Speaker A thereby concurs with speaker B's understanding that the figur ative turn was designed to conclude a n d terminate the prior topic. T h u s the standard topic transition sequence is evidence of a symmetry be tween speakers' production of figurative idioms a n d their interpretation by recipients - w h e r e "interpretation" refers not to what the expressions themselves " m e a n " (in the w a y that psycholinguists have investigated the "interpretation" of idioms), b u t to the conversational activity they perform (topical closing). This symmetry, or mutual orientation to the topically terminal implicativeness of a turn in which a figurative expression has b e e n produced, is generally embedded in this topic transition sequence. It is particularly visible, however, in cases where the co-participants SIMULTANEOUSLY a n d INDEPENDENTLY m o v e to step 4 of the se quence, the opening of a next topic. This h a p p e n s in ex. 10. Recall that Robbie summarizes the difficulties she's b e e n having at the school w h e r e she teaches, sneakine fieiirativelv: "B't take this with a dnllon 'v salt voti knn::w". Lesley -
13
DREW A N D HOLT
FIGURES OF SPEECH
109
are characteristic of the standard transition sequence, steps 2 a n d 3. After this, each of t h e m simultaneously changes topic - at least, it appears that Lesley is about to, w h e n she drops out in response to the overlapping, topically initial enquiry from Robbie. I n h e r arrowed turn, Lesley does a n emphatic inbreath (characteristic of some disjunctive next move), followed b y what is clearly recog nizable as the beginning of " W h a t " ("Wuh-"). which is almost certainly the start of " W h a t I called about w a s . . . " But her m o v e to that n e x t topic coincides with Robbie (in her arrowed turn) simultaneously enquiring: "Okay, W'ddiyou w a n n a talk f m e abou-(h)f (which also begins with increased amplitude). H e n c e , Lesley and Robbie simultaneously perform the same action: m o v i n g to a n e w n e x t topic. As it h a p p e n s , they are probably also b o t h orienting to the relevance of not just any n e w topic, but specifically the matter of w h y Lesley has telephoned (this occurs about 15 minutes into the call); i.e., it appears that Lesley was about to announce w h y she called, at the same time as R o b b i e asks w h y she called. At any rate, they are simultaneously treating this as a place w h e r e it is relevant to introduce a n e w topic. T h e significance of their simultaneous m o v e to a n e w topic is that it provides evidence for the shared intersubjective "reality" of this position (to post the pro duction of a figurative idiom) as a point w h e r e it is appropriate to change topic. Instances in which b o t h participants e m b a r k o n the same action simultaneously (in overlap) serve as particularly strong evidence that a n observable pattern is the product of a shared understanding (which, of course, n e e d n o t b e conscious) in this context, that figurative expressions can b e used to summarize and close down topics, a n d thereby to occasion a m o v e to next topics. For this reason, we might regard the use of figurative expressions as a n intersubjectively avail able practice or device for topic termination.
Topic Transition: I n t r o d u c i n g a N e w Topic
I
Recall that w e are describing the c o m p o n e n t s of the sequence in which the use of a figurative expression results in the termination of that topic and the transition to a new topic. We h a v e discussed h o w idioms can work to summarize the pre ceding talk/topic; a n d w e h a v e outlined the transition sequence in which par ticipants collaboratively disengage themselves from a prior topic. T h e final component - the introduction of a n e x t / n e w topic - is of central importance: It is only w h e n participants m o v e to a n e w topic that w h a t h a s b e e n an incipient Dransition sequence, until that point, b e c o m e s fully realized or ratified as a ter mination of the prior topic a n d a transition to a n e x t topic. We have b e e n relying until n o w on a n intuitive, but nonetheless robust, sense of how one participant or the other introduces a n e x t or n e w topic. For example, in 10, the topic that R o b b i e draws to a conclusion with the figurative expression "with a dollop V salt" has b e e n her teaching experiences. As w e h a v e seen, she then changes the subject b y inquiring what Lesley has telephoned h e r about -
DISCOURSE STUDIES
110
of a mutual acquaintance. W h e n Lesley subsequently announces "Anyway we h a d a very good evening o n Sjj.turda:y.," that is rather clearly a change of topic. I n each of exx. 4 - 1 0 , the change of topic is similarly "obvious." T h e difficulties associated with defining what constitutes a topic, and with specifying analytically the various referential a n d other linguistic m e a n s (shared reference, propositional content, pro-terming, a n a p h o r a a n d deixis, lexical repetition etc.) through which topical connections a n d hence discourse cohesion are m a n a g e d , are well e n o u g h k n o w n n o t to n e e d recapitulation h e r e (but cf. Li 1976, Schank 1977, M a y n a r d 1980, Brown & Yule 1983 chap. 3, Levinson 1983: 312-16, Schegloff 1990). But note that one important respect in which the concept of topic m a y b e difficult to apply to the analysis of natural discourse is the prob lem of identifying topical boundaries. Generally, one topic of conversation merges almost i m p e r c e p t i b l y into the n e x t , i n a seamless "stepwise" progression (Jefferson 1984, Sacks 1992:2.291-302). H e n c e the precise points of topic shifts or changes, a n d h o w such shifts w e r e m a n a g e d , are often analytically opaque. Of course, if o n e cannot identify the boundaries of particular topics, then it is difficult to offer an account of the linguistic basis for topical coherence, either in principle or for actual topics. T h e changes of topic that follow figurative summaries of the prior topic in exx. 4 - 1 0 are not, however, m a n a g e d in a stepwise fashion. T h e y are "obvious" changes, not only in the intuitive sense of their involving a dramatic change in reference or content, b u t m o r e particularly because, in most cases, speakers m a r k that they are a b o u t to change the topic of conversation. T h e y d o so, gen erally, in the design of the prefatory (turn-initial) components/features of the turn in which they introduce the n e w topic (arrowed in the following fragments). (from 4) Mum: -» Lesley: (from #6) Steven: Lesley: -> Lesley: (from 10) Lesley: Robbie: -» Lesley: -> Robbie: (11)
Marvellous. .tk.hhhh Anyway we had a very good evening on Saturda:y.... .hhhh [Yeah [Yg:h. •• , . (0.2) .hhh Alright. Well I'dl get my husband then; to get in touch with the address. 1
Yes: that's ri:ght,= ' =Ye[s. i [Yes. Ye[h .tch .hhhhhh Wuh[Q k a y, W'ddiyou wanna talk t' me abou(h)t
[Field:88U:l:9:ll] (Simplified) Dana: I can't really say hey Mum I've got prob ['ms Gordon: [hhh ehhh hgh hgh Dana: She'd run [ amide.](0.3) [( )
DREW AND HOLT
Dana: —> Dana:
FIGURES OF SPEECH
ill
There you go:. (0.9) WeTdl- (0.4) uh::m,hh 1(h) can't think V anything else u-reanlly exciting to say?
When, for example, Lesley announces in 4 that she a n d her h u s b a n d " h a d a very good evening o n S a t u r d a y , " she marks this as a change to a n e w topic b y beginning her turn with an audible inbreath; she then produces a prefatory dis continuity m a r k e r "Anyway," the beginning of which is p r o d u c e d with increased amplitude or stress. Such prefatory c o m p o n e n t s as "Anyway", "Alright" (6), "Well" (11), " O k a y " (10), either alone or in combination ("But anyway" in 7, "Alright well" in 6), are disjunctive in that they work to disengage the forthcoming turn from being tied or connected to, or coherent with, its prior turn; i.e., such components are p r o d u c e d to disengage this n e x t turn, topically, from its prior. Hence, b y producing such c o m p o n e n t s - often in conjunction with some com bination of audible inbreaths a n d raised amplitude - the speakers in such cases signal that what they are about to say will b e u n c o n n e c t e d with what they have previously b e e n discussing, and h e n c e that they are about to introduce a new topic. This is clear evidence that participants themselves treat what they are doing as disengaging from the prior topic, a n d as introducing a n e w a n d quite different topic. In most instances in our collection of topical transitions following the use of a figurative idiom, the introduction of the n e w topic is signaled b y (some combination of) increased amplitude, raised pitch/amplitude, a n d self-editing or hesitancy (including inbreath) - a n d , most importandy, b y those prefatory discontinuity markers that suspend the relevance of the prior topic. (That is, they instruct the co-participant n o t to try to look for any connection b e t w e e n what the two of t h e m h a v e just b e e n speaking about, a n d w h a t is about to b e said.) Again, this serves as evidence that the production of a figurative expression, along with the sequence of brief agreements that a figurative turn generates, constitutes a shared practice for closing down a topic and m o v i n g o n to a n e x t topic: a practice to which participants mutually or intersubjectively orient in conversation.
Failure to A c h i e v e Topical Closure Some cases in our collection of figurative expressions did not r u n off in such a straightforward or tighdy organized topic-termination/transition sequence. T h e topic being talked about did n o t e n d after, or in the close vicinity of, the p r o d u c tion of a figurative summary. I n particular, we h a v e two types of cases in which a speaker's attempt to close a topic with a figurative expression did not result in closing that topic - instances that nevertheless relate to the topic termination pat tern we h a v e b e e n describing. T h e first t y p e involves instances w h e r e t h e expression itself haonpns tn nrr-nd™ — ^— • "
112
DISCOURSE STUDIES
Touched-off Topical Developments Devices or practices in conversation do not work in an automatic or mechanistic fashion: The practices evident in conversational patterns are R E S O U R C E S that enable speakers to engage, recurrendy, in certain activities, using means by which those activities will be coherent, recognizable, and meaningful to co-participants. But the use of those resources D O E S N O T D E T E R M I N E the course of the interaction. At any point in an interaction, participants may orient to the possibilities that a conversational practice occasions; nevertheless, they are not obliged or con strained to follow the sequential track implicated in those possibilities. In short, they m a y choose to take a different direction - to suspend the sequential track implicated in an object and instead take a different track. So it is with figurative expressions and the topically terminal possibility their use occasions. In the first set of cases in which the use of a figurative summary does not result in topic ter mination and transition, it appears that one speaker finds something further connected to or touched off b y the figurative expression - so that the talk moves along from it in a more "step-wise" fashion, characteristic of topical development in conversation. (12)
[Field:X(C):2:l:2:7] Joan: YIs. 'aven't even bought a Christmas tree-: we've dragged an old one down that Kenneth use to 'ave in iz bedroom, [.hh Lesley: [ihYe[:s:. Joan: [An o:ld (0.3) sorta make believe one. You[know, .hh[An' made do with= Lesley: [Ye:s, [yes. 1 -» Joan: =tha:t. We're 'ailing to really (.) tighten our 1 -> (0.7) belts. 2&3—> Lesley: ihYes well (0.3) so. are. we:. Because uh(.) it seems to me every body else seems t' be doing ss, we::ll. .((Continues about her husband's doing badly because "nobody's buying anything for agriculture"))
Prior to this extract, Joan has been telling Lesley about the financial difficulties her family faces because of her husband's reduced income, and the economies they are having to make as a consequence - economies which she details, then summarizes figuratively (arrow 1). Lesley responds to this figurative summary, first b y doing a minimal acknowledgment (arrow 2), and then (arrow 3) by con tinuing and developing the topic of reduced income. She shifts from talking about Joan's husband's income to her own family's difficulties in this respect: "well (0.3) SQ are we.:" (i.e., "having to tighten our belts"). Thus there is something of a topical shift (in reference or focus) followingjoan's figurative summary, but n o t the kind of clear and marked topic change characteristic of the sequences
DREW A N D HOLT
FIGURES OF SPEECH
113
in which the figurative expression is PIVOTAL: It is used to summarize the topicthus-far, but it serves to touch off further related matters (see also Jefferson t984:203). In such instances, the figurative expression might h a v e terminated the topic about which the participants h a d b e e n speaking. Recipients p r o d u c e minimal agreement tokens to these summaries, a n d at this point they are potentially dis engaged from that topic. However, one of t h e m t h e n finds in the figurative turn the opportunity to continue somewhat " o n topic." H e n c e the figurative expres sions in such cases b r i n g the participants to a point at which they C O U L D H A V E disengaged from a previous topic and introduced a n e w topic; however, they O E C U N E to d o so, pulling back from disengagement a n d opting instead to con tinue the topic (albeit in a stepwise move). This, then, is the first way in which the standard sequence for closure a n d change of topic is n o t realized, because one participant chooses to exploit the topical opportunities that h a p p e n to b e touched off b y the figurative expression. u
Disagreement, Disaffiliation, and Multiple Figurative Idioms The second type of case in which the production of a figurative s u m m a r y does not result in topic termination a n d transition involves some manifest lack of accord, affiliation, or agreement between the participants. Recall that the standard sequence described above is characterized b y agreement or a measure of accord between them. I n exx. 1 a n d 4 - 1 1 , the recipient of the turn in which a figurative expression has b e e n p r o d u c e d (speaker B in the schematic model) concurs with speaker A's figurative s u m m a r y assessment. These agreements m a y take quite niinimal forms (e.g. "Yes"; or " N o " where appropriate, as in 9, seeJefferson 1994); or they m a y consist of appropriate expressions of sympathy (ex. 5), affiliation (exx. 8 and 10), etc. Whatever form of agreement or accord is used b y the recipi ent, that speaker is in effect n o t only agreeing to the prior speaker's s u m m a r y assessment: H e or she is also implicitiy accepting or confirming the prior speaker's move to close d o w n the previous topic, b y declining to take the opportunity in thai slot to add anything or say anything m o r e substantive about the matter being discussed. I n other words, substantive agreement to the prior speaker's assessment also conveys a kind of "procedural" agreement, to d r a w that topic to an end. T h e recipient, in producing in this slot a minimal agreement (etc.), simultaneously concurs b o t h with the prior speaker's assessment and with his or her move to close that topic (through the figurative summary). I n this fashion, topics can b e closed with a g r e e m e n t b e t w e e n the participants, t h r o u g h t h e relatively brief sequence o u d i n e d schematically a b o v e . 15
To begin with, there are cases that seem to b e a variant of the "standard" topic transition sequence. (13) |Heritage: 1:6:5-6] (Talking about clipping Mrs H's dog's daws) Mrs. H: En that's botherin' me yp [u know 1= Bene: rw>-«
114
DISCOURSE STUDIES
Ilene: Mrs. H:
Well that's it because you've gjnly just got to cut the tips off. Mnu:.
Ilene: Mrs. H: Ilene:
You know you- you musn't cut it very far down, Yes. You musn't cut it onto the biaxk because it's li:ke cutting into our own ouick. Yes of gourse it ti[s. [En they'll scream blue ' rnurder if you d(h)o th(h)a(h) [t [hihhihY(h)ehah Well I've given it up as a bad job anywajy. [eh heh heh hih You knpjw, .h But l:lpok ah I wz (0.2) Fm havin:g stnll a big pro:blem with my: sister's baxk . . .
Mrs. H: 1 —> Ilene: 2 -» Mrs. H: Ilene: 4 -> Mrs. H:
V)
J u s t prior to this extract, Mrs. H . has b e e n explaining to Ilene - w h o evidently breeds dogs, and w h o m Mrs. H . treats as something of an expert - that she has b e e n having difficulties clipping h e r dog's claws. She has asked Ilene if she can r e c o m m e n d a reliable vet in the n e i g h b o r h o o d w h o would do it for her. In b r o a d terms, the extract resembles earlier cases: Speaker A (Ilene) produces a figurative s u m m a r y (arrow 1, "they'll scream blue murder"), in response to which M r s H . produces a minimal agreement form (arrow 2) and subsequently opens a n e w topic (arrow 4) - which h a p p e n s to b e h e r p r i m a r y reason for calling, to ask Ilene's son, a physiotherapist, to treat her sister's back. But in o n e respect, at least, this differs from earlier cases. M r s . H ' s response to Ilene's production of a figurative expression is not restricted to concurring with Ilene's prior turn. Instead, she continues b y adding to h e r agreement d o n e with a slight chuckle, reciprocating the chuckling laughter in Ilene's prior turn - a figure of speech of h e r own: "Well I've given it up. as a b a d j o b anyway.'' Certainly this does n o t extend the termination sequence very far; nevertheless, it results in a termination sequence which is n o t quite so attenuated as the "stand ard" sequence described above. T h e r e is perhaps some slight difference between the positions of the coparticipants at this point; the difference just surfaces w h e n Mrs. H says, "IVg given it up. as a b a d j o b anyway." At the beginning of this extract, Ilene has been asked to r e c o m m e n d a vet w h o might clip the dog's claws, a n d she is warning about the dangers of trying to clip t h e m oneself - a danger she graphically sum marizes figuratively. However, it appears that M r s . H is displaying that she does n o t n e e d to b e advised n o t to attempt to clip t h e m herself: She has already decided it's too difficult for her. H e n c e she is asking Ilene to recommend a vet (data n o t shown), a n d this is m a d e explicit in h e r figurative summary. To a degree, then, Mrs. H RESISTS the advice implicit in Ilene's warning. (On the sensitivities of giving a n d receiving advice, a n d resistance to advice more generally, see Jefferson & Lee 1992, Heritage & Sell 1992.) W h e n Mrs. H adds that she has "given it up. as a b a d job anyway," she conveys that the warning, and J
DREW A N D HOLT
FIGURES OF SPEECH
115
This kind of case, in which the recipient responds to her co-participant's figurative s u m m a r y with one of her own, is fairly benign, in the sense that the difference in the positions expressed figuratively is p e r h a p s only incipient; the difference does not approach overt disagreement or conflict (in this respect, note both Mrs. H ' s initial minimal agreement a n d h e r reciprocal chuckling). Significantly, the co-participants then m o v e successfully to o p e n a n e w topic. In other instances, however, the differences between participants b e c o m e more manifest at the interactional surface of the talk. Although one speaker may attempt to close the topic b y producing a figurative summary, this is not accepted b y the recipient. T h e recipient withholds the kinds of minimal agree ment associated with the standard topic transition sequence, with the result that the co-participants do not achieve a topic termination and transition to n e x t topic. Instead, the topic b e c o m e s protracted until further attempts are m a d e to get the other's agreement, often through additional figurative summaries. (14) [SBL:3:1:3-4] (Talking about asking the committee of a women's organization, of which Marylou and Claire are members, about showing some giftware at an event) 1 Marylou: Maybe we sh' do tha:t. 2 Claire: Mm hnig AN' AN' EH- A:SK how many: MIGHT BE 3 inTRESTED duh came BEC'Z IF YEE DON'T HAVE 4 ENOUGH THET'S INTERESTED W'L THEN (.) fojjey 5 ONITYIH[KNOW. 6 Marylou: [We:ll I don't think it's a matter 'v it 7 having t' be. right no_:w. I think it's something 8 thet will snowba.'ll. 9 (0.7) 10 Marylou: Yihknoi[w? 11 Claire [Mm::[h m . 12 Marylou: [I think it's somethin' thet (.) , 13 '11 haftih be worked up 14 (1.6) 15 Marylou: W[hich is alright en it can be done at 16 Claire [(Wgh-) 17 Marylou: any ti:me en [I don't care whether a: lot 'v'm 18 Claire: [Mmhm ? 19 Marylou: come 'r not because 20 (0.7) 21 Marylou: uh: if they don't wan' to. 22 (0.7) 23 Marylou: But the thing i:s ah: the more the better, 24 (0.3) 25 Marylou: Bu [t they don't haftuh (place order) in the club = 26 Claire: [WE:LL YOU KNOW EVERY O N E A'T H E 27 Marylou: =they c['n u-have their own frie.:nd[s (and uh) 28 Claire: [Mh [.hhh Ever' one 29 a' those oifficers yih know Ham woiFo —
116
DISCOURSE STUDIES
For several minutes before this, they h a v e b e e n talking about Marylou's want ing to show some giftware, which people might purchase, at a n event being organized b y a w o m e n ' s committee; b u t she's n o t sure of the propriety of sug gesting it to them. Claire, another m e m b e r of the committee, has b e e n en couraging Marylou to ask t h e m if they'd like a showing. T h e extract begins at the point w h e r e M a r y l o u concludes (14:1) that " m a y b e w e should" ask the committee about showing the giftware, with which Claire agrees. I n a general sense, Claire is affiliating with Marylou; she is encouraging Marylou to go ahead a n d ask (the committee) "how m a n y might b e interested to c o m e " to a sale of the giftware, o n the grounds that she can b e sure it will sell well (14:26,28-30). T h e r e is, however, a difference between t h e m at another level. Claire agrees that Marylou should ask the committee, in order to j u d g e the likely interest in such a sale - h e r point evidendy b e i n g that if there is insufficient interest, then "fooey O N I T " (14:3-5), presumably conveying that that would b e the e n d of i t Marylou does not assent to this position, taking instead the position: "We:ll I don't think it's a matter 'v it having t' b g right no.:w." (lines 6-7). T h e difference between t h e m is that, for Claire, it would b e useful to get a n indication of whether there is sufficient interest; but Marylou's position is that she could broach it with the committee in the h o p e that interest will build over time. Marylou summarizes h e r position figuratively in 14:7-8, "I think it's some thing thet will snp_wba:ll." This does not elicit from Claire any form of agreement/ concurrence; instead there is a pause (14:9), following which M a r y l o u further solicits Claire's agreement, i n response to which Claire p r o d u c e s only the non-committal token " M m : : h m . " (14:11)- which manifestly withholds agree ment. (Cf. Pomerantz 1984a; and for a further account of the difference between various response tokens, including " M m h m " , a n d their interactional and sequential implicativeness, see G a r d n e r 1997.) This absence or withholding of agreement on the part of Claire to Marylou's p r o p o s e d figurative s u m m a r y re sults in the continuation of the topic. Specifically, Marylou pursues h e r case for believing that it does not matter if few are interested to begin with, because it's something that will gather m o m e n t u m ; in the course of this, Claire continues to withhold agreement (in 14:14, 18, 20, a n d possibly 22). T h e n (14:23) Marylou attempts another idiomatic summary, "But the thing i:s a h : the m o r e the better." (For a n account of such formal idioms, employing a general syntactic pattern here, "the X-er the Y-er" - see Fillmore et al. 1988.) I n continuing to explain the case for h e r position, Marylou can b e seen to b e pursuing C l a i r e ' s agreemeni (Pomerantz 1984b), culminating in h e r producing a second idiomatic summary.' This also fails to elicit Claire's agreement, a n d the matter r e m a i n s unresolved, at least in this extract. Earlier we n o t e d that, in the standard topic transition sequence, there is a conjunction between (a) the recipient's agreement or concurrence with the prior speaker's figurative s u m m a r y assessment, a n d (b) the recipient's accepting the opportunity to terminate that topic a n d m o v e to a next. Recipients' minimal agreements in response to the figurative summaries (sympathy tokens, etc.) conjoin those substantive a n d procedural agreements. Ex. 14 begins rr> illnsrraifl
DREW AND HOLT
FIGURES O F SPEECH
117
production of a figurative expression does not result in topical closure and tran sition. Quite simply, the A B S E N C E of topic closure is associated with a difference or disaffiliation between speakers - in effect, with a form of incipient disagree ment. Such disagreement is not overtly expressed; rather, it is implicit in the absence of agreement in the recipient's response to the figurative summary. The prior speaker (the one who produced that summary) treats that absence as a withholding of agreement by the recipient, as Marylou does in 14; this results in a continuation of the topic and a pursuit of some resolution, perhaps through subsequent attempts to secure agreement. At any rate, when the production of a figurative summary of a topic fails to lead to the termination of that topic and the transition to a next, then the failure and its attendant topic continuation are associated with a lack of accord, and perhaps with incipient disagreement, be tween the participants. A final instance rather dramatically illustrates this association between (a) the failure of a figurative summary to achieve topic closure, and (b) disaffili ation between speakers. It appears, from what has been said immediately before this extract, that Ilene's son is interested in buying a house which Raybee is selling. The estate agents (realtors) acting on Raybee's behalf have claimed that they introduced Ilene's son to the property, through their having sent him details of the house (the documents referred to here) - a claim Ilene disputes. What hangs on this is whether Ilene's son is free to negotiate a price directly with Raybee, or whether he should instead negotiate with the agents ("Moss and Company"). (15)
[Heritage:OI: 1:2-3] 1 Raybee: Well as far as I'm con:cer:ned i:t's: that um I'll 2 haf to accg:pt Moss 'n Company's argument that (0.3) 3 your son was introduced to the property via them. = 4 Ilene: =Yg:s w.ell no.:w .h obviously pjie's going to have 5 to do that but I can asgujre youi .hh that he was 6 not. 7 (•) 8 Ilene: .hhh We've checked now on all the papers 'e has an' 9 Moss 'n Comp'ny said they were sent through the 10 pest we have had n:nothing from Moss 'n Comp'ny II through the post. 12 (0.3) 13 Ilene: Anyway, (.) Tha.'t's th- uh you know you can't (.) 14 argue in it's like (.) uh:[m 15 Raybee: [Well 16 (•) 17 Ilene: banging yer he.ad against a[brick wadl. 18 Raybee: [E z far as I'm 19 concerned on this situation, all private 20 negotiations between us must cejr.se. 91
118
DISCOURSE STUDIES
23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
Ilene: Mmfhm Raybee: [Ah:nd (.) any negotiations you: wish to enter in on the property you have to go via Moss 'n Co. Ilene: Mm:. Raybee: .hh I been on t' th' solicitor (he thought that) yihknow give me s'm: legal gufcdanfce Ilene: [Yeah:. Yah. Raybee: A::nd I'm really left between th' devil 'n deep blue sea: I have no. option BTJ:T.h (0.2) to revert to that. (•)
Raybee: Ilene: Ilene: Raybee: Ilene:
U_h: [: because of the c- the cost involved an:d. [Mm Ye[ah. [w'n it beek- ended up in an argument, Ye:s well it (.) we've just hadda terrific argument wi:th Mistuh Michael. (0.5) Raybee: Oh: U.h::m an.ywa::y uh:m, (0.2) Now he rexkons thez Ilene: been another offer put iyi, Ilene: I don't know whethuh that's true or not.
At the point where this extract begins, R a y b e e informs Ilene that she (Raybee) "has to" accept her agents' version, that they did send the necessary details to Ilene's son. While recognizing that R a y b e e is constrained to d o this (15:4-5), Ilene very directly contests the agents'version (lines 5-11). She then summarizes figuratively the position in which she finds herself: "like banging yer head against a brick wa:ll" (15:14, 17). T h e r e is plainly n o expression of concurrence or agreement b y Raybee, w h o instead continues (in overlap, 15:17-18) to set oui her position (15:18-29), which she then summarizes as b e i n g "left between th' devil'n d e e p blue seal" (15:32-33). Although Ilene's response to that is initially a minimal token of concurrence, "Yeah" (15:38), R a y b e e h a p p e n s to complete her s u m m a r y turn in a fashion which Ilene treats as a n opportunity to continue discussing the dispute between her and Raybee's agents. Neither of the figur ative expressions with which each summarizes her position elicits the other's agreement or concurrence, a n d h e n c e neither results in topical closure. The lack of affiliation between participants is quite manifest in the way that neither accedes to or concurs with the other's figurative s u m m a r y of h e r position. Parenthetically, we can discern from this a rather different account for the p h e n o m e n o n of multiple idioms from that provided particularly b y writers iD stylistics. Noting that idioms often occur in flurries, a n d true to their dismissal o< idioms as belonging to the unoriginal a n d therefore degraded forms of language (e.g. Black 1972:169), they h a v e explained the occurrence of multiple idioms as <~» ~-~„„l,„». , . qti i r l i n m this infecii l
c
Q
O
DREW A N D HOLT
FIGURES OF SPEECH
119
such "contamination," w e can see in these cases that multiple (figurative) idioms may instead b e associated with the pursuit b y each speaker of her position one with which h e r co-participant is reluctant to agree or concur. H e n c e one or the other m a y subsequently p r o d u c e another (alternative) figurative s u m m a r y of her position, possibly in an attempt to close the topic with some minimal agreement b e t w e e n them. Such instances as these in exx. 13-15 illustrate the other major type of cases in our corpus, in which the production of a figurative assessment does n o t result in disengagement from a n d closure of a topic, with the introduction of a n e w topic. This occurs w h e n there is some conflict or non-alignment between partici pants, manifest in part through the recipient's unwillingness to agree to - or perhaps more apdy, in 14-15, to concede to - the prior speaker's figurative assessment.
Conclusion We should m a k e it clear what we h a v e n o t m e a n t to claim in this account of the association between speakers' use of figurative expressions a n d topic termination/ transition in conversation. First, we do n o t suggest that this is a general account of where or w h y figurative expressions are used; this was simply the most strik ing and recurrent pattern or organization evident in our corpus. Undoubtedly, there are other organizations that m a y b e discovered through further research, and that m a y reveal other discourse or pragmatic functions of figures of speech. Second, we are even further from offering a general account of the way that speakers in conversation close o n e topic a n d m o v e on to a next. It is clear that there are other devices or m e t h o d s that speakers can use to summarize a n d bring a topic to a close, including repetition a n d other s u m m a r y objects (some of which have b e e n discussed in the literature; cf. M a y n a r d 1980, Jefferson l 8 4 . and Button 1990). I n neither of these respects, then, are w e proposing general accounts, if "general" is taken to m e a n that this is the only w a y that figurative expressions are used in conversation, or the only w a y that participants manage topical closure in conversation. It is, however, a general account in a different sense: It is an account which cuts across the other kinds of discourse variables with which it is thought figurative idioms are associated (in the very sparse research literature that is relevant here), such as their uses in narratives and stories, in talking about third parties (rather than about self or co-participant, Strassler 1982), or in complaining (Drew & Holt 1988). u
The "logic" of our investigation has b e e n to treat a linguistic object, with whatever syntactic a n d other linguistic properties such objects m a y possess, as an interactional device. T h a t is, the properties of figurative expressions, par ticularly their non-empirical or "general" applicability, are a resource for m a n aging certain interactional tasks in conversation. O n e of those tasks has b e e n identified here, namely summarizing a n d drawing a topic to a close. T h e sequen tial pattern associated with the occurrence of figurative expressions is the product frhp Ifinrt r%f ravnnrr*** J .... ' 1
120
DISCOURSE STUDIES
the sequential distribution of a linguistic object (here, figurative expressions), along with its distribution in topic termination a n d transition sequences, gives us access to the interactional functions of that object, i.e. to the kinds of inter actional "problems" for which participants n e e d to have solutions, a n d the w a y that this object is fitted to m a n a g i n g that task. Focusing o n the sequential dis tribution of a linguistic object helps us, then, to appreciate its social interactional function(s). I m p o r t a n t questions about language processing a n d the cognitive correlates of such processing m a y n e e d to b e addressed through experimental studies of, e.g., the speed of recognition of the literal or figurative m e a n i n g of certain idio matic expressions. But there are equally important questions concerning w h y speakers of a natural language use figurative idioms at all. We h a v e suggested that we should investigate W H E R E they are used, to determine the orderliness of their being used F R O M T I M E T O T I M E (recalling our o p e n i n g sentence). F r o m this orderliness we can detect the social functions (rather than cognitive aspects) of the use of figures of speech in interaction. Finally, w e h a v e shown, through certain details of the topic termination a n d transition sequence, that participants themselves orient to these properties a n d functions of figurative expressions, even in instances w h e r e the production of a figurative expression fails to result in topical closure.
Notes We have presented versions of this paper, over several years, to conferences and groups in many countries. We wish to thank all those, too numerous to list here, who have com mented on this research, and who have made suggestions, observations etc. that have helped us clarify and refine or analysis. We are particularly indebted to Auli Hakulinen and Joerg Bergmann for their incisive and helpful comments. 1. In some cases shown in this paper, the turn in which a figurative expression occurs consists just of that expression, in a single unit turn, as here in ex. 1. However, in other cases, a figurative expression occurs with other surrounding material, in a multiunit turn. 2. A very significant subset of this research effort focuses on linguistic competence and the appropriate comprehension of idioms - or rather, on the "deficient processing of proverbs and related nonliteral expressions" (Van Lancker 1990), either by those whose competencies are not fully developed, especially children (Gibbs 1987), or by those who are impaired, such as aphasics and those with certain neurological ab normalities (Van Lancker & Kempler 1987, Van Lancker 1990, 1991). 3. There are some exceptions, notably Strassler 1982 and McCarthy 1998. These studies explore idioms in naturally occurring spoken English; however, their focus on the pragmatics of idiom usage concerns broader patterns of association or correlations of idiom usage, e.g. between idioms and talk about a third person, rather than about self (Strassler 1982:103). Nevertheless, some of McCarthy's findings (1998, ch. 4), narticularlv concerning the "shifting from the event line to the evaluation line . . . ,
DREW A N D HOLT
121
FIGURES OF SPEECH
4. On grammar as a feature or property of turn design, and thereby sequentially bounded, see Ochs et al. 1996. 5. However, it should be noted that many seemingly "lexical" idioms have their origins in standardized collocations and are really quasi-lexical, e.g."pushover" (meaning "very easy"). 6. But on issues concerning the modification or syntactic flexibility of idioms, see Green 1975 and Nicolas 1995. Ex. 10, below, illustrates the kind of lexical flexibility that some idioms may allow: Here a "dollop" of salt is a recognizable and orderly version of the standard idiom "take with a pinch (or grain) of salt" - orderly, in that "dollop" plainly conveys an upgraded sense of skepticism. 7. In this respect, the kind of expressions we are investigating probably correspond to those which Fillmore et al. (1988:504-5) call "decoding idioms". 8. We gratefully acknowledge the assistance of the Field family and their friends in allowing us to record their conversations; and of Gail Jefferson, who transcribed all these data. 9. By "new," we mean a next topic, different from that which has been concluded following the production of a figurative expression. Hence, speakers may return to a topic previously talked about in the conversation, e.g. in exx. 5-6. The point is, though, that they CHANGE topic. 10. This is only a broad generalization, which does not apply to all cases - particularly since, in many instances where co-participants have been talking about some third party, they may have contributed equally to the "tellings" that make up the topic. (Nevertheless, at the points where figurative expressions are used, the on